Mark Kurt Schutze Thesis 2

The significance of genetic and ecological
diversity in a wide-ranging insect pest,

Paropsis atomaria Olivier
(Coleoptera: Chrysomelidae)
Mark Kurt Schutze
B.Sc. (Hons), University of Queensland
School of Natural Resource Sciences
Queensland University of Technology
Gardens Point Campus
Brisbane, Australia
This thesis is submitted as a requirement for the degree of Doctor of Philosophy

Keywords: cryptic species, local adaptation, phenotypic plasticity, seasonal
plasticity, host specialisation, population genetics, Eucalyptus, forestry, predictive
modelling, body size, Bergmann’s Rule.
Abstract to thesis
Paropsis atomaria (Coleoptera; Chrysomelidae) is a eucalypt feeding leaf beetle
endemic to southern and east coast Australia, and it is an emergent pest of the
eucalypt hardwood industry. Paropsis atomaria was suspected to be a cryptic species
complex based on apparent differences in life history characteristics between
populations, its wide geographical distribution, and extensive host range within
Eucalyptus. In this study genetic and ecological characters of P. atomaria were
examined to determine the likelihood of a cryptic complex, and to identify the nature
and causes of ecological variation within the taxon.
Mitochondrial sequence variation of the gene COI was compared between

populations from the east coast of Australia (South Australia to central Queensland)
to assess genetic divergence between individuals from different localities and host
plant of origin. Individuals from four collection localities used for the molecular
analysis were then compared in a morphometric study to determine if observed
genetic divergence was reflected by morphology, and common-garden trials using
individuals from Lowmead (central Qld) and Canberra (ACT) were conducted to
determine if morphological (body size) variation had a genetic component. Host
plant utilisation (larval survival, development time, and pupal weight) by individuals
from Lowmead and Canberra were then compared to determine whether differential
host plant use had occurred between populations of P. atomaria; individuals from
each population were reared on an allopatric and sympatric host eucalypt species (E.
cloeziana and E. pilularis). Finally, developmental data from each population was
compared and incorporated into a phenology modelling program (Dymex™) using
temperature as the principle factor explaining and predicting population phenology
under field conditions.
Molecular results demonstrated relatively low genetic divergence between

populations of P. atomaria which is concomitant with the single species hypothesis,
however, there is reduced gene flow between northern and southern populations, but
no host plant related genetic structuring. Morphometric data revealed insufficient
evidence to separate populations into different taxa; however a correlation between
latitude and size of adults was discovered, with larger beetles found at lower latitudes
(i.e., adhering to a converse Bergmann cline). Common garden experiments revealed
body size to be driven by both genetic and environmental components. Host plant
utilisation trials showed one host plant, E. cloeziana, to be superior for both northern
and southern P. atomaria populations (increased larval survival and reduced larval
development time). Eucalyptus pilularis had a negative effect on pupal weight for
Lowmead (northern) individuals (to which it is allopatric), but not so for Canberra
(southern) individuals. DYMEX™ modelling showed voltinism to be a highly plastic
trait driven largely by temperature.
Results from across all trials suggest that P. atomaria represents a single species with
populations locally adapted to season length, with no evidence of differential host
plant utilisation between populations. Further, voltinism is a seasonally plastic trait
driven by temperature, but with secondary influential factors such as host plant
quality. These data, taken combined, reveal phenotypic variability within P.
atomaria as the product of multiple abiotic and biotic factors and representing a
complex interplay between local adaptation, phenotypic plasticity, and seasonal
plasticity. Implications for pest management include an understanding of population
structure, nature of local adaptation and host use characteristics, and predictive
models for development of seasonal control regimens.
Table of contents
List of figures ....................................................................................................... i
List of tables......................................................................................................... v
List of publications.............................................................................................. viii
Statement of original authorship ....................................................................... ix
Acknowledgements.............................................................................................. x
CHAPTER 1: General introduction and literature review............................. 1
Patterns and processes of genotypic and phenotypic variation

Variation and cryptic species complexes ................................................ 1
Causes of intraspecific variation ............................................................ 6
Study system ......................................................................................................... 9
Objectives of study................................................................................................ 10
Thesis style............................................................................................................ 13
References ............................................................................................................. 14
CHAPTER 2: Literature review on P. atomaria .............................................. 21
Notes on taxonomy and nomenclature.................................................................. 22
Biology and ecology ............................................................................................. 22
Notes on host range............................................................................................... 24
Symptoms – description........................................................................................ 25
Morphology
Eggs......................................................................................................... 26
Larvae ..................................................................................................... 27
Pupae ...................................................................................................... 28
Adults ...................................................................................................... 29
Similarities to other species .................................................................................. 30
Detection and inspection methods ........................................................................ 30

Geographical distribution......................................................................................31
Invasiveness ..........................................................................................................32
Phytosanitary risk ..................................................................................................32
Means of movement and dispersal
Natural dispersal.....................................................................................32
Silvicultural practices .............................................................................32
Movement in trade ..................................................................................32
Notes on natural enemies ......................................................................................33
Control
Biological control ...................................................................................34
Host plant resistance...............................................................................34
Pheromonal control ................................................................................34
Chemical control.....................................................................................34
Economic impact...................................................................................................35
Environmental impact ...........................................................................................36
References .............................................................................................................37
CHAPTER 3: Species status and population structure of the

Australian Eucalyptus pest Paropsis atomaria Olivier (Coleoptera:
Chrysomelidae)....................................................................................................40
Abstract .................................................................................................................41
Introduction ...........................................................................................................42
Materials and methods ..........................................................................................45
Insect collection and identification .........................................................45
DNA sequencing......................................................................................46
Statistical analysis...................................................................................47
Results ...................................................................................................................48
Sequence variation ..................................................................................48

Variation between sites within P. atomaria............................................. 50
Variation between host plants within P. atomaria .................................. 53
Discussion ............................................................................................................. 53
Acknowledgements ............................................................................................... 56
References ............................................................................................................. 57
CHAPTER 4: Converse Bergmann cline in a Eucalyptus herbivore,

Paropsis atomaria Olivier (Coleoptera: Chrysomelidae): phenotypic
plasticity or local adaptation? ............................................................................ 65
Abstract ................................................................................................................. 66
Introduction ........................................................................................................... 67
Materials and methods .......................................................................................... 69
Temperature data.................................................................................... 69
Body size of field collections................................................................... 69
Common-garden experiments ................................................................. 71
Results ................................................................................................................... 72
Temperature – latitude correlation......................................................... 72
Body size of field collections................................................................... 73
Common-garden experiments ................................................................. 75
Discussion ............................................................................................................. 80
References ............................................................................................................. 83
CHAPTER 5: Larval development of two geographically isolated

populations of Paropsis atomaria on two species of Eucalyptus ...................... 86
Abstract ................................................................................................................. 87
Introduction ........................................................................................................... 88
Materials and methods .......................................................................................... 90
Field collection of adult beetles.............................................................. 91

Host plants ..............................................................................................91
Experiments.............................................................................................92
Results ...................................................................................................................93
Survival to pupation ................................................................................93
Development time....................................................................................94
Pupal weight ...........................................................................................95
Discussion .............................................................................................................97
Acknowledgements ...............................................................................................99
References .............................................................................................................100
CHAPTER 6: Thermal requirements, field mortality and

population phenology modelling of Paropsis atomaria Olivier, an
emergent pest in subtropical hardwood plantations........................................103
Abstract .................................................................................................................104
Introduction ...........................................................................................................105
Materials and methods ..........................................................................................107
Thermal requirements and thresholds for immature P. atomaria
life stages.................................................................................................107
Mortality of immature stages ..................................................................109
Population modelling..............................................................................110
Field validation .......................................................................................114
Results ...................................................................................................................114
Thermal requirements and thresholds for immature P. atomaria
life stages.................................................................................................114
Field mortality of immature life stages ...................................................117
Population modelling and field validation..............................................118
Discussion ...............................................................................................121
Development and mortality.....................................................................121

Population modelling.............................................................................. 123
Future improvements .............................................................................. 123
Conclusion .............................................................................................. 124
References ............................................................................................................. 126
CHAPTER 7: General discussion...................................................................... 132
Thesis summary .................................................................................................... 132
Local adaptation .................................................................................................... 133
Selective pressures relevant to phytophagous insects ............................ 134
Is there evidence of local adaptation in P. atomaria?............................. 135
Implications for pest management ........................................................................ 139
References ............................................................................................................. 142
Appendix 1: Morphometric study of Paropsis atomaria (supplement

to Chapter 3)............................................................................................ 145
Appendix 2: Cross mating trial between Canberra (A.C.T.) and
Lowmead (Qld) individuals.................................................................... 151
List of Figures
Chapter 2
Figure 1. Paropsis atomaria egg batches laid on a) branchlet of
Eucalyptus sp., and b) leaf tip. Photos by Mark Schutze (not included
in CABI datasheet).....................................................................................................26
Figure 2. First instar Paropsis atomaria larvae. Photo by Mark Schutze..................27
Figure 3. Third instar Paropsis atomaria larvae. Photo by Amy Carmichael ...........27
Figure 4. Fourth instar Paropsis atomaria larvae showing distinctive

black lateral and dorsal markings. Photo by Amy Carmichael..................................28
Figure 5. Paropsis atomaria pupa. Photo by Amy Carmichael.................................29
Figure 6. Pair of Paropsis atomaria adults. Photo by Mark Schutze ........................29
Figure 7. Species morphologically similar to Paropsis atomaria: a) P.

elytrura from south-west Western Australia, and b) P. deboeri from
Tasmania. Photos by a) Mark Schutze and b) David de Little. (not
included in CABI datasheet) ......................................................................................30
Figure 8. Characteristic leaf ‘scalloping’ damage of eucalypt foliage by

adults of P. atomaria. Photo by Mark Schutze ..........................................................31
Chapter 3
Figure 1. Map of eastern Australia showing the five sampling locations
for Paropsis atomaria investigated in this study. ......................................................45
Figure 2. 95% parsimony network of 23 haplotypes obtained by

sequencing a 508 bp fragment of the mtDNA COI gene for 93
individuals of Paropsis atomaria generated by TCS. Arbitrary
haplotype numbers refer to haplotypes listed in Table 2. Size of oval
represents relative numbers of individuals possessing the haplotype.
Small filled circles represent hypothetical intermediate haplotypes.
Shading denotes what proportion of each haplotype is represented by
individuals from a specific collection site..................................................................49
i
Figure 3. 95% parsimony network of 23 haplotypes obtained by
sequencing a 508 bp fragment of the mtDNA COI gene for 93
individuals of Paropsis atomaria as generated by TCS. Arbitrary
haplotype numbers refer to haplotypes listed in Table 2. Size of oval
represents relative numbers of individuals possessing the haplotype.
Small filled circles represent hypothetical intermediate haplotypes.
Shading denotes what proportion of each haplotype is represented by
individuals collected from a specific species of Eucalyptus. .................................... 50
Chapter 4
Figure 1. Geographical locations where Paropsis atomaria individuals were
collected for this study. ............................................................................................. 70
Figure 2. Average daily temperatures for Paropsis atomaria field season

plotted against latitude for each of the four field collection sites: Lowmead,
Beerburrum, Bangalow and Canberra (left-right); Pearson correlation co-
efficient r = -0.992, P < 0.05. See text for calculation of averages........................... 73
Figure 3. Plot of pronotum width against latitude for de novo field collected
Paropsis atomaria (circles = females; squares = males). Latitude rounded to
nearest degree. Numbers above plots = number of individuals measured.
Different letters denote statistically significant difference (P < 0.05) in
pronotum width between locations for each sex (females lower case; males
upper case). Points slightly offset for clarity............................................................. 74
Figure 4. Relationship between pronotum width and latitude for historically

collected (i.e. collected prior to this study) Paropsis atomaria females
(circles) and males (squares) demonstrating converse Bergmann cline. Gap
between clusters due to lack of specimens collected from those latitudes................ 75
Figure 5. Mean development time (days) for each population of Paropsis

atomaria (Canberra 35ºS = grey and Lowmead 24ºS = black) reared at four
temperatures (16 ºC, 20 ºC, 24 ºC, and 27 ºC). Numbers above plots =
number of replicates. Different letters denote significant difference (Tukey
post hoc comparisons, P < 0.05) in total development time between each
temperature (Canberra upper case and Lowmead lower case). Asterisk
denotes significant difference (pairwise ANOVA, P < 0.05) in development
ii
time between populations reared at the same temperature. Points slightly
offset for clarity..........................................................................................................77
Figure 6. Effects of larval rearing temperatures (16 ºC, 20 ºC, 24 ºC, and 27
ºC) on adult body size (pronotum width, mm) for females (circles) and
males (squares) of Paropsis atomaria from two population origins
(Canberra indicated in black and Lowmead indicated in grey). Numbers
above plots = number of individuals measured. Different letters denote
significant differences (Tukey post hoc comparisons, P < 0.05) in average
pronotum width between temperature trials for each population (Canberra
upper case and Lowmead lower case). Asterisks denote significant
difference (pairwise ANOVA, P < 0.05) in pronotum width between
populations reared at the same temperature. Points slightly offset for clarity. ..........79
Chapter 5
Figure 1. Map showing location of two source populations of Paropsis
atomaria used in the current study: Lowmead and Canberra; and
distributions of Eucalyptus cloeziana and E. pilularis (data sourced from
Australia’s Virtual Herbarium: http://www.anbg.gov.au/avh/). ................................92
Figure 2. Mean ± 1 S.E. development time (days) of Paropsis atomaria

from egg eclosion to adult emergence for two populations (Canberra: circle
and Lowmead: square) reared on two host plants: Eucalyptus cloeziana and
E. pilularis. Asterisk denotes significant difference (P < 0.05) for between
population comparisons within a host plant trial; different lower case letters
denote significant difference (P < 0.05) for between host-plant comparisons
calculated within each beetle population. Numbers above plots equal
number of replicates...................................................................................................94
Figure 3. Mean ± 1 S.E. pupal weights (grams) of Paropsis atomaria

females and males, for two populations (Canberra: circles, and Lowmead:
squares) reared on two host plants: Eucalyptus cloeziana and E. pilularis.
Asterisk denotes significant difference (P < 0.05) in pupal weight between
beetle populations reared on the same host plant. Different letters denote
significant difference (P < 0.05) in pupal weights within a beetle population
iii
between host plants. Numbers above plots represent number of individuals
measured.................................................................................................................... 96
Chapter 6
Figure 1. Schematic diagram of the life cycle of Paropsis atomaria used in
the DYMEX™ model. .............................................................................................. 111
Figure 2. Mean ± s.e. mortality rate between Li and adult Paropsis atomaria
at four constant temperatures in the laboratory. Different letters denote
means that differ significantly................................................................................... 116
Figure 3 Stage-specific average ± s.e. mortality of immature Paropsis

atomaria at four temperatures: 16 °C (diamonds), 20 °C (squares), 24 °C
(triangles) and 27 °C (circles).. ................................................................................. 117
Figure 4. Field validation of ParopSys model between 28 Sept 2005 and 10

May 2006 for Site III, South East Queensland - (A) Adults (B) Eggs (C)
Total larvae. Solid lines represent field phenological data collected
fortnightly (right y-axis); dotted lines represent DYMEXTM model data
using Data Drill meteorological data for Site III (left y-axis). DYMEXTM
population numbers are dependent on numbers initialised and so relate
relatively to field data (numbers per shoot). ............................................................. 119
Figure 5. DYMEXTM model predictions for adult, egg and total larval
populations of Paropsis atomaria for ACT (Canberra) and Queensland
(Lowmead) between 28 Sept 2005 and 10 May 2006. DYMEXTM
population numbers are dependent on numbers initialised and so relate
relatively to field data (numbers per shoot). ............................................................. 120
Appendix 1
Figure 1. All measurements take for Paropsis atomaria individuals collected
from the field and reared in common-garden trials................................................... 146
Figure 2. Canonical discriminant results for morphometric data based

on fifteen measurements taken from female (top) and male (bottom).
Paropsis atomaria samples sourced from four collection localities
(Beerburrum, Qld; Bangalow, N.S.W.; Lowmead, Qld; and Canberra,
A.C.T.) and also for individuals of P. deboeri (close relative of P.
iv
atomaria) from Tasmania. Note that only a single male specimen was
available for P. deboeri analysis. .............................................................................149
v
List of Tables
Chapter 2
Table 1. Host plants from which Paropsis atomaria has been collected. ................. 24
Chapter 3
Table 1. Location, date, sample size (n) and host species of P. atomaria
investigated in this study. .......................................................................................... 46
Table 2. Number of Paropsis atomaria screened for each study with

relative haplotype frequencies and absolute number of individuals
sampled for each haplotype (in parenthesis). Left-hand of table denotes
which haplotypes are associated with a particular sampling location and
in what proportion. Right-hand of table denotes which haplotypes are
associated with a particular Eucalyptus host and in what proportion.
E.g. X E.c. = Eucalyptus grandis X E. camaldulensis hybrid. ................................. 51
Table 3. Number of Paropsis atomaria individuals screened for

location and Eucalyptus host molecular analyses with number of
haplotypes, nucleotide diversity and haplotype diversity. Standard
deviations (SD) are provided for nucleotide and haplotype diversity.
(E.g. X E.c. = Eucalyptus grandis X E. camaldulensis hybrid). n =
sample size. ............................................................................................................... 52
Table 4. Pairwise genetic differentiation between populations of

Paropsis atomaria from four sample locations in eastern Australia.
Below diagonal: pairwise estimates of Fst calculated by AMOVA
employing Kimura 2 parameter distances among haplotypes. Asterisk
denotes statistical significance (p < 0.05). Above diagonal: distance in
kilometres (km) between sample locations. .............................................................. 53
Table 5. Estimation of genetic differentiation of populations of

Paropsis atomaria from three different Eucalyptus host plants.
Pairwise estimates of Fst calculated by AMOVA employing Kimura 2
parameter distances among haplotypes. .................................................................... 53
vi
Chapter 4
Table 1. Two-way ANOVA of the effect of sex and location (and their
interaction) on Paropsis atomaria pronotum width for de novo
collected field material...............................................................................................73
Table 2. Two-way ANOVA of the effect of location and temperature

(and the interaction) on total development time in days (egg – adult) for
larvae reared at 4 temperatures (16 ºC, 20 ºC, 24 ºC, and 27 ºC). .............................76
Table 3. Three-way ANOVA results for the effect of sex, location, and
temperature (and interactions) on adult body size (pronotum width,
mm) for larvae reared at 4 temperatures (16 ºC, 20 ºC, 24 ºC, and 27
ºC). .............................................................................................................................78
Chapter 5
Table 1. Two-way ANOVA of the effects of location, host plant (and
interaction) on pupal weights (g) attained by female and male Paropsis
atomaria from two source populations (Lowmead and Canberra) reared
on Eucalyptus cloeziana and E. pilularis...................................................................95
Chapter 6
Table 1. Developmental thresholds (T0), thermal requirements (DD) and
proportion of development time for immature lifestages of Paropsis atomaria .. . .115
Table 2. Estimated mortality (proportion of eggs and larvae lost) at
each immature life stage of Paropsis atomaria in the field at two sites
(Li = first instar, Lii = second instar, Liii = third instar, Liv = fourth
instar)..........................................................................................................................118
Appendix 1
Table 1. Fifteen measurements selected for morphometric study of
Paropsis atomaria individuals collected from four sites. ..........................................147
Table 2. Measurements for each of the fifteen selected measurements

(mm ± standard deviation) for the morphometric analysis of Paropsis
atomaria collected from the following four sites: Beerburrum, Qld
vii
(BBR); Bangalow, N.S.W. (BAN); Lowmead, Qld (LOW); and
Canberra, A.C.T. (CAN). .......................................................................................... 148
Appendix 2
Table 1. Total duration of copulation (minutes) between paired
Paropsis atomaria individuals from one of two source populations
(Can = Canberra, A.C.T.; Low = Lowmead, Qld). Numbers with ‘+’
denote those pairs who were still copulating at the termination of the
experiment (observations made over 300 minutes). Total number of
replicates (including those that did not mate) given in parentheses.......................... 152
viii
List of publications
1. CAB International (2005) Forestry Compendium. CAB International, U.K.
2. Schutze, M.K., Mather, P.B. & Clarke, A.R. (2006) Species status and population
structure of the Australian Eucalyptus pest Paropsis atomaria Olivier (Coleoptera:
Chrysomelidae). Agricultural and Forest Entomology, 8, 323-332.
3. Schutze, M.K. and Clarke, A.R. Converse Bergmann cline in a Eucalyptus

herbivore, Paropsis atomaria Olivier (Coleoptera: Chrysomelidae): phenotypic
plasticity or local adaptation? Global Ecology and Biogeography, 17 (3), 424-431.
4. Schutze, M.K. and Clarke, A.R. Larval development by two geographically

isolated populations of Paropsis atomaria on two species of Eucalyptus. In prep for
submission to: Australian Journal of Entomology.
5. Nahrung, H.F., Schutze, M.K., Clarke, A.R., Duffy, M.P., Dunlop, E.A. and
Lawson, S.A. (2008) Thermal requirements, field mortality and population
phenology modelling of Paropsis atomaria Olivier, an emergent pest in subtropical
hardwood plantations. Forest Ecology and Management, 255 (8), 3515-3523.
ix
Statement of original authorship
This work has not previously been submitted for a degree or diploma at any other
educational institution. To the best of my knowledge, this thesis contains no material
from any other source, except where due reference is made
Mark K. Schutze
x
Acknowledgements
First and foremost I extend my greatest thanks to my principal supervisor, Anthony
R. Clarke. Without Tony’s sagely guidance, consistent support, and unending
patience, I would never have made it this far – Tony is a mentor in every sense of the
word.
Thanks also to Peter B. Mather, my associate supervisor. Peter has always been there
to provide another view, perspective, and opinion, which is something that has
always been valued and appreciated.
My parents, Kurt, Maureen, and Len, all deserve special mention. While my father,
Kurt, is sadly now gone, he always encouraged me to pursue my interests and to do
my best at whatever I undertook. I hope I would have made him proud. My mum,
Maureen, and stepfather, Len, have been an unending source of support throughout
the entire PhD process, always helping me see the brighter side during darker times,
and reminding me of the important things – even when I insisted that they ‘didn’t
understand’; I now know that they did.
Without my friends, too, I would be nowhere. Those who have supported and helped
me, even when they didn’t know they were doing so, are too numerous to mention,
but particular thanks must go to (in no particular order) Katarina Mikac, Helen
Nahrung, J. Paul Cunningham, Angela Duffy, Andrew Ridley, Luis Fernando
Vargas, Amanda Mergler, Stephen Montieth, Mike Duffy, Jo Kent, Alexsis Wilson,
Ana Pavasovic, Peter Prentis, Corinna Lange, and Daniel Jackson.
Thank you also to the following colleagues and professional organisations that
contributed to this project: Gunter Maywald (CSIRO), Simon Lawson (QDPI&F),
Richard Lunney (ITC Plantations), David De Little, Mamoru Matsuki, ANIC,
Orange Agricultural Institute, University of Sydney, State Forests of NSW, Forestry
SA, WA Museum, and of course, QUT, for funding me through a QUT Postgraduate
Research Award.
xi
Chapter 1
General introduction and literature review
Variation within and among individual organisms can result from a number of
influences. Such variation may be due to individuals belonging to two or more
‘good’ biological species, even if they are not currently recognised by taxonomists.
Alternatively, differences may be due to intraspecific variation within a single
genome, differential phenotypic expression of the same genome (phenotypic
plasticity), or a combination of the two. The current study addresses these issues as
they apply to a widely dispersed Australian eucalypt leaf beetle, Paropsis atomaria
Olivier (Coleoptera: Chrysomelidae).
This introduction provides a background to the importance of accurate species level

identification, regardless of whether the direction of the research is theoretical or
applied. Initially, it focuses on identifying possible cryptic species complexes,
followed by a brief appraisal of abiotic and biotic environmental influences on
geographic variation in phytophagous ectotherms. There follows a brief overview of
the study organism, the chrysomelid beetle P. atomaria, and a final section that
outlines the thesis objectives and flow of research chapters.
Patterns and processes of genotypic and phenotypic variation
Variation and cryptic species complexes

For more than 150 years, biologists have been interested in how and why species
vary over their geographical distributions. To this end, physiological, morphological,
developmental, or behavioural studies have the potential to address questions of local
adaptation and speciation, and hence provide the foundation from which to explain
such variation. Geographical variation is widespread across many phytophagous
insects, with high diversity often explained by specialisation on one or a few species
of host plant (Ballabeni et al., 2003). Indeed, after closer inspection, many
‘generalist’ herbivores are often found to consist of host adapted populations (= host
races) or even cryptic species (Jaenike, 1990). While adaptation to host plants can be
a driver of variation in herbivore taxa, other factors can also play a role in
influencing diversity across a species’ range. Climatic conditions, for example, can
Chapter 1: Introduction 1
shape the characteristics of individuals in a population, especially for ectotherms that
have rates of development closely correlated with temperature (Chown and Gaston,
1999).
The first step toward explaining variation, however, is to identify the biological
system under study. That is, are we observing a single biological species, or a
complex of previously unrecognised species masquerading under a single taxonomic
identity, i.e., a cryptic species complex? This question is paramount, as future
research must be based on a confident appraisal of the biological unit (or units)
concerned. While many studies have been undertaken without initial recourse to
rigorous species confirmation, the absence of fundamental taxonomic groundwork
has the potential to result in many years of wasted effort and resources (Walter,
2003). This is especially relevant for taxa of economic or conservation significance.
In 1982, for example, isoenzyme analysis revealed a Western Australian moth, the
Jarrah Leaf Miner (Perthida glyphopa), to consist of up to three genetically distinct,
but morphologically indistinguishable taxa (Mahon et al., 1982). This rendered the
previous 50 years of related ecological research largely uninterpretable. The
biological control of Karoo caterpillar, Loxostege frustalis, in South Africa provides
a more dramatic example. Between 1942 and 1952 large numbers (up to 6 million) of
the exotic parasitoid wasp, Chelonus texanus, were released to control the pest
caterpillar. In 1951, however, researchers discovered that all parasitism-related
caterpillar deaths were not caused by the introduced C. texanus, but by the
morphologically similar native wasp, C. curvimaculatus (Annecke and Moran,
1977). Clearly, the above cases would have benefited from accurate, meaningful
species identification at the beginning of the study, and they consequently serve as a
salient reminder to initiate such research prior to undertaking any larger programme.
Cryptic species complexes were once thought uncommon, as morphologically

defined species were assumed to closely match the biological reality (Dobzhansky,
1941). Today, however, it is widely recognised that previously defined strains, host
races, or biotypes often constitute ‘good’ species (Clarke and Walter, 1995), with
examples of cryptic complexes now common throughout the literature (Bickford et
al., 2007). But what defines a ‘good’ species? There is perhaps no more vexed
question in evolutionary biology than how to define species. With regularity, new
books or proceedings appear debating existing species concepts and may even
2 Chapter 1: Introduction
suggest new ones (Hey, 2001). Allopatric versus sympatric speciation, the relative
importance of pre- and post-mating isolating mechanisms, and the role of selection
versus genetic drift are just some of the issues that are revisited continually (see
individual chapters in Hey (2001) for examples). In spite of the intensity of the
debate, most biologists accept a genetic view, in which a species is regarded as “a
field for gene recombination” (Carson, 1957). That is, individuals within a species
(under normal conditions) freely exchange and recombine genes with other members
of the same gene pool. What limits the “field”, e.g., isolation (sensu Mayr) versus
recognition (sensu Paterson) versus cohesion (sensu Templeton), is a related, yet
another, critical and ongoing debate (Harrison, 1998). Given the lack of resolution as
to how to define biological species, and the potential conflict between definitions, it
is important to state the definition being used in any particular study. For the current
study, Paterson’s (1993) Recognition Concept of species will be the reference point
because it focuses on critical factors that unite individuals within a species without
the need to relate to other species (unlike the Biological / Isolation Species Concept)
and hence has value in helping to understand variation within and among species
(Lambert & Spencer 1995).
According to the Recognition Concept, species are, “that most inclusive population
of individual biparental organisms which share a common fertilisation system”
(Paterson, 1982). In theory, cryptic species represent such populations, but whose
identification and separation from other groups (with different fertilisation systems)
is confounded by high levels of among taxon morphological similarity. Hence, they
cannot be easily identified using traditional taxonomic (i.e., morphological)
techniques (Paterson, 1991). Furthermore, the discovery of cryptic species
complexes is likely to be non-random, and we may expect to find them in some
groups more so than in others (Bickford et al., 2007). For instance, taxa with a wide
host range across a large distribution, varying life history traits from one region to
another (e.g., obligate diapause in one region, facultative in the other), or those that
possess critical non-visual components in the mate recognition system (e.g., auditory,
chemical, or tactile) may represent groups that we may a priori suspect of more
likely constituting a cryptic species complex (Paterson, 1991; Bickford et al., 2007).
In line with the Recognition Concept of species, cryptic species – like any other –
should be identifiable by elements of their fertilisation system which are discrete
within a specific gene pool and different from those among other gene pools.
Furthermore, as a critical component of the fertilisation system is the specific mate
recognition system (the series of signals and reciprocal responses between potential
mates), mating trials under controlled conditions should therefore, in theory, aid in
species identification. Mating trials are, however, time consuming, logistically
challenging, and results may be inconclusive (due to inappropriate experimental
design, the confounding effects of unnatural laboratory conditions, or uninterpretable
results) (Walter, 2003). Indirect methods, including molecular approaches, may
provide a more expedient pathway for estimating the extent of a gene pool, as
members of the same species should possess relatively high levels of genetic
similarity (compared with other taxa) due to sharing a common fertilisation system.
Additionally, molecular approaches glean information from individuals collected
directly from the wild, providing a snap-shot of the natural situation, and thus
eliminating the potential confounding influences of mating trials conducted under
artificial conditions.
There are several molecular marker approaches that have been used for species
identification, each providing indirect measures of gene flow within and among
populations. Allozyme gel electrophoresis, for example, has been in use since the late
1960’s (Hsiao, 1989) and has been applied extensively for cryptic species studies
(see, for example, Krafsur and Obrycki, 2000; Navajas et al., 2000; Aguin-Pombo,
2002; Mutebi et al., 2002; Zhu et al., 2002; Mattiucci et al., 2003; Martin-Sanchez et
al., 2003; Naumova et al., 2003). This method documents variation in a diverse array
of soluble enzyme products, determining relative frequencies of alleles in sample
populations, thus estimating levels of contemporary gene flow within and among
populations. There are limitations to this technique, however, as it is time consuming
and can produce ambiguous results if populations do not exist in sympatry, as
geographical barriers between populations may explain any lack of gene flow rather
than existence of discrete mating systems (Walter, 2003).
Analysis of historical gene flow using a direct sequence analysis of DNA (especially
of the mitochondrial genome) is now a frequently used method for assisting cryptic
species identification, largely due to its recent reduction in cost, and the rapidity and
ease of its execution (Bickford et al., 2007). Mitochondrial DNA (mtDNA) is
maternally inherited, it is not altered by recombination, there are many copies per
cell, and its mutation rates are often up to 20 times faster than nuclear DNA.
Furthermore, well studied mtDNA genes (e.g., cytochrome c oxidase I) consist of
variably constrained regions, thus permitting their use across a range of taxonomic
levels (Lunt et al., 1996; Loxdale and Lushai, 1998). Such characteristics allow
questions of maternal ancestry, population genetic structure, and gene flow to be
addressed at the species level (Avise, 1986; Avise et al., 1987; Simon et al., 1994).
However, as mtDNA is maternally inherited, the degree of paternal gene flow
remains unknown, and may have consequences in studies of species exhibiting sex-
biased dispersal patterns which can generate results similar to those expected for a
cryptic species complex. Additionally, due to the reduced variability of mtDNA
sequence data compared with enzyme data (in certain cases), fewer numbers of
individuals can substantiate a cryptic complex study. For example, in the case of the
Sugarcane Weevil, Rhabdoscelus obscurus, two cryptic species were confirmed from
sequence data obtained from only six individuals, which was subsequently
corroborated with ecological data (Giblin-Davis et al., 2000). Consequently, many
recent cryptic complex studies have, and continue to assess, mtDNA sequence data
(Mattiucci et al., 2003; Hebert et al., 2004; Quicke et al., 2006).
One of the principle cautions when using mtDNA to assess intraspecific and
interspecific relationships is that there is no standard level of divergence that allows
certainty with regard to species identification. Relative levels of divergence between
populations remains the most common – albeit arbitrary – measure used, with
mtDNA distances between species considered to range upwards of 3%. For example,
the earwig, Forficula auricularia, was considered a single species consisting of a
mosaic of populations that differed in reproductive biology (unlikely considering the
recognition concept as outlined above). Mitochondrial DNA sequence data, however,
revealed inter-population sequence divergence at 5.82%, approximately six times
greater than intra-population divergence (1.07% and 0.66% among each of two
groups). When coupled with the data on reproductive biology, mtDNA analysis
provided indirect support for the presence of two cryptic species (Wirth et al. 1998).
Therefore, sequence distance data used in isolation can not define species, but rather
it provides a level of relatively easily acquired information that can contributes to
their identification. The coupling of sequence data with other information (e.g.,
ecological, behavioural, or physiological data) is desirable and has been
demonstrated across a range of studies incorporating molecular data with information
on host utilisation (Kaneshiro and Kambysellis, 1999; Sembene and Delobel, 1998),
variable life history and behavioural traits (Morrow et al., 2000; Knio et al., 2001;
Willmott et al., 2001), symbiont associations (Six and Paine, 1997), elements of
specific mate recognition systems (Jeraj and Walter, 1998; Mousseau and Howard,
1998; Schul, 1999; Kimura et al., 2002; Henry et al., 2003), and traditional
taxonomic appraisals (Palmer, 2002; Maingon et al., 2003).
Causes of intraspecific geographic variation

Assuming appropriate studies have addressed questions on gene flow, population
structure, and species limits, and there is confidence that a variable taxon constitutes
a single biological species (versus a cryptic complex of species), then it would be
appropriate to investigate factors that contribute to intra-specific variation. For
phytophagous insects – as in other species – phenotypic variation is determined by
either genetic factors, environmental influences, or their interaction. Furthermore,
expression of specific phenotypes may be due to either local adaptation of specific
traits or phenotypic plasticity.
Local adaptation is generally reserved for patterns and processes observed among
populations of a single species connected by dispersal and gene flow; it is defined as
the acquisition of a suite of locally suited traits resulting from the genetic
differentiation of a population relative to other populations (Kawecki and Ebert,
2004). Phenotypic plasticity, on the other hand, is the result of a single genotype
producing multiple phenotypes as a direct response to environmental conditions
experienced by an individual (West-Eberhard, 1989). Additionally, phenotypic
plasticity itself may be adaptive and selected for by species inhabiting highly
heterogeneous environments subject to predictable change (Via et al., 1995). A
species of frog in Sweden, Rana temporaria, for example, experiences variable pool-
drying regimens across different islands and has evolved phenotypic plasticity suited
to changing conditions (Lind and Johansson, 2007). Therefore, a study investigating
variation within a species must seek to understand the underlying mechanisms
driving phenotypic diversity; namely, are geographically related differences the
product of local adaptation or phenotypic plasticity?
Local adaptation can be hindered by temporal variation in selective forces or habitat
quality, together with the homogenising effects of high levels of gene flow among
populations (Kawecki and Ebert, 2004). Consequently, reduced temporal but greater
spatial heterogeneity in habitat quality and low levels of gene flow between
phenotypically distinct populations may present an a priori reason for suspecting
local adaptation (Kawecki and Ebert, 2004). Phenotypic plasticity can, however,
occur in species with populations that experience high levels of gene flow and be
more likely if: (i) there is a strong match between individual phenotypes and
respective local environments; (ii) there is an associated low cost of plasticity; (iii)
there are equal frequencies of alternative environments; and (iv) environments vary
temporally rather than spatially (Moran, 1992). With this background in mind, the
relative contribution of either local adaptation or phenotypic plasticity can be
determined relative to abiotic and biotic influences affecting geographic variation
within taxa
Abiotic factors can play a significant role toward influencing variation in insects.
Temperature, for example, often directly influences rates of metabolism and hence
larval development in insects and other ectothermic taxa (Sibly and Atkinson, 1994;
Van der Have and De Jong, 1996). The observed strength of this relationship has
resulted in the formulation of theories such as the Temperature-Size Rule, whereby
lower developmental temperatures produce larger individuals (Carleton, 1960;
Vannote and Sweeny, 1980; Lonsdale and Levinton, 1985; Atkinson, 1994; Partridge
et al., 1994; Van der Have and De Jong, 1996; Atkinson and Sibly, 1997; Chown and
Gaston, 1999; Ramsden and Elek, 1998; Reeve et al., 2000). Large body size for
certain species can confer an advantage, and hence is a possible source of local
adaptation. Advantages associated with large body size may include greater potential
fecundity for females (Carne, 1966), increased reproductive success in males (Reeve
et al., 2000), or improved starvation resistance for individuals that experience
frequent adverse environmental conditions (Arnett and Gotelli, 2003).
Season length can also play a pivotal role in influencing body size in ectotherms.
Extended season length provides increased time for individuals to reach a larger
body size at maturation, and hence may increase their potential fecundity. Shorter
seasons, however, provide less time for individuals to reach maturity, which may
produce smaller adult body sizes (Roff, 1980). As both season length and
temperature are usually highly correlated with latitude, they can produce large-scale
geographical trends in ectotherm variation. Indeed, body size of both endotherms and
ectotherms over latitudinal gradients has been documented widely in the literature
(see Blanckenhorn and Demont (2004) for a list of arthropod examples) and has
resulted in the formulation of numerous rules to explain such variation, the most
famous of which was proposed by Carl Bergmann, and is known as Bergmann’s
Rule (Bergmann, 1847).
Bergmann’s Rule states that members of a species tend to be larger at higher

latitudes (Blanckenhorn and Demont, 2004). The original mechanism proposed to
explain this observation was that large size conferred greater heat retention in
mammals found in colder climates (i.e., higher latitudes) due to a reduction in the
surface area-volume ratio (Bergmann, 1847). This mechanism is now largely
dismissed for many endothermic species exhibiting such a cline, and is even less
probable for ectotherms. As the body temperature of ectothermic animals fluctuates
relative to ambient conditions, heat conservation, regardless of ectotherm size, is
therefore not usually possible (McNab, 1971; Atkinson and Sibly, 1997).
Additionally, the reverse of a classical Bergmann cline, a converse Bergmann cline
(Park, 1949) (i.e., larger sizes at lower latitudes), is equally as common amongst
arthropod taxa and is often considered to be under genetic control (Masaki, 1978;
Mousseau and Roff, 1989; Blanckenhorn and Demont, 2004). Indeed, it is often for
the reasons outlined above (i.e., shorter season lengths providing less time to mature
and hence smaller adults) that converse Bergmann clines in ectothermic species are
thought to evolve, and is often hypothesised to result from local adaptation of
individuals to local season length (Roff, 1980; Blanckenhorn and Demont, 2004),
Biotic influences may also play a strong role in shaping variation, as host plant
specialisation in phytophagous insects is widespread, with many herbivores having a
close association with one or few host plant species (Fox and Morrow, 1981; Joshi
and Thompson, 1995). This contributes to increased diversity amongst plant-feeding
clades relative to their non-phytophagous sister groups (Jaenike, 1990).
Consequently, host plant quality, distribution, and composition can have profound
effects on variation in herbivorous species, as specialisation on host plants, if given
sufficient time and isolation between populations, may lead to the evolution of host
races, and potentially new species (Ballabeni et al., 2003).
As outlined above, commencing a study of biological variation requires careful
consideration of several key components. To begin, the very nature of the taxon
under study requires clarification: does it likely consist of single biological species,
or is it probably a complex of morphologically indistinquishable cryptic taxa? This
question may be addressed via multiple approaches, ranging from molecular
characterisation to rigorous morphological assessments, or investigating ecological
aspects of the study group. Only from this foundation can studies of patterns and
processes relating to geographic and phenotypic variation proceed, regardless of
whether they are driven by biotic influences, abiotic factors, or a complex interaction
of the two.
Study System
Paropsis atomaria is a leaf beetle endemic to Australia. This species feeds
principally on plants that belong to the genus Eucalyptus and has been recorded from
at least 20 host tree species (CABInternational, 2005), a degree of polyphagy rarely
seen for herbivorous insects (Fox and Morrow, 1981; Jermy, 1984; Claridge et al.,
1997). The geographical distribution of P. atomaria extends from the temperate
south-east of Australia to the tropical north of coastal Queensland (Schutze et al.,
2006), covering a wide latitudinal gradient across varied environmental conditions.
Existing information on P. atomaria (detailed in Chapter 2 of this thesis) identifies
biological attributes such as wide host range, variable voltinism (e.g., increased
number of generations per season in northern, lower latitudes) and developmental
physiology. Important differences in the biology of P. atomaria populations from
different geographical locations have also been reported. Carne (1966) measured
growth rates of P. atomaria larvae collected from Canberra and found optimum
developmental time was achieved between 21ºC and 24ºC, whereas Bailey (2001)
determined 27.5ºC to be the optimum developmental temperature for larvae from
sub-tropical Queensland. Furthermore, populations in Canberra undergo diapause
during the colder months of the year (May – August) (Carne, 1966), a behaviour not
clearly demonstrated for individuals from warmer climates (such as Queensland) (S.
Lawson, [Qld] Department of Primary Industries & Fisheries, pers. comm.).
These biological differences, when coupled with the wide host range, extensive
geographical distribution, and unresolved taxonomic history of P. atomaria, give
strong a priori reasoning for suspecting a cryptic species complex within the taxon.
Furthermore, as P. atomaria is an emerging pest of plantation forestry in Australia, it
is likely to be of growing interest in the future and it is therefore timely to review
both the extent and underlying causes of its biological variation.
Objectives of the study

Given that previous studies had identified different life history characteristics
between southern (temperate) and northern (sub-tropical) populations, the principle
aim was to address the question as to whether P. atomaria constitutes a cryptic
species complex. To this end, the current state of knowledge about the taxonomy and
biology of P. atomaria was examined (Chapter 2), to form a basis for subsequent
molecular and ecological studies. Subsequently, beetles were collected from multiple
locations across the natural distribution, ranging from temperate south-eastern
Australia (Victoria, Canberra, and South Australia), to the sub-tropics of central
Queensland, for a population genetic study of the species (Chapter 3). Due to its
utility, the COI mtDNA gene was selected for sequence analysis to assess haplotype
diversity within and among collection localities. From these data, gene flow was
measured among populations and an assessment made as to whether P. atomaria, as
currently recognised, constitutes a cryptic species complex. Results of this analysis
showed that populations were sufficiently homogeneous to argue that P. atomaria
constitutes a single taxon, but that gene flow was reduced between southern
(Canberra, A.C.T.) and northern (Bangalow, N.S.W. to Lowmead, Qld) populations.
Following the genetic study, individuals from four of the locations used in the
molecular analysis were examined in a morphometric analysis (Chapter 4). This was
conducted to determine if morphometric data supported the single species hypothesis
proposed by the molecular study. Initial morphometric results of field collected
beetles demonstrated insufficient differentiation between populations to render them
different species, however there was a clear trend of increasing body size at lower,
northern latitudes (= a converse Bergmann cline). To investigate the importance and
nature of the cline, further historical P. atomaria collections were included to assess
the correlation between body size and latitude over time. In addition, common
garden experiments using beetles collected from the full extent of the natural range
(Canberra, A.C.T. and Lowmead, Qld) were conducted with individuals reared under
four constant temperature conditions. Results demonstrated the converse Bergmann
cline was also evident in historical collection material, that P. atomaria populations
conformed to the Temperature-Size rule (i.e., inverse relationship between
developmental temperature and adult body size), and that body size variation across
the latitudinal gradient was under genetic control, and not the product of phenotypic
plasticity (reflecting the results of the initial population genetic analysis).
Given that genetic variation among populations also influenced morphology in the
form of a converse Bergmann cline, physiological differences among populations
were examined focussing on larval host plant utilisation on different host plants
(Chapter 5). Additional common garden experiments were conducted, also using
Canberra and Lowmead individuals, with larvae from both populations reared on two
eucalypt host species, E. cloeziana and E. pilularis. Host species were chosen as each
occurs sympatrically with one or the other beetle population while being allopatric
with the other: E. cloeziana occurs in sympatry with the northern Lowmead
population whilst E. pilularis occurs in sympatry with the southern Canberra
population. The nature of host plant distribution allowed larval fitness on the
sympatric host plant to be assessed relative to that for the allopatric host species.
Results of this study reveal that E. cloeziana is a better host plant compared to E.
pilularis, as it resulted in increased survival and reduced development time for both
populations, and increased pupal mass for individuals from Lowmead – further
supporting the single species hypothesis.
Finally, constant temperature development data generated in Chapter 4 was

incorporated into a DYMEXTM based cohort phenology model for P. atomaria
developed in collaboration with colleagues (Chapter 6). This model uses different
biological data sets, including laboratory development data, field mortality,
fecundity, and diapause characteristics, to predict P. atomaria field phenology based
on local field temperatures. Output of the final model closely matched field
observation data not used in its development and is thus potentially a valid tool for
testing different hypotheses concerning phenological variation across P. atomaria’s
geographical range. The model shows reduced voltinism for southern temperate
populations (due to reduced season length) when compared with northern
populations, suggesting disjunct seasonal phenologies among P. atomaria
populations are a plastic response largely driven by temperature.
In the final discussion (Chapter 7), findings of previous chapters are reviewed and
the broader significance discussed, particularly relating to determining species limits,
causes and influences of intraspecific variation in ectotherm species such as insects,
and how the information can be applied for practical management strategies.
Thesis style
The structure of this thesis follows QUT rules for a PhD by publication, which
allows thesis examination to be based on the presentation of a body of related
published or submitted works, linked together by introduction and discussion
chapters. Consequently, all figures and tables are reinitialised for each chapter.
References
Aguin-Pombo, D. (2002). Genetic differentiation among host-related Alebra
leafhoppers (Hemiptera: Cicadellidae). Heredity 88, 415-422.
Annecke, D. P., and Moran, V. C. (1977). Critical reviews of biological control in
South Africa. Journal of the Entomological Society of Southern Africa 40,
127-145.
Arnett, A. E., and Gotelli, N. J. (2003). Bergmann'
s rule in larval ant lions: testing
the starvation resistance hypothesis. Ecological Entomology 28, 645-650.
Atkinson, D. (1994). Temperature and organism size - a biological law for
ectotherms. Advances in Ecological Research 25, 1-58.
Atkinson, D., and Sibly, R. M. (1997). Why are organisms usually bigger in colder
environments? Making sense of a life history puzzle. Trends in Ecology and
Evolution 12, 235-239.
Avise, J. C. (1986). Mitochondrial DNA and the evolutionary genetics of higher
animals. Philosophical Transactions of the Royal Society of London, Series B,
Biological Sciences. 312, 325-342.
Avise, J. C., Arnold, J., Ball, R. M., Bermingham, E., Lamb, T., Neigel, J. E., Reeb,
C. A., and Saunders, N. C. (1987). Intraspecific phylogeny: the mitochondrial
DNA bridge between population genetics and systematics. Annual Review of
Ecology and Systematics. 18, 489-552.
Bailey, K. E. (2001). The effect of temperature on development and growth of
Paropsis atomaria: Using Dymex to predict population trends of P. atomaria
for future management, Queensland University of Technology.
Ballabeni, P., Gotthard, K., Kayumba, A., and Rahier, M. (2003). Local adaptation
and ecological genetics of host-plant specialization in a leaf beetle. OIKOS
101, 70-78.
Bergmann, C. (1847). Über die Verhältnisse der Wärmeökonomie der Thiere zu ihrer
Grösse. Gottinger studien 3, 595-708.
Bickford, D., Lohman, D. J., Sodhi, N. S., Ng, P. K. L., Meier, R., Winkler, K.,
Ingram, K. K., and Das, I. (2007). Cryptic species as a window on diversity
and conservation. Trends in Ecology and Evolution 22, 148-155.
Blanckenhorn, W. U., and Demont, M. (2004). Bergmann and converse Bergmann
latitudinal clines in arthropods: Two ends of a continuum? Integrative and
Comparative Biology 44, 413-424.
CABInternational (2005). Forestry Compendium. CAB International, Wallingford
UK.
Carleton, R. (1960). The application of Bergmann'
s and Allen'
s rules to the
poikilotherms. Journal of Morphology 106, 85-108.
Carne, P. B. (1966). Ecological characteristics of the eucalypt defoliating
chrysomelid Paropsis atomaria Ol. Australian Journal of Zoology 14, 647-
672.
Carson, H. L. (1957). The species as a field for gene recombination. In "The Species
Problem" (E. Mayr, ed.), pp. 23-38. American Association for the
Advancement of Science, Pub. no 50, Washington, DC.
Chown, S. L., and Gaston, K. J. (1999). Exploring links between physiology and
ecology at macro-scales: the role of respiratory metabolism in insects.
Biological Reviews of the Cambridge Philosophical Society 74, 87-120.
Claridge, M. F., Dawah, H. A., and Wilson, M. R. (1997). Species in insect
herbivores and parasitoids - sibling species, host races and biotypes. In
"Species: The Units of Biodiversity" (M. F. Claridge, Dawah, H.A., and
Wilson, M.R, ed.), pp. 247-272. Chapman & Hall, London.
Clarke, A. R., and Walter, G. H. (1995). "Strains" and the classical biological control
of insect pests. Canadian Journal of Zoology 73, 1777-1790.
Dobzhansky, T. (1941). "Genetics and the Origin of Species," Colombia University
Press, New York.
Fox, L. R., and Morrow, P. A. (1981). Specialization: Species property or local
phenomenon? Science 211, 887-893.
Giblin-Davis, R. M., Gries, R., Crespi, B., Robertson, L. N., Hara, A. H., Gries, G.,
O'
Brien, C. W., and Pierce, H. D. J. (2000). Aggregation pheromones of two
geographical isolates of the New Guinea sugarcane weevil, Rhabdoscelus
obscurus. Journal of Chemical Ecology 26, 2763-2780.
Harrison, R.G. (1998) Linking evolutionary pattern and process: The relevance of
species concepts for the study of speciation. Pp 19-31 (Chp 2) in: Endless
Forms: species and speciation (Howard, D.J. & Berlocher, S.H. eds). Oxford
University Press, Oxford.
Hebert, P. D. N., Penton, E. H., Burns, J. M., Janzen, D. H., and Hallwachs, W.
(2004). Ten species in one: DNA barcoding reveals cryptic species in the
neotropical skipper butterfly Astraptes fulgerator. Proceedings of the
National Academy of Science of the United States of America 101, 14812-
14817.
Henry, C. S., Brooks, S. J., Duelli, P., and Johnson, J. B. (2003). A lacewing with the
wanderlust: The European song species '
Maltese'
, Chrysoperla agilis, sp.n.,
of the carnea group of Chrysoperla (Neuroptera: Chrysopidae). Systematic
Entomology 28, 131-147.
Hey, J. (2001). "Genes, catgegories and species: The evolutionary and cognitive
cause of the species problem," Oxford University Press, Oxford.
Hsiao, T. H. (1989). Estimation of genetic variability amongst Coleoptera. In
"Electrophoretic Studies on Agricultural Pests" (H. D. Loxdale and J. den
Hollander, eds.), Vol. 39, pp. 143-180. Clarendon Press, Oxford.
Jaenike, J. (1990). Host specialization in phytophagous insects. Annual Review of
Jeraj, M., and Walter, G. H. (1998). Vibrational communication in Nezara viridula:
Response of Slovenian and Australian bugs to one another. Behavioural
Processes 44, 51-58.
Jermy, T. (1984). Evolution of insect/host plant relationships. American Naturalist
124, 609-630.
Joshi, A., and Thompson, J. N. (1995). Trade-offs and the evolution of host
specialization. Evolutionary Ecology 9, 82-92.
Kaneshiro, K. Y., and Kambysellis, M. P. (1999). Description of a new allopatric
sibling species of Hawaiian picture-winged Drosophila. Pacific Science 53,
208-213.
Kawecki, T. J., and Ebert, D. (2004). Conceptual issues in local adaptation. Ecology
Letters 7, 1225-1241.
Kimura, T., Sakai, J., and Honda, H. (2002). Microstructure and pheromone
producing function of male hair-pencils in the yellow peach moth,
Conogethes punctiferalis (Lepidoptera: Pyralidae). Entomological Science 5,
237-247.
Knio, K. M., Goeden, R. D., and Headrick, D. H. (2001). Resource utilisation in the
sibling species Trupanea nigricornis (Coquillett), a polyphage, and the
narrowly oligophagous T. bisetosa (Coquillett) (Diptera: Tephritidae) in
southern California. Proceedings of the Entomological Society of Washington
103, 946-961.
Krafsur, E. S., and Obrycki, J. J. (2000). Coleomegilla maculata (Coleoptera:
Coccinellidae) is a species complex. Annals of the Entomological Society of
America 93, 1156-1163.
Lambert, D. M., and Spencer, H. G. (eds.) (1995). Speciation and the Recognition
Concept: Theory and Application. John Hopkins University Press,

Baltimore, Maryland U.S.A.
Lind, M. I., and Johansson, F. (2007). The degree of adaptive phenotypic plasticity is
correlated with the spatial environmental heterogeneity experienced by island
populations of Rana temporaria. Journal of Evolutionary Biology 20, 1288-
1297.
Lonsdale, D. J., and Levinton, J. S. (1985). Latitudinal differentiation in copepod
growth: An adaptation to temperature. Ecology 66, 1397-1407.
Loxdale, H. D., and Lushai, G. (1998). Molecular markers in entomology. Bulletin of
Entomological Research 88, 577-600.
Lunt, D. H., Zhang, D.-X., Szymura, J. M., and Hewitt, G. M. (1996). The insect
cytochrome oxidase I gene: evolutionary patterns and conserved primers for
phylogenetic studies. Insect Molecular Biology 5, 153-165.
Mahon, R. J., Miethke, P. M., and Mahon, J. A. (1982). The evolutionary
relationships of three forms of the Jarrah Leaf Miner, Perthida glyphopa
(Common) (Lepidoptera: Incurvariidae). Australian Journal of Zoology 30,
243-249.
Maingon, R. D. C., Ward, R. D., Hamilton, J. G. C., Noyes, H. A., Souza, N., Kemp,
S. J., and Watts, P. C. (2003). Genetic identification of two sibling species of
Lutzomyia longipalpis (Diptera: Psychodidae) that produce distinct male sex
pheromones in Sobral, Ceara State, Brazil. Molecular Ecology 12, 1879-
1894.
Martin-Sanchez, J., Diaz, M., Artacho, M. E., and Valero, A. (2003). Molecular
arguments for considering Hysterothylacium fabri (Nematoda: Anisakidae) a
complex of sibling species. Parasitology Research 89, 214-220.
Masaki, S. (1978). Climatic adaptation and species status in the lawn ground cricket.
Oecologia 35, 343-356.
Mattiucci, S., Cianchi, R., Nascetti, G., Paggi, L., Sardella, N., Timi, J., Webb, S. C.,
Bastida, R., Rodriguez, D., and Bullini, L. (2003). Genetic evidence for two
sibling species within Contracaecum ogmorhini Johnston & Mawson, 1941
(Nematoda: Anisakidae) from otariid seals of boreal and austral regions.
Systematic parasitology 54, 13-23.
McNab, B. K. (1971). On the ecological significance of Bergmann'
s Rule. Ecology
52, 845-854.
Moran, N. (1992). The evolutionary maintenance of alternative phenotypes. The
American Naturalist 139, 971-989.
Morrow, J., Scott, L., Congdon, B., Yeates, D., Frommer, M., and Sved, J. (2000).
Close genetic similarity between two sympatric species of tephritid fruit fly
reproductively isolated by mating time. Evolution 54, 899-910.
Mousseau, T. A., and Howard, D. J. (1998). Genetic variation in cricket calling song
across a hybrid zone between two sibling species. Evolution 52, 1104-1110.
Mousseau, T. A., and Roff, D. A. (1989). Adaptation to seasonality in a cricket:
patterns of phenotypic and genotypic variation in body size and diapause
length along a cline in season length. Evolution 43, 1483-1496.
Mutebi, J. P., Tripet, F., Alexander, J. B., and Lanzaro, G. C. (2002). Genetic
differentiation among populations of Lutzomyia longipalpis (Diptera:
Psychodidae) in Central and South America. Annals of the Entomological
Society of America 95, 740-752.
Naumova, E. S., Bulat, S. A., Mironenko, N. V., and Naumov, G. I. (2003).
Differentiation of six sibling species in the Saccharomyces sensu stricto
complex by multilocus enzyme electrophoresis and UP-PCR analysis.
Antonie van Leeuwenhoek 83, 155-166.
Navajas, M., Tsagkarakov, A., Lagnel, J., and Perrot, M. J. M. (2000). Genetic
differentiation in Tetranychus urticae (Acari: Tetranychidae): Polymorphism,
host races or sibling species? Experimental and Applied Acarology 24, 365-
376.
Palmer, M. (2002). Landmark-based morphometric analysis of two sibling species of
the genus Asida (Coleoptera, Tenebrionidae). Contributions to Zoology 70,
213-220.
Park, O. (1949). Application of the converse Bergmann principle to the carabid
beetle, Dicaelus purpuratus. Physiological Zoology 22, 359-372.
Partridge, L., Barrie, B., Fowler, K., and French, V. (1994). Evolution and
development of body size and cell size in Drosophila melanogaster in
response to temperature. Evolution 48, 1269-1276.
Paterson, H. E. H. (1982). Perspectives on speciation by reinforcement. South
African Journal of Science 78, 53-57.
Paterson, H. E. H. (1991). The recognition concept of cryptic species among
economically important insects. In "Research methods and prospects" (M. P.
Zalucki, ed.), pp. 1-10. Springer-Verlag, New York.
Paterson, H. E. H. (1993). Evolution and the recognition concept of species.
Collected writings. The Johns Hopkins University Press. Baltimore, U.S.A.
Quicke, D. L. J., Mori, M., Zaldivar-Riverson, A., Laurenne, N. M., and Shaw, M. R.
(2006). Suspended mummies in Aleiodes species (Hymenoptera : Braconidae
: Rogadinae) with descriptions of six new species from western Uganda based
largely on DNA sequence data. Journal of Natural History 40, 2663-2680.
Ramsden, N., and Elek, J. A. (1998). Life cycle and development rates of the leaf
beetle Chrysophtharta agricola (Chapuis) (Coleoptera: Chrysomelidae) on
Eucalyptus nitens at two temperature regimens. Australian Journal of
Reeve, M. W., Fowler, K., and Partridge, L. (2000). Increased body size confers
greater fitness at lower experimental temperature in male Drosophila
melanogaster. Journal of Evolutionary Biology 13, 836-844.
Roff, D. A. (1980). Optimizing development time in a seasonal environment: The
ups and downs of clinal variation. Oecologia 45, 202-208.
Schul, J. (1999). Neuronal basis for spectral song discrimination in the bushcricket
Tettigonia cantans. Journal of Comparative Physiology A 184, 457-461.
Schutze, M. K., Mather, P. B., and Clarke, A. R. (2006). Species status and
population structure of the Australian Eucalyptus pest Paropsis atomaria
Olivier (Coleoptera: Chrysomelidae). Agricultural and Forest Entomology 8,
323-332.
Sembene, M., and Delobel, A. (1998). Genetic differentiation of groundnut seed-
beetle populations in Senegal. Entomologia Experimentalis et Applicata 87,
171-180.
Sibly, R. M., and Atkinson, D. (1994). How rearing temperature affects optimal adult
size in ectotherms. Functional Ecology 8, 486-493.
Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H., and Flook, P. (1994).
Evolution, weighting, and phylogenetic utility of mitochondrial gene
sequences and a compilation of conserved polymerase chain reaction primers.
Annals of the Entomological Society of America 87, 651-701.
Six, D. L., and Paine, T. D. (1997). Ophiostoma clavigerum is the mycangial fungus
of the Jeffrey pine beetle, Dendroctonus jeffreyi. Mycologia 89, 858-866.
Van der Have, T. M., and De Jong, G. (1996). Adult size in ectotherms: temperature
effects on growth and differentiation. Journal of Theoretical Biology 183,
329-340.
Vannote, R. L., and Sweeny, B. W. (1980). Geographic analysis of thermal
equilibria: a conceptual model for evaluating the effect of natural and
modified thermal regimes on aquatic insect communities. American
Naturalist 115, 667-695.
Via, S., Gomulkiewicz, R., De Jong, G., Scheiner, S. M., Schlichting, C. D., and Van
Tienderen, P. H. (1995). Adaptive phenotypic plasticity: consensus and
controversy. Trends in Ecology and Evolution 10, 212-217.
Walter, G. H. (2003). "Insect Pest Management and Ecological Research,"
Cambridge University Press, Cambridge, UK.
West-Eberhard, M. J. (1989). Phenotypic plasticity and the origins of diversity.
Annual Review of Ecology and Systematics. 20, 249-278.
Willmott, K. R., Constantino, L. M., and Hall, J. P. W. (2001). A review of Colobura
(Lepidoptera: Nymphalidae) with comments on larval and adult ecology and
description of a sibling species. Annals of the Entomological Society of
America 94, 185-196.
Wirth, T., Le Guellec, R., Van Cassel, M., and Veuille, M. (1998). Molecular and
reproductive characterization of sibling species in the European earwig
(Forficula auricularia). Evolution 52, 260-265.
Zhu, X. Q., D'
Amelio, S., Palm, H. W., Paggi, L., George, N. M., and Gasser, R. B.
(2002). SSCP-based identification of members within the Pseudoterranova
decipiens complex (Nematoda: Ascaridoidea: Anisakidae) using genetic
markers in the internal transcribed spacers of ribosomal DNA. Parasitology
124, 615-623.
Chapter 2
Literature Review on Paropsis atomaria
This chapter has been published as the ‘Datasheet for Paropsis atomaria’ within:
CAB International (2005) Forestry Compendium. CAB International, U.K.
The purpose of the Compendium is to provide forestry workers with a detailed

background on the current state of knowledge regarding important pests to the
industry. Relevant information covers taxonomy, biology, morphology, similar taxa,
diagnostics, natural enemies, risk assessment, and control measures. This chapter is,
therefore, a comprehensive appraisal of the literature specific to Paropsis atomaria
up to the commencement of this thesis and I am the sole author.
Chapter 2: Literature review of Paropsis atomaria 21

Literature review on Paropsis atomaria
Notes on taxonomy and nomenclature

The family Chrysomelidae contains an extreme variety of beetles, with the
Chrysomelinae representing the largest subfamily in Australia, with more than 50
described genera and over 600 species (Lawrence and Britton, 1996). The most
important group, Paropsina, contains the genus Paropsis, erected by Oliver (1807)
when he described 15 new species of chrysomelid within this new genus, of which P.
atomaria was one. Upon re-examination of the genus by Selman (1963), P. obsoleta
was selected as the type species since a number of Olivier’s original Paropsis species
had subsequently been reassigned to different genera.
Paropsis atomaria was first described by Olivier (1807). Marsham (1808) also
described a species under the name P. atomaria, that was subsequently synonymised
with P. charybdis Stål, now a significant pest in New Zealand, where it was
accidentally introduced from Australia. Furthermore, Marsham described another
species, P. reticulata, which is now synonymised with P. atomaria Olivier (Selman,
1963). Other synonyms of P. atomaria include P. granulosa Boisd., P.
sanguinipennis Germar. and P. incarnata Erich. Paropsis incarnata has since been
recognized as a different species (Blackburn, 1901), and has been renamed P.
deboeri Selman (Selman, 1983).
Additionally, P. elytrura Blackburn and P. deboeri Selman, found in Western

Australia and Tasmania, respectively, are morphologically very similar to P.
atomaria and their taxonomic status regarding their relationship with P. atomaria
remains to be fully resolved.
Biology and ecology

Paropsis atomaria has been studied most comprehensively in the Australian Capital
Territory (ACT) and it is from here that most of its biological traits have been
characterized.
In the ACT, P. atomaria is bivoltine, with adults actively flying, feeding and mating
over the periods October to December and mid-late January to March (Carne, 1966).
The period between April and September sees a hibernation stage, where sexually
mature adults overwinter. Diapause commences in response to a reduction in day
22 Chapter 2: Literature review of Paropsis atomaria

length, whilst the exhaustion of fat bodies stimulates its termination (Carne, 1966).
The egg, larval and pupal stages continue beyond March, however, fourth instar
larvae are not encountered after the beginning of May.
Females lay eggs in a distinctive fashion. Eggs are deposited upright around the stem
of a young eucalypt shoot, forming a ringed cluster with each of the eggs projecting
radially (Fig. 1a,b) (Cumpston, 1939). The number of eggs per cluster varies between
20-100 (Elliott et al., 1998), with the diameter of the stem of high importance with
regard to selection of the oviposition site (Tanton and Khan, 1978a). Eggs are also
occasionally deposited on the apex of leaves (Fig 1b). Larvae hatch after 10-14 days
(Carne, 1966) by means of concentrated pressure in the thorax expanding hatching
spines on each side of the body (Cumpston, 1939). Newly-emerged larvae consume
their egg shells before moving to suitable foliage to feed.
One of the key factors determining suitable larval food is leaf toughness. This has
been demonstrated in a number of studies where leaf toughness versus larval
development was measured (Larsson and Ohmart, 1988; Ohmart et al., 1987).
Previous ideas that nitrogen and secondary plant compounds such as tannins, phenols
and oils may also be critical factors in choice of leaf have been largely dispelled after
studies demonstrated these to be much less important factors with regard to feeding
and development than leaf toughness (Fox and Macauley, 1977). Some work has also
looked at the ways in which P. atomaria sequesters secondary plant compounds
(Morrow and Fox, 1980).
The gregarious larval stage lasts for a period of 3-4 weeks, with a total of four
instars. Optimum larval developmental time occurs between 21°C-24°C. A notable
behaviour of the larvae is their defence mechanism. When disturbed, they elevate
their posterior end and evert defensive glands from between terminal segments.
Attacking insects, e.g., ants, have been observed to die within a few minutes of
contact with these glands (Carne, 1966). The defensive chemical secreted contains
hydrogen cyanide, benzaldeyhyde and glucose (Moore, 1967).
During the fourth instar and towards the end of the larval phase, individuals cling
less tenaciously to the leaves, and ultimately drop to the ground, where they form
cells several inches below the surface. About five days later, cell pupation
commences, and about ten days later the adult emerges. Three weeks following adult
emergence, females are competent to oviposit (Carne, 1966).

Notes on host range
Paropsis atomaria occurs on multiple species of the genus Eucalyptus L’Her
(Myrtaceae) (Table 1). It generally occurs in low numbers under natural conditions;
however, beetle populations occasionally outbreak in plantations, especially in
susceptible eucalypt species such as E. cloeziana (Shepherd, 2001), which is native
to the sub-tropics of Australia and is utilized by forestry. In the ACT, however, the
preferred host species of P. atomaria is E. blakelyi (Ohmart et al., 1985). Because E.
blakelyi does not extend far into the distribution of E. cloeziana, it remains unclear
which of these two hosts P. atomaria would preferentially use given the opportunity
to select one over the other.
Other commercial plantation eucalypts on which this species is found include E.

pilularis, E. grandis, E. dunnii and E. camaldulensis. Of these, E. grandis can be
subjected to intensive infestation, and occurs from within the distribution of E.
blakelyi, up into the tropics, within the distribution of E. cloeziana.
Table 1. Host plants from which Paropsis atomaria has been collected.
HOST PLANTS or CROPS AFFECTED MAIN HOST OTHER
(Please write scientific name) (main host on HOSTS
which pest causes
economic damage)
Eucalyptus cloeziana X
Eucalyptus pilularis X
Eucalyptus grandis X
Eucalyptus dunnii X
Eucalyptus cladocalyx X
Eucalyptus blakelyi X
Eucalyptus melliodora X
Eucalyptus polyanthemos X
Eucalyptus leucoxylon X
Eucalyptus conica X
Eucalyptus fastigata X
Eucalyptus rossi X
Eucalyptus macrorhyncha X
Eucalyptus radiata X
Angophora floribunda X
Eucalyptus pauciflora X
Eucalyptus divei X
Eucalyptus camaldulensis X

Symptoms - description
Larval aggregations may be seen on leaves of varying ages on affected trees. Young
larvae, particularly 1st and 2nd instar are usually located on the newest foliage, as they
are incapable of consuming the tougher, older leaves. Aggregations may comprise
individuals of varying instars. Heavy infestations are readily identifiable by the loss
of foliage and high numbers of larvae and adults. Lower level infestations may be
detected by the characteristic ‘scalloping’ of leaves caused by adult feeding, and/or
the loss of young foliage in the growth areas of the tree caused by larval attack.
Whilst tree death occurs under heavy infestation, the direct effect on the tree from
heavy infestation is more likely to be reduction in growth rate and wood quality. Tree
death is more likely in younger trees, which are more susceptible to infestation.

Morphology
Eggs
Eggs are elongate and laid upright with their longitudinal axis perpendicular to the
substrate, in a ringed cluster around the stem of a young shoot, or occasionally on the
leaf-tip (Fig 1a,b). They are distinctive in that they possess external ornamentation,
comprising four apical projecting horns and four longitudinal ridges. Colour varies
from almost white to mauve, with ornamentation tending to be more golden or
purplish in colouration. The number of eggs per egg cluster varies from 40-100, but
usually consists of around 60-80 individual eggs (Cumpston, 1939).
a)
b)
Figure 1. Paropsis atomaria egg batches laid on a) branchlet of Eucalyptus sp., and
b) leaf tip. Photos by Mark Schutze (not included in CABI datasheet)

Larvae
First instar larvae (Fig. 2) possess large, kidney-shaped tubercles on the meso- and
meta-thorax. General body colour is shiny yellow, with contrasting black pigmented
areas.
Figure 2. First instar Paropsis atomaria larvae. Photo by Mark Schutze.
Second and third instar (Fig. 3) larvae have lost pigmentation of the body tubercles
and prothoracic shield, so that the body is glistening yellow, with black head capsule
and terminal abdominal segments. Legs and spiracles are brown. Larvae of the
closely related species Chrysophtharta variicollis may be distinguished from P.
atomaria in that the former are dull cream in colour with black spiracles.
Figure 3. Third instar Paropsis atomaria larvae. Photo by Amy Carmichael
The fourth (final) instar is very distinctive (Fig. 4). Pigmented areas show increased
intensity, in stark contrast to the general yellow colouration of the body. The

prothoracic shield is black, and legs brown. A black median longitudinal line extends
from the prothoracic shield to the 7th abdominal segment, and large lateral black
areas on abdominal segments 1-6, with each partially enclosing a white spot. Lateral
tubercles enclosed in these areas are prominent and bear numerous setae (Cumpston,
1939).
Figure 4. Fourth instar Paropsis atomaria larvae showing distinctive black lateral
and dorsal markings. Photo by Amy Carmichael
Pupae
Pupae are pale to bright yellow in colouration, with light brown pubescence (Fig. 5).
Male pupae are generally smaller than females, with male body length averaging
13.9mm and females 15mm (Reid and Ohmart, 1989). The terminal portion carries a
ventral bilobed dark brown shield and two rows of small brown tubercles. Legs and
wing sheaths are pallid and translucent. Just prior to emergence, the hindwings
appear black, and the elytra pink (Cumpston, 1939). Pupae can be sexed via
examination of the ventral part of the abdominal apex (Reid and Ohmart, 1989). The
hind margin of sternite VIII is with a small median incision in males, and with a
deep, median cleft to the base in females. Also, lobes of sternite IX are ovate and
separated by their diameter in males, whereas in females they are transverse and
contiguous.

Figure 5. Paropsis atomaria pupa. Photo by Amy Carmichael
Adults
Strongly convex body (Fig. 6). Antennae moderately robust and filiform, consisting
of 11 segments (including scape and pedicel). Dorsal colouration: yellow with
orange/pale sanguineous markings, more intense on elytra. Elytra may also possess
darker markings, consisting of peripheral longitudinal line on either side and one-two
dots per elytra. Ventral colouration: pale fulvous yellow. Legs pale fulvous yellow.
Males: 10mm long, approximately 7-8 mm wide. Fore and mid basitarsi possess
uniform ventral discs of setae, which have an adhesive quality and are used for
gripping the elytra of the female during mating. Hind basitarsi do not possess such
setae, but rather a narrow glabrous line.
Females: Generally larger than males, 12-13mm long, and 8-9mm wide. All basitarsi
lack ventral disc of setae as seen in fore and mid basitarsi of males.
Figure 6. Pair of Paropsis atomaria adults. Photo by Mark Schutze

Similarities to other species
Other paropsine species may exist within eucalypt plantations; however their
morphology is sufficiently disparate to prevent misidentification. There exist two
closely related sibling species of P. atomaria: P. elytrura Blackburn (Fig. 7a) and P.
deboeri Selman (Fig. 7b). The former is restricted to south-west Western Australia,
the latter to Tasmania. These species are very similar morphologically, however due
to their allopatric distribution they are unlikely to cause confusion in the field.
a) b)
Figure 7. Species morphologically similar to Paropsis atomaria: a) P. elytrura from

south-west Western Australia, and b) P. deboeri from Tasmania. Photos by a) Mark
Schutze and b) David de Little. (not included in CABI datasheet)
Second and third instar larvae of P. atomaria may be confused with those of
Chrysophtharta variicollis Chap., however P. atomaria larvae are much brighter and
possess brown spiracles, whereas C. variicollis larvae are dull cream in colour and
have black spiracles.
Detection and inspection methods

Paropsis atomaria infestation is typified by extensive damage to young and
coppicing leaf-growth. This is principally caused by the larvae (especially early
instars) preferentially feeding upon softer, younger leaves. Adults and later instars
are capable of consuming older, tougher material, and may therefore be located on
other parts of affected trees.
Leaf damage is particular to life-stage, and it is possible to discern larval from adult
attack. For example, adult damage generally has the appearance of multiple semi-
circular ‘bite-marks’ along the perimeter of leaves, known as ‘scalloping’ (Fig. 8).
Larval attack is gregarious and nocturnal, consisting of consuming one entire leaf
before moving onto the next, leaving nothing but the bare twig (Cumpston, 1939).

Figure 8. Characteristic leaf ‘scalloping’ damage of eucalypt foliage by adults of
P. atomaria. Photo by Mark Schutze
Adults and larvae are typically observed over the warmer months, active from
October to March in the colder climates of Australia (e.g., ACT), however they may
be observed in the field a little beyond March in warmer, more northerly regions
such as south-east Queensland.
Overwintering adults conceal themselves between crevices and any leaves that may
be bound together (e.g., from spider silk), and may be found outside their active
months.
The distinctive egg batches are usually located on young branchlets, typically around
1-1.5mm in diameter. Egg clusters are occasionally laid on leaf tips.
Pupae are located in the soil under affected trees, and can only be detected by
extensive soil sampling.
Geographical distribution
Paropsis atomaria has a wide distribution within Australia. Records place it from
eastern central Queensland, along the east coast of Australia, to southern Victoria,
extending west to South Australia. There is one record of this species existing as far
north as Townsville. It extends as far west inland as Orange, NSW, however
significant populations are usually located in more coastal regions.

Invasiveness
In some countries this species is considered a potential invasive pest. See
‘Phytosanitary Risk’ for details.
Phytosanitary risk
This is an endemic pest to Australia, and has not been recorded as impacting on any
other countries. Any concern regarding P. atomaria should be restricted to industries
where Eucalyptus species are the commodity in question. Other species of paropsine
beetles have been accidentally introduced to New Zealand and South Africa from
Australia where they cause considerable defoliation to commercial hardwood forests.
Paropsis atomaria is listed as of high risk potential for importation of unprocessed

logs into the United States (Kleinjunas et al., 2003). In New Zealand, where the
accidental introduction of paropsine beetles has caused considerable damage to
eucalypt plantation productivity (Withers, 2001), P. atomaria is listed as a regulated
pest for imports from Australia, including pests potentially associated with bark,
wood packing and sawn wood (Ormsby, 2001). Phytosanitary treatment options
include fumigations, heat treatment, reshipment, or destruction (Ormsby, 2001).
Means of movement and dispersal
Natural dispersal (non-biotic)

There remains a risk for P. atomaria eggs and larvae to be transported if plant
material from an infested area is moved to a new, un-infested area. Likewise, there is
a small risk of pupae being relocated if soil from an infested area is also moved.
Adults and pre-pupae may be transported inadvertently in camping gear.
Silvicultural practices
It is possible that overwintering adults may be transported in logs, as they may be
located beneath bark or in splits and cracks in the wood (Simmul & deLittle, 1999)
Movement in trade
There is potential for movement of individuals through trade if untreated logs are
transported containing overwintering adults beneath bark.

Notes on natural enemies
Paropsis atomaria is attacked by numerous hymenopteran and dipteran parasitoids
and hyperparasitoids, as well as several predators. Tanton and Khan (1978a)
highlighted numerous such species.
Eggs are primarily parasitized by hymenopteran species, including Aphanomerella

ovi Dodd (Platygasteridae), Neopolycystus insectifurax Gir. (Pteromalidae), and
Enoggera sp. (Pteromalidae). Baeoanusia albifunicle (Girault) (Platygasteridae) is a
hyperparasitoid of Enoggera sp. (Jones & Withers, 2003). Additionally, unidentified
species of Epiencyrtus (Encyrtidae) and Trissolcus (Scelionidae) emerged from eggs.
Parasitism causes up to 20% mortality, with parasitized eggs possessing a dull brown
appearance rather than the typically glossy, yellow, unaffected eggs.
Larvae are parasitised by hymenopteran and dipteran species. Dipteran species

include members of the Tachinidae: Froggattimyia anguliventris Mall., F. tillyardi
Mall., and Paropsivora sp. These species are common each year. Hymenopteran
parasitoids include Eadya paropsidis (Huddleston & Short) (Braconidae), Bracon sp.
(Braconidae) and Tetrastichus sp. (Eulophidae), with E. paropsidis representing the
dominant larval parasitoid (Simmul & deLittle, 1999), usually laying six eggs per
host and causing up to 93% parasitisation (Tanton & Epila, 1984). Furthermore, two
hyperparasitoids, Mesochorus sp. (Ichneumonidae), and Perilampus tasmanicus
(Cameron) (Pteromalidae), hyperparasitise tachinid puparia. Parasitised larvae show
symptoms in the pre-pupal stages, taking on a darker colouration, with tachinid
larvae boring a hole out of the body to pupate. Tachinid larvae that remain within the
larvae during pupation are attacked by hyperparasitoids. High levels of
parasitiasation of larvae occur in Febuary-March. Rates of parasitisation range from
0%-20% for tachinids, and 0%-25% for hymenopterans. Combined parasitisation
rates range from 0%-41%.
Adult parasitisation is less common, and in cases where it has been observed, a
protozoan, Pleistophora sp. is found to have been the cause. Adults parasitized by
this species exhibit lack of coordination and ability to maintain adequate contact to
the substrate.
Predation is typically seen on eggs and larvae. Principle insect predators include
coleopterans and hemipterans. Coleopteran predators consist of coccinellids, such as

Cryptolaemus montrouzieri Muls., Harmonia (Leis) conformis (Boisd.), Rhyzobius
discolor Er., R. ventralis Er, and Cleobora mellyi (Mulsant). Only first and second
instar larvae are particularly susceptible to attack. Hemipteran predators include
species from the family Pentatomidae, Cermatulus nasalis (West.) and Oechalia
schellenbergii (Guer.-Men.). The former species, C. nasalis feeds on sluggish,
fourth-instar larvae, whist O. schellenbergii feeds on eggs as well as larvae. A
possible predator, Rayieria basifer (Walk.) (Miridae), has been observed on P.
atomaria larvae, but not attacking them.
Control
Biological control
No biological control programmes have been developed for this pest species.
Host-Plant Resistance
Even though P. atomaria exists on multiple eucalypt species, certain eucalypts are
more susceptible to attack than others. Where possible, less susceptible alternatives
should be considered for forestry programs. See ‘Notes on Host Range’ for details.
Pheromonal control
Little is known concerning pheromonal attractants in this species, so as a result no
such means of control have been developed.
Chemical control
Two chemicals were assessed by Tanton and Khan (1978b,c,d): fenitrothion and
aminocarb. With regard to egg mortality, fenitrothion proved more effective at lower
does, being about 30 times more effective than aminocarb. It was found that a 20ppm
concentration of fenitrothion achieved 100% mortality, with concentrations as low as
0.625ppm achieving high levels of mortality. In contrast, aminocarb needed
applications of at least 62.5ppm in order to achieve comparable effectiveness.
Similarly, fenitrothion had an increased effect on larval mortality over aminocarb.
However, as Tanton and Khan (1978b) outlined, aminocarb was more effective than
fenitrothion when considering growth rate of 2nd instar larvae. In contrast, 4th instar
larval growth rate was more severely affected by fenitrothion than aminocarb.

Insecticide application did not affect food consumption by larvae of any given
weight, but rather the effect was a disruption of digestion and utilization of food. In
treated larvae, there was disintegration of cellular structures of the digestive system,
curtailing digestion and absorption (Tanton and Khan, 1978c). Consequently, there
was increased utilization of fat bodies. Fourth instar larvae were the most capable of
recovery after insecticide application.
Adults produced from treated larvae showed increased levels of deformity; however
they fed and excreted normally. Fecundity was reduced in all adults whose larval
stage had been treated, with aminocarb-treated individuals demonstrating lower
fecundity than fenitrothion-treated individuals. There were no adverse effects on
parasitoids in treated individuals. Furthermore, a later study demonstrated that
parasitized larvae were more susceptible to fenitrothion or DDT had higher mortality
rates if they had been parasitized (Tanton and Epila, 1984).
Paropsine populations can also be controlled using a pyrethroid-based insecticide ( -

cypermethrin) (Elliott et al. 1998), which has been used against Tasmanian
eucalyptus leaf beetles in quantities of 100g/L, and applied at a rate of 250mL/ha.
Additionally, maldison (500g/L) has been used against leaf beetles, amongst other
pests, in eucalypt and native plant situations in Western Australia. Maldison is
effective on contact or after digestion. Carbaryl is also effective.
In Queensland, two insecticides are used against leaf-beetle populations, namely

Dominex 100 ( -cypermethrin) and Saboteur 400 (systemic insecticide dimethoate).
Application of these chemicals is only recommended when levels of infestation are
severe (>50%), as infestation rates below this level are unlikely to cause significant
loss of growth to the tree.
Application of any of these insecticides should coincide with presence of the first
two instars, as it is at this stage that the pest is noticeable without having caused
serious damage.
Economic impact
Because the intensive cultivation of eucalypts is a fairly recent forestry initiative,
there exists little to no information on the precise economic impact P. atomaria has
on the industry.

Environmental impact
Environmental impact is negligible as this is an endemic species to Australia, and
adverse effects are only demonstrated within the forestry industry.

References
Baly JS, 1862. Descriptions of the species belonging to the genus Paropsis Journal of
Entomology, 2:291-310.
Blackburn T, 1901. Revision of the genus Paropsis Part VI Proceedings of the
Linnean Society of NSW, 26:159-196.
Carne PB, 1966. Ecological characteristics of the eucalypt defoliating chrysomelid
Paropsis atomaria Ol. Australian Journal of Zoology, 14:647-672.
Cumpston M, 1939. Observations on the bionomics and morphology of seven of the
tribe Paropsini (Chrysomelidae). Proceedings of the Linnean Society of
NSW, 64:353-366.
Elliott HJ, Ohmart CP, Wylie FR, 1998. Insect pests of Australian forests: ecology
and management Inkata Press, Melbourne.
Fox LR, Macauley BJ, 1977. Insect grazing on Eucalyptus in response to variation in
leaf tannins and nitrogen. Oecologia, 29:145-162.
Jones DC, Withers TM, 2003. The seasonal abundance of the newly established
parasitoid complex of the Eucalyptus Tortoise Beetle (Paropsis charybdis).
New Zealand Plant Protection, 56:51-55.
Kliejunas JT, Burdsall HH, DeNitto GA, Eglitis A, Haugen DA, Harverty MI,
Micales JA, Tkacz BM, Powell MR, 2003. Pest risk assessment of the
importation into the United States of unprocessed logs and chips of eighteen
Eucalypt Species from Australia. General Technical Report FPL-GTR-137.
Madison, WI: U.S. Department of Agriculture, Forest Service, Forest
Products Laboratory. 206 p.
Larsson S, Ohmart CP, 1988. Leaf age and larval performance of the leaf beetle
Paropsis atomaria. Ecological Entomology, 13:19-24.
Lawrence JF, Britton EB, 1996. Coleoptera (Beetles). In: The Insects of Australia: A
textbook for students and research workers. Melbourne University Press,
Melbourne.
Moore BP, 1967. Hydrogen cyanide in the defensive secretions of larval Paropsini
(Coleoptera: Chrysomelidae). Journal of the Australian Entomological
Society, 6:36-38.
Morrow PA, Fox LR, 1980. Effects of variation in Eucalyptus essential oil yield on
insect growth and grazing damage. Oecologia, 45:209-219.

Ohmart CP, Stewart LG, Thomas JR, 1985. Effects of food quality, particularly
nitrogen concentrations, of Eucalyptus blakelyi foliage on the growth of
Paropsis atomaria larvae (Coleoptera: Chrysomelidae). Oecologia, 65:543-
549.
Ohmart CP, Thomas JR, Stewart LG, 1987. Nitrogen, leaf toughness and the
population dynamics of Paropsis atomaria Olivier (Coleoptera:
Chrysomelidae) - A hypothesis. Journal of the Australian Entomological
Society, 26:203-207.
Olivier GA, 1807. Paropside. Hist. Nat. Mus., 92:596-598.
Ormsby M, 2001. Import Health Standard (draft). Ministry of Agriculture and
Forestry Biosecurity Authority, Wellington, NZ. 13 pp.
Phillips C, 1996. Insects, diseases and deficiencies associated with eucalypts in
South Australia. 1st edn. Primary Industries, SA Forests.
Reid CAM, Ohmart CP, 1989. Determination of the sex of pupae of Paropsis
atomaria Olivier, and the related paropsina (Coleoptera: Chrysomelidae).
Journal of the Australian Entomological Society, 28:29-30.
Selman BJ, 1963. A reappraisal of the genus Paropsis Ol. (Chrysomelidae,
Coleoptera), with particular reference to the species introduced into New
Zealand. The Annals and Magazine of Natural History, 13:43-47.
Selman BJ, 1983. The naming of the Tasmanian species of Paropsis Olivier
(Coleoptera: Chrysomelidae) Journal of the Australian Entomological
Society, 22:333-339.
Shepherd B, 2001. Chrysomelid leaf-eating beetle Paropsis atomaria (Olivier): its
damage to Eucalyptus cloeziana (F. Muell), with identification of potential
parasitic control agents in recently established southeast Queensland
commercial hardwood forests, Sunshine Coast University.
Simmul TL, de Little DW, 1999. Biology of the Paropsini (Chrysomelidae:
Chrysomelinae) Advances in Chrysomelidae Biology ML Cox Leiden,
Backhuys Publ. 463-477.
Tanton MT, Epila JSO, 1984a. Parasitization of larvae of Paropsis atomaria Ol.
(Coleoptera: Chrysomelidae) in the Australian Capital Territory Australian
Journal of Zoology, 32:251-259.
Tanton MT, Epila JSO, 1984b. Effects of DDT and Fenitrothion on field collected
larvae of a eucalypt-defoliating beetle, Paropsis atomaria Ol.: II. Effects of

treatment of the host host on the parasitoids. Australian Journal of Zoology,
32:337-344.
Tanton MT, Khan SM, 1978a. Aspects of the biology of the eucalypt-defoliating
chrysomelid beetle Paropsis atomaria Ol. in the Australian Capital Territory.
Australian Journal of Zoology, 26:113-120.
Tanton MT, Khan SM, 1978b. Effects of Fenitrothion and Aminocarb, at doses
giving low mortailty, on surviving eggs and larvae of the eucalypt-defoliating
chrysomelid beetle Paropsis atomaria Ol.: I. Methods, mortality and relative
toxicity. Australian Journal of Zoology, 26:121-126.
Tanton MT, Khan SM, 1978c. Effects of Fenitrothion and Aminocarb, at doses
chrysomelid beetle Paropsis atomaria Ol.: II. Biology of survivors. Australian
Journal of Zoology, 26:127-138.
Tanton MT, Khan SM, 1978d. Effects of Fenitrothion and Aminocarb, at doses
chrysomelid beetle Paropsis atomaria Ol.: III. Histological changes in treated
larvae. Australian Journal of Zoology, 26:139-146.
Withers TM, 2001. Colonisation of eucalypts in New Zealand by Australian insects.
Austral Ecology, 26:467-476.

Chapter 3
Species status and population structure of the Australian Eucalyptus

pest Paropsis atomaria Olivier (Coleoptera: Chrysomelidae)
This chapter has been published as:
Schutze, M.K., Mather, P.B. & Clarke, A.R. (2006) Species status and population
structure of the Australian Eucalyptus pest Paropsis atomaria Olivier (Coleoptera:
Chrysomelidae). Agricultural and Forest Entomology, 8, 323-332.
Statement of Joint Authorship
Schutze, M.K.
Designed and developed experimental protocol. Carried out field and laboratory
work, and analysed data. Wrote manuscript and acted as corresponding author
Mather, P.B.
P.B. Mather co-supervised the study design and experimental protocols, assisted in
the interpretation of data, and contributed to editing and structure of the manuscript.
Clarke, A.R.
A.R. Clarke was the principal supervisor of study design and experimental protocols,
and assisted in the interpretation of data and the construction of the manuscript.
40 Chapter 3: Species status and population structure of Paropsis atomaria

Species status and population structure of the Australian Eucalyptus
pest Paropsis atomaria Olivier (Coleoptera: Chrysomelidae)
Abstract
1 Paropsis atomaria Olivier represents an emergent pest of Eucalyptus plantations
in Queensland and New South Wales, Australia. Most prior studies on the
biology and control of P. atomaria have centred on populations from Canberra
in the Australian Capital Territory, but the biological relationship between
beetles from Canberra and those from up to 1500 km further north are unknown.
2 DNA markers were used to determine whether P. atomaria from Canberra are
the same biological species as those from Eucalyptus forestry plantations in
northern New South Wales and Queensland, where the beetle has become an
important pest. Using the mitochondrial gene, cytochrome c oxidase I (COI),
individuals collected from across the distribution of P. atomaria were
investigated for haplotype diversity and levels of mitochondrial divergence.
3 Within P. atomaria, genetic distance averaged 0.5% across 23 unique haplotypes
for 93 individuals, with an average of 14% difference between P. atomaria and
the outgroup species, Paropsis obsoleta. Significant genetic structure was
observed relative to geographical distribution, but not with respect to host plant
species of origin. Greatest divergence was between the southern-most sample
site (Canberra) and northern sites in New South Wales and Queensland,
indicating reduced gene flow between these regions.
4 Individuals from across eastern Australia belong to the same genetic species
with population substructuring evident. Consequently, there is no evidence to
suggest cryptic species complexes exist within the currently defined taxon.
Continued implementation of control strategies for P. atomaria across its
distribution is appropriate.
Keywords Cryptic species, cytochrome c oxidase I, forestry, leaf beetle, population
structure
Chapter 3: Species status and population structure of Paropsis atomaria 41

Introduction
The implementation of control measures and management programmes for pest
species relies on accurate, biologically meaningful identification. However, this
process rests upon original taxonomic designations from which identifications are
based. If the link between taxonomically defined and biological species (sensu
Paterson, 1991) is incongruous, management efforts may be misdirected, resulting in
wastage of time, effort and resources (Annecke & Moran, 1977; Mahon et al., 1982;
Walter, 2003). Unlike some historical beliefs that biological species would nearly
always match their taxonomically defined counterparts (Dobzhansky, 1941),
disparity between the two is now commonplace, with many previously identified
groups such as ‘strains’, ‘biotypes’ or ‘populations’ of one species probably
representing separate ‘good species’ (Clarke & Walter, 1995). Therefore, it is of
paramount importance that economically important species be investigated at the
outset of a control programme to ensure control measures are directed towards a
cohesive biological entity, rather than a collection of different biological species
masquerading under a single taxonomic identity, the so-called cryptic species
complex (Walter, 2003). Molecular studies over the last 5 years have demonstrated
that cryptic species occur much more frequently than previously thought, whereas
theory has predicted for many years that cryptic species may exist where courtship
relies on transitory signals (e.g. pheromones, calls, optical signals), rather than fixed
morphological features (Paterson, 1991).
Suspected species complexes are expensive to investigate in time and resources. It is

therefore desirable first to determine the likelihood of a cryptic complex and then
make informed decisions as to whether further investigation is a priority. A number
of clues may indicate if a cryptic complex exists, usually centred on disparate
biological observations within a taxonomically defined species. Broad polyphagy, for
example, is considered rare, because most insect species are host specific (Fox &
Morrow, 1981; Jermy, 1984; Claridge et al., 1997), particularly in the Lepidoptera,
Coleoptera (Singer, 2001) and Chrysomelidae (Mardulyn & Mililkovitch, 2005).
Therefore, highly polyphagous systems may consist of complexes of multiple
specialist species, rather than a single generalist species (Walter & Benfield, 1994;
Milne & Walter, 1998, 2000). Disparity in life history characteristics may also
suggest complexes, as demonstrated by variation in insecticide resistance (Subbarao

et al., 1988; Hemingway et al., 1999; Umina & Hoffman, 1999), reproductive
strategies (Wirth et al., 1998; Guillet et al., 2000a, b), prey suitability (Krafsur &
Obrycki, 2000), diapause characteristics (Kenis & Mills, 1998) or fecundity
(Frohlich et al., 1999).
Using the above criteria, the Eucalyptus leaf beetle, Paropsis atomaria Olivier
(Coleoptera: Chrysomelidae), is a candidate for a priori suspicion of a cryptic
species complex. The beetle is endemic to eastern Australia and is widely distributed
from the temperate south (Victoria and South Australia) to the tropical north
(Queensland), a distance greater than 2000 km. Paropsis atomaria has been studied
predominantly in the Australian Capital Territory (towards the southern part of its
range) and, as such, most data describing life history characteristics and other
biological attributes have been based on individuals sourced from this region. Such
studies include early natural history work under the synonym Paropsis reticulata
(Cumpston, 1939), assessment of ecological characteristics (Carne, 1966a, b),
research into causes of mortality and effect of control measures (Tanton & Khan,
1978a, b, c, d; Tanton & Epila, 1984a, b, c, d), and investigation of growth rate and
fecundity responses to variable plant compounds (Ohmart et al., 1985a, b; Ohmart et
al., 1987; Larsson & Ohmart, 1988; Ohmart & Larsson, 1989). Studies on P.
atomaria outside of the Australian Capital Territory are scant, particularly in
northern parts of its natural range, due to its historical nonpest status in the
subtropics. For example, two studies of emergent eucalypt pests, one from New
South Wales (Stone, 1993) and the other from Queensland (Wylie & Peters, 1993),
demonstrate that although P. atomaria was an established pest in New South Wales
(which entirely surrounds the Australian Capital Territory), the beetle did not rate a
mention in the Queensland appraisal. It is only after recent large-scale establishment
of eucalypt plantations in the Australian subtropics, and the consequent emergence of
P. atomaria as a pest in these plantations, that studies into biological characteristics
of P. atomaria from regions outside the Australian Capital Territory have been
undertaken (Bailey, 2001, unpublished Honours thesis). This increased pest status is
probably due to the emergence of P. atomaria from adjacent natural forests into
newly-established forestry stands, rather than a range expansion from southern
populations. These investigations have yielded differences in life-history
characteristics between populations collected from south-east Queensland compared

with previous studies from the Australian Capital Territory (Carne, 1966a),
particularly with regard to the optimal developmental temperatures of larvae.
Other traits also suggest that P. atomaria may be polytypic. The beetle possesses a
wide host range and utilizes at least 21 Eucalyptus species across its current
distribution (CAB International, 2005), in spite of it belonging to a taxonomic group
that typically possesses much narrower host affinities (Edwards & Wanjura, 1990).
Additionally, across the broad geographical distribution of P. atomaria, there is a
marked change in climatic conditions, from the temperate south to the tropical north,
and the eucalypts on which P. atomaria feed rarely occur naturally across this entire
range, with nearly all possessing limited regional distributions (Brooker & Kleinig,
1983, 1994). Paropsis atomaria therefore may represent either a single biological
species, tolerant of wide ranging environmental variables and possessing a large host
range, or multiple independently evolving populations, which may include cryptic
species, adapted to local environmental conditions and hosts.
Molecular studies utilizing the mitochondrial genome have demonstrated

effectiveness when assessing potential cryptic species complexes for groups from
varied taxa (Yeh et al., 1997; Danforth et al., 1998; Funk, 1998; Wirth et al., 1998;
Frohlich et al., 1999; Giblin et al., 2000; Guillet et al., 2000b; Uribe et al., 2001;
Salvato et al., 2002; Mattiucci et al., 2003). Calculation of FST values from sequence
data and subsequent determination of gene flow estimates enable a relatively
straightforward assessment of potential evolutionary divergence between sampled
populations. The protein coding mtDNA gene cytochrome c oxidase I (COI) is
commonly used in such studies due to its versatility across multiple taxa because
both its structure and function are well understood (Simon et al., 1994; Lunt et al.,
1996). For cryptic species studies, the highly variable regions of the gene are often
targeted because they are more likely to reveal differences between recently evolved
sibling species should they occur.
The present study aimed to utilize the highly variable region of the molecular marker
COI to determine: (i) whether there is evidence for cryptic species or locally adapted
populations within P. atomaria and (ii) the level of genetic structure within P.
atomaria and assess potential causes for such intraspecific variation if such cryptic
complexes are not apparent.

Materials and methods
Insect collection and identification

Adult and larval beetles were hand collected into > 70% alcohol over two seasons
(2003 and 2005) from eucalypt plantation and revegetation sites across the east coast
of Australia (Table 1, Fig. 1). Due to the gregarious nature of P. atomaria, and the
tendency of individuals to remain sedentary in the presence of abundant resources
(Carne, 1966a), samples were collected from multiple locations within a field site to
reduce the chance of collecting related individuals from the same broods. Because of
the lack of reliable systematic keys, field identification of P. atomaria by forestry
researchers is currently based on experience, gross morphological comparison to
illustrations from handbooks (Waterson & Urquhart, 1995), and by the presence of
highly distinctive egg masses (Cumpston, 1939). Consequently, material for this
study was collected in the same manner. The outgroup species for genetic analysis
was Paropsis obsoleta Olivier, collected from Rosedale, central Queensland (24° 38
S, 151° 55 E).
Figure 1: Map of eastern Australia showing the five sampling locations for Paropsis
atomaria investigated in this study.

Table 1. Location, date, sample size (n) and host species of P. atomaria investigated
in this study.
Location and date n Eucalyptus host

Lowmead Queensland, 14 E. grandis X E. camaldulensis
24º29’22”S, 151 º42’14”E
February 2005
Beerburrum Queensland, 24 E. cloeziana and E. pilularis
26 º58’02”S, 153 º03’06”E
March 2003
Bangalow N.S.W., 21 E. grandis and E. pilularis
28 º43’11”S, 153 º31’07”E
April 2003
Canberra A.C.T., 28 E. spp.
35 º18’51”S, 149 º09’16”E
March 2003
Mount Gambier S.A., 6 E. cladocalyx
37 º 54’42”S, 140 º 53’13”E
March 2003
DNA sequencing
Genomic DNA was extracted from larvae using a Chelex extraction technique
(Walsh et al., 1991). DNA of adults was extracted from three legs in a standard
proteinase K phenol : chloroform extraction method (Fukatsu, 1999).
A highly variable fragment of COI was polymerase chain reaction (PCR) amplified
using primers UEA7 (5 -TAC AGT TGG AAT AGA CGT TGA TAC-3 ) and
UEA10 (5 -TCC ATG CAC TAA TCT GCC ATA TTA-3 ) (Lunt et al., 1996).
PCR amplification was carried out in a 25-µL final volume reaction containing 3.1
µL Biotech 10 × PCR buffer, 3 mM MgCl2, 0.4 mM dNTPs, 0.4 µM of each primer,
1.5 U Biotech Taq polymerase, 1 µL tDNA and 15.6 µL ddH2O. PCRs were run on
an Eppendorf mastercycler gradient thermocycler, with the profile: 94 °C for 5 min;
39 cycles of 95 °C for 40 s, 48 – 56 °C for 1 min, 72 °C for 40 s; 72 °C for 8 min;
held at 4 °C.

PCR products were visualized on 1.5% agarose gels run in TBE buffer and stained
with ethidium bromide prior to DNA purification using the Roche High PCR product
purification kit (Roche, Germany). Gel verification and spectrophometer
quantification were conducted prior to sequencing. Sequencing PCR was conducted
according to manufacturer’s specifications using the BigDye Terminator mix version
3.1 and the UEA7 primer. Sequencing was carried out in 12-µL final volume
reactions, consisting of 1 µL BigDye Terminator (PE Applied Biosystems, Foster
City, California), 3 µL dilution buffer, 0.27 µM primer and 4 µL of tDNA (at 5 – 20
ng / µL). The sequencing profile was: 94 °C for 5 min; 29 cycles of 96 °C for 10 s,
50 °C for 5 s and 60 °C for 4 min; held at 4 °C. Purification of sequencing reactions
was accomplished via the salt method, and samples were analysed on an ABI 3730xl
automated DNA sequencer by the Australian Genome Research Facility Ltd.
Statistical analysis
Sequence data were aligned and edited in Biomanager, version 2.0 (Australian
National Genomic Information Service; http://biomanager.angis.org.au) after
confirmation via National Centre for Biotechnology Information GenBank
(http://www.ncbi.nlm.nih.gov/). Sequence alignment was achieved using CLUSTAL W
(Thompson et al., 1994), followed by manual cross-verification. Unique haplotypes
were determined in the program COLLAPSE, version 1.2 (D. Posada, available at
http://darwin.uvigo.es/) and imported into MEGA, version 2.1 (Kumar et al., 2001),
where the number of variable sites and nucleotide and amino acid composition were
calculated. Sequences of each P. atomaria haplotype are available in GenBank under
accession numbers DQ335220 – 42, and the sequenced fragment for outgroup P.
obsoleta under accession number DQ338533.
The computer program TCS version 1.2 (Clement et al., 2000) was used to construct
a statistical parsimony haplotype network using the 95% parsimony criterion. This
procedure provides an overall visual impression of how divergent haplotypes are
from each other with regard to number of base pair changes, allowing for a
qualitative assessment of haplotype distribution with regard to sample site and host
plant of origin.
Genetic differentiation within and among groups incorporating haplotype frequencies

together with evolutionary divergences among haplotypes was determined by

analysis of molecular variance (Excoffier et al., 1992) using ARLEQUIN, version
2.000 (Schneider et al., 2000). Distances among haplotypes were estimated using the
Kimura 2 parameter distance method that accounts for differing rates of transition
versus transversion mutations (Kimura, 1980). The significance of resulting global F
statistics and corresponding pairwise FST were determined using a nonparametric
permutation procedure incorporation 1000 permutations (Excoffier et al., 1992).
Estimation of the number of migrants between populations (Nm) were calculated
from FST but the results are not presented due to the inherent unreliability of such
estimates due to the likely violation of key assumptions (Whitlock & McCauley,
1999). Two separate analyses were undertaken: one to determine geographical
substructure and the other to assess substructure based on host plant of origin.
For analysis based on geography, individuals collected from a district were grouped
together for: Lowmead, Queensland; Beerburrum, Queensland; Bangalow, New
South Wales; and Canberra, Australian Capital Territory (Fig. 1). Samples from
Mount Gambier were excluded from geographical analysis due to low sample size.
The second analysis, based on host plant of origin, consisted of grouping individuals
based on the species of eucalypt from which they were sampled. In some cases, the
host plant species of origin was unknown, especially for material collected from
Canberra. Additionally, material collected from Lowmead was sourced from a single
species of eucalypt, a hybrid of E. grandis and E. camaldulensis. To avoid
confounding host plant effect by geography, only those sample sites where at least
two host species occur sympatrically were included. The host plant species assessed
for this analysis were E. cloeziana, E. pilularis, and E. grandis.
Results
Sequence variation
Sequences of a 508-bp fragment of the COI gene were obtained for 93 P. atomaria
individuals, representing 23 unique haplotypes. Of the 508 sites, 26 were variable. A
single homologous fragment was amplified for the outgroup species P. obsoleta.
Pairwise distances demonstrated reduced intraspecific distance compared with
interspecific distance. Pairwise distances among P. atomaria populations were in the
range 0 – 1.4% (average 0.5%), whereas those between P. atomaria and P. obsoleta
were in the range 13.7 – 14.7% (average 14%). The amplified region was AT rich

(A, 31%; C, 16.1%; G, 13.7%; T, 39.2%), with a larger number of transition to
transversion mutations present. Translation into amino acid identity revealed four P.
atomaria individuals that each had a single unique amino acid substitution with
respect to the remaining identical 89 individuals, compared with eight amino acid
substitutions between P. atomaria and the single P. obsoleta sequence.
The 95% haplotype network generated in TCS (Figs 2 and 3) exhibited no evidence
for homoplasy due to the absence of any loops (Posada & Crandall, 2001) and the
most common haplotypes in the network were Ha 1 and Ha 6 (Table 2, Figs 2 and 3).
When collection site was mapped onto the network, no haplotypes clustered with any
specific site; instead well-represented haplotypes occurred at multiple collection sites
(Fig. 2). Similarly, when host plant data were mapped onto the network, most
haplotypes representing more than a single individual did not associate exclusively
with a single host species, but consisted of individuals collected from multiple
eucalypt hosts (Fig. 3).
Figure 2: 95% parsimony network of 23 haplotypes obtained by sequencing a 508 bp

fragment of the mtDNA COI gene for 93 individuals of Paropsis atomaria generated
by TCS. Arbitrary haplotype numbers refer to haplotypes listed in Table 2. Size of
oval represents relative numbers of individuals possessing the haplotype. Small filled
circles represent hypothetical intermediate haplotypes. Shading denotes what
proportion of each haplotype is represented by individuals from a specific collection
site.

Figure 3: 95% parsimony network of 23 haplotypes obtained by sequencing a 508
bp fragment of the mtDNA COI gene for 93 individuals of Paropsis atomaria as
generated by TCS. Arbitrary haplotype numbers refer to haplotypes listed in Table 2.
Size of oval represents relative numbers of individuals possessing the haplotype.
Small filled circles represent hypothetical intermediate haplotypes. Shading denotes
what proportion of each haplotype is represented by individuals collected from a
specific species of Eucalyptus.
Variation between sites within P. atomaria

An assessment of relative haplotype frequencies revealed ten common haplotypes at
two sample sites, with 54% of haplotypes unique to a specific site (Table 2).
However, unique haplotypes were represented predominantly by single individuals
(85%), with haplotypes Ha 12, Ha 18 and Ha 21 represented by multiple individuals
from Bangalow (n = 4), Beerburrum (n = 5) and Lowmead (n = 2), respectively.
Haplotypes Ha 1 and Ha 6 were most frequently represented, being the common

haplotypes in Beerburrum (0.333) and Canberra (0.464), respectively. The two
common haplotypes were not unique to any site, but were sampled from across the
distribution, with Ha 1 found in all sites except Mount Gambier and Ha 6 found in all
sites except Lowmead.

LOCATION HOST
Mount E. E. E. E.g. X
Site/Host Lowmead Beerburrum Bangalow Canberra Gambier pilularis cloeziana grandis E.c.
n 14 24 21 28 6 18 15 12 14
Haplotype
Ha 1 0.14 (2) 0.33 (8) 0.10 (2) 0.21 (6) 0.33 (6) 0.27 (4) 0.14 (2)
Ha 2 0.05 (1) 0.04 (1) 0.08 (1)
Ha 3 0.05 (1) 0.04 (1) 0.06 (1)
Ha 4 0.14 (2) 0.04 (1) 0.05 (1) 0.04 (1) 0.07 (1) 0.08 (1) 0.14 (2)
Ha 5 0.04 (1)
Ha 6 0.04 (1) 0.05 (1) 0.46 (13) 0.50 (3) 0.07 (1) 0.08 (1)
Ha 7 0.08 (2) 0.11 (3) 0.06 (1) 0.07 (1)
Ha 8 0.29 (4) 0.04 (1) 0.05 (1) 0.07 (2) 0.11 (2) 0.29 (4)
Ha 9 0.05 (1) 0.06 (1)
Ha 10 0.07 (1) 0.08 (2) 0.14 (3) 0.11 (2) 0.07 (1) 0.15 (2) 0.07 (1)
Ha 11 0.04 (1) 0.14 (3) 0.11 (2) 0.07 (1) 0.08 (1)
Ha 12 0.19 (4) 0.06 (1) 0.23 (3)
Ha 13 0.05 (1) 0.08 (1)
Ha 14 0.05 (1) 0.08 (1)
Ha 15 0.05 (1) 0.08 (1)
Ha 16 0.04 (1) 0.06 (1)
Ha 17 0.04 (1) 0.07 (1)
Ha 18 0.21 (5) 0.06 (1) 0.27 (4)
Ha 19 0.04 (1) 0.07 (1)
Ha 20 0.07 (1) 0.07 (1)
Ha 21 0.14 (2) 0.50 (3) 0.14 (2)
Chapter 3: Species status and population structure of Paropsis atomaria

Ha 22 0.07 (1) 0.07 (1)
Ha 23 0.07 (1) 0.07 (1)
TABLE 2: Number of Paropsis atomaria screened for each study with relative
haplotype frequencies and absolute number of individuals sampled for each haplotype
(in parenthesis). Left-hand of table denotes which haplotypes are associated with a
particular sampling location and in what proportion. Right-hand of table denotes
51
which haplotypes are associated with a particular Eucalyptus host and in what
proportion. E.g. X E.c. = Eucalyptus grandis X E. camaldulensis hybrid.
Individuals from Bangalow possessed the highest nucleotide (0.007) and haplotype
diversity (0.938) (Table 3). Conversely, individuals from Canberra possessed the
lowest nucleotide diversity (0.003) and Mount Gambier the lowest haplotype
diversity (0.600), albeit with high standard deviation (due to low sample size).
Table 3: Number of Paropsis atomaria individuals screened for location and

Eucalyptus host molecular analyses with number of haplotypes, nucleotide diversity
and haplotype diversity. Standard deviations (SD) are provided for nucleotide and
haplotype diversity. (E.g. X E.c. = Eucalyptus grandis X E. camaldulensis hybrid). n
= sample size.
Populations/ Nucleotide Number of Haplotype

host species n
diversity ± SD haplotypes diversity ± SD
Lowmead 14 0.004 ± 0.003 8 0.901 ± 0.058
Beerburrum 24 0.005 ± 0.003 11 0.855 ± 0.054
Bangalow 21 0.007 ± 0.004 13 0.938 ± 0.032
Canberra 28 0.003 ± 0.002 8 0.743 ± 0.070
Mt Gambier 6 0.004 ± 0.003 2 0.600 ± 0.129
E. pilularis 18 0.006 ± 0.003 10 0.882 ± 0.063
E. cloeziana 15 0.005 ± 0.003 9 0.886 ± 0.061
E. grandis 13 0.007 ± 0.004 10 0.949 ± 0.051
E. cladocalyx 6 0.004 ± 0.003 2 0.600 ± 0.129
E.g. X E.c. 14 0.004 ± 0.003 8 0.901 ± 0.058
Significant population differentiation was found among the four sites examined (FST
= 0.0853, P < 0.05). The majority of variation was partitioned within sites (91.47%,
d.f. = 83, P < 0.05), rather than among sites (8.53%, d.f. = 3, P < 0.05). FST
estimates were highest for all pairwise Canberra comparisons (mean = 0.144, P <
0.05 for all comparisons), suggesting reduced gene flow between Canberra and
northern sample sites relative to the level of gene flow observed amongst northern
sites Beerburrum, Bangalow, and Lowmead (mean pairwise FST = 0.022, P > 0.05
for all comparisons) (Table 4).

Table 4: Pairwise genetic differentiation between populations of Paropsis atomaria
from four sample locations in eastern Australia. Below diagonal: pairwise estimates
of Fst calculated by AMOVA employing Kimura 2 parameter distances among
haplotypes. Asterisk denotes statistical significance (p < 0.05). Above diagonal:
distance in kilometres (km) between sample locations.
Beerburrum Bangalow Canberra Lowmead

Beerburrum - 200 km 1000 km 320 km
Bangalow 0.012 - 800 km 520 km
Canberra 0.123* 0.153* - 1,320 km
Lowmead 0.029 0.026 0.157* -
Variation between host plants within P. atomaria

Genetic differentiation based on host plant data was not significant (FST = 0.00197, P
> 0.05). Pairwise FST values were generally low (mean = 0.009, P > 0.05 for all
comparisons), indicating an absence of genetic structuring related to host plant of
origin (Table 5) with a low percentage of variation explained by the among-group
comparison (0.2%, d.f. = 2, P > 0.05) compared with within-site comparisons
(99.8%, d.f. = 42, P < 0.05).
Table 5: Estimation of genetic differentiation of populations of Paropsis atomaria

from three different Eucalyptus host plants. Pairwise estimates of Fst calculated by
AMOVA employing Kimura 2 parameter distances among haplotypes.
E. pilularis E. cloeziana
E. cloeziana 0.002
E. grandis 0.011 -0.014
Discussion
Observed levels of intraspecific variation within P. atomaria were low (mean =
0.5%) compared with outgroup, P. obsoleta (mean = 14%), supporting the
hypothesis that the P. atomaria populations sampled constitute a single species.
Comparable levels of intra- and interspecific variation of the COI fragment used in
the present study were found compared with studies of other insect taxa that have
examined the same region (Jamnongluk et al., 2003; Otranto et al., 2003).
The parsimony haplotype network revealed no evidence for divergent haplotypes

strictly associated with either collection locality or host plant of origin, supporting
the contention that there were no geographically or host plant restricted races within

the taxon. The implications of these findings are that investigations carried out on P.
atomaria in the Australian Capital Territory can be applied to populations elsewhere
in the north of the range, particularly from northeast New South Wales to central
Queensland, because individuals from the northern extent of the range are expected
to possess similar biological attributes to those found in Canberra because they
belong to the same gene pool. Local differentiation, if observed, may therefore be the
result of phenotypic plasticity responding to local environmental conditions, rather
than underlying genetic differences between regions or host plants, as would be
expected in a case of divergent evolutionary lineages such as cryptic species
complexes. However, it is important to note that the gene under study (COI)
represents a potentially unlinked neutral marker, which may not reveal modes of
local adaptation to prevailing conditions should they be apparent.
Assessment of gene flow among regions reveals historical dispersal between the
northern sites, Lowmead, Beerburrum, and Bangalow. However, pairwise FST values
were one order of magnitude higher (combined with corresponding low estimated
migration rate, data not shown) between Canberra and the northern sites, suggesting
reduced gene flow between the southern site and its northern counterparts. Specific
isolation-by-distance (IBD) tests (e.g. Mantel tests) were not used due to inadequate
number of populations sampled, resulting in an unacceptable risk of Type II error
(Peterson & Denno, 1998). However, IBD is a potential explanation for this system
because, similar to most leaf beetle species (Mardulyn & Mililkovitch, 2005), P.
atomaria in the field is assumed to lead a moderately sedentary existence, with adults
rarely leaving an area if resources are locally abundant (Carne, 1966a), and larvae
completing their entire development on or very near to the host plant where they
were deposited as eggs. As a consequence, long-distance migration is probably only
likely when local resources are depleted, necessitating dispersal to new areas.
Considering the broad host range of P. atomaria on such abundant hosts as
eucalypts, we consider that populations in any one area will rarely encounter such a
reduction in available resources that there would be a resulting need to disperse large
distances to find new host plants. An alternative explanation is provided by
population expansion. As with the case in testing for IBD, low numbers of
populations sampled resulted in the inability to conduct rigorous statistical analyses
to test this theory. Regardless, IBD is proposed as being more probable due to the

life-history characteristics of P. atomaria and the low likelihood that this species has
expanded into new areas due to long historical associations between P. atomaria and
its many host tree species over much of the Australian landscape.
No differentiation was revealed between host plant species, with populations on E.

cloeziana, E. pilularis and E. grandis, essentially the same genetically. Although
these three species encompass only a small number of the known host plants for P.
atomaria, the results support the hypothesis that P. atomaria represents an
oligophagous species on Eucalyptus species, with no strict host-associated races or
sibling-species. Paropsis atomaria possesses a high tolerance for secondary eucalypt
metabolites, such as variable concentrations in phenols (Fox & Macauley, 1977) and
the ability of larvae to readily absorb a major proportion of ingested terpenoids
(Ohmart & Larsson, 1989) may assist in explaining its capacity for oligophagy, with
the predominant factors affecting fecundity and larval growth being leaf nitrogen
levels and leaf toughness (Ohmart et al., 1985a). Nitrogen levels are negatively
correlated with leaf toughness and although a decrease in nitrogen results in a
decrease in P. atomaria fecundity, the corresponding increase in leaf toughness
results in increased mortality in early instars due to the inability of the larvae to
physically chew the leaves (Ohmart et al., 1987). Principle factors deterring feeding
may therefore include surface chemicals (i.e. waxes) or volatiles (i.e. essential oils)
(Ohmart et al., 1987). However, if such deterrents are not present, or their impact on
host selection is low [as is the case for the related paropsine Chrysophtharta
bimaculata (Olivier) (Steinbauer et al., 1998)], then it is unlikely P. atomaria would
discriminate between host species, therefore resulting in a polyphagous herbivore.
This investigation into the potential existence of a cryptic complex within P.

atomaria has revealed that the current taxonomic definition of P. atomaria remains
sound. Although population analyses have revealed intraspecific structuring,
suggestive of IBD, this intraspecific variation remains minor. Further sampling,
especially for intermediate populations, is recommended to resolve the likely
mechanism responsible for observed population structuring. Individuals collected
from within the sampled distribution, and identified using gross external
morphology, do represent the same biological species. Therefore, new work on P.
atomaria, undertaken in the northern limits of its distribution where it is a serious

threat to the forestry industry, may reliably build upon studies previously undertaken
in the southern part of the species’ range.
Acknowledgements
We thank Stephen Monteith, Helen Nahrung, Martin Henery and Angela Duffy for
assistance with field collection, Simon Lawson, Richard Lunney, QDPI-Forestry and
New South Wales State Forests for access to plantation sites, David Hurwood for
assistance with molecular analysis, and Katarina Mikac for useful comments on the
manuscript.

References
Annecke, D. P. & Moran, V. C. (1977) Critical reviews of biological control in South
Africa. Journal of the Entomological Society of Southern Africa, 40, 127 –
145.
Brooker, M. I. H. & Kleinig, D. (1983) F ield Guide to Eucalypts: South-Eastern

Australia. Inkata Press, Australia.
Brooker, M. I. H. & Kleinig, D. (1994) F ield Guide to Eucalypts: Northern

Australia. Inkata Press, Australia.
CAB International (2005) Forestry Compendium. CAB International, U.K.
Carne, P. B. (1966a) Ecological characteristics of the eucalypt defoliating

chrysomelid Paropsis atomaria Ol. Australian Journal of Zoology, 14, 647 –
672.
Carne, P. B. (1966b) Growth and food consumption during the larval stages of
Paropsis atomaria (Coleoptera: Chrysomelidae). Entomologia
Experimentalis et Applicata, 9, 105 – 112.
Claridge, M. F., Dawah, H. A. & Wilson, M. R. (1997) Species in insect herbivores

and parasitoids – sibling species, host races and biotypes. Species: The Units
of Biodiversity (ed. by M. F. Claridge, H. A. Dawah and M. R. Wilson), pp.
247 – 272. Chapman & Hall, U.K.
Clarke, A. R. & Walter, G. H. (1995) ‘Strains’ and the classical biological control of
insect pests. Canadian Journal of Zoology, 73, 1777 – 1790.
Clement, M., Posada, D. & Crandall, K. A. (2000) TCS: a computer program to

estimate gene genealogies. Molecular Ecology, 9, 1657 – 1659.
Cumpston, M. (1939) Observations on the bionomics and morphology of seven of

the tribe Paropsini (Chrysomelidae). Proceedings of the Linnean Society of
New South Wales, 64, 353 – 366.
Danforth, B. N., Mitchel, P. L. & Packer, L. (1998) Mitochondrial DNA

differentiation between two cryptic Halictus (Hymenoptera: Halictidae)
species. Annals of the Entomological Society of America, 91, 387 – 391.

Dobzhansky, T. (1941) Genetics and the Origin of Species. Colombia University
Press, New York, New York.
Edwards, P. B. & Wanjura, W. J. (1990) Physical attributes of eucalypt leaves and

the host range of chrysomelid beetles. Symposia Biologica Hungarica, 39,
227 – 236.
Excoffier, L., Smouse, P. E. & Quattro, L. M. (1992) Analysis of molecular variance

inferred from metric distances among DNA haplotypes: application to human
mitochondrial DNA restriction data. Genetics, 131, 479 – 491.
Fox, L. R. & Macauley, B. J. (1977) Insect grazing on Eucalyptus in response to

variation in leaf tannins and nitrogen. Oecologia, 29, 145 – 162.
Fox, L. R. & Morrow, P. A. (1981) Specialization: species property or local

phenomenon? Science, 211, 887 – 893.
Frohlich, D. R., Torres, J. I., Bedford, I. D., Markham, P. G. & Brown, J. K. (1999)
A phylogeographical analysis of the B emisia tabaci species complex based
on mitochondrial DNA markers. Molecular Ecology, 8, 1683 – 1691.
Fukatsu, T. (1999) Acetone preservation: a practical technique for molecular

analysis. Molecular Ecology, 8, 1935 – 1945.
Funk, D. J. (1998) Isolating a role for natural selection in speciation: host adaptation
and sexual isolation in Neochlamisus bebbianae leaf beetles. Evolution, 52,
1744 – 1759.
Giblin, D. R. M., Gries, R., Crespi, B. et al. (2000) Aggregation pheromones of two
geographical isolates of the New Guinea sugarcane weevil, Rhabdoscelus
obscurus. Journal of Chemical Ecology, 26, 2763 – 2780.
Guillet, S., Guiller, A., Deunff, J. & Vancassel, M. (2000b) Analysis of a contact
zone in the Forficula auricularia L. (Dermaptera: Forficulidae) species
complex in the Pyrenean Mountains. Heredity, 85, 444 – 449.
Guillet, S., Josselin, N. & Vancassel, M. (2000a) Multiple introductions of Forficula

auricularia L. species complex (Dermaptera: Forficulidae) in eastern North
America. Canadian Entomologist, 132, 49 – 57.

Hemingway, J., Karunaratne, S. H. P. P. & Claridge, M. F. (1999) Insecticide
resistance spectrum and underlying resistance mechanisms in tropical
populations of the brown planthopper (Nilaparvata lugens) collected from
rice and the wild grass Leersia hexandra. International Journal of Pest
Management, 45, 15 – 223.
Jamnongluk, W., Baimai, V. & Kittayapong, P. (2003) Molecular evolution of

tephritid fruit flies in the genus Bactrocera based on the cytochrome oxidase
I gene. Genetica, 119, 19 – 25.
Jermy, T. (1984) Evolution of insect/host plant relationships. American Naturalist,

124, 609 – 630.
Kenis, M. & Mills, N. J. (1998) Evidence for the occurrence of sibling species in
Eubazus spp. (Hymenoptera: Braconidae), parasitoids of Pissodes spp.
weevils (Coleoptera: Curculionidae). Bulletin of Entomological Research, 88,
149 – 163.
Kimura, T. (1980) A simple method for estimating evolutionary rates of base

substitutions through comparative studies of nucleotide sequences. Journal of
Molecular Evolution, 16, 111 – 120.
Krafsur, E. S. & Obrycki, J. J. (2000) Coleomegilla maculata (Coleoptera:

Coccinellidae) is a species complex. Annals of the Entomological Society of
America, 93, 1156 – 1163.
Kumar, S., Tamura, K., Jakobson, I. B. & Nei, M. (2001) MEGA2: molecular
evolutionary genetics analysis software. Bioinformatics, 17, 1244 – 1245.
Larsson, S. & Ohmart, C. P. (1988) Leaf age and larval performance of the leaf
beetle Paropsis atomaria. Ecological Entomology, 13, 19 – 24.
Lunt, D. H., Zhang, D.- X., Szymura, J. M. & Hewitt, G. M. (1996) The insect
cytochrome oxidase I gene: evolutionary patterns and conserved primers for
phylogenetic studies. Insect Molecular Biology, 5, 153 – 165.
Mahon, R. J., Miethke, P. M. & Mahon, J. A. (1982) The evolutionary relationships

of three forms of the Jarrah Leaf Miner, Perthida glyphopa (Common)
(Lepidoptera: Incurvariidae). Australian Journal of Zoology, 30, 243 – 249.

Mardulyn, P. & Mililkovitch, M. C. (2005) Inferring contemporary levels of gene
flow and demographic history in a local population of the leaf beetle
Gonioctena olivacea from mitochondrial DNA sequence variation. Molecular
Ecology, 14, 1641 – 1653.
Mattiucci, S., Cianchi, R., Nascetti, G. e t al. (2003) Genetic evidence for two sibling
species within Contracaecum ogmorhini Johnston & Mawson, 1941
(Nematoda: Anisakidae) from otariid seals of boreal and austral regions.
Systematic Parasitology, 54, 13 – 23.
Milne, M. & Walter, G. H. (1998) Host species and plant part specificity of the
polyphagous onion thrips, Thrips tabaci Lindeman (Thysanoptera:
Thripidae), in an Australian cotton-growing area. Australian Journal of
Entomology, 37, 115 – 119.
Milne, M. & Walter, G. H. (2000) Feeding and breeding across host plants within a
locality by the widespread thrips Frankliniella schultzei, and the invasive
potential of polyphagous insects. Diversity and Distributions, 6, 243 – 257.
Ohmart, C. P. & Larsson, S. (1989) Evidence for absorption of eucalypt essential oils
by Paropsis atomaria Olivier (Coleoptera: Chrysomelidae). Journal of the
Australian Entomological Society, 28, 201 – 205.
Ohmart, C. P., Stewart, L. G. & Thomas, J. R. (1985a) Effects of food quality,

particularly nitrogen concentrations, of Eucalyptus blakelyi foliage on the
growth of Paropsis atomaria larvae (Coleoptera: Chrysomelidae). O
ecologia, 6 5, 543 – 549.
Ohmart, C. P., Stewart, L. G. & Thomas, J. R. (1985b) Effects of nitrogen

concentrations of Eucalyptus blakelyi foliage on the fecundity of Paropsis
atomaria (Coleoptera: Chrysomelidae). Oecologia, 68, 41 – 44.
Ohmart, C. P., Thomas, J. R. & Stewart, L. G. (1987) Nitrogen, leaf toughness and
the population dynamics of Paropsis atomaria Olivier (Coleoptera:
Chrysomelidae) – a hypothesis. Journal of the Australian Entomological
Society, 26, 203 – 207.
Otranto, D., Colwell, D. D., Traversa, D. & Stevens, R. (2003) Species identification
of Hypoderma affecting domestic and wild ruminants by morphological and

molecular characterization. Medical and Veterinary Entomology, 17, 316 –
325.
Paterson, H. E. H. (1991) The recognition concept of cryptic species among

economically important insects. Research Methods and Prospects (ed. by M.
P. Zalucki), pp. 1 – 10. Springer-Verlag, New York, New York.
Peterson, M. A. & Denno, R. F. (1998) The influence of dispersal and diet breadth on
patterns of genetic isolation by distance in phytophagous insects. American
Naturalist, 152, 428 – 446.
Posada, D. & Crandall, K. A. (2001) Intraspecific gene geneologies: trees grafting

onto networks. Trends in Ecology and Evolution, 16, 37 – 44.
Salvato, P., Battisti, A., Concato, S., Masutti, L., Patarnello, T. & Zane, L. (2002)
Genetic differentiation in the winter pine processionary moth (Thaumetopoea
pityocampa – wilkinsoni complex), inferred by AFLP and mitochondrial
DNA markers. Molecular Ecology, 11, 2435 – 2444.
Schneider, S., Roessli, D. & Excoffier, L. (2000) ARLEQUIN Computer Program,

Version 2.000. Genetics and Biometry Laboratory, University of Geneva,
Switzerland.
Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H. & Flook, P. (1994)
Evolution, weighting, and phylogenetic utility of mitochondrial gene
sequences and a compilation of conserved polymerase chain reaction primers.
Annals of the Entomological Society of America, 87, 651 – 701.
Singer, M. S. (2001) Determinants of polyphagy by a wooly bear caterpillar: a test of

the physiological efficiency hypothesis. Oikos, 93, 194 – 204.
Steinbauer, M. J., Clarke, A. R. & Madden, J. L. (1998) Oviposition preference of a

Eucalyptus herbivore and the importance of leaf age on interspecific host
choice. Ecological Entomology, 23, 201 – 206.
Stone, C. (1993) Insect pest problems of eucalypt plantations in Australia: 2. New

South Wales. Australian Forestry, 56, 363 – 369.
Subbarao, S. K., Vasantha, K. & Sharma, V. P. (1988) Responses of Anopheles

culicifacies sibling species A and B to DDT and HCH in India: implications
in malaria control. Medical and Veterinary Entomology, 2, 219 – 223.

Tanton, M. T. & Epila, J. S. O. (1984a) Parasitization of larvae of Paropsis atomaria
Ol. (Coleoptera: Chrysomelidae) in the Australian Capital Territory.
Australian Journal of Zoology, 32, 251 – 259.
Tanton, M. T. & Epila, J. S. O. (1984b) Effects of DDT and Fenitrothion on field

collected larvae of a eucalypt-defoliating beetle, Paropsis atomaria Ol. I.
Mortality, relative toxicity, development of treated larvae, and effect of the
primary parasitoids. Australian Journal of Zoology, 32, 325 – 336.
Tanton, M. T. & Epila, J. S. O. (1984c) Effects of DDT and Fenitrothion on field

collected larvae of a eucalypt-defoliating beetle, Paropsis atomaria Ol. II.
Effects of treatment of the host host on the parasitoids. Australian Journal of
Zoology, 32, 337 – 344.
Tanton, M. T. & Epila, J. S. O. (1984d) Description of the planidium of Perilampus

tasmanicus Cameron (Hymenoptera: Chalcidoidea), a hyperparasitoid of
larvae of Paropsis atomaria Olivier (Coleoptera: Chrysomelidae). J ournal of
the Australian Entomological Society, 23, 149 – 152.
Tanton, M. T. & Khan, S. M. (1978a) Aspects of the biology of the eucalypt-

defoliating chrysomelid beetle Paropsis atomaria Ol. in the Australian
Capital Territory. Australian Journal of Zoology, 26, 113 – 120.
Tanton, M. T. & Khan, S. M. (1978b) Effects of Fenitrothion and Aminocarb, at

doses giving low mortailty, on surviving eggs and larvae of the eucalypt-
defoliating chrysomelid beetle Paropsis atomaria Ol. I. Methods, mortality
and relative toxicity. Australian Journal of Zoology, 26, 121 – 126.
Tanton, M. T. & Khan, S. M. (1978c) Effects of Fenitrothion and Aminocarb, at

defoliating chrysomelid beetle Paropsis atomaria Ol. II. Biology of
survivors. Australian Journal of Zoology, 26, 127 – 138.
Tanton, M. T. & Khan, S. M. (1978d) Effects of Fenitrothion and Aminocarb, at

defoliating chrysomelid beetle Paropsis atomaria Ol. III. Histological
changes in treated larvae. Australian Journal of Zoology, 26, 139 – 146.

Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994) CLUSTAL W: improving
the sensitivity of progressive multiple sequence alignment through sequence
weighting, position-specific gap penalties and weight matrix choice. Nucleic
Acids Research, 22, 4673 – 4680.
Umina, P. A. & Hoffman, A. A. (1999) Tolerance of cryptic species of blue oat mites
(Penthalus spp.) and the redlegged earth mite (Halotydeus destructor) to
pesticides. Australian Journal of Experimental Agriculture, 39, 621 – 628.
Uribe, S. S. I., Lehmann, T., Rowton, E. D., Velez, B. I. D. & Porter, C. H. (2001)
Speciation and population structure in the morphospecies Lutzomyia
longipalpis (LutZ. and Neiva) as derived from the mitochondrial ND4 gene.
Molecular Phylogenetics and Evolution, 18, 84 – 93.
Walsh, P. S., Metzger, D. A. & Higuchi, R. (1991) Chelex 100 as a medium for
simple extraction of DNA for PCR-based typing of forensic material.
Biotechniques, 10, 506 – 513.
Walter, G. H. (2003) Insect Pest Management and Ecological Research. Cambridge

University Press, U.K.
Walter, G. H. & Benfield, M. D. (1994) Temporal host plant use in three

polyphagous Heliothinae with special reference to Helicoverpa punctigera
(Wallengren) (Noctuidae: Lepidoptera). Australian Journal of Ecology, 19,
458 – 465.
Waterson, D. & Urquhart, C. A. (1995) Forest Protection, Research Division Series

Number E6: Leaf Beetles (Coleoptera: Chrysomelidae). State Forests of New
South Wales, Australia.
Whitlock, M. C. & McCauley, D. E. (1999) Indirect measures of gene flow and

migration: F ST is not equal to 1/(4 Nm + 1). Heredity, 82, 117 – 125.
Wirth, T., Le Guellec, R., Van Cassel, M. & Veuille, M. (1998) Molecular and
reproductive characterization of sibling species in the European earwig
(Forficula auricularia). Evolution, 52, 260 – 265.
Wylie, F. R. & Peters, B. C. (1993) Insect pest problems of eucalypt plantations in

Australia. 1. Queensland. Australian Forestry, 56, 358 – 362.

Yeh, W. B., Yang, C. T. & Kang, S. C. (1997) Identification of two sibling species,
Ephemera formosana and E. sauteri (Ephemeroptera: Ephemeridae), based
on mitochondrial DNA sequence analysis. Zhonghua Kunchong, 17, 257 –
268.

Chapter 4
Converse Bergmann cline in a Eucalyptus herbivore, Paropsis

atomaria Olivier (Coleoptera: Chrysomelidae): phenotypic plasticity
or local adaptation?
Schutze, M.K. and Clarke, A.R. Converse Bergmann cline in a Eucalyptus herbivore,
Paropsis atomaria Olivier (Coleoptera: Chrysomelidae): phenotypic plasticity or
local adaptation? Global Ecology and Biogeography. 17 (3), 424-431.
Schutze, M.K.
Clarke, A.R.
N.b. While a single measurement of body size (pronotum width) is reported in this
chapter, fifteen body size measurements were taken as part of the overall thesis. The
remaining fourteen measurements were not included in the final publication due to
referee insistence. Supplementary results (including all measurements taken) are
presented in Appendix 1 to this thesis.
Chapter 4: Converse Bergmann cline in Paropsis atomaria 65

Converse Bergmann cline in a Eucalyptus herbivore, Paropsis
atomaria Olivier (Coleoptera: Chrysomelidae): phenotypic plasticity
or local adaptation?
Abstract
Aim To measure latitude-related body size variation in field collected Paropsis
atomaria Olivier (Coleoptera: Chrysomelidae) individuals and to conduct common-
garden experiments to determine whether such variation is due to phenotypic
plasticity or local adaptation.
Location Four collection sites from the east coast of Australia were selected for
contemporary field collections: Canberra (latitude 35º19’S); Bangalow (latitude
28º43’S); Beerburrum (latitude 26º58’S); and Lowmead (latitude 24º29’S). Museum
specimens collected over the past 100 years and covering the same geographic area
as contemporary field collections came from one state, one national, and one private
collection.
Methods Body size (pronotum width) was measured for 118 field collected beetles
and 302 specimens from collections. We then reared larvae from the latitudinal
extremes (Canberra and Lowmead) to determine whether the size cline was the result
of phenotypic plasticity or evolved differences (= local adaptation) between sites.
Results Beetles decrease in size with increasing latitude, representing a converse

Bergmann cline. A decrease in developmental temperature produced larger adults for
both Lowmead (low latitude) and Canberra (high latitude) individuals, and those
from Lowmead were larger than those from Canberra when reared under identical
conditions.
Main conclusions The converse Bergmann cline in P. atomaria is likely the result
local adaptation to season length.
Keywords Leaf beetle, latitude, temperature, body size, converse Bergmann’s Rule,
season length
66 Chapter 4: Converse Bergmann cline in Paropsis atomaria

Introduction
Body size variation in animals is perhaps the single most important quantitative
character measure as it strongly influences most physiological and fitness traits
(Blanckenhorn & Demont, 2004). Given this importance, large-scale systematic
variation of body size over latitudinal gradients has been of interest to biologists for
over 150 years (Bergmann, 1847; Atkinson, 1994; Blackburn et al., 1999;
Blanckenhorn & Demont, 2004). Reports of such trends in nature are so pervasive in
the literature (e.g., see Blanckenhorn & Demont 2004 for a review of arthropod
examples) that they have resulted in the construction of numerous rules which
attempt to provide mechanistic explanations for the observed phenomenon of body
size variation with latitude.
The environmental variable most often associated with changing latitude is

temperature. Consequently, temperature has been at the core of most attempts to
relate variation in body size to geographical gradients, with Bergmann’s Rule the
most often invoked (Bergmann, 1847; Blackburn et al., 1999). Put simply,
Bergmann’s Rule states that individuals are larger at higher latitudes. Bergmann’s
original hypothesis was based on the observation of mammals and explained in terms
of heat-conservation; the larger the animal in a colder climate, the lower its surface
area-to-volume ratio, and hence the greater its capacity to retain heat (thus conferring
an advantage). This explanation has been largely dismissed for endotherms and is
even less likely for ectotherms, whose body temperatures fluctuate rapidly and are
highly dependant on ambient conditions (heat conservation, therefore, is not usually
possible) (McNab, 1971; Atkinson & Sibly, 1997). Furthermore, the reverse
scenario, converse Bergmann’s Rule – individuals are smaller at higher latitudes
(Park, 1949) – has been documented for many species and was equally represented
as conventional Bergmann clines in a recent study of 48 arthropod species
(Blanckenhorn & Demont, 2004).
For ectotherms, temperature directly influences body size through its effects on
metabolic rates and development time (Sibly & Atkinson, 1994; Van der Have & De
Jong, 1996): lower developmental temperatures typically result in larger individuals
(temperature-size rule; Carleton, 1960; Vannote & Sweeny, 1980; Lonsdale &
Levinton, 1985; Atkinson, 1994; Partridge et al., 1994; Van der Have & De Jong,
1996; Atkinson & Sibly, 1997; Ramsden & Elek, 1998; Chown & Gaston, 1999;

Reeve et al., 2000; but see Walters & Hassall, 2006 for a reverse trend in the
grasshopper, Chorthippus brunneus). Consequently, if a Bergmann cline is the result
of phenotypic plasticity in a species which follows the temperature-size rule, we may
expect larger individuals at higher latitudes (colder temperatures), thus conforming to
a conventional Bergmann’s cline. Importantly, however, this does not imply a
Bergmann cline in an ectotherm species is the inherent result of phenotypic
plasticity, as adaptive explanations – such as starvation resistance as documented in
the ant lion Myrmeleon immaculatus (Arnett & Gotelli, 2003) – may be equally
valid.
To determine whether a Bergmann cline or its converse is the product of adaptive

mechanisms or phenotypic plasticity, common-garden experiments are required. For
instance, should an ectotherm species adhere to the temperature-size rule during
developmental trials and also conform to a converse Bergmann cline in the wild, we
may justifiably conclude genetic differences between populations are driving
latitudinal body size variation (i.e., an adaptive mechanism) (Masaki, 1978;
Mousseau & Roff, 1989; Blanckenhorn & Fairbairn, 1995; Blanckenhorn & Demont,
2004). The next step is to determine the likely adaptive mechanism driving such a
cline. In many cases converse Bergmann clines in an ectotherm species are proposed
to be mediated by the interaction of temperature effects on growth, season length,
and average development time for the organism (Blanckenhorn & Demont, 2004),
with higher latitude seasons providing less time (and resources) for individuals to
mature, thereby producing smaller adults (Carleton, 1960; Roff, 1980; Fischer &
Fiedler, 2002; Blanckenhorn & Demont, 2004).
Paropsis atomaria Olivier (Coleoptera: Chrysomelidae) is a widely distributed

Australian endemic leaf beetle. Adults and larvae feed on the foliage of at least 20
species of Eucalyptus L’Her (Myrtaceae) (CABInternational, 2005) and the beetle is
an emergent pest of eucalypt plantations in Queensland (Qld) and New South Wales
(N.S.W.). Paropsis atomaria consists of at least two partially isolated populations
along the east coast of Australia, from the temperate south (latitude ~35º) to the sub-
tropical north (latitudes 28º – 24º) (Schutze et al., 2006). Beetles from the southern
areas of the distribution (Canberra, Australian Capital Territory (A.C.T.)) are active
from October until March / April (Carne, 1966), whereas northern populations living
in south east Queensland experience a longer field season and adults are active from

as early as September through to at least mid-April (Nahrung, 2006). The large
geographical range, together with genetic differentiation between populations along
that range, renders P. atomaria ideal for investigating a Bergmann cline and its
underlying mechanism.
In this study, we first describe a converse Bergmann cline in P. atomaria body size
across latitude using both recent and historical collections. To determine whether the
cline results from phenotypic plasticity in response to temperature or is the product
of local adaptation, we conducted common-garden experiments with wild caught
beetles from the two extremes of the latitudinal gradient. We conclude the converse
Bergmann cline seen for P. atomaria is likely the product of local adaptation to
season length.
Temperature data
To determine the degree of correlation between latitude and temperature, long-term
climatic data was sourced from the Australian Bureau of Meteorology. We calculated
the average daily temperature (ºC) between the months of October and April (P.
atomaria field season) based on the closest data record site for each collection
locality: Lowmead: Gladstone Radar (23°51’36”S, 151°13’36”E; averages based on
data from 1957 – 2004); Beerburrum: Caloundra signal station (26°48’00”S,
153°09’00”E; averages from 1899 – 1992); Bangalow: Lismore central street
(28°48’36”S, 153°17’24”E; averages from 1884 – 2003); and Canberra: Canberra
airport (35°17’60”S, 149°11’60”E; averages from 1939 – 2004).
Body size of field collections
Material
Collection sites were selected based on the following criteria: 1) they occurred across
a significant part of the species range, ensuring tropical and temperate locations were
included; and 2) sufficient numbers of individuals were present for analysis.
Consequently, the following four sites were chosen: tropical/sub-tropical Lowmead
(central Qld, 24º29’22”S, 151°42’14”E), Beerburrum (south-east Qld, 26º58’02”S,
153°03’06”E), Bangalow (north-east N.S.W., 28º43’11”S, 153°31’07”E) and
temperate Canberra (A.C.T., 35º18’51”S, 149°09’16”E) (Fig. 1) (n.b. for brevity, site

latitude information hereafter only given to nearest degree except for specific
collection localities). Paropsis atomaria was collected from trees within forestry
plantations at every site except Canberra (collected within the Jerrabomberra
wetlands).
Figure 1. Geographical locations where Paropsis atomaria individuals were

collected for this study.
Adult beetles were identified based on gross morphology (Waterson & Urquhart,
1995) and hand collected into 70% ethanol. Due to the gregarious and largely
sedentary nature of P. atomaria (Carne, 1966), every effort was taken to sample from
multiple sites within each sampling location in order to reduce the possibility of
collecting directly related individuals. Beetles from Beerburrum, Bangalow, and
Canberra were collected during March and April of 2003, whilst beetles from
Lowmead were collected in February of 2005. Our analysis also included research
collection specimens from the Agricultural Scientific Collections Unit (Orange
Agricultural Institute), the Australian National Insect Collection (A.N.I.C.), and one
private collection.

Character selection
Width of pronotum was selected as the relative measure of adult body size from an
original survey of 15 body parts, as it was straightforward to measure, present in all
specimens, and not subject to distortion (as is the case for total body length or width,
which is inaccurate due to distortion in the resting elytra). Measurements were made
by a single observer (MKS) for each individual from the four collection sites using a
calibrated eye-piece micrometer on a stereomicroscope to the nearest 0.1 mm.
All statistical procedures were run in SPSS v. 14.0 for WINDOWS.
We analysed the relationship between temperature and latitude using Pearson

correlation analysis.
For de novo collections, we treated latitude as a categorical variable and used a two-
way ANOVA to determine the effects of sex, latitude, and their interaction on body
size (pronotum width): size ~ sex + latitude + sex*latitude. For historical collection
material, we treated latitude as a continuous variable (collection sites varied
considerably) and used a Pearson correlation analysis to determine the direction and
strength of the relationship between latitude and body size (pronotum width).
Common-garden experiments
Study insects
We collected beetles from the two latitudinal extremes of the current study:
Canberra, A.C.T. (35º18’51”S, 149°09’16”E) and Lowmead, central Qld
(24º29’22”S, 151°42’14”E) during December 2005 and January 2006.
Approximately 50-100 beetles from each site were maintained on E. tereticornis

foliage in outdoor cultures in Brisbane Queensland for the duration of the trial. Eggs
from stock cultures were allowed to hatch and larvae permitted to consume the egg
chorion prior to rearing in controlled temperature cabinets at four temperatures: 16
ºC, 20 ºC, 24 ºC, and 27 ºC. Larvae were supplied daily with fresh E. pilularis leaves
taken from potted or plantation trees. On any one day all leaves supplied to larvae
came from one source, with individual shoots randomised before being placed in
rearing containers so as to minimise any potential for diet to bias treatments.

Twenty neonate individuals were placed in each petri dish with foliage and
moistened filter paper. Third and fourth instar larvae were transferred to larger
containers for the remainder of development. Pre-pupal larvae were removed from
rearing containers and placed in petri dishes until adult eclosion. The number of 20-
larvae replicates ranged between 10 – 13 for each location (Canberra and Lowmead)
and temperature. Not all replicates survived through to adult eclosion, especially
Lowmead individuals reared at 27 ºC. Total development time was the number of
days from egg eclosion until 50% of the surviving cohort emerged as adults. After 2
– 3 days adults were placed into 70% alcohol for preservation, from which pronotum
width was measured (as for field material).
The following statistical models were conducted for common garden experiments: a
two-way ANOVA testing development time (days) ~ rearing temperature + location +
rearing temperature*location, with Tukey post hoc tests for temperature for each
location, followed by pairwise ANOVA between locations for each temperature trial;
and a three-way ANOVA for body size (pronotum width) ~ sex + rearing temperature
+ latitude + interactions, with Tukey post hoc tests for temperature for each location,
followed by pairwise ANOVA for each temperature trial between each location for
both sexes.
Results
Temperature – latitude correlation

The average daily temperature for each of the four collection localities during the
October to April P. atomaria field season was as follows: Lowmead (latitude 24ºS):
25.70 ºC; Beerburrum (latitude 27ºS): 23.09 ºC; Bangalow (latitude 29ºS): 22.77 ºC;
and Canberra (latitude 35ºS): 17.48 ºC. Pearson correlation analysis of temperature
against latitude revealed a strong significant negative correlation (r = -0.992, P =
0.008) (Fig. 2).

27
Average temperature (ºC) Oct - Apr

25 r = -0.992
P = 0.008
23
21
19
17
15
24 26 28 30 32 34 36
Latitude (º)
Figure 2. Average daily temperatures for Paropsis atomaria field season plotted
against latitude for each of the four field collection sites: Lowmead, Beerburrum,
Bangalow and Canberra (left-right); Pearson correlation co-efficient r = -0.992, P <
0.05. See text for calculation of averages.
Body size of field collections

For de novo material, sex and latitude significantly affected body size, but their
interaction did not (Table 1). Average male pronotum width across all locations (N =
63) was significantly less than that of females (N = 55) ( = 5.5 ± 0.5 mm, = 6.0 ±
0.4 mm; ANOVA d.f. = 1, M.S. = 7.132, F = 36.823, P < 0.001). For females,
Canberra (latitude 35ºS) individuals were significantly smaller than those collected
from Bangalow (latitude 29ºS) and Beerburrum (latitude 27ºS), which in turn were
significantly smaller than Lowmead (latitude 24ºS) females. For males, Canberra,
Bangalow, and Beerburrum beetles were significantly smaller than Lowmead beetles
(Fig. 3).
Table 1. Two-way ANOVA of the effect of sex and location (and their interaction) on
Paropsis atomaria pronotum width for de novo collected field material.
Effect d.f. M.S. F P value
Sex 1 8.990 97.532 < 0.001
Location 3 3.927 42.602 < 0.001
Sex*Location 3 0.027 0.297 0.83
Error 110 0.092

12
6.50 a
Mean pronotum width +/- 1 SE (mm)

6.25 17
22 b 11
6.00
A b
5.75 15
12 c
14
5.50
B
B
5.25 15
B
5.00
Lowmead Beerburrum Bangalow Canberra

24ºS 27ºS 29ºS 35ºS
Collection site / latitude
Figure 3. Plot of pronotum width against latitude for de novo field collected Paropsis
atomaria (circles = females; squares = males). Latitude rounded to nearest degree.
Numbers above plots = number of individuals measured. Different letters denote
statistically significant difference (P < 0.05) in pronotum width between locations for
each sex (females lower case; males upper case). Points slightly offset for clarity.
One-hundred and forty-nine males and 153 females from historical collection
material were measured for pronotum width (latitudes ranged from 19°11'
60”S to
37°38'
60”S). Pronotum widths for males was again significantly less than that of
females ( = 5.6 ± 0.4 mm, = 6.1 ± 0.3 mm; ANOVA d.f. = 1, M.S. = 20.262, F =
173.015, P < 0.001). Pearson’s correlation analysis revealed a significant negative
relationship between pronotum width and latitude for males (r = -0.357, P < 0.001),
but a weaker, non-significant negative correlation for females (r = -0.110, P = 0.185)
(Figure 4).

Females
r = -0.110
P = 0.185
7.3
Pronotum width (mm)

6.8
6.3
5.8
5.3
4.8
18 23 28 33 38
Latitude (º)
Males
r = -0.357
P = 5.87 × 10-6
6.5
Pronotum width (mm)
6.0
5.5
5.0
4.5
19 24 29 34 39
Latitude (º)
Figure 4. Relationship between pronotum width and latitude for historically collected
(i.e., collected prior to this study) Paropsis atomaria females (circles) and males
(squares) demonstrating converse Bergmann cline. Gap between clusters due to lack
of specimens collected from those latitudes.
Common garden experiments
Development time
All individuals for each trial were analysed together regardless of sex (impossible to
sex larvae or rear individually due to gregarious feeding behaviour). There was a
significant effect of both location and temperature on development time of larvae,
but no interaction between the two (Table 2). Post hoc tests revealed a significant
difference in development time between all developmental temperatures for Canberra

individuals, and similarly so for Lowmead individuals except for between 24 ºC and
27 ºC, for which mean development time was not significantly different between
temperature treatments (Fig. 5). Developmental times between populations for each
temperature were the same except for the extremes (16 ºC and 27 ºC), in which
Canberra individuals developed faster than Lowmead individuals (significant for the
16 ºC trial, P = 0.028; and close to statistical significance for the 27 ºC trial, P =
0.059) (Fig. 5).
Table 2. Two-way ANOVA of the effect of location and temperature (and the
interaction) on total development time in days (egg – adult) for larvae reared at 4
temperatures (16 ºC, 20 ºC, 24 ºC, and 27 ºC).
Effect d.f M.S. F P value
Location 1 10.405 5.41 0.023
Temperature 3 4516.934 2348.699 < 0.001
Location*Temperature 3 3.544 1.843 0.149
Error 59 1.923

70
* 10
10
a
A
Mean development time +/- 1 SE (days)

60
50
10 8
40
B b
9 6
6
8
30 C c c
D
16 20 24 27
Developmental temperature (ºC)
Figure 5. Mean development time (days) for each population of Paropsis atomaria
(Canberra 35ºS = grey and Lowmead 24ºS = black) reared at four temperatures (16
ºC, 20 ºC, 24 ºC, and 27 ºC). Numbers above plots = number of replicates. Different
letters denote significant difference (Tukey post hoc comparisons, P < 0.05) in total
development time between each temperature (Canberra upper case and Lowmead
lower case). Asterisk denotes significant difference (pairwise ANOVA, P < 0.05) in
development time between populations reared at the same temperature. Points
slightly offset for clarity.
Body size
Sex, location, and temperature all significantly affected adult body size (none of the
interactions were significant) (Table 3). Females were, again, significantly larger
than males; northern beetles (Lowmead; 24ºS) were significantly larger than southern
beetles (Canberra; 35ºS) when reared at the same temperature for all comparisons
except between Lowmead and Canberra males reared at 24 ºC; and post hoc tests
revealed higher temperatures produced significantly smaller adult females, and
smaller (but not significant) males (Fig. 6).

Table 3. Three-way ANOVA results for the effect of sex, location, and temperature
(and interactions) on adult body size (pronotum width, mm) for larvae reared at 4
temperatures (16 ºC, 20 ºC, 24 ºC, and 27 ºC).
Effect d.f. M.S. F P value
Sex 1 9.714 196.185 < 0.001
Location 1 3.061 61.646 < 0.001
Temp 3 0.284 5.719 0.001
Sex*Location 1 0.092 1.848 0.176
Sex*Temperature 3 0.043 0.858 0.464
Location*Temperature 3 0.016 0.326 0.807
Sex*Location*Temperature 3 0.006 0.120 0.948
Error 212 0.050

6.00 15
Females
15
5.90 15
a ab
7
5.80
* * Lowmead (24ºS)
15 15 b
5.70 ab
*
5.60 15 *
A
AB
20
5.50
AB Canberra (35ºS)
5.40
B
5.30
16 20 24 27
5.60 Males
15
15 15
5.40
15
* a 5
a 15 Lowmead (24ºS)
* a
16 a
5.20 A
15
*
A
A
Canberra (35ºS)
5.00
A
4.80
16 20 24 27
Figure 6. Effects of larval rearing temperatures (16 ºC, 20 ºC, 24 ºC, and 27 ºC) on
adult body size (pronotum width, mm) for females (circles) and males (squares) of
Paropsis atomaria from two population origins (Canberra indicated in black and
Lowmead indicated in grey). Numbers above plots = number of individuals
measured. Different letters denote significant differences (Tukey post hoc
comparisons, P < 0.05) in average pronotum width between temperature trials for
each population (Canberra upper case and Lowmead lower case). Asterisks denote
significant difference (pairwise ANOVA, P < 0.05) in pronotum width between
populations reared at the same temperature. Points slightly offset for clarity.

Discussion
Paropsis atomaria conforms to both a converse Bergmann cline and the temperature-
size rule. Wild-caught adults demonstrate a clear trend of decreasing size with
increasing latitude (Figs 3 & 4), and common-garden experiments showed that
rearing larvae at higher temperatures generally resulted in smaller adults (Fig. 6).
Furthermore, northern beetles were consistently larger than southern beetles,
regardless of rearing temperature (Fig. 6). Therefore, we conclude the observed
converse Bergman cline is due to genetic differences between populations,
representing a case of local adaptation rather than phenotypic plasticity.
Adaptive explanations for Bergmann clines in ectotherms include starvation

resistance (Arnett & Gotelli, 2003), adaptive co-variation among life-history traits
(Angilletta et al., 2004), adaptive phenotypic plasticity (Partridge et al., 1994), and
voltinism mediated by season length (Roff, 1980; Blankenhorn & Demont, 2004).
The starvation resistance hypothesis, as seen for the ant-lion Myrmeleon
immaculatus, is an example whereby particular geographic regions within a
latitudinal gradient experience unpredictable seasonal conditions resulting in highly
variable food availability for some populations (Arnett & Gotelli, 2003).
Consequently, it becomes an adaptive advantage for populations in these
unpredictable environments (usually at higher latitudes) to have an increased body
size as a means to resist starvation during periods of food unavailability. Whilst valid
for the above example, we do not believe this mechanism applies to P. atomaria, as
we observe a converse Bergmann cline, counter to what may be expected under a
starvation resistance hypothesis should resource predictability relate similarly to such
latitudinal variation. Additionally, there is no reason to believe any particular region
lacks available resources compared to another, as Eucalyptus is the dominant genus
inhabiting Australian forests and constitutes at least 92 % of native forests and
woodlands (Morrow, 1976), with P. atomaria recorded from more than 20 species
(CABInternational, 2005).
Adaptive co-variation among life-history traits may produce a conventional

Bergmann cline in ectotherm species with populations experiencing large differences
in age at maturation. This is the case for Sceloporine lizards, in which larger
individuals living in colder climates (higher latitudes) may mature a full year later
than smaller body-sized animals exposed to warmer conditions (Angilletta et al.,

2004). Whilst P. atomaria is similar to lizards in that it is an ectotherm that adheres
to the temperature size rule, P. atomaria individuals reach maturity within a single
season (significantly less than a year). Therefore, season length across a latitudinal
gradient may play a more important role in determining rate to maturation and
consequent adult size in a rapidly developing insect such as P. atomaria. Indeed,
season length is considered important in determining the direction of the cline
(Bergmann or converse Bergmann) for arthropods in particular (Roff, 1980;
Mousseau, 1997; Blanckenhorn & Demont, 2004). Populations of Teleogryllus
cricket in Japan, for example, demonstrate a converse Bergmann cline in which
northern (lower latitude) individuals mature earlier (and are correspondingly smaller)
than their southern counterparts (Masaki, 1972) and this is regarded an adaptive
response to season length.
Several factors indicate season length is an important factor producing a converse

Bergmann cline in P. atomaria. Our results demonstrate development time and size
at maturation is genetically controlled, with development time for southern (high
latitude) populations generally shorter across the four developmental temperatures
studied (Fig. 5), with an associated decrease in adult body size compared to northern,
low latitude beetles (Fig. 6). We propose that reduced development time is an
adaptive response to shortened season length in southern temperate regions
compared to sub-tropical regions in the north, where increased season length permits
a longer growth period resulting in increased average adult body sizes (= increased
potential fecundity). Such local adaptation is considered possible due to restricted
gene flow between temperate and sub-tropical populations of P. atomaria, with high
pairwise FST comparisons between the southern population (Canberra) and all
northern collection localities (Bangalow, Beerburrum, and Lowmead) (Schutze et al.,
2006).
Furthermore, other factors, especially host plant quality, may influence final adult
body size in P. atomaria within and between populations across the latitudinal
gradient. Indeed, P. atomaria develops at variable rates depending on which host it is
reared, resulting in correspondingly variable adult body sizes (Carne, 1966; Schutze
in prep.). We believe host plant response contributes to the weaker correlation of
body size for historical collection material with latitude compared to our analysis of
de novo material. Measurements for historical material were taken from specimens

collected over 100 years, which had consequently developed under a wide range of
biotic and abiotic environmental conditions. This individual-to-individual variation
in turn will have masked, but not hidden, the converse cline effect. We recommend
further trials comparing host-plant response between populations be conducted to
determine if locally adapted populations have also acquired variable developmental
responses to different species of Eucalyptus.
Acknowledgements
We thank Stephen Monteith, Helen Nahrung, Alexsis Wilson, Martin Henery and
Angela Duffy for assistance with field collections. We also thank Simon Lawson of
QDPI&F (Forestry), Richard Lunney of Integrated Tree Crops (ITC) and NSW State
Forests for access to plantation sites. We also thank the curators of the Orange
Agricultural Institute, the Australian National Insect Collection and Gunter Maywald
for the generous loaning of material used in this study.

References
Angilletta, M.J.Jr., Niewiarowski, P.H., Dunham, A.E., Leaché, A.D. & Porter, W.P.
(2004) Bergmann’s clines in ectotherms: illustrating a life-history perspective
with Sceloporine lizards. The American Naturalist, 164, E168-E183.
Arnett, A.E. & Gotelli, N.J. (2003) Bergmann'
s rule in larval ant lions: testing the
starvation resistance hypothesis. Ecological Entomology, 28, 645-650.
Atkinson, D. (1994) Temperature and organism size - a biological law for
ectotherms. Advances in Ecological Research, 25, 1-58.
Atkinson, D. & Sibly, R.M. (1997) Why are organisms usually bigger in colder
environments? Making sense of a life history puzzle. Trends in Ecology and
Evolution, 12, 235-239.
Bergmann, C. (1847) Über die Verhältnisse der Wärmeökonomie der Thiere zu ihrer
Grösse. Gottinger studien, 3, 595-708.
Blackburn, T.M., Gaston, K.J. & Loder, N. (1999) Geographic gradients in body
s rule. Diversity & Distributions, 5, 165-
size: a clarification of Bergmann'
174.
Blanckenhorn, W.U. & Demont, M. (2004) Bergmann and converse Bergmann
latitudinal clines in arthropods: Two ends of a continuum? Integrative and
Comparative Biology, 44, 413-424.
Blanckenhorn, W.U. & Fairbairn, D.J. (1995) Life history adaptation along a
latitudinal cline in the water strider Aquarius remigis (Heteroptera: Gerridae).
Journal of Evolutionary Biology, 8, 21-41.
CABInternational (2005) Forestry Compendium. CAB International, Wallingford
UK.
Carleton, R. (1960) The application of Bergmann'
s and Allen'
s rules to the
poikilotherms. Journal of Morphology, 106, 85-108.
Carne, P.B. (1966) Ecological characteristics of the eucalypt defoliating chrysomelid
Paropsis atomaria Ol. Australian Journal of Zoology, 14, 647-672.
Chown, S.L. & Gaston, K.J. (1999) Exploring links between physiology and ecology
at macro-scales: the role of respiratory metabolism in insects. Biological
Reviews of the Cambridge Philosophical Society, 74, 87-120.

Fischer, K. & Fiedler, K. (2002) Reaction norms for age and size at maturity in
response to temperature: a test of the compound interest hypothesis.
Evolutionary Ecology, 16, 333-349.
Lonsdale, D.J. & Levinton, J.S. (1985) Latitudinal differentiation in copepod growth:
An adaptation to temperature. Ecology, 66, 1397-1407.
Masaki, S. (1972) Climatic adaptation and photoperiodic response in the band-legged
ground cricket. Evolution, 26, 587-600.
Masaki, S. (1978) Climatic adaptation and species status in the lawn ground cricket.
Oecologia, 35, 343-356.
McNab, B.K. (1971) On the ecological significance of Bergmann'
s Rule. Ecology,
52, 845-854.
Morrow, P.A. (1976) The significance of phytophagous insects in the Eucalyptus
forests of Australia. The role of Arthopods in Forest Ecosystems (ed. by W.J.
Mattson), pp. 19-29. Springer-Verlag, New York.
Mousseau, T.A. & Roff, D.A. (1989) Adaptation to seasonality in a cricket: patterns
of phenotypic and genotypic variation in body size and diapause length along
a cline in season length. Evolution, 43, 1483-1496.
Mousseau, T.A. (1997) Ectotherms follow the converse to Bergmann'
s Rule.
Evolution, 51, 630-632.
Nahrung, H.F. (2006) Paropsine beetles (Coleoptera: Chrysomelidae) in south-
eastern Queensland hardwood plantations: identifying potential pest species.
Australian Forestry, 69, 270-274.
Park, O. (1949) Application of the converse Bergmann principle to the carabid
beetle, Dicaelus purpuratus. Physiological Zoology, 22, 359-372.
Partridge, L., Barrie, B., Fowler, K. & French, V. (1994) Evolution and development
of body size and cell size in Drosophila melanogaster in response to
temperature. Evolution, 48, 1269-1276.
Ramsden, N. & Elek, J.A. (1998) Life cycle and development rates of the leaf beetle
Chrysophtharta agricola (Chapuis) (Coleoptera: Chrysomelidae) on
Eucalyptus nitens at two temperature regimens. Australian Journal of
Entomology, 37, 238-242.

Reeve, M.W., Fowler, K. & Partridge, L. (2000) Increased body size confers greater
fitness at lower experimental temperature in male Drosophila melanogaster.
Journal of Evolutionary Biology, 13, 836-844.
Roff, D.A. (1980) Optimizing development time in a seasonal environment: The ups
and downs of clinal variation. Oecologia, 45, 202-208.
Schutze, M.K., Mather, P.B. & Clarke, A.R. (2006) Species status and population
structure of the Australian Eucalyptus pest Paropsis atomaria Olivier
(Coleoptera: Chrysomelidae). Agricultural and Forest Entomology, 8, 323-
332.
Sibly, R.M. & Atkinson, D. (1994) How rearing temperature affects optimal adult
size in ectotherms. Functional Ecology, 8, 486-493.
Van der Have, T.M. & De Jong, G. (1996) Adult size in ectotherms: temperature
effects on growth and differentiation. Journal of Theoretical Biology, 183,
329-340.
Vannote, R.L. & Sweeny, B.W. (1980) Geographic analysis of thermal equilibria: a
conceptual model for evaluating the effect of natural and modified thermal
regimes on aquatic insect communities. The American Naturalist, 115, 667-
695.
Walters, R.J. & Hassall, M. (2006) The temperature-size rule in ectotherms: may a
general explanation exist after all? The American Naturalist, 167, 510-523.
Waterson, D. & Urquhart, C.A. (1995) "Forest Protection, Research Division Series
Number E6: Leaf Beetles (Coleoptera: Chrysomelidae)." State Forests of
New South Wales, Beecroft.

Chapter 5
Larval development of two geographically isolated populations of

Paropsis atomaria on two species of Eucalyptus
This chapter is in prep for submission as:
Schutze, M.K. and Clarke, A.R. Larval development of two geographically isolated
populations of Paropsis atomaria on two species of Eucalyptus. Australian Journal
of Entomology.
Schutze, M.K.
Clarke, A.R.
86 Chapter 5: Larval development of Paropsis atomaria on two hosts

Larval development of two geographically isolated populations of
Paropsis atomaria on two species of Eucalyptus
Abstract
Paropsis atomaria Olivier (Coleoptera: Chrysomelidae) is an endemic, Eucalyptus
L’Her. (Myrtaceae) feeding beetle with a distribution that extends along nearly the
entire east coast of Australia. Beetle populations from across the range display low
gene flow between northern and southern populations, together with variation in
adult body size. This study investigated potential for differential host plant utilisation
by geographically isolated populations of P. atomaria by measuring southern
(Canberra, Australian Capital Territory) and northern (Lowmead, central
Queensland) larval developmental response to rearing on two host species:
Eucalyptus cloeziana (that have a distribution sympatric with the northern beetle
population), and E. pilularis (sympatric with the southern population).
Developmental characteristics studied included the following: larval survival, larval
development time, and pupal weight. Eucalyptus cloeziana (northern host) was
superior compared with E. pilularis, showing increased survival rates for both
northern and southern P. atomaria individuals, together with reduced developmental
times. Southern P. atomaria pupal weights were generally lower than northern pupae
regardless of the host plant larvae were reared on. Northern P. atomaria individuals
took significantly longer to mature than did southern beetles and were significantly
heavier when reared on E. cloeziana. No such difference existed for populations
reared on E. pilularis. Pupal weight was significantly reduced for northern P.
atomaria individuals reared on E. pilularis (southern host). Our results demonstrate
similar host plant utilisation for northern and southern populations of P. atomaria,
with both populations performing better on E. cloeziana. The only difference
between populations was pupal mass attained, which may be due to other factors
such as underlying genetic differences between populations resulting from adaptation
to other factors.
Keywords: Chrysomelidae, Coleoptera, Myrtaceae, host specialisation, larval

performance, leaf beetle, local adaptation
Chapter 5: Larval development of Paropsis atomaria on two hosts 87

Introduction
Phytophagous insect species that possess a broad host range and that occur over large
geographical areas may consist of multiple localised populations with specialist host
affiliations (Jaenike, 1990). Each population, if sufficiently isolated from other
populations and subjected to local selective pressures is predicted, over time, to
develop adaptations particular to local environmental conditions (Fujiyama et al.,
2005). If a tight association exists between populations and their respective host
plants, this may lead to the evolution of host races and potential for new species
(Ballabeni et al., 2003). Further to this, adaptation to one suite of host plants may
coincide with a decrease in the suitability of other host plants not found within the
range of a localised population, as explained by trade-off theory (Dethier, 1954;
Mackenzie, 1996). There are, however, many examples where such a straightforward
trade-off does not exist, as some herbivore populations exhibit increased fitness on
novel compared with natal hosts (Ballabeni et al., 2003; Joshi and Thompson, 1995).
Additionally, differential host plant use by herbivore populations may not necessarily
result from adaptation to a particular host species, but may reflect pleiotropic effects
of genes involved in host plant utilisation which are influenced by a wide range of
non-host plant related selective pressures (Fujiyama et al., 2005).
For an individual to become specialised on a host plant – or a suite of hosts –

requires host-related behaviours that increase an individual’s realised fecundity and
enhance survival and fitness of offspring (Jaenike, 1990). Detection of host
specialisation can be a complex biological process, that is further complicated
conceptually, by continued refinement and debate over definitions of host races and
intermittent stages of host specialisation (see Dres and Mallet (2002) for a review).
Generally, however, current theory requires populations of a species meet a number
of criteria before they can be referred to as host races. These requirements include:
populations are genetically differentiated, sympatric, use different hosts, and there is
appreciable (but not panmictic) gene flow among them (Dres and Mallet, 2002).
Additionally, while not considered mandatory to host race identification, improved
fitness on natal hosts over alternative hosts is considered supporting evidence for
existence of host races (Dres and Mallet, 2002). Measures of fitness often focus on
those related to adult behaviour (oviposition and host food preference) and larval
development (growth, development rate, survival, and potential fecundity) (Gratton

and Welter, 1998; Harris et al., 2001; Moon and Stiling, 2006; Steinbauer et al.,
1998; Verdon et al., 2007), with increased fitness on one host relative to another host
being indicative of some degree of host specialisation.
Paropsine leaf beetles (Coleoptera: Chrysomelidae) are common in Australia, with

both adult and larval beetles feeding externally on the leaves of Eucalyptus L’Her
(Myrtaceae) hosts (Simmul & de Little 1999). Several species are recognised pests of
Eucalyptus plantations, where feeding causes production losses (Baker et al., 2002;
Clarke et al., 1997; Nahrung and Allen, 2003; Schutze et al., 2006). One such
species, Paropsis atomaria Olivier, is an emergent pest of plantations in New South
Wales and Queensland and is recorded from at least 20 species of Eucalyptus
(CABInternational, 2005). As noted for other paropsine species (Baker et al., 2002;
Patterson et al., 1996), different Eucalyptus species vary in their suitability as food
plants for P. atomaria, with different host species inducing variable developmental
responses relating to larval survival, larval development time, and pupal weight
(Carne, 1966a, b).
The geographical distribution of P. atomaria is extensive and ranges from the

temperate south (South Australia, Victoria, and the Australian Capital Territory
(A.C.T.)) to the tropics of Queensland, with molecular evidence demonstrating
reduced gene flow between temperate and tropical populations (Schutze et al., 2006).
Furthermore, the distribution of Australia’s eucalypt flora is highly fragmented, with
many species possessing a limited geographical range (Brooker and Kleinig, 1983,
1994), and this potentially exposes local populations of P. atomaria to a restricted
suite of Eucalyptus species. Whilst Carne (1966a) had studied host utilisation by P.
atomaria of different Eucalyptus species within the Canberra region, no study has
examined potential differential host utilisation between P. atomaria individuals
collected from other regions across its wide distribution. Consequently, there remains
a lack of information for this species regarding the potential of either differential host
utilisation through pleiotropic effect on host-related genes, or the presence of host
races adapted to sympatric eucalypt species.
Population structuring in P. atomaria is evident across its range, with geographically

isolated populations having revealed differential haplotype composition (Schutze et
al., 2006). The same study, however, found no evidence for strict haplotype / host-
plant association, therefore indicating an absence of host races. While this conclusion

remains sound based on genetic evidence, the COI gene used by Schutze et al.
(2006) is a potentially unlinked neutral marker and may have failed to reveal local
host adaptation should it exist. It therefore remains possible that even though genetic
data revealed no evidence of host races, local adaptation to host plants may yet have
occurred in P. atomaria populations.
Furthermore, adult body size of P. atomaria over a latitudinal gradient had

previously been found to conform to a converse Bergman cline (i.e., smaller beetles
at higher latitudes), and this cline was found to be under genetic control and the
likely result of local adaptation to season length rather than phenotypic plasticity
(Schutze & Clarke, 2008). This finding supports the analysis of Schutze et al. (2006),
in that genetic variation between northern and southern populations of P. atomaria
exists, but it goes further by demonstrating how this variation is not simply neutral,
but does have the potential to affect physiological traits within the species. Such
variation may extend to differential host plant use as outlined above.
We have therefore chosen to continue our earlier studies of P. atomaria to examine

the influence of different eucalypt hosts on larval development of two populations
(northern and southern) of P. atomaria. We aim to test for potential differential larval
development between populations and to examine the possibility of either (i) local
adaptation to host plant or (ii) pleiotropic effects altering developmental
characteristics as influenced by adaptation to non-host related selective pressures. If
physiological response to host – as observed through larval growth characteristics –
remains unchanged between beetle populations, then further support would be lent to
an absence of host races (and consequently no local adaptation to host species or
host-associated pleiotropic effect). However, differential larval development on host
species between populations would suggest either the presence of incipient host races
or a pleiotropic effect on host plant utilisation.

To test differential host-plant use between populations of P. atomaria, we reared two
populations of beetles identified by Schutze et al. (2006) as having low levels of
gene flow between them (northern population, Lowmead, central Queensland; and
southern population, Canberra, A.C.T.), on two species of host eucalypt chosen for
their distributional characteristics (E. cloeziana, sympatric with northern P. atomaria

population; and E. pilularis, sympatric with the southern P. atomaria population).
Measures of larval fitness taken included: survival to pupation, total development
time, and pupal weight.
Field collection of adult beetles

We collected adult P. atomaria beetles from two source populations: a mixed
Eucalyptus species forest at Jerrabomberra wetlands in Canberra (35º18’51”S,
149º09’16”E), Australian Capital Territory (A.C.T.) (southern beetles), and from a
forestry plantation site (E. grandis X E. camaldulensis hybrids) at Lowmead
(24º29’22”S, 151º42’14”E), central Queensland (northern beetles). These locations
were selected as they contained abundant numbers of beetles and it was from each of
these sites that previous studies revealed disparate population genetic structure
(Schutze et al., 2006) and adult body size (Schutze & Clarke, 2008). A minimum of
50 adults were collected from each locality and maintained at ambient temperature
on E. tereticornis foliage in outdoor cultures in Brisbane, south-east Queensland
(27º27’S, 152º58’E) for the duration of the trial.
Host plants
Eucalyptus cloeziana and E. pilularis were the study host plants chosen to measure
larval development. Each host plant possesses a different natural distribution (Fig 1):
E. pilularis is a southern eucalypt species sympatric with the southern (Canberra)
population of P. atomaria, whilst E. cloeziana has a restricted northern distribution
sympatric with the northern (Lowmead) population of P. atomaria.
Eucalyptus cloeziana and E. pilularis were obtained from the Queensland

Department of Primary Industries & Fisheries and were established as saplings
maintained in outdoor pots. Supplemental foliage was required occasionally for both
species and was collected from a forestry plantation located at Beerburrum, south-
east Queensland (26º58’02”S, 153º03’06”E). Supplemental foliage was fed to all
replicates equally to avoid bias.

Figure 1. Map showing location of two source populations of Paropsis atomaria
used in the current study: Lowmead and Canberra; and distributions of Eucalyptus
cloeziana and E. pilularis (data sourced from Australia’s Virtual Herbarium:
http://www.anbg.gov.au/avh/).
Experiments
Three aspects of larval development were measured for each population on each host
plant: survival to pupation, development time (days), and average pupal weight
(grams).
Egg batches were removed daily from field cages and maintained at 24 ºC prior to
emergence. Neonate larvae were permitted to consume their egg chorion, after which
they were segregated into groups of 20 individuals and placed in petri dishes with a
supply of newly emerged flush foliage (E. cloeziana or E. pilularis). Larvae of P.
atomaria are obligate group feeders in their early instars (Carne 1966a) and must
establish on young foliage (Larsson & Ohmart 1988), which dictated the
experimental design. Larvae from the same egg batch were separated to control for
potential maternal effects. Ten 20-larvae replicates for both Lowmead and Canberra
were established for E. cloeziana and E. pilularis trials. All trials were maintained in
constant temperature cabinets at 24 ºC with a 16D: 8L regimen. Foliage was replaced
and petri dishes were cleaned of frass daily, taking care not to disturb feeding larvae.

Survival was recorded as the number of individuals surviving to pupation for each
trial. To test the possibility that host plant produced a differential survival rate
between males and females for each source population, a chi-squared test on total
numbers reaching pupation was conducted for both sexes on each host. As no host
plant effect was apparent (see Results), males and females were grouped with
proportion surviving arcsin square root transformed prior to two-way ANOVA to
determine effect of host plant, population, and interaction between the two on
survival.
Development time was total number of days from egg eclosion to 50% adult
emergence. Results were analysed using two-way ANOVA to determine effects of
source population, host plant, and their interaction on development time (days).
Once all individuals had pupated, pupae were sexed based on characters detailed in
Reid and Ohmart (1989) and weighed (grams). Males and females are significantly
different in weight (data not shown) and were analysed separately with two-way
ANOVA to determine the effect on pupal weight by source population, host plant, and
their interaction. Where significant interaction effects of host and location were
detected by the two-way ANOVA, a sliced ANOVA was used to split the analysis to
examine the effects of each factor at individual levels of the other.
Results
Survival to pupation
There was no difference in the ratio of males and females surviving to pupation
2
between host plants for either population (Canberra = 0.165, d.f. = 1, P = 0.685;
2
Lowmead = 0.019, d.f. = 1, P = 0.890), hence males and females were combined
for each ANOVA.
Two-way ANOVA revealed a significant host plant effect on survival to pupation (F1,
36 = 13.826, P = 0.001), but no effect of source population (F1, 36 = 2.254, P = 0.142)
or the interaction between host plant and source population (F1, 36 = 1.764, P =
0.193). Survival to pupation for Canberra larvae was significantly higher on E.
cloeziana (59 %) compared with E. pilularis (21.5 %) (F1, 18 = 26.67, P < 0.001), and
whilst Lowmead survival was similarly reduced on E. pilularis, it was not

statistically significant (survival to pupation on E. cloeziana = 38.5 %; E. pilularis =
25 %; F1, 18 = 1.88., P = 0.188).
Development time
Two-way ANOVA revealed a significant host plant effect on development time (F1, 30
= 16.693, P < 0.001), but development time was not affected by origin of population
(F1, 30 = 1.832, P = 0.186) or the interaction between host plant and population origin
(F1, 30 = 1.832, P = 0.186). Canberra larvae reared on E. cloeziana developed
significantly faster than those reared on E. pilularis (F1, 17 = 17.725, P = 0.001),
whereas there was no significant difference in development time between host plants
for Lowmead individuals (F1, 13 = 3.146, P = 0.100) (Fig. 2).
33.5
6
Mean development time (days) +/- 1 SE
33.0 9
32.5
B a
32.0
9
31.5
a
31.0
30.5 10
30.0 A
29.5
Eucalyptus cloeziana Eucalyptus pilularis

Host plant
Figure 2. Mean ± 1 S.E. development time (days) of Paropsis atomaria from egg
eclosion to adult emergence for two populations (Canberra: circle and Lowmead:
square) reared on two host plants: Eucalyptus cloeziana and E. pilularis. Different
letters denote significant difference (P < 0.05) for between host-plant comparisons
calculated within each beetle population. Numbers above plots equal number of
replicates.

Pupal weight
There was a significant effect of location, host plant, and their interaction on pupal
weight for both males and females (Table 1) and hence both sexes were analysed
separately in subsequent analyses.
Lowmead males and females reared on E. cloeziana were significantly larger than
Lowmead males and females reared on E. pilularis ( F1, 65 = 8.515, P = 0.005; F1,
58 = 19.075, P < 0.001), whereas Canberra male and female pupal weights did not
differ between host plants ( F1, 92 = 0.012, P = 0.912; F1, 65 = 0.110, P = 0.741)
(Fig. 3).
As there was a significant interaction effect of host and location on pupal weight for
both males and females (see Table 1), sliced ANOVA revealed a significant difference
in pupal weight between Canberra and Lowmead individuals when reared on E.
cloeziana ( F1, 157 = 27.07, P < 0.001; F1, 123 = 22.30, P < 0.001), but not so for
those reared on E. pilularis (Fig. 3).
Table 1. Two-way ANOVA of the effects of location, host plant (and interaction) on
pupal weights (g) attained by female and male Paropsis atomaria from two source
populations (Lowmead and Canberra) reared on Eucalyptus cloeziana and E.
pilularis.
Effect d.f. MS F P value
Location 1 0.005 15.096 < 0.001
Host plant 1 0.002 4.869 0.029
Location*Host plant 1 0.001 4.221 0.042
Error 157 < 0.001
Location 1 0.001 6.819 0.010
Host plant 1 0.001 5.621 0.019
Location*Host plant 1 0.001 8.395 0.004
Error 123 < 0.001

Females
41
0.160
a
0.155
Mean pupal weight (g) +/- 1 SE
0.150
26
*
0.145
b
24
70
0.140
A A
0.135
0.130

Host plant
Males
0.120 36
a
Mean pupal weight (g) +/- 1 SE
0.115
0.110 *
19
24
48
0.105 A
b
A
0.100

Host plant
Figure 3. Mean ± 1 S.E. pupal weights (grams) of Paropsis atomaria females and
males, for two populations (Canberra: circles, and Lowmead: squares) reared on two
host plants: Eucalyptus cloeziana and E. pilularis. Asterisk denotes significant
difference (P < 0.05) in pupal weight between beetle populations reared on the same
host plant. Different letters denote significant difference (P < 0.05) in pupal weights
within a beetle population between host plants. Numbers above plots represent
number of individuals measured.

Discussion
At least two genetically differentiated populations of P. atomaria are present along
the east coast of Australia (Schutze et al., 2006), with individuals from northern
populations (northern NSW to central Qld) inherently larger in body size compared
with those from the south (Canberra, ACT) (Schutze & Clarke, 2008). The current
study addressed whether there is also differential host plant utilisation between
individuals from northern (Lowmead) and southern (Canberra) populations of P.
atomaria. Further, considering that P. atomaria is a largely sedentary species (Carne,
1966a), combined with the fragmented distributions of Australia’s eucalypt flora
(Brooker and Kleinig, 1983, 1994), we tested the hypothesis that respective northern
and southern populations may have adapted to local, sympatric eucalypt species.
Paropsis atomaria is a polyphagous species that feeds on at least 20 recognised hosts

within the genus Eucalyptus (CABInternational, 2005). Eucalyptus is the dominant
plant genus in Australia and constitutes 92 % of native forests and woodlands
(Morrow, 1976), with most environments containing mixed-species stands belonging
to multiple eucalypt subgenera (Noble, 1989). Our trial included only two host plant
species that represent a small proportion of the many potential hosts to which wild
populations of beetles are exposed. Therefore, it may be argued a priori, that beetles
from each population would be unlikely to experience strong selective pressures to
become specially adapted to the two hosts we used as many other potential host
species are present in the field, and in some cases at greater densities. This scenario
contrasts with those systems with tight associations between herbivore populations
and locally available host plant species. For example, different populations of the
chrysomelid Oreina elongata Suffrian are naturally exposed to specific hosts and, as
a consequence, each population has evolved specialisation relative to its available
host (Ballabeni et al., 2003). As populations of P. atomaria do not have such strict
associations with local host species, it is less likely to evolve host races – a
hypothesis supported by previous genetic studies (Schutze et al., 2006).
Nevertheless, we have demonstrated similar host plant utilisation between the two
study populations. While the northern host, E. cloeziana, produced higher survival
rates and shorter development time for larvae from both populations, especially those
from Canberra, a significant difference in pupal weight was found between
populations. Whilst there was no difference in average pupal weight of southern

individuals reared on either E. cloeziana or E. pilularis, there was a significant
reduction in pupal weight for northern Lowmead individuals reared on the allopatric
host, E. pilularis (relative to its sympatric host, E. cloeziana). As increased pupal
weight confers increased potential fecundity (Carne, 1966a), the significant reduction
in pupal weight for Lowmead individuals reared on E. pilularis may therefore imply
reduced fitness of P. atomaria on this host, and consequently a diminished suitability
of E. pilularis compared with E. cloeziana.
Alternatively, influences other than direct host plant adaptation may produce the
differential host use observed. Differential host use can result from pleiotropic
effects, such as adaptation to other environmental influences that affect genes
associated with host use. Furthermore, host specialisation on eucalypt species not
included in our study may also occur. As previously mentioned, adult body size in P.
atomaria varies over a latitudinal gradient with northern adults inherently larger than
southern adults (Schutze & Clarke, in press). Consequently, pleiotropic effects
influencing host plant use may flow on from local adaptation to environmental
factors affecting adult body size (such as season length). Due to local adaptation to
season length, Canberra individuals may be genetically constrained to reach a
maximum body size which can not be exceeded even when presented with a superior
host species, whereas inherently larger northern beetles are capable of greater body
sizes given such higher quality resources. Examination of developmental response
across a wider range of Eucalyptus species may reveal tighter host specialisation
occurring on other species; however we believe this unlikely due to the large number
of Eucalyptus species used by P. atomaria resulting in a low likelihood of host
specialistion (as outlined above).
While we have demonstrated some differential host plant utilisation by

geographically isolated populations, we see little evidence for host adapted local
populations. Most of the traits measured demonstrated consistent improved
performance on E. cloeziana over E. pilularis, a result consistent with a fixed,
species-wide preference for E. cloeziana not influenced by local adaptation. We
recommend extending trials to include aspects of adult behaviour, especially
oviposition and adult feeding preferences. Preliminary unpublished observations by
the authors indicate, for example, that whilst E. cloeziana appears the better host for
larval development compared with E. pilularis, it was the least favoured host by

females for oviposition by either population. An inverse relationship between
oviposition preference and larval development has been shown previously in P.
atomaria, albeit for beetles sourced from a single population (Carne, 1966a).
Inclusion of further data will contribute to elucidating the mechanisms that explain
disparity between populations and can aid in determining whether P. atomaria
consists of populations adapted to specific host eucalypt species.
Acknowledgements
We thank those who assisted with collection of field material, namely Martin
Henery, Katarina Mikac, and Angela Duffy. We thank ITC Plantations for access to
eucalypt plantations, Queensland DPI&F for access to eucalypt plantations, saplings,
and facilities, and Alexsis Wilson and Helen Nahrung for assistance with lab-work
and intellectual input on this project. We acknowledge QUT for financial support to
M.K.S. through the QUTPRA.

References
Baker, S. C., Elek, J. A., and Candy, S. G. (2002). Comparison of feeding efficiency,
development time and survival of Tasmanian eucalyptus leaf beetle larvae
Chrysophtharta bimaculata (Olivier) (Coleoptera: Chrysomelidae) on two
hosts. Australian Journal of Entomology 41, 174-181.
and ecological genetics of host-plant specialization in a leaf beetle. Oikos
101, 70-78.
Brooker, M. I. H., and Kleinig, D. (1983). "Field guide to Eucalypts: South-eastern
Australia," Inkata Press, Sydney.
Brooker, M. I. H., and Kleinig, D. (1994). "Field Guide to Eucalypts: Northern
Australia," Inkata Press, Sydney.
CABInternational (2005). Forestry Compendium. CAB International, Wallingford
UK.
Carne, P. B. (1966a). Ecological characteristics of the eucalypt defoliating
672.
Carne, P. B. (1966b). Growth and food consumption during the larval stages of
Paropsis atomaria (Coleoptera: Chrysomelidae). Entomologia
Experimentalis et Applicata 9, 105-112.
Clarke, A. R., Zalucki, M. P., Madden, J. L., Patel, V. S., and Paterson, S. C. (1997).
Local dispersion of the Eucalyptus leaf-beetle Chrysophtharta bimaculata
(Coleoptera: Chrysomelidae), and implications for forest protection. Journal
of Applied Ecology 34, 807-816.
Dethier, V. G. (1954). Evolution of feeding preferences in phytophagous insects.
Evolution 8, 33-54.
Dres, M., and Mallet, J. (2002). Host races in plant-feeding insects and their
importance in sympatric speciation. Philosophical Transactions of the Royal
Society London Series B 57, 471-492.
Fujiyama, N., Matsumoto, K., Kobayashi, N., Ohta, Y., and Katakura, H. (2005).
Differentiation in the ability to utilize Pterostyrax hispida (Ebenales:
Styracaceae) among four local populations of the phytophagous ladybird

beetle Henosepilachna yasutomii (Coleoptera: Coccinellidae). Population
Ecology 47, 91-98.
Gratton, C., and Welter, S. C. (1998). Oviposition prefence and larval performance
of Liriomyza helianthi (Diptera: Agromyzidae) on normal and novel host
plants. Environmental Entomology 27, 926-935.
Harris, M. O., Sandanayaka, M., and Griffin, W. (2001). Oviposition preferences of
the Hessian fly and their consequences for the survival and reproductive
potential of offspring. Ecological Entomology 26, 473-486.
Larsson, S., and Ohmart, C. P. (1988) Leaf age and larval performance of the leaf
beetle Paropsis atomaria. Ecological Entomology 13, 19-24.
Mackenzie, A. (1996). A trade-off for host plant utilization in the Black Bean Aphid,
Aphis fabae. Evolution 50, 155-162.
Moon, D. C., and Stiling, P. (2006). Trade-off in oviposition strategy: choosing poor
quality host plants reduces mortality from natural enemies for a salt marsh
planthopper. Ecological Entomology 31, 236-241.
Morrow, P. A. (1976). The significance of phytophagous insects in the Eucalyptus
forests of Australia. In "The role of Arthropods in Forest Ecosystems" (W. J.
Mattson, ed.), pp. 19-29. Springer-Verlag, New York.
Nahrung, H. F., and Allen, G. R. (2003). Geographic variation, population structure
and gene flow between populations of Chrysophtharta agricola (Coleoptera:
Chrysomelidae), a pest of Australian eucalypt plantations. Bulletin of
Entomological Research 93, 137-144.
Noble, I. R. (1989). Ecological traits of the Eucalyptus L'
Herit. subgenera
Monocalyptus and Symphyomyrtus. Australian Journal of Botany 37, 207-
224.
Patterson, K. C., Clarke, A. R., Raymond, C. A., and Zalucki, M. P. (1996).
Performance of first instar Chrysophtharta bimaculata larvae (Coleoptera:
Chrysomelidae) on nine families of Eucalyptus regnans (Myrtaceae).
Chemoecology 7, 94-106.

Reid, C. A. M., and Ohmart, C. P. (1989). Determination of the sex of pupae of
Paropsis atomaria Olivier, and the related paropsina (Coleoptera:
Chrysomelidae). Journal of the Australian Entomological Society 28, 29-30.
Schutze, M. K., Mather, P. B., and Clarke, A. R. (2006). Species status and
population structure of the Australian Eucalyptus pest Paropsis atomaria
Olivier (Coleoptera: Chrysomelidae). Agricultural and Forest Entomology 8,
323-332.
Schutze, M.K., and Clarke, A.R. (2008). Converse Bergmann cline in a Eucalyptus
herbivore, Paropsis atomaria Olivier (Coleoptera: Chrysomelidae):
phenotypic plasticity or local adaptation? Global Ecology and Biogeography.
DOI: 10.1111/j.1466-8238.2007.00374.x
Simmul, T. L., and de Little, D. W. (1999). Biology of the Paropsini (Chrysomelidae:
Chrysomelinae). In Cox, M. L. (ed.), Advances in Chrysomelidae Biology 1,
Backhuys Publishers, Leiden, The Netherlands, pp. 463-477.
Steinbauer, M. J., Clarke, A. R., and Madden, J. L. (1998). Oviposition preference of
a Eucalyptus herbivore and the importance of leaf age on interspecific host
choice. Ecological Entomology 23, 201-206.
Verdon, A., Margraf, N., Davison, A. C., Rahier, M., and Naisbit, R. E. (2007).
Conserved oviposition preferences in alpine leaf beetle populations despite
host shifts and isolation. Ecological Entomology 32, 62-69.

Chapter 6
Thermal requirements, field mortality and population phenology modelling of

Paropsis atomaria Olivier, an emergent pest in subtropical hardwood
plantations
Nahrung, H.F., Schutze, M.K., Clarke, A.R., Duffy, M.P., Dunlop, E.A. and Lawson,
S.A. (2008) Thermal requirements, field mortality and population phenology
modelling of Paropsis atomaria Olivier, an emergent pest in subtropical hardwood
plantations. Forest Ecology and Management, 255 (8), 3515-3523.
Nahrung, H.F.
Contributed towards experimental design, field data collection, and data interpretation.
Wrote parts of manuscript, co-ordinated manuscript construction and acted as corresponding
author.
Schutze, M.K.
Designed and developed experimental protocols (developmental temperature data). Carried

out field and laboratory work, and analysed data. Contributed to interpretation of data and
writing of manuscript.
Clarke, A.R.
A.R. Clarke was the principal supervisor of study design and experimental protocols, and
assisted in the interpretation of data and the construction of the manuscript.
Duffy, M.P.
Designed and undertook field trials and contributed to interpretation of results. Contributed
to manuscript content.
Dunlop, E.A.
Implemented field and laboratory data into the construction of DYMEX model.
Lawson, S.A.
Contributed to experimental design, interpretation, and manuscript construction.
Chapter 6: Population phenology modelling of Paropsis atomaria 103

Abstract
Paropsis atomaria is a recently emerged pest of eucalypt plantations in subtropical

Australia. Its broad host range of at least 20 eucalypt species and wide geographical
distribution provides it the potential to become a serious forestry pest both within
Australia and, if accidentally introduced, overseas. Although populations of P.
atomaria are genetically similar throughout its range, population dynamics differ
between regions. Here we determine temperature-dependent developmental
requirements using beetles sourced from temperate and subtropical zones by
calculating lower temperature thresholds, temperature-induced mortality, and day-
degree requirements. We combine these data with field mortality estimates of
immature life stages to produce a cohort-based model, ParopSys, using DYMEXTM
that accurately predicts the timing, duration, and relative abundance of life-stages in
the field and number of generations in a spring-autumn (September to May) field
season. Voltinism was identified as a seasonally plastic trait dependent upon
environmental conditions, with two generations observed and predicted in the
Australian Capital Territory, and up to four in Queensland. Lower temperature
thresholds for development ranged between 4 and 9 °C, and overall development
rates did not differ according to beetle origin. Total immature development time
(egg – adult) was approximately 769.2 ± s.e. 127.8 DD above a lower temperature
threshold of 6.4 ± s.e. 2.6 °C. ParopSys provides a basic tool enabling forest
managers to use the number of generations and seasonal fluctuations in abundance of
damaging lifestages to estimate the pest risk of P. atomaria prior to plantation
establishment, and predict the occurrence and duration of damaging lifestages in the
field. Additionally, by using local climatic data the pest potential of P. atomaria can
be estimated to predict the risk of it establishing if accidentally introduced overseas.
Improvements to ParopSys’ capability and complexity can be made as more
biological data become available.
Keywords: eucalypt, DYMEXTM, voltinism, seasonal plasticity
104 Chapter 6: Population phenology modelling of Paropsis atomaria

1. Introduction
Commercial hardwood production forests are a recent initiative in subtropical

Australia, with large-scale eucalypt planting recently exceeding 90 000 ha (Parsons
et al., 2006). The concomitant emergence of insect pests associated with plantations,
and the growth and economic losses they cause, are among the most serious
problems faced by plantation managers (Ohmart, 1990). For example, paropsine
chrysomelid beetles cause significant defoliation that can affect the growth rate,
height, volume, and possibly pulpwood quality of trees (Candy et al., 1992; Elek,
1997; Elliott et al., 1998), and are major pests in the commercial eucalypt-growing
regions of Australia (de Little,1989; Simmul & deLittle, 1999), South Africa (Tribe,
2000) and New Zealand (Withers, 2001).
Paropsis atomaria Olivier (Coleoptera: Chrysomelidae) is one such paropsine pest of

eucalypts. This species has four larval instars and long-lived adults that all feed on
the new growth of trees, removing apical leaves and resulting in a characteristic
broom-topped appearance to trees (Cumpston, 1939; Carne, 1966a). Paropsis
atomaria is the most abundant paropsine beetle in plantations of Eucalyptus
cloeziana (F. Muell.) (Nahrung, 2006), and other eucalypt species (Lawson, pers.
obs.), and as such poses a risk to hardwood productivity in New South Wales (NSW)
and Queensland (Qld) (Stone, 1993; Lawson & King, 2002; Schutze et al., 2006).
Although initially not considered a major problem in commercial eucalypt
plantations (see Wylie and Peters, 1993; Elliott et al., 1998; Strauss, 2001), P.
atomaria is now a documented pest of E. grandis (Hill ex Maiden), E. cloeziana and
E. pilularis Smith in Qld and NSW, and of E. camaldulensis Dehnh., E. dunnii
Maiden and E. pilularis Smith in NSW (Simmul & de Little 1999), and is associated
with Corymbia citriodora subsp. variegata (F. Muell.) A.R. Bean & M.W.
McDonald in Queensland (Nahrung, 2006), and several eucalypt species in Victoria
(Collett, 2001) and South Australia (Phillips, 1996). Its pest potential is further
evidenced by its specific inclusion as a containment hazard in the United States
(Eisler, 1999; Kliejunas et al., 2003) and as a regulated pest in New Zealand for
eucalypt-associated imports from Australia (MAF, 2003).
Paropsis atomaria represents a single genetic species (Schutze et al., 2006)

throughout its geographical distribution from South Australia and Victoria to
northern Qld, and it has a broad host range of around 20 eucalypt species

(CABInternational, 2005). Published research on P. atomaria originates
predominantly from the Australian Capital Territory (ACT), and although we can
now reliably use these data in relation to subtropical populations because populations
are genetically similar (Schutze et al., 2006), climatic variation between regions may
result in significant differences in population dynamics. For example, in the ACT, P.
atomaria is bivoltine (Carne, 1966a), but in south-east Qld (SEQ) it can undergo up
to four generations each year (Nahrung, 2006; Duffy, 2007).
Understanding the life history and population dynamics of pests is paramount to

achieving their long-term management (Cox, 1994; Nylin, 2001), while seasonal
predictability of the appearance and duration of susceptible life-stages is essential for
effective application of control measures. Further, the requirements for accurate
forest health reporting (Stone & Coops, 2001), certification of forests for
sustainability (Stone & Coops, 2001; Govender, 2002), and the contentiousness of
pesticide use (Jenkin & Tomkins, 2006), mean that effective targeted management
strategies are important. Here we present underpinning research and a population
phenology model, ParopSys, to help deliver such targeted management for P.
atomaria.
Laboratory trials were used to determine thermal requirements of immature stages of

P. atomaria and these results were integrated with estimations of field mortality
through the DYMEXTM modelling programme (Maywald et al. 2004), which has
been used to produce predictive models for other hardwood forestry pests, including
gumleaf skeletoniser (Farr 2002) and autumn gum moth (Steinbauer et al. 2004).
Lower temperature thresholds, development rates, and mortality were calculated
using beetles originating from temperate and subtropical regions of Australia.
Phenological sampling (not used in model construction) was conducted to assess the
model’s ability to predict voltinism, lifestage peaks and durations.

2. Materials and Methods
2.1. Thermal requirements and thresholds for immature P. atomaria lifestages
Although Carne (1966a) reported development rates of an ACT population of P.
atomaria at 5 – 6 constant temperatures (7.2 - 29.4 °C), we conducted our own
development trials for this study. We considered that inherent inaccuracies in
reading from the development time curves presented and subjectively determining
the linear portion of Carne’s results may cumulatively render DD estimates
unreliable. Further, the question of local adaptation and isolation-by-distance (see
Schutze et al., 2006) and differences in voltinism (Nahrung, 2006) between
temperate and subtropical populations may also mean that results from the ACT do
not apply to subtropical populations. We therefore conducted a new series of
constant temperature development trials using P. atomaria collected from the ACT
and Queensland. These experiments also formed part of a larger study (see Schutze
& Clarke, in press) examining the species status, causes of intraspecific body size
variation, and host plant utilisation of P. atomaria throughout its geographical range.
Paropsis atomaria were collected from two field sites (ACT (Canberra) 35 º18’51”S,
149 º09’16”E and Qld (Lowmead) 24º29’22”S, 151 º42’14”E) in December 2004
and January 2005, and 50-100 beetles from each site were maintained in separate
outdoor colonies on E. tereticornis Smith foliage.
Egg batches were collected daily from rearing colonies, placed in Petri dishes (one
egg batch per dish), and maintained at one of four trial temperatures: 16 ºC, 20 ºC, 24
ºC and 27 ºC. Between nine and thirteen replicate egg batches were used for each
temperature/population origin treatment, with egg development time recorded as the
number of days from egg batch laying to larval eclosion.
For larval development trials, larvae hatched from egg batches in each colony were
divided between temperature treatments to control for possible maternal effects.
Twenty neonate larvae (which had been allowed to feed on their egg chorion) were
placed in each replicate container (Petri dish: 10 – 13 replicates per population),
together with foliage and moistened filter paper. Temperatures used were the same
as for the egg development trials. Larvae were supplied daily with fresh E. pilularis
leaves taken from potted or plantation trees. To control for possible diet effects, on
any one day all leaves supplied to larvae came from one source, with individual
shoots randomised before being placed in rearing containers.

Replicates were checked daily and instar changes noted: instar duration was
calculated based on when 50% of surviving individuals had moulted into the next
stage. Once greater than 50% of larvae in any one replicate reached Liii, all larvae
were transferred to larger containers for the remainder of larval development. When
individuals reached the pre-pupal stage (characterised by cessation of activity and
longitudinal compression, see Cumpston, 1939; Carne, 1966a), they were removed
from rearing containers and placed in clean Petri dishes until adult eclosion.
Between 10 and 13 replicates were conducted for each population (ACT and Qld)
and treatment temperature, but not all replicates survived through to adult eclosion,
especially Qld individuals reared at 27 ºC (due to increased mortality during
development).
Six immature developmental stages were used in DD and T0 calculations (egg, Li,
Lii, Liii, Liv, pre-pupa+pupa), and development time for each stage was considered
as the number of days until 50% of the surviving cohort reached the subsequent
stage. Pre-pupal and pupal stages were combined at the outset because they are non-
feeding, difficult to sample in the field, and ecologically inactive. Data were
analysed using mean development rate (the reciprocal of development time) for each
developmental stage for each treatment temperature. A linear regression model was
fitted to the development rate for each of the six immature life stages described
above, yielding for each an equation in the form y = a + bx, where y is the rate of
development (1/days), x is temperature, a is the intercept and b is the slope. An
Analysis of Covariance (ANCOVA) was conducted for each developmental stage to
determine whether development rate differed between population origin. Total
immature development time did not differ between ACT and Qld populations except
at 16 ºC (Schutze & Clarke, in press); nor was there any difference in development
rate for each immature stage separately (see Results) so mean development data for
each site were pooled for DD and T0 calculations. Because early instars (Li and Lii)
are difficult to differentiate in the field (Duffy, 2007), they were combined into an
additional developmental stage to enable model development and validation, and
permit application of field-based mortality estimates. The lower temperature
thresholds (T0) for development were estimated by solving the regression equation
for development rate = 0 (x-intercept, ie the temperature below which no
development occurs), and the number of DD required for each life stage was

estimated by 1/b for each immature life stage (as in Nahrung et al. 2004). Standard
errors for T0 and DD estimates were calculated using the methods of Campbell et al.
(1974).
2.2. Mortality of immature life stages
2.2.1. Laboratory estimates Using data from the development rate experiments
outlined above, mortality was compared between beetle origin (ACT and Qld) and
treatment temperature, following arcsine-square-root transformation of mortality
rates, using a two-way ANOVA, with post-hoc differences between temperature
treatments identified using Fishers LSD test. Stage-specific mortality as a function
of temperature was also determined, and compared using a two-way ANOVA
(temperature*development stage). Overall laboratory egg-Li mortality data were
estimated using the average hatch rate of unparasitised field-collected egg batches
reported by Duffy (2007) and Duffy et al. (in press).
2.2.2. Field estimates To compare laboratory mortality estimates and mortality in

the presence of natural enemies (see Nahrung et al., in press), field surveys counting
the number of eggs, early instar larvae (Li+Lii), Liii, and Liv were conducted at two-
weekly intervals between September 2004 and April 2005 at two E. cloeziana
plantation sites as follows: Site I 26°04 30.72 S 152°44 8.88 E Mean average daily
temperature was 22.48 ± 0.19 °C, maximum mean 28.75 °C and minimum mean
13.75 °C; and Site II 26°11 20.4 S 152°29 40.2 E Mean average daily temperature
was 22.66 ± 0.20 °C, maximum mean 29.5 °C and minimum mean 13.75 °C.
Eight sections across each plantation were representatively selected on each census
date (different sections and trees each time), and three branches from each of six
trees within each section were visually searched for P. atomaria lifestages. The
number of egg batches and larvae of each instar on these 144 branches was counted
and used to determine the average number of every lifestage present per branch
throughout the field season. The population difference between egg and final instar
larvae was calculated to estimate overall field mortality rates of immature stages.
Mortality between each developmental stage was estimated using the total of all life-
stages recorded during the season at each site. Proportional mortality was calculated
using the difference between the number of individuals in successional stages, and
between egg and final instar larvae for overall immature mortality. An average egg

batch size of 76 eggs per batch (Duffy, 2007) was used to estimate subsequent
mortality rates.
2.3. Population modelling
2.3.1. Model description and overview
The model, ParopSys, describes the life cycle processes and population dynamics of
P. atomaria in relation to climate. ParopSys was created using DYMEXTM V2
(Maywald et al., 2004). The life cycle processes for ParopSys were modelled on a
daily time step for 236 days, representing the phase of the P. atomaria life cycle
when the beetles are active on foliage (i.e. flying, feeding and mating) during the
months of September to May (Duffy, 2007). The first time step (Day 1) begins on
the 21st of September (when beetles were first observed in the field) and concludes
on the 15th May (after which no adults were observed in the field, Duffy, 2007).
Immigration and emigration are not explicitly considered, as they are assumed to be
equal with no net effect on the numbers of eggs.
ParopSys identifies eight discrete life stages in the P. atomaria life cycle: egg, early-
instar larvae (1st and 2nd instar), 3rd instar larvae, 4th instar larvae, pupae (pre-pupae
and pupae), pre-reproductive adults, overwintering adults, and sexually mature adults
(reproductive) (Figure 1). A series of functions describe the lifecycle processes,
including development and mortality rates for each life stage, as well as the transfer
of individuals from one life stage to the next, and adult fecundity and rates of
reproduction. Field and laboratory data presented within this paper were used
wherever possible to derive life cycle functions. Where this data set was insufficient,
data published in Carne (1966a) or unpublished laboratory estimates were used.
Relationships of P. atomaria to environmental factors other than temperature were
not explicitly considered.

Egg
Larva 1 & 2
Larva 3
Larva 4
Pupa
Pre-reproductive Adult
Adult
Overwintering Adult
Figure 1: Schematic diagram of the life cycle of Paropsis atomaria used in the
DYMEXTM model.
2.3.2. Egg and larval development
The thermal thresholds and functions for rates of egg and larval development used in
ParopSys are presented in Table 1. T0 were entered as the threshold temperature,
above which a linear relationship between temperature and development time
occurred. High temperature-induced reduction in development rate was not
considered in this model.
2.3.3. Mortality
Table 2 shows the average estimated mortality of each immature P. atomaria life
stage experienced under field conditions. However, Table 2 does not quantify field
mortality of pupae and adult beetles. Mortality experienced in laboratory trials was

used instead, with a correction for increased mortality that would be experienced in
the field by these stages. This was derived by calculating the average mean
difference between field and laboratory mortality data for the egg through to 4th
larval instar stages, and multiplying the pupal and adult lab mortality rates by this
average value. In the model, pupae experience a 0.08 mortality rate on exit, while
adult mortality and longevity was modelled using a constant mortality parameter of
0.006 per day (mortality associated with field predation and other biotic and abiotic
factors) combined with mortality due to age. Beetles were assumed to live for 85.6
days based on mean longevity of adult beetles (average of Carne, 1966 and Nahrung,
unpubl. data). Rates for each possible cause of mortality in the field were not
quantified explicitly. Therefore, in ParopSys, P. atomaria cohorts experience a
combined total mortality on exit from each life stage.
2.3.4. Stage transfer
Stage transfer of egg and larval stages in ParopSys was determined by analysing the
relationship between accumulated degree days and the proportion of individuals
developing into each life stage at each temperature. The resulting pattern of transfer
is predominantly linear (Maywald, et al. 2004; our results not shown) and hence
ParopSys uses a linear-above-threshold transfer function to determine transfer rate
(i.e. the daily proportion of individuals in a cohort moving into a new stage). This
function results in a spread of individuals transferring between stages whereby
transfer commences at the lower heat threshold (accumulated degree days) and is
completed at the upper heat threshold. Prepupal and pupal stages were considered
together for the purposes of development rate estimation and therefore also in the
model. As with the larval stages, development and transfer from the fourth larval
instar used a linear above threshold function (Table 1).
2.3.5. Pre-reproductive adult development and fecundity
Development rates for pre-reproductive adult beetles (pre-oviposition period) were

derived from Carne (1966a). Newly-emerged adults undergo a period of maturation
before being capable of oviposition, modelled as a linear above threshold function in
the model (development rate = 0.0018, T0 = 2.7 °C). Potential reproductive capacity
of 640 eggs per female (Carne, 1966a) was used to describe fecundity. Because
DYMEXTM does not specifically model sex ratios, a mean fecundity of 320 eggs per
beetle was applied in the model based on the observed 1:1 operational sex ratio of

this beetle (Duffy, 2007). A pulse function was used to describe the temporal
distribution of egg-laying, with batches of 32.5 eggs per adult being laid every seven
days (Carne, 1966a).
2.3.6. Overwintering adults
Carne (1966a) reported that P. atomaria overwintering in ACT populations is

triggered by daylength and terminated in response to temperature, but he did not
provide specific data that we could use in our model. Furthermore, the translation to
SEQ populations may not be accurate. Arbitrary days of the year were therefore
used in ParopSys to initiate (27th April) and terminate (21st September)
overwintering. These dates coincide with changes in beetle activity observed in the
field (Duffy, 2007). Adults are the only lifestage that overwinter, and the model
assumes that all pre-reproductive individuals present will overwinter between the
27th of April and the 21st September, with half (Nahrung, unpubl. data) surviving to
reproduce. In the model lifecycle, surviving overwintered adults resume activity as
pre-reproductive adults and thus undergo a pre-oviposition period before eggs are
produced (Carne, 1966a) (Figure 1).
2.3.7. Meteorological Data
Meteorological data used in ParopSys were obtained from the Silo Data drill website
(http://www.nrw.qld.gov.au/silo/datadrill/). This is spatially interpolated data and
may not equate exactly with localised conditions. A circadian temperature model
was used to drive temperature related functions in ParopSys. This enables hourly
calculations of the average temperature to be derived, which are based on the
interpolation of the daily maximum and minimum temperatures using a composite
sine and exponential function.
2.3.8. Model initialization
For all sites, the model was run for the period from the 21st September 2005 to the
15th May 2006. It was assumed that only adults survive over winter, so the lifecycle
module was initialised with 0.5 pre-reproductive adults per day for ten days from the
start date. Pre-reproductive adults were used to initialise the model instead of
reproductive adults because overwintering adults need to feed for a period of time
before reproduction (Carne, 1966; see above).

2.4. Field validation
2.4.1. South East Queensland
To provide data to validate ParopSys, a third plantation, Site III, 26°05 97.2 S
152°43 7.54 E, was sampled during the 2005/2006 season. Mean average daily
temperature was 23.00 ± 0.19°C, with a maximum mean of 29.75°C and minimum
mean of 14.25°C. Samples as in 2.2.2 were taken every two weeks between October
2005 and April 2006 to provide phenological data that were then used to check the
accuracy of the model in predicting the onset, duration and peaks of each
developmental stage in the field, and to predict the number of beetle generations.
2.4.2. Temperate vs. subtropical conditions
To test the model’s validity over a range of environmental conditions that occur in
the extremely wide natural distribution of P. atomaria and for the populations where
developmental data were compared in this study, the model was run over the same
time period (September 2005 to May 2006) using climatic data for Canberra
(temperate south eastern Australia) and Lowmead (subtropical central Qld).
3. Results
3.1. Thermal requirements and thresholds for immature P. atomaria lifestages
Development rates did not differ according to source of beetle origin for any
developmental stage (ANCOVA, F1,5 = 0 – 0.62, P = 0.47 – 0.99), suggesting that the
differences in voltinism reported between them (compare Carne, 1966, Nahrung,
2006) is probably a seasonally plastic trait dependent upon field conditions. Data
from ACT and Qld populations were therefore combined to produce developmental
thresholds (T0) and DD requirements for ParopSys (Table 1).

Table 1. Developmental thresholds (T0), thermal requirements (DD) and proportion
of development time for immature lifestages of Paropsis atomaria.
Immature Regression R2 T0 ± s.e. DD ± s.e. Proportion
stage equation* P-value (°C) of
immature
life cycle**
Egg y = 0.008x – 0.98 8.9 ± 0.7 125.0 ± 0.20
0.0713 <0.001 7.0
Li y = 0.0115x – 0.887 5.3 ± 2.4 87.0 ±
0.0607 <0.001 12.7
Lii y = 0.0156x – 0.884 5.6 ± 2.4 64.1 ± 9.5
0.0862 <0.001
Li+Lii y = 0.0066x – 0.887 5.4 ± 2.9 166.7 ± 0.21
0.0358 <0.001 26.8
Liii y = 0.0145x – 0.927 6.0 ± 1.8 69.0 ± 7.9 0.09
0.0876 <0.001
Liv y = 0.0037x – 0.703 4.0 ± 6.9 270.3 ± 0.11
0.0147 0.01 72.2
pp+p y = 0.0051x – 0.964 8.0 ± 1.1 196.1 ± 0.40
0.041 <0.001 15.5
*Temperature range 16 – 27 °C. Linear regression model y = a + bx where y is the
rate of development (1/days), x is temperature, a is the intercept and b is the slope.
** at an average spring/summer temperature of 23 °C
Total immature development time (egg – adult) was approximately 769.2 ± s.e. 127.8
DD above T0 6.4 ± s.e. 2.6 °C: about 49 days at the average field temperature of 23
°C. As a proportion of total development time under average field conditions, the
longest stage durations were for fourth instar larvae and pre-pupae+pupae (Table 1),
while the shortest was Lii and Liii.

Overall Li-adult mortality in the laboratory was higher from Qld-sourced P.
atomaria (2-way ANOVA, F1,72 = 4.4, P = 0.04) with ACT larvae exhibiting greater
survival at 20 °C and 27 °C than those from Qld; survival at 16 °C and 24 °C did not
differ according to origin. Overall Li-adult mortality (Figure 2) increased with
temperature (2-way ANOVA, F3,72 = 13.9, P < 0.001), best described (R2 = 0.98) by
the polynomial function y = - 0.028x2 + 0.28x + 0.17. The interaction between beetle
origin and temperature on overall immature mortality was almost significant (2-way
ANOVA, F3,72 = 2.8, P = 0.05). Nevertheless, to obtain stage-specific mortality as a
function of temperature for ParopSys, we combined data from ACT and Qld
populations.
1 c
c
0.8 b
mortality rate
a
0.6
0.4
0.2
0
16 20 24 27
temperature (C)
Figure 2: Mean ± s.e. mortality rate between Li and adult Paropsis atomaria at four
constant temperatures in the laboratory. Different letters denote means that differ
significantly.

Stage-specific transfer mortality differed according to developmental stage and
temperature (2-way ANOVA, stage: F3,361 = 63.1, P < 0.001; temperature: F4,361 = 14.2,
P < 0.001; Figure 3), but with a significant interaction between factors
(stage*temperature: F12,361 = 5.03, P < 0.001). Mortality at temperatures 24 °C and
above did not differ significantly (Fishers LSD post-hoc test).
0.8
0.6
mortality rate
0.4
0.2
0
Li-Lii Lii-Liii Liii-Liv Liv-p p-A
Figure 3: Stage-specific average ± s.e. mortality of immature Paropsis atomaria at

four temperatures: 16 °C (diamonds), 20 °C (squares), 24 °C (triangles) and 27 °C
(circles).
3.2. Field mortality of immature life stages
Less than 8% of eggs survived to become fourth instar larvae (Table 2). The highest
mortality occurred between egg and early instar larvae (first and second instars) at
Site I, and between early instar larvae and third instar larvae at Sites II and III. These
data do not reflect loss from larval parasitoids which generally emerge from fourth
instar larvae or pre-pupae, and nor is loss from Liv onwards determined.

Table 2: Estimated mortality (proportion of eggs and larvae lost) at each immature
life stage of Paropsis atomaria in the field at two sites (Li = first instar, Lii = second
instar, Liii = third instar, Liv = fourth instar).
Life Stage Site I Site II Average ± se
Egg to Li+Lii 0.81 0.68 0.75 ± 0.1
Li+Lii to Liii 0.60 0.74 0.67 ± 0.1
Liii to Liv 0.38 0.04 0.21 ± 0.2
all larvae (Li to 0.75 0.75 0.75 ± 0
Liv)
Egg - Liv 0.95 0.92 0.94 ± 0.02
3.3. Population modelling and field validation
3.3.1. South East Queensland
To validate the model’s predictive capability, we compared field-derived

phenological data obtained from Site III for September 2005 to May 2006 and
predictions made by ParopSys (Figure 4). The model shows a very good fit with the
field data in terms of number of generations, timing of generational peaks, and the
relative sizes and shapes of each peak. Three P. atomaria generations were observed
during the active beetle season at Site III and the model was accurate in predicting
the same number of generations. Timing of each peak in the model differed
somewhat from the field data, but was mostly within the margin associated with
fortnightly collection of field data (Figure 4). Relative sizes of each population peak
were very well predicted by the model for all life stages, with the greatest variation
occurring in the timing and size of the first generation for both eggs and total larvae.
The size and timing of the final, highest population peak was very close to that of the
observed field data for all life stages (Figure 4).

90 12
80 (A)
10
Field Population (Mean No. per shoot)

70
Dymex Population size
60 8
50
6
40
30 4
20
2
10
0 0
8000 6
7000
(B)
5

6000
4
5000
4000 3
3000
2
2000
1
1000
0 0
2000 12
1800 (C)
10

1600
1400
8
1200
1000 6
800
4
600
400
2
200
0 0
28/09/05
12/10/05
26/10/05
9/11/05
23/11/05
7/12/05
21/12/05
4/01/06
18/01/06
1/02/06
15/02/06
1/03/06
15/03/06
29/03/06
12/04/06
26/04/06
10/05/06
Sample/Simulation Date
Figure 4: Field validation of ParopSys model between 28 Sept 2005 and 10 May
2006 for Site III, South East Queensland - (A) Adults (B) Eggs (C) Total larvae.
Solid lines represent field phenological data collected fortnightly (right y-axis);
dotted lines represent DYMEXTM model data using Data Drill meteorological data
for Site III (left y-axis). DYMEXTM population numbers are dependent on numbers
initialised and so relate relatively to field data (numbers per shoot).

3.3.2. Temperate vs subtropical conditions
Results of the simulations comparing temperate and subtropical populations are

shown in Figure 5. ParopSys correctly predicted bivoltinism in Canberra (see Carne,
1966a) and that P. atomaria would have had four generations during the season at
Lowmead. The model initialized both locations with the same number of pre-
reproductive adults but peak adult populations were around ten-fold higher at
Lowmead than at Canberra. The model predicts one more generation per year at
Lowmead compared to Site III in SEQ.
20 250
18
16 200
14
Canberra Lowmead
Adult Population
Adult Population
12 150
10
8 100
4 50
0 0
3500 20000
3000
15000
2500
Egg Population
Egg Population
2000
10000
1500
1000
5000
500
0 0
800 2000
700
600 1500
Larval Population
Larval Population
500
400 1000
300
200 500
100
0 0
28/09/2005
12/10/2005
26/10/2005
9/11/2005
23/11/2005
7/12/2005
21/12/2005
4/01/2006
18/01/2006
1/02/2006
15/02/2006
1/03/2006
15/03/2006
29/03/2006
12/04/2006
26/04/2006
10/05/2006
28/09/2005
12/10/2005
26/10/2005
9/11/2005
23/11/2005
7/12/2005
21/12/2005
4/01/2006
18/01/2006
1/02/2006
15/02/2006
1/03/2006
15/03/2006
29/03/2006
12/04/2006
26/04/2006
10/05/2006
Simulation Date Simulation Date
TM
Figure 5: DYMEX model predictions for adult, egg and total larval populations of
Paropsis atomaria for ACT (Canberra) and Queensland (Lowmead) between 28 Sept
2005 and 10 May 2006. DYMEXTM population numbers are dependent on numbers
initialised and so relate relatively to field data (numbers per shoot).

4. Discussion
4.1. Development and mortality
Our temperature-development results coincide with those originally calculated by

Carne (1966a) for ACT beetles, and they are also consistent across populations
sourced from within relative extremes of P. atomaria’s geographical distribution.
Such low variation in development rates between populations over spatial and
temporal scales supports the finding that P. atomaria is genetically similar
throughout its range (Schutze et al., 2006). Therefore, the difference in the number of
P. atomaria generations observed between regions suggests that voltinism is a
seasonally plastic trait influenced by environmental conditions such as temperature.
This is supported by the high degree of accuracy provided by our model in which
temperature was the only environmental variable included. We do, however, consider
other factors – especially photoperiod and host plant quality – contribute towards
determining voltinism in P. atomaria (see Carne, 1966a).
Photoperiod is an important factor that can indirectly influence voltinism through its
role as a trigger in diapause initiation and termination. Photoperiod is considered the
most influential and seasonally reliable diapause cue in insects, while temperature is
considered the second most important environmental regulator (Tauber & Tauber,
1976, Tauber et al., 1986). In the paropsine Chrysophtharta agricola (Chapuis)
(Coleoptera: Chrysomelidae), for example, whilst temperature was secondarily
responsible for inducing diapause under controlled conditions, photoperiod was the
dominant environmental factor (Nahrung & Allen, 2004b). Further, the interplay
between photoperiod and temperature may be critical for the induction of diapause,
as demonstrated for the flea beetle, Argopistes coccinelliformis Csiki (Coleoptera:
Chrysomelidae) (Inoue, 2001). Carne (1966a) reported that newly-emerged P.
atomaria adults are responsive to photoperiodic cues for reproductive diapause: in
the ACT, adults emerging from pupation in February attain reproductive maturity,
whereas those that emerge after the first week of March enter diapause without
reproductive development. Further work is required to elucidate diapause cues under
subtropical conditions: a limitation of ParopSys is our use of an arbitrary date, rather
than a specific environmental cue, for simulating the initiation of reproductive
diapause. Indeed, altering the start date by one week to 14 September gave a better
fit with field data for the timing of peaks (output not shown), suggesting that we

didn’t accurately identify the start of the season. Nevertheless, ParopSys generates
field-corroborated voltinism accurately for subtropical and temperate regions.
Host plant quality and availability is also likely to contribute to the number of
generations produced in a season, whereby if host plant quality is poor, or suitable
hosts are not present, oviposition may be delayed until suitable larval food sources
appear (see Carne, 1966a). Host plant influence is an important factor for paropsine
population dynamics (Ohmart 1991) as for Chrysophtharta bimaculata (Olivier), an
important paropsine pest of Tasmanian eucalypt plantations (Steinbauer et al., 1998).
In this case, host plant quality (amount of flush foliage present) plays a more
important role in stimulating oviposition than does host species (Steinbauer et al.,
1998). Further, P. atomaria populations located merely 20 km apart in SEQ
exhibited variable voltinism within the same field season (Nahrung, 2006; Duffy,
2007) – a result unexplainable using temperature or photoperiod data alone. Delayed
oviposition by females due to poor early-season host plant quality (i.e. less flush
foliage) was proposed as the driving cause influencing variable voltinism in that case
(Duffy 2007). The potential importance of flush foliage in driving P. atomaria
populations is not unexpected considering the importance of host plant quality on
successful P. atomaria larval establishment, with first instars suffering extremely
high levels of mortality on older, tougher leaves (Ohmart et al., 1987, Larsson &
Ohmart, 1988).
Early instars suffer the greatest mortality in the field despite their gregarious
behaviour potentially increasing defence (Sillen-Tulberg 1988) and feeding
establishment (Nahrung et al. 2001). Early paropsine instars experience high
mortality under laboratory conditions in C. bimaculata (Baker et al., 2002) and C.
agricola (Nahrung et al., 2001), and our overall egg-Liv field mortality estimates for
P. atomaria were similar to these temperate species (deLittle et al., 1990; Nahrung &
Allen, 2004a, respectively). Our laboratory trials revealed relatively low survival
rates for fourth instar larvae, exacerbated by higher rearing temperatures, especially
above 24 °C. Increased heat stress under experimental conditions may have caused
high mortality at this life stage in the laboratory.

4.2. Population modelling
ParopSys is a simple DYMEXTM model based on temperature-dependent

development thresholds, field- and laboratory-derived mortality data and general
ecological knowledge of the beetle’s behaviour. Despite this simplicity, validation
against field data showed that ParopSys was accurate in predicting the number,
timing, and relative size of P. atomaria generations. This varied somewhat with
beetle stage: timing of adult population peaks most closely matched that of the field
data, with larval peaks showing the greatest discrepancies with timing of egg peak
heights intermediate between these (Figure 4). The model also suggested that an
early field egg peak may have been missed because field sampling did not commence
until 25 October 2005. For all stages, relative peak heights for each generation
closely matched that of the field data, with the largest populations occurring from
early March (adults) to late March – early April (larvae). Height of this final peak
for all stages was approximately three-fold higher than for the earlier two peaks.
This agrees with field observations of severe beetle damage at this time.
The model was also robust across climatic zones by correctly predicting bivoltinism
at Canberra (ACT), and one extra generation at Lowmead (central Qld) compared to
SEQ. Since 2004 forestry plantation companies in this area have reported severe P.
atomaria defoliation of young Eucalyptus taxa in mid- to late-May, coinciding with
peak larval and adult populations predicted by ParopSys (Lawson, unpubl. data),
although phenological field data for Lowmead are not currently available. The model
also predicts that P. atomaria is likely to be a more serious pest in the subtropics,
with much larger populations compared with temperate areas.
4.3. Future improvements
At its current state of development, ParopSys does not incorporate sophisticated

environmental and life cycle parameters that could improve its predictive ability.
The interaction between leaf dynamics, herbivory, and rainfall (see e.g. Stone &
Bacon, 1995), as well as tree growth rates could be incorporated to provide estimates
of stand productivity and losses. Although the good fit of the model with field data
suggests that beetle immigration and emigration is absent or equal within seasons,
any such movement could potentially be linked to plantation proximity to native
vegetation acting as a beetle and natural enemy source or sink (see Strauss, 2001).

Future improvements to ParopSys may also incorporate egg parasitism (Duffy et al.,
in press; Nahrung et al., in press; Nahrung & Duffy, 2008); basking behaviour of
beetle stages that may affect development rates (e.g. Maddox, 1995); differential
performance on different host species, and specific factors that determine diapause
induction and termination (particularly in relation to daylength). The relationships
between beetle size with geographic origin, host plant species/quality, and fecundity
also deserves further investigation (see Carne, 1966a; Schutze & Clarke 2008).
Future versions of ParopSys will incorporate ‘event scenarios’ where the efficacy of
number and timing of control measures, such as insecticides, can be evaluated as
desktop studies. The model can also be linked with tree growth models to produce
cost-benefit analyses of management strategies for P. atomaria when impact data
become available. Linking the model with GIS software may also enhance its
attractiveness to plantation managers. DYMEXTM versions 2 and 3 incorporate a
Climex-type mapping function (Maywald et al., 2004) that can be used in regional
and global risk modelling for plantations and a climate change function that can be
used to predict how the risk of serious damage by P. atomaria may change with
currently available global warming scenarios.
5. Conclusion
The robustness of ParopSys for predicting numbers of generations, timing of

population peaks (in particular the late season population peak, where the most
severe and long-term impact on tree growth rates occurs), and size of the final
population peak suggest that ParopSys may be helpful to plantation managers in
developing risk models for current and future plantations. The model may also assist
in predicting year to year fluctuations in P. atomaria damage and thus assist
managers in planning forest health surveys, monitoring, and management responses.
Acknowledgments
Our sincere thanks to Nikki Sims & Andy Hulthen (both QUT) for laboratory
assistance; Jacinta Hodnett, Janet McDonald, Daniel Hancox & Rebeccah Aigner (all
Department of Primary Industries and Fisheries) for field assistance. Sincere thanks
also to Gunter Maywald (CSIRO Entomology) for assistance with DYMEXTM. MKS
was in receipt of a QUT Postgraduate Research Award, and parts of this work were
carried out under Australian Research Council Linkage Projects Program

(LP0454856) in conjunction with Forestry Plantations Queensland (formerly DPI-
Forestry). We gratefully acknowledge all organisations for their support.

References
Baker, S.C., Elek, J.A., Candy, S.G., 2002. A comparison of feeding efficiency,
development time and sutvival of Tasmanian eucalyptus leaf beetle larvae
Chrysophtharta bimaculata (Olivier) (Coleoptera: Chrysomelidae) on two
hosts. Australian Journal of Entomology 41, 174-181.
CABInternational, 2005. Forestry Compendium. Wallingford UK: CAB
International.
Campbell, A., Frazer, B.D., Gilbert, N., Gutierrez, A.P., Mackauer, M., 1974.
Temperature requirements of some aphids and their parasites. Journal of
Applied Ecology 11, 431-438.
Candy, S.G., Elliott, H.J., Bashford, R., Greener, A., 1992. Modelling the impact of
defoliation by the leaf beetle, Chrysophtharta bimaculata (Coleoptera:
Chrysomelidae), on height growth of Eucalyptus regnans. Forest Ecology and
Management 54, 69-87.
Carne, P.B., 1966a. Ecological characteristics of the eucalypt-defoliating
chrysomelid Paropsis atomaria Ol. Australian Journal of Zoology 14, 647-672.
Carne, P.B., 1966b. Growth and food consumption during the larval stages of
Paropsis atomaria (Coleoptera: Chrysomelidae). Ent. exp. & appl. 9, 105-112.
Collett, N., 2001. Insect pests of young eucalypt plantations. Agriculture Notes 6,
December 2001. Department of Natural Resources and Environment,
Heidelberg.
Cox, M.L., 1994. Diapause in the Chrysomelidae. In Jolivet, PH, Cox ML and
Petitpierre, E. (Eds), Novel aspects of the biology of Chrysomelidae. Kluwer
Academic Publishers, Dordrecht, pp. 469-502.
Cumpston M., 1939. Observations of the bionomics and morphology of seven
species of the tribe Paropsini (Chrysomelidae). Proceedings of the Linnean
Society of New South Wales 64, 353-366.
de Little, D.W., Elliott, H.J., Madden, J.L., Bashford, R., 1990. Stage-specific
mortality in two field populations of immature Chrysophtharta bimaculata
(Oliver) (Coleoptera: Chrysomelidae). Journal of the Australian
Entomological Society 29, 51-55.
deLittle, D.W., 1989. Paropsine chrysomelid attack on plantations of Eucalyptus
nitens in Tasmania. New Zealand Journal of Forestry Science 19, 223-227.

Duffy, M.P., Nahrung, H.F., Lawson, S.A., Clarke, A.R., in press. Direct and
indirect effects of egg parasitism by Neopolycystus Girault sp. (Hymenoptera:
Pteromalidae) on Paropsis atomaria Olivier (Coleoptera: Chrysomelidae).
Australian Journal of Entomology.
Duffy, M.P., 2007. Population phenology and natural enemies of Paropsis
atomaria in South-East Queensland. MSc thesis, School of Natural Resource
Sciences, Queensland University of Technology. 100pp.
Eisler, R., 1999. Containment Hazard Reviews Report No. 35 Cumulative index to
chemicals and to common and scientific names of species listed in containment
hazard reviews 1 through 34. Biological Science Report USGS/BRD/BSR--
1999-0002. Pauxtent Wildlife Research Center, Laurel, MD. 80 pp.
Elek, J.A., 1997. Assessing the impact of leaf beetles in eucalypt plantations and
exploring options for their management. Tasforests 9, 139-154.
Elliott, H.J., Ohmart, C.P., Wylie, F.R., 1998. Insect Pests of Australian Forests:
ecology and management. Melbourne, Inkata Press.
Elliott, H.J., Bashford, R., Greener, A., Candy, S.G., 1992. Integrated pest
management of the Tasmanian Eucalyptus leaf beetle, Chrysophtharta
bimaculata (Olivier) (Coleoptera: Chrysomelidae). Forest Ecology and
Management 53, 29-38.
Elliott, H.J., Bashford, R., Greener, A., Candy, S. G., 1992. Integrated pest
management of the Tasmanian Eucalyptus beetle, Chrysophtharta bimaculata
(Olivier) (Coleoptera: Chrysomelidae). Forest Ecology and Management 53,
29-38.
Farr, J. D., 2002. Biology of the gumleaf skelentoniser, Uraba lugens Walker
(Lepidoptera: Noctuidae), in the southern jarrah forest of Western Australia.
Australian Journal of Entomology 41, 60-69.
Govender, P., 2002. Management of insect pests: have the goalposts changed with
certification? Southern African Forestry Journal 195, 39-45.
Hopkins, R. J., Ekbom, B., 1999. The pollen beetle, Meligethes aeneus, changes to
egg production rate to match host quality. Oecologia 120, 274-278.
Inoue, T., 2001. Life history of the flea beetle, Argopistes coccinelliformis Csiki
(Coleoptera: Chrysomelidae) VII. Effects of photoperiod and temperature on
induction of reproductive diapause in newly emerged adults. Applied
Entomology and Zoology 36, 53-58.

Jenkin, B.M., Tomkins, B., 2006. Chemical pesticide usage by Australian
plantation industries. FWPRDC Project PN06.4016. Forest and Wood Products
Research and Development Corporation, Melbourne.
Kliejunas, J.T., Burdsall, H.H.J., deNitto, G.A., Eglitis, A., Haugen, D.A., Haverty,
M.I., Micales, J.A., Tkacz, B.M., Powell, M.R., 2003. Pest risk assessment of
the importation into the United States of unprocessed logs and chips of eighteen
Eucalypt species from Australia. General Technical Report FPL-GTR-137.
Madison, WI: U.S. Department of Agriculture, Forest Service, Forest Products
Laboratory. 206 pp.
Larsson, J., Ohmart, C.P., 1988. Leaf age and performance of the leaf beetle
Paropsis atomaria. Ecological Entomology 13, 19-24.
Lawson, S.A., King, J., 2002. Leaf beetles (Paropsines). Queensland Forest
Research Institute, Agency for Food and Fibre Sciences, DPI. Brisbane: DPI
Note FOR0027.
Maddox, C.D.A., 1995. Aspects of the biology of Paropsisterna tigrina (Chapuis)
(Coleoptera: Chrysomelidae) the major pest species of Melaleuca alernifolia
(Cheel) (Myrtaceae), Masters thesis, Department of Entomology, The
University of Queensland.
M.A.F., 2003. Import Health Standard: Sawn wood from all countries. New
Zealand Ministry for Agriculture and Forestry, Forest Biosecurity: 21 pp.
Maywald, G.F., Sutherst, R.W., Kriticos, D.J., 2004. DYMEX Professional:
Modelling Natural Systems, Version 2.0, Hearne Scientific Publishing,
Melbourne.
Nahrung, H.F., 2006. Paropsine beetles (Coleoptera: Chrysomelidae) in South-East
Queensland hardwood plantations: identifying potential pest species.
Australian Forestry 69, 270-274.
Nahrung, H.F., Allen, G.R., 2004a. Population dynamics of the chrysomelid leaf
beetle Chrysophtharta agricola (Chapuis), a pest of Eucalyptus nitens
plantations in Tasmania. TasForests 15, 67-84.
Nahrung, H.F., Allen, G.R., 2004b. Overwintering ecology of Chrysophtharta
agricola: mechanisms of reproductive diapause induction and termination.
Australian Journal of Zoology 52, 505-520.
Nahrung, H.F., Allen, G. R., Patel, V. S., 2004. Day-degree development and
phenology modelling of the immature stages of Chrysophtharta agricola

(Chapuis) (Coleoptera: Chrysomelidae), a pest of eucalypt plantations.
Australian Journal of Entomology 43, 177-183.
Nahrung, H.F., Duffy, M.P., 2008. Exploring the fourth trophic level: do
hyperparasitoids influence biocontrol of a forestry pest in a native system? The
New Zealand Entomologist 31, 6-12.
Nahrung, H.F., Duffy, M.P., Lawson, S.A., Clarke, A.R., in press. Natural enemies
of Paropsis atomaria (Coleoptera: Chrysomelidae) in South-East Queensland
eucalypt plantations. Australian Journal of Entomology.
Nahrung, H.F., Dunstan, P.K., Allen, G.R., 2001. Larval gregariousness and
neonate establishment of the eucalypt-feeding beetle Chrysophtharta agricola
(Coleoptera: Chrysomelidae: Paropsini). Oikos 94, 358-364.
Nylin, S., 2001. Life history perspectives on pest insects: What’s the use? Austral
Ecology 26, 507-517.
Ohmart, C. P., 1990. Insect pests in intensively-managed eucalypt plantations in
Australia: some thoughts on this challenge to a new era in forest management.
Australian Forestry 53, 7-12.
Ohmart, C.P., 1991. Role of food quality in the population dynamics of
chrysomelid beetles feeding on Eucalyptus. Forest Ecology and Management
39, 35-46.
Ohmart, C.P., Thomas, J.R., Stewart, L.G., 1987. Nitrogen, leaf toughness and the
population dynamics of Paropsis atomaria Olivier (Coleoptera:
Chrysomelidae) – a hypothesis. Journal of the Australian Entomological
Society 26, 203-207.
Parsons ,M., Gavran, M., Davidson, J., 2006. Australia’s plantations 2006 National
Report. Department of Agriculture, Fisheries and Forestry. 24pp.
Philips, C., 1996. Insects, diseases and deficiencies associated with eucalypts in
South Australia. Primary Industries South Australia, Mt Gambier.
Schutze, M.K., Clarke, A.R. (2008). Converse Bergmann cline in a Eucalyptus
herbivore, Paropsis atomaria Olivier (Coleoptera: Chrysomelidae): phenotypic
plasticity or local adaptation? Global Ecology and Biogeography. DOI:
10.1111/j.1466-8238.2007.00374.x
Schutze, M.K., Mather, P.B., Clarke, A.R., 2006. Species status and population
structure of the Australian Eucalyptus pest Paropsis atomaria Olivier
(Coleoptera: Chrysomelidae). Agricultural and Forest Entomology 8, 323-332.

Sillen-Tullberg, B., 1988. Evolution of gregariousness in aposematic butterfly
larvae: a phylogenetic analysis. Evolution 42, 293-305.
Simmul, T.L., deLittle, D.W., 1999. Biology of the Paropsini (Chrysomelidae:
Chrysomelinae). In Advances in Chrysomelidae Biology. Cox, M. L. (Ed.)
Leiden, Backhuys Publishers, pp 463-477.
Steinbauer, M. J., Clarke, A. R., Madden, J. L., 1998. Oviposition preference of a
Eucalyptus herbivore and the importance of leaf age on interspecific host
Steinbauer, M. J., Kriticos, D. J., Lukacs, Z., Clarke, A. R., 2004. Modelling a
forest lepidopteran: phonological plasticity determines voltinism which
influences population dynamics. Forest Ecology and Management 198, 117-
131.
Stone, C., 1993. Insect problems of eucalypt plantations in Australia. Australian
Forestry 56, 363-369.
Stone, C., Coops, N., 2004. Assessment and monitoring of damage from insects in
Australian eucalypt forests and commercial plantations. Australian Journal of
Entomology 43, 283.
Stone, C., Bacon, P. E., 1995. Leaf dynamics and insect herbivory in a Eucalyptus
camaldulensis forest under moisture stress. Australian Journal of Ecology 20,
473-481.
Stone, C., Matsuki, M., Carnegie, A., 2003. Pest and disease assessment in young
eucalypt plantations: field manual for using the Crown Damage Index, editor
Parsons, M. (Ed.) Canberra, National Forest Inventory, Bureau of Rural
Sciences.
Strauss, S.Y., 2001. Benefits and risks of biotic exchange between Eucalyptus
plantations and native Australian forests. Austral Ecology 26, 447-457.
Tauber, M.J., Tauber, C.A., 1976. Insect seasonality, diapause maintenance,
termination and post-diapause development. Annual Review of Entomology
21, 81-107.
Tauber, M. J., Tauber, C.A., Masaki, S., 1986. Seasonal adaptations of insects.
Oxford University Press, New York. 411pp.
Tribe, G.D., 2000. Ecology, distribution and natural enemies of the Eucalyptus-
defoliating beetle Trachymela tincticollis (Blackburn) (Chrysomelidae:

Chrysomelini: Paropsina) in southwestern Australia, with reference to its
biological control in South Africa. African Entomology 8, 23-45.
Withers, T.M., 2001. Colonisation of eucalypts in New Zealand by Australian
insects. Austral Ecology 26, 467-476.
Wylie, F.R., Peters, B.C., 1993. Insect pest problems of eucalypt plantations in
Australia. 1. Queensland. Australian Forestry 56, 358-362.

Chapter 7
General Discussion
Thesis summary
In this thesis, different aspects of genetic and phenotypic variation within Paropsis
atomaria were investigated as an approach to assess if this economically important
taxon represented a single genetic species or a complex of undetected cryptic species.
Comparisons of work by previous authors had revealed evidence for differences in
life history characteristics among individuals sourced from geographically distant
populations, including different optimum development temperatures and seasonal
phenologies. Also, wide host range suggested that P. atomaria was unusual among
the Chrysomelidae, a group that predominantly have narrow host ranges.
Consequently, as P. atomaria is an emergent pest of forestry, it was important to
determine if populations from different parts of the natural geographical range, or
that use different host plants in the same location, were conspecific. Determining this
would therefore permit the use of all historical data accumulated for P. atomaria,
regardless of from where the data was collected.
Detailed empirical studies were used to investigate underlying population structure

of the species, morphological variation between populations, developmental
responses with respect to host plant utilisation, and population phenology across the
range, to determine if cryptic species were present, and if not, what explained
intraspecific variation in P. atomaria. First, molecular techniques were used to
indicate that P. atomaria, as currently defined, is essentially a single gene pool with
partial geographical isolation among populations from the northern extent of the
range compared with the southern extreme (Chapter 3). Second, while morphometric
analysis further supported the single taxon hypothesis (see Appendix 1), variation in
body size was present among individuals, with body size correlated with latitude in
accordance with a converse Bergmann cline. Some of this variation did have a
genetic component (Chapter 4), reinforcing the recognition of some intraspecific
genetic differentiation between southern and northern populations. Third, variation in
the insect’s biotic environment (specifically host plant) also influenced body size
differentially between populations (Chapter 5), but that populations did not differ in
132 Chapter 7: Final discussion

their overall performance when reared on two eucalypt species. Finally, experimental
data and computer modelling prediction reveal voltinism to be a plastic trait largely
mediated by temperature (Chapter 6). Consequently, each line of investigation,
whether genetic, morphological, physiological, or developmental, shows no evidence
of a cryptic species complex within P. atomaria. Furthermore, preliminary mating
trials (see Appendix 2) further demonstrated a lack of evidence for different species
within the taxon, as individuals from the extent of the range recognised each other as
mates under artificial conditions, and hence the results are uninterpretable (see
Walter, 2003). Phenotypic variation in body size, developmental characteristics, and
life histories is therefore a complex interaction between genotype, and both abiotic
(climatic) and biotic (host plant) factors. Specifically, temperature imposed a direct
effect on individual body size as explained by the Temperature-Size rule, whereas
latitudinal variation in body size is most likely an adaptive response to season length.
Moreover, the nature of phenotypic plasticity differed among individuals from
different localities, with host plant-related pupal mass variation for northern
individuals affected more so than for southern populations, however seasonal
plasticity in voltinism was unchanged across the distribution. Therefore, phenotypic
differences in P. atomaria are intraspecific, and not evidence of a cryptic species
complex.
These data show that restriction of gene flow operates in concert with phenotypic
differentiation among populations. The temptation can be, therefore, to invoke local
adaptation as the chief explanatory mechanism for observed variation across P.
atomaria’s wide geographical distribution. The data presented here reveal, however,
how multiple intimate interactions between the organism and its environment
produce a more complex scenario, and that alternative explanations to local
adaptation exist. As a consequence, the evidence both for and against local
adaptation explaining the observed patterns will be developed below, taking into
consideration important factors that could also contribute to the variation reported
here.
Local adaptation
Local adaptation can be defined as the pattern of resident genotypes within a
population possessing, on average, higher relative fitness within their local habitat
compared with genotypes originating from other habitats (Kawecki and Ebert, 2004).
Chapter 7: Final discussion 133

As the habitat within which an individual resides can directly influence the relative
fitness of its specific genotype (Falconer, 1952; Via and Lande, 1985), and the
process by which genotypes vary between environments is largely, but not
exclusively, the result of selective processes operating on different local populations
(Mayr, 1942), local genotypes are predicted to perform less well when removed from
their natal habitat and placed in a foreign environment. Importantly, however, local
adaptation by populations to their immediate environment is not a simple, inevitable
process (Walter, 2003). While directional selection or random genetic drift may
affect organisms, those individuals are also invariably under strong stabilising
selection for other traits, such as the necessity to maintain a specific fertilisation
system within a given environment (Paterson, 1993). Thus, local adaptation, if it
occurs, should therefore be viewed as a process brought about in response to strong
localised selective pressures imposed on a geographically constrained population.
Issues pertinent to local adaptation in phytophagous insects, such as P. atomaria, are
discussed below.
Selective pressures relevant to phytophagous insects
Phytophagous insects are, by definition, ectothermic animals that feed on plants.

Consequently, two important factors are immediately apparent to determining
variation and potential local adaptation in phytophagous insects, and these are:
climate (abiotic) and the host plant (biotic). The increased abundance of
phytophagous insects compared with their non-plant feeding relatives is testament to
the importance of the variety of plant life in promoting such diversity among plant-
feeding insect clades (Jaenike, 1990), while the importance of temperature and
relative humidity in affecting insect fitness is also widely acknowledged (Chown and
Gaston, 1999).
Host plant specialisation is widespread across phytophagous insect groups, with

generalist species often consisting of either locally specialised populations (Fox and
Morrow, 1981) or potentially cryptic species complexes (Clarke and Walter, 1995).
Further, the relative dietary breadth of a species (from monophagous to polyphagous)
is considered, by some, to be important in affecting the relative rate of gene flow
among populations – specialist phytophagous insects are hypothesised to show more
restricted gene flow compared with generalist species, and hence be more prone to
speciation (Futuyma and Moreno, 1988; however see contrary evidence proposed by

Peterson and Denno, 1998). Considering the wide host range, the potential for host
specialisation in P. atomaria would be predicted by the argument above.
Climatic variables including season length, temperature, humidity, and photoperiod

can all impose marked effects on insect behaviour, physiology, and related ecological
characteristics (Chown and Gaston, 1999; Chown and Klok, 2003; Inoue, 2001;
Sibly and Atkinson, 1994). As a consequence, seasonal climatic fluctuations (both
long term predictability and short term local unpredictability) have the potential to
impose pressure therefore on regional populations of non-migrating insects to
complete development and become reproductively active within a given season,
thereby resulting in local adaptation to proximate climatic factors (Blanckenhorn and
Fairbairn, 1995). The extensive range of P. atomaria over a wide latitudinal gradient
therefore implies that local adaptation to seasonal pressures is likely to cause
adaptive differentiation.
As previously emphasised, strong directional selective forces are required for

significant ecological change to take place within populations. Without biotic and
abiotic environmental influences resulting in effective differential fitness between
genotypes, there is no impetus for local adaptation to occur, with local adaptation
more likely in cases of strong selection against genotypes adapted to alternative
environments (Kawecki and Ebert, 2004).
Is there evidence of local adaptation in P. atomaria?
Gene flow – One of the principle considerations for determining the potential for
local adaptation is the degree of gene flow among populations (Ballabeni et al.,
2003), with restricted gene flow considered necessary for local adaptation to occur,
as any potential for fixation of locally advantageous traits will generally be swamped
by high levels of migration between populations (Kawecki and Ebert, 2004; Peterson
and Denno, 1998). Factors affecting gene flow can include: geographic distance
between populations, dispersal capability, ecological specialisation, phenological
isolation, habitat patchiness, habitat persistence, and the frequency and nature of
extinction / re-colonisation events (Hastings, 1983; Peterson and Denno, 1998). For
P. atomaria, each of these influences may contribute differentially to affecting gene
flow between populations, and consequently are of varying significance.

The effect of geographical distance on gene flow is tied intimately with the dispersal
capabilities of a species. For those taxa with great dispersal potential over wide
geographical distances, the rate of gene flow among populations is large; however
sedentary species are likely to exhibit heightened genetic differentiation at a much
smaller spatial scales (Peterson and Denno, 1998). Paropsis atomaria population
structure revealed higher levels of gene flow among northern populations compared
with gene flow between northern and southern populations (Chapter 3). This may be
due to isolation by distance, as dispersal by P. atomaria is considered relatively poor
(Carne, 1966). Furthermore, the accuracy of our computer-generated phenology
model, ParopSys, provides additional support for poor dispersal within P. atomaria
and consequently low gene flow between populations (Chapter 6). If, for example,
migration rates between populations during a given season were high, we would
expect greater stochasticity in the predictions relative to field validation results.
Instead, predicted and observed phenologies are tightly correlated, indicating a single
cohort of individuals remains at the same location for the duration of the season.
There is little evidence of ecological specialisation affecting gene flow in P.

atomaria, as shown by a lack of significant population genetic association with host
plant species, together with similar host utilisation characteristics for northern and
southern populations. Therefore, the effects of geographical distance and dispersal
capability are considered the principle determinants of gene flow between
populations of P. atomaria. Thus, limited gene flow between geographically distant
populations does provide the potential for local adaptation to occur.
Genetic maintenance of phenotypic variation – Local adaptation is often considered

to be an alternative explanation to phenotypic plasticity (Kawecki and Ebert, 2004).
Whilst local adaptation is the fixation within a population of specific genetically
determined traits in response to consistent environmental selective pressures,
phenotypic plasticity enables individuals to express a range of differential phenotypic
traits in direct response to specific environmental conditions (Nylin and Gotthard,
1998; Via et al., 1995). Further, phenotypic plasticity may be favoured in species that
reside within heterogeneous environments that experience significant temporal
variation (Kawecki and Ebert, 2004; Via and Lande, 1985). As a consequence, local
adaptation may be more likely for species in situations with little temporal variation,
but greater spatial diversity across habitats.

For P. atomaria, trials investigating effect of temperature on body size clearly
demonstrated a genetic component to body size that correlated with latitudinal
variation in the form of a converse Bergmann cline (Chapter 4). This result is
supportive of local adaptation over phenotypic plasticity. Considering the
environments within which P. atomaria lives, local adaptation to seasonal constraints
is considered possible as there is comparatively little temporal variation in a given
habitat between years, but greater spatial variation between habitats, particularly with
respect to latitude-associated season length.
Affinity for sympatric host plants – Given reduced gene flow and partial isolation
among P. atomaria populations, local adaptation within a population may
theoretically result in increased fitness of individuals to their local food source, as
local specialisation to proximate biotic resources, such as host plants, is often
considered likely for generalist herbivores (Fox and Morrow, 1981). Furthermore,
given strong directional selective pressures, populations may become rapidly adapted
to their proximate environment (Carroll et al., 2001; Lee, 2002). The Soapberry Bug,
Jadera haematoloma, for example, is hypothesised to consist of two recently
diverged (~100 generations) races in Florida, U.S.A., one constituting individuals
adapted to their original hosts (Balloon Vine, Cardiospermum corindum), and
another geographically isolated population that adapted rapidly to the introduced
southeast Asian Goldenrain Tree, Koelreuteria elegans (Carroll and Boyd, 1992;
Carroll et al., 2001). Such adaptation was explained as strong directional selection
for beak length in order to access seeds within fruit of varying radii; long beaked
individuals were found on thicker native fruits, with short beaked individuals on
small introduced fruits. Furthermore, there was evidence for evolved trade-offs, with
each population performing less well on non-natal hosts during reciprocal rearing
experiments on the two host plants (Carroll et al., 2001). Trade-offs, while often
considered important during local adaptation are, however, not always apparent –
especially in studies of host plant specialisation as phytophagous insects are often
shown to perform better on novel hosts compared with natal species (Joshi and
Thompson, 1995).
For P. atomaria, there is no evidence for local host plant specialisation. No

significant association was detected between haplotype diversity and host plant of
origin (Chapter 3), and while northern populations reach greater adult body sizes,

with reduced development time, and having higher survival rate on the sympatric
host species (Eucalyptus cloeziana), this host is also better for the southern
population of P. atomaria (compared to E. pilularis) (see Chapter 5). Indeed there is
little reason to consider P. atomaria to have become locally specialised on any host
plant species, as the influence of host plant species and their respective geographical
distributions are unlikely to be an influential selective pressure for P. atomaria.
While the geographic distributions of different eucalypt host species utilised by P.
atomaria are fragmented – a condition favouring local host plant specialisation in
phytophagous insect species (Fox and Morrow, 1981) – host quality seems less likely
to depend on which species is utilised by the insect, but, similarly to other Australian
paropsines, more so the quality of foliage present as defined by age, toughness, and
nitrogen concentration (Ohmart et al., 1987; Steinbauer et al., 1998). Indeed, P.
atomaria individuals seem capable of feeding on many eucalypt species due to a
capacity for metabolising defensive terpenoids (major secondary compounds
contained within Eucalyptus foliage) (Ohmart and Larsson, 1989) and as such are
capable of feeding on most eucalypt species. Nevertheless, northern beetles are
significantly larger than southern individuals when reared on the superior quality
host (E. cloeziana); a difference not apparent when individuals from both
populations are reared on the poorer host, E. pilularis. This may be due to the poorer
nutritional quality of the E. pilularis preventing northern beetles from maximising
their pupal weight given a better host (such as E. cloeziana). Southern beetles may,
however, be incapable of growing larger even when presented with a better quality
host, as revealed by their failure to attain a higher pupal mass when reared on the
superior host, E. cloeziana – a circumstance potentially due to genetic constraints
preventing individuals from growing larger and possibly due to local adaptation to
season length (Chapter 5). In summary, there seems little evidence to suggest local
adaptation to sympatric host plant species in P. atomaria.
Summation – For this thesis, I have demonstrated that there is no evidence to indicate
P. atomaria constitutes a cryptic species complex, and that variation in a simple
observable character like body size appears to be under the influence of both genetic
and environmental factors. Furthermore, underlying causes of intraspecific variation
have potentially very different origins. Adult body size in P. atomaria is likely a by-
product of both genetic determinants and developmental processes that occur during

larval development. Smaller adult body size, for instance, results from poor nutrition
or high stress during development, or reduced time (i.e., short season) to reach
maturity. Paropsis atomaria body size changes as a consequence of direct
temperature effect, with a negative correlation between increasing rearing
temperature and adult body size (plastic effect) (Chapter 4, page 74, Fig. 6).
However, as larger P. atomaria occur at lower latitudes (Chapter 4, page 69, Fig. 3),
body size also has a genetic component that I postulate to be explained by local
adaptation to season length. Furthermore, body size in P. atomaria is influenced
additionally by host plant quality (Chapter 5, page 91, Fig. 3), which imposes a
differential effect on host plant utilisation among P. atomaria populations (the
northern population responds differently compared with the southern population) and
this represents a combination between environmental and genetic effects on
development and ultimately body size in this beetle with respect to host plant.
Finally, seasonal plasticity was demonstrated within P. atomaria, and that it is
largely a function of direct temperature and seasonal effect imposing limits to the
number of generations possible within a field season. Thus, in this widely dispersed
beetle, with limited gene flow between disjunct populations, variation in
morphological and physiological traits are a combination of local adaptation,
phenotypic plasticity, and phenological plasticity; not just local adaptation alone.
This study emphasises the need to look more critically at within species variation and
not to automatically assume local adaptation as an explanation for all variation
observed between geographically and genetically disjunct populations.
Implications for pest management

The work described here has a number of potentially important considerations for
future management of P. atomaria populations, as some ecological characteristics
are shared across populations, while other characteristics are different among them.
Such observations are relevant to considerations pertaining to insecticide resistance,
population modelling, and host susceptibility. Most importantly, however, the
knowledge that P. atomaria constitutes a single species with largely shared
biological attributes across the distribution means historical research into the taxon
may be applied to populations across its geographical range.
The ability of local populations to disperse, together with associated probability of

local adaptation, has practical applications for pest management. Resistance to

control measures (e.g., insecticide resistance), is often considered a form of
adaptation by populations to local selective pressures (i.e., chemical pesticides), and
by understanding relative rates of dispersal and gene flow, we can better predict the
potential for resistance to evolve. By examining the population genetic structure of
the destructive Tasmanian forestry pest, Chrysophtharta bimaculata, for example, it
was revealed that due to high levels of gene flow among populations across the
island, localised resistance was unlikely if low levels of control agents were applied
(Congdon et al., 1997). Similarly for P. atomaria, a greater understanding of gene
flow in the species may provide insight into the potential for resistance to build up
(i.e., high levels of gene flow may reduce the probability of localised resistance).
Such assumptions must, however, be approached with due caution, as resistance to
insecticides may occur through relatively minor changes involving single allelic
substitutions, and hence be easily lost due to disadvantageous pleiotropic effect
(Walter, 2003).
The consistency of developmental characteristics as evidenced by comparisons

between northern and southern populations shows that populations have not evolved
different phenologies. The application of this knowledge, combined with the
understanding phenology to be largely under the influence of ambient temperature
therefore permits accurate modelling of the life cycle for P. atomaria across the
season (see Chapter 6). The input of further information into the model, such as
pesticide application events, may therefore be undertaken with confidence, as any
predictions can be applied across the species’ range.
Finally, evidence of differential host plant utilisation by different populations across

the geographical distribution of P. atomaria is noteworthy. As northern beetles are
genetically predisposed to reach a greater body size given a better quality host plant,
this may have implications for the plantation of specific eucalypt host species in
northern regions of the distribution. The continued planting of better hosts in lower
latitudes may allow northern individuals to reach larger sizes, which may increase
their potential fecundity (see Carne, 1966). Such increased fecundity and subsequent
population number consequently elevates the damage potential of P. atomaria, and
can affect population density throughout the season. This prediction could be tested
using the current phenology computer model, ParopSys. The knowledge that some
host species are poorer hosts and will reduce potential for large bodied northern

populations to reach their maximum size may therefore emerge as an effective future
management option for the eucalypt hardwood industry.

References
and ecological genetics of host-plant specialization in a leaf beetle. Oikos
101, 70-78.
Blanckenhorn, W. U., and Fairbairn, D. J. (1995). Life history adaptation along a
latitudinal cline in the water strider Aquarius remigis (Heteroptera: Gerridae).
Journal of Evolutionary Biology 8, 21-41.
Carne, P. B. (1966). Ecological characteristics of the eucalypt defoliating
672.
Carroll, S. P., and Boyd, C. (1992). Host race radiation in the Soapberry Bug:
Natural history with the history. Evolution 46, 1053-1069.
Carroll, S. P., Dingle, H., Famula, T. R., and Fox, C. W. (2001). Genetic architecture
of adaptive differentiation in evolving host races of the soapberry bug, Jadera
haematoloma. Genetica 112-113, 257-272.
Chown, S. L., and Gaston, K. J. (1999). Exploring links between physiology and
ecology at macro-scales: the role of respiratory metabolism in insects.
Biological Reviews of the Cambridge Philosophical Society 74, 87-120.
Chown, S. L., and Klok, C. J. (2003). Altitudinal body size clines: latitudinal effects
associated with changing seasonality. Ecography 26, 445-455.
Clarke, A. R., and Walter, G. H. (1995). "Strains" and the classical biological control
of insect pests. Canadian Journal of Zoology 73, 1777-1790.
Congdon, B. C., Lange, C. L., and Clarke, A. R. (1997). Geographical variation and
gene flow in the eucalyptus defoliating beetle Chrysophtharta bimaculata
(Coleoptera: Chrysomelidae). Journal of Applied Ecology 34, 1287-1292.
Falconer, D. S. (1952). The problem of environment and selection. The American
Naturalist 86, 293-298.
Fox, L. R., and Morrow, P. A. (1981). Specialization: Species property or local
phenomenon? Science 211, 887-893.
Futuyma, D. J., and Moreno, G. (1988). The evolution of ecological specialization.
Annual Review of Ecology and Systematics. 19, 207-233.
Hastings, A. (1983). Can spatial variation alone lead to selection for dispersal?
Theoretical Population Biology 24, 244-251.

Inoue, T. (2001). Life history of the flea beetle, Argopistes coccinelliformis Csiki
(Coleoptera: Chrysomelidae) VII. Effects of photoperiod and temperature on
induction of reproductive diapause in newly emerged adults. Applied
Entomology and Zoology 36, 53-58.
Kawecki, T. J., and Ebert, D. (2004). Conceptual issues in local adaptation. Ecology
Letters 7, 1225-1241.
Lee, C. E. (2002). Evolutionary genetics of invasive species. TREE 17, 386-391.
Mayr, E. (1942). "Systematics and the Origin of Species," Colombia University
Press, New York.
Nylin, S., and Gotthard, K. (1998). Platicity in life-history traits. Annual Review of
Ohmart, C. P., and Larsson, S. (1989). Evidence for absorption of eucalypt essential
oils by Paropsis atomaria Olivier (Coleoptera: Chrysomelidae). Journal of
the Australian Entomological Society 28, 201-205.
Ohmart, C. P., Thomas, J. R., and Stewart, L. G. (1987). Nitrogen, leaf toughness
and the population dynamics of Paropsis atomaria Olivier (Coleoptera:
Chrysomelidae) - A hypothesis. Journal of the Australian Entomological
Society 26, 203-207.
Paterson, H. E. H. (1993). "Evolution and the recognition concept of species.
Collected writings," The Johns Hopkins University Press, Baltimore.
Peterson, M. A., and Denno, R. F. (1998). The influence of dispersal and diet breadth
on patterns of genetic isolation by distance in phytophagous insects. The
American Naturalist 152, 428-446.
Sibly, R. M., and Atkinson, D. (1994). How rearing temperature affects optimal adult
size in ectotherms. Functional Ecology 8, 486-493.
Steinbauer, M. J., Clarke, A. R., and Madden, J. L. (1998). Oviposition preference of
a Eucalyptus herbivore and the importance of leaf age on interspecific host

Via, S., Gomulkiewicz, R., De Jong, G., Scheiner, S. M., Schlichting, C. D., and Van
Tienderen, P. H. (1995). Adaptive phenotypic plasticity: consensus and
controversy. Trends in Ecology and Evolution 10, 212-217.
Via, S., and Lande, R. (1985). Genotype-Environment Interaction and the Evolution
of Phenotypic Plasticity. Evolution 39, 505-522.

Appendix 1
Morphometric study of Paropsis atomaria (supplement to Chapter 4)
Introduction
Morphometric variability between populations is a traditional technique used to
identify cryptic species complexes, and while the advent of molecular approaches
has revolutionised cryptic complex studies, morphometric techniques continue to be
applied (Hoberg et al, 1999; Bain et al., 2003). However, while molecular
approaches have become affordable, morphometric analysis remains a key tool in
cryptic species identification due to its ease of application and minimal requirement
for specialised tools and training. Consequently, a morphometric analysis was
undertaken as an additional approach towards identifying whether Paropsis atomaria
constituted a cryptic species complex. Results of this analysis demonstrate no clear
evidence for a cryptic species complex; however latitudinal body-size trends
revealed a converse Bergmann cline which is detailed in Chapter 4 of this thesis.
Fifteen body measurements (Fig 1, Table 1) were selected for examination across P.
atomaria individuals collected from the following four locations in Australia:
Lowmead, central Queensland; Beerburrum, south-east Queensland; Bangalow,
north-east NSW; and Canberra, ACT. See Chapter 4 for measurement methodology.
Also included for comparison were samples of P. deboeri, a Tasmanian species
closely allied with P. atomaria (see Chapter 2).
Discriminant function analysis in SPSS was used to examine the fifteen body
measurements and to make a qualitative assessment of morphometric similiarity
among populations.
Appendix 1: Supplementary morphometric data 145

Figure 1. All measurements take for Paropsis atomaria individuals collected from
the field and reared in common-garden trials.
146 Appendix 1: Supplementary morphometric data

Table 1: Fifteen measurements selected for morphometric study of Paropsis
atomaria individuals collected from four sites.
Abbreviation Character description

AN Length of first antennal segment
AT Distance between base of antennae
CL Length of clypeus
CW Width of clypeus
EL Length of elytra
EW Width of both elytra (maximum width of body)
EY Distance between eyes
FL Length of fore tibia
HL Length of hind tibia
ML Length of mid tibia
PD Length of pedicel
PL Length of pronotum
PW Width of pronotum
TL Total length from tip of head to tip of elytra/abdomen
Results and Discussion

Table 2 shows the measurements for each of the body measurements of P. atomaria
samples from each location (mm, means and standard deviations). Note that males
and females are analysed separately due to significant differences in size between the
sexes.
Discriminant function analysis (Fig 2) reveales no clear evidence for any population
of P. atomaria emerging as different from any other. While each population occupies
a slightly different region in space, there is significant overlap between them. This is
in contrast to the close relative, P. deboeri, which is clearly different from all
populations of P. atomaria (Fig 2). Note, however, that only multiple samples of
female P. deboeri were examined, with only a single male specimen available. This
lack of P. deboeri specimens for the male analysis may explain the increased
disparity between populations seen for the male results, as a similar pattern emerges
for females when P. deboeri samples are removed from the female analysis (results
not shown).
Due to the overlap between populations in the discriminant analysis, there is

insufficient evidence for cryptic species based on a morphometric approach, however
as outlined above, the latitudinal gradient trend is elaborated upon in Chapter 4.

SC PD AN EY AT CW CL PW PL EW EL TL FL ML HL
Sample Sex N mean±SD mean±SD mean±SD mean±SD mean±SD mean±SD mean±SD mean±SD mean±SD mean±SD mean±SD mean±SD mean±SD mean±SD mean±SD
BBR M 12 0.64±0.06 0.27±0.03 0.43±0.05 2.01±0.14 1.62±0.11 1.35±0.09 0.64±0.06 5.45±0.48 2.40±0.21 7.70±0.70 7.07±0.58 10.15±0.89 1.98±0.19 2.15±0.19 2.38±±0.18
F 17 0.71±0.04 0.29±0.02 0.48±0.04 2.25±0.11 1.81±0.09 1.52±0.07 0.71±0.07 6.10±0.28 2.63±0.17 8.61±0.46 9.12±0.67 12.08±0.78 2.26±0.11 2.44±0.18 2.67±0.13
BAN M 14 0.62±0.05 0.28±0.02 0.46±0.04 2.04±0.09 1.64±0.06 1.35±0.05 0.64±0.03 5.39±0.25 2.34±0.13 7.51±0.31 7.16±0.29 10.34±0.47 2.04±0.11 2.21±0.14 2.43±0.14
F 11 0.72±0.04 0.29±0.02 0.49±0.05 2.25±0.10 1.80±0.08 1.50±0.05 0.68±0.04 5.93±0.30 2.56±0.11 8.39±0.28 8.95±0.50 12.51±0.71 2.25±0.15 2.48±0.17 2.66±0.17

LOW M 22 0.68±0.05 0.28±0.02 0.48±0.02 2.24±0.11 1.78±0.09 1.46±0.07 0.66±0.15 5.96±0.33 2.57±0.14 8.31±0.42 7.81±0.73 10.94±0.73 2.15±0.13 2.28±0.16 2.51±0.17
F 12 0.74±0.03 0.30±0.01 0.51±0.01 2.42±0.08 1.93±0.06 1.60±0.04 0.73±0.05 6.53±0.15 2.80±0.14 8.96±0.21 9.75±0.46 13.00±0.50 2.36±0.12 2.54±0.11 2.76±0.12
CAN M 15 0.64±0.05 0.28±0.02 0.44±0.04 2.00±0.11 1.58±0.09 1.31±0.07 0.61±0.04 5.13±0.28 2.21±0.12 7.06±0.58 7.27±0.64 9.99±0.97 1.93±0.15 2.12±0.16 2.33±0.16
F 15 0.69±0.03 0.29±0.03 0.48±0.03 2.21±0.08 1.78±0.06 1.49±0.05 0.66±0.04 5.63±0.27 2.45±0.11 7.90±0.44 8.44±0.45 11.59±0.55 2.09±0.10 2.32±0.15 2.51±0.15
Table 2. Measurements for each of the fifteen selected measurements (mm

± standard deviation) for the morphometric analysis of Paropsis atomaria
collected from the following four sites: Beerburrum, Qld (BBR);
Bangalow, N.S.W. (BAN); Lowmead, Qld (LOW); and Canberra, A.C.T.
(CAN).
Figure 2. Canonical discriminant results for morphometric data based on fifteen
measurements taken from female (top) and male (bottom). Paropsis atomaria
samples sourced from four collection localities (Beerburrum, Qld; Bangalow,
N.S.W.; Lowmead, Qld; and Canberra, A.C.T.) and also for individuals of P.
deboeri (close relative of P. atomaria) from Tasmania. Note that only a
single male specimen was available for P. deboeri analysis.

References
Bain R.H., Lathrop A., Murphy R.W., Orlov N.L., and Cuc H.T. (2003). Cryptic
Species of a Cascade Frog from Southeast Asia: Taxonomic Revisions and
Descriptions of Six New Species. American Museum Novitates 3417(1): 1-
60.
Hoberg E.P., Monsen K.J., Kutz S., and Blouin M.S. (1999). Structure, Biodiversity,
and Historical Biogeography of Nematode Faunas in Holarctic Ruminants:
Morphological and Molecular Diagnoses for Teladorsagia boreoarcticus n.
sp. (Nematoda: Ostertagiinae), a Dimorphic Cryptic Species in Muskoxen
(Ovibos moschatus). The Journal of Parasitology 85(5): 910-934.

Appendix 2
Cross mating trial between Canberra (A.C.T.) and Lowmead (Qld)

individuals
Introduction
Whether individuals accept each other as mates can contribute towards assessing the
species status of a study taxon. Consequently, mating trials investigating the potential
for individuals to mate can then be regarded as a key component to cryptic species
identification. The problem arises, however, in the interpretation of results. Very few
experimental outcomes are capable of unequivocally determining species status
given the recognition concept of species (Walter, 2003). Due to the confounding
effects of laboratory conditions, only in a case of a) controls mating and b) crosses
not mating, can relative confidence be assigned to the existence of multiple
biological species. If, however, controls and crosses freely accept each other as
mates under laboratory conditions, a case for or against cryptic species can not be
resolved, regardless of the viability of resultant offspring (Walter, 2003).
Consequently, as cross-mating trials between geographically disjunct populations of
Paropsis atomaria revealed both control and experimental crosses mated freely,
there is no evidence of a cryptic species.

Individuals collected from two localities (Canberra, A.C.T. and Lowmead, Qld) were
paired in plastic containers to observe mating behaviour. The following pairings
were made: Canberra male X Canberra female; Canberra male X Lowmead female;
Lowmead male X Lowmead female; and Lowmead male X Canberra female.
Pairs were observed for 300 minutes at a time, recording behaviour every five
minutes. Behaviour was recorded as either: no mating behaviour; mounting; or
copulation. Copulation was recorded when the male genitalia was clearly inserted
into the female, whereas mounting was recorded as when the male was mounted but
with no copulation. Only total duration of copulation is presented (Table 1).
Appendix 2: Mating trials 151

Results and Discussion
Table 1 reveals that for each cross, whether control or experimental, there was
acceptance between males and females as potential mates with at least some pairs
engaging in extended periods of copulation. Consequently, as both control and
experimental crosses were successful, results can not be interpreted to either support
or refute a case of a cryptic species complex within P. atomaria. While experimental
crosses do mate under laboratory conditions, field conditions may pose a very
different reality as other elements of the currently largely unknown mate recognition
system may act to segregate populations and result in assortative mating within their
natural environment. Evidence for a cryptic species complex would exist should a
lack of mate recognition between populations occur in the wild; however, laboratory
trials as presented fail to do so.
Table 1. Total duration of copulation (minutes) between paired Paropsis atomaria

individuals from one of two source populations (Can = Canberra, A.C.T.; Low =
Lowmead, Qld). Numbers with ‘+’ denote those pairs that were still copulating at the
termination of the experiment (observations made over 300 minutes). Total number
of replicates for each trial (including those that did not mate) given in parentheses.
Can X Can Can X Low Low X Low Low X Can

(n = 17) (n = 17) (n = 21) (n = 8)
85 140+ 200 165+
85 145 105 140+
145 300+ 265+ 135
175 95 130 185
10 185 165 285+
80+ 130 25
180 190
45 200
35 185
250
180
References
152 Appendix 2: Mating trials

Mark Kurt Schutze Thesis 2

Uploaded by

Copyright:

Available Formats

Mark Kurt Schutze Thesis 2

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mark Kurt Schutze Thesis 2

Uploaded by

Copyright:

Available Formats

The significance of genetic and ecological

diversity in a wide-ranging insect pest,

Mark Kurt Schutze

B.Sc. (Hons), University of Queensland

School of Natural Resource Sciences

Queensland University of Technology

Gardens Point Campus

This thesis is submitted as a requirement for the degree of Doctor of Philosophy

Mitochondrial sequence variation of the gene COI was compared between

Molecular results demonstrated relatively low genetic divergence between

List of publications.............................................................................................. viii

Statement of original authorship ....................................................................... ix

CHAPTER 1: General introduction and literature review............................. 1

Patterns and processes of genotypic and phenotypic variation

Causes of intraspecific variation ............................................................ 6

Study system ......................................................................................................... 9

CHAPTER 2: Literature review on P. atomaria .............................................. 21

Notes on taxonomy and nomenclature.................................................................. 22

Biology and ecology ............................................................................................. 22

Notes on host range............................................................................................... 24

Similarities to other species .................................................................................. 30

Detection and inspection methods ........................................................................ 30

Phytosanitary risk ..................................................................................................32

Means of movement and dispersal

Silvicultural practices .............................................................................32

Movement in trade ..................................................................................32

Notes on natural enemies ......................................................................................33

Biological control ...................................................................................34

Host plant resistance...............................................................................34

Pheromonal control ................................................................................34

Environmental impact ...........................................................................................36

CHAPTER 3: Species status and population structure of the

Materials and methods ..........................................................................................45

Insect collection and identification .........................................................45

Sequence variation ..................................................................................48

Variation between host plants within P. atomaria .................................. 53

CHAPTER 4: Converse Bergmann cline in a Eucalyptus herbivore,

Materials and methods .......................................................................................... 69

Body size of field collections................................................................... 69

Common-garden experiments ................................................................. 71

Temperature – latitude correlation......................................................... 72

Body size of field collections................................................................... 73

Common-garden experiments ................................................................. 75

CHAPTER 5: Larval development of two geographically isolated

Materials and methods .......................................................................................... 90

Field collection of adult beetles.............................................................. 91

Survival to pupation ................................................................................93

Pupal weight ...........................................................................................95

CHAPTER 6: Thermal requirements, field mortality and

Materials and methods ..........................................................................................107

Thermal requirements and thresholds for immature P. atomaria

Mortality of immature stages ..................................................................109

Field validation .......................................................................................114

Thermal requirements and thresholds for immature P. atomaria

Field mortality of immature life stages ...................................................117

Population modelling and field validation..............................................118

Development and mortality.....................................................................121

Future improvements .............................................................................. 123

Conclusion .............................................................................................. 124

Acknowledgements ............................................................................................... 124