Journal of Coastal Conservation (2022) 26: 24

https://doi.org/10.1007/s11852-022-00870-7

Assessment method of the sea turtle-nesting habitat of small reef

islands
Manuel Garcin1 · Myriam Vende-Leclerc2 · Tyffen Read3 · Marc Oremus4 · Hugo Bourgogne4

Received: 25 June 2021 / Revised: 2 May 2022 / Accepted: 10 May 2022 / Published online: 8 June 2022
© The Author(s) 2022

Abstract
The coral islets in the lagoon of New Caledonia are a major breeding and nesting site for various species. Many species
reproduce and nest there, whether they are Chelonians (green turtles and loggerhead turtles), offshore seabirds (Procel-
lariidae: shearwaters, petrels …) or more coastal (Laridae: terns, ospreys …) or sqamata (sea krait: Laticauda saintgironsi
and Laticauda laticaudata). These species are dependent on the persistence of the islets in which they find the necessary
conditions for nesting. Previous research has shown that these islets can evolve very rapidly at a yearly to decadal times-
cale. These geomorphological evolutions lead to surface area variations as well as geomorphological, topographical and
land cover changes. These changes may lead to an evolution of the sea turtle nesting habitat: reduction of suitable areas
for nesting and changes of the accessibility to the nesting sites. The goal of our work is to propose a method that allows
to determine, map and quantify the more or less favourable character of the islets coasts to the nesting of sea turtles. Using
several parameters, an indicator of the nesting attractiveness of coast is constructed and mapped. This method was applied
to 13 coral islets in the southern lagoon of New Caledonia. Comparison of the results with the location of sea turtle nests
acquired during the 2020 campaign attest the relevance of the method.

Keywords Sea turtle · Nesting habitat · Coral islet · Geomorphology · Sedimentology · indicator · South Pacific · New
Caledonia

Introduction

Problematic
Manuel Garcin
m.garcin@brgm.fr
Preferred sea turtles nesting sites are sandy backshores,
Myriam Vende-Leclerc
myriam.vende-leclerc@gouv.nc
which are very sensitive to environmental changes and
hydrometeorological events such as storms and cyclones.
Tyffen Read
tyffen_read@hotmail.com
Evolution of the environmental conditions (geodynamics,
wave climate, short and long term sea level rise, inten-
Marc Oremus
moremus@wwf.nc
sity and frequency of storms and cyclones, tidal and wave
induced currents) lead to morphological changes of the
1
BRGM, 3 avenue Claude Guillemin, 45060 Orléans Cedex 2, fore-beach, erosion of the beaches and the backshore (Ballu
France et al. 2011; Dickinson 1999; Faure et al. 2010; Ford and
2
DIMENC/SGNC, 1 Ter rue Unger, Vallée du Tir BP M2, Kench 2014; Ford 2012, 2013; Ouillon et al. 2010; Ramsay
98849 Nouméa Cedex, New-Caledonia 2011; Stoddard 1990; Testut et al. 2015; Woodroffe 2008;
3
BP 9424, 98807 Nouméa, New-Caledonia Xue 1997; Yates et al. 2013).
4
WWF, Parc Zoologique et Forestier Michel Corbasson, In the context of sea level rise induced by climate change,
BP692—98845 Nouméa, New-Caledonia coastline retreat and coastal erosion are expected conse-
quences Nicholls and Cazenave 2010; Nunn 2013; Nurse

13
24 Page 2 of 13 M. Garcin et al.

et al. 2014; Roy and Connel, 1991; Webb and Kench, 2010; 2006; IFRECOR.NC1, SHOM2 data). The Southern and
Woodroffe et al., 2008). It is in this perspective that some Northern parts of the lagoon are very wide, with distances
work has been carried out, to determine the impact of the between the barrier reef and Grande-Terre ranging from 20
latter on turtle nesting habitats (e.g. Fish et al. 2005, Long to 31 km. In the middle part of the island, the lagoon nar-
et al. 2011). rows to just 1.7 km.
New Caledonia is firstly of interest in terms of biodiver- There are two types of islands in the lagoon: (i) elevated
sity Bozec et al. 2005; Brischoux and Bonnet 2009; Debenay islands with rock outcrops basement; (ii) low-lying islets
and Fernandez 2009; Jimenez et al. 2011; Pandolfi and Bret- (cayes; Garcin et al., 2016) made up of bioclastic sands
agnolle, 2002) and is also a major breeding and nesting area associated with coral reef patches. The rocky islands vary in
for sea turtles in the South Pacific (Read et al. 2015). Turtles size and generally have a relatively abrupt landform, while
come to nest on the beaches of the Grande-Terre of New the low-lying islets are smaller in area and very low in alti-
Caledonia (Read et al. 2013) and the Loyalty Islands (Read tude (generally up to 2 m). Sea turtles’ nests are mainly
et al. 2015) but also on coral islets beaches (Mounier 2007; observed on beaches of the second type.
Oremus and Mattei 2017). The islets of Entrecasteaux and
Chesterfieds are nesting sites for green turtles (Chelonia
mydas) while those of the lagoon of Grande-Terre are fre- Method
quented by loggerhead turtles (Caretta caretta).
However, these islets are fragile, threatened and their Determining the turtles’ nesting habitat
morphological evolution and surface areas decrease over requirements
time (Garcin et al., 2016). These morphological evolutions
have an impact on the beaches’ topography as well as on the The loggerhead and green turtles, although of different sizes,
land cover (herbaceous, shrubby or wooded), the beachrock have comparable requirements in terms of nesting criteria.
outcropping, the backshore granulometry and accessibility. The first phase of study focuses on the analysis of islets’
All these factors contribute to the modification of the apti- characteristics in relation with turtles’ nesting habitat. This
tude of each islet to accommodate nesting sites for numer- concerns the geomorphology, sedimentology, land cover
ous species, such as chelonians or offshore birds. and the general context of the islet. Once the inventory of
Sea turtles do not nest on all available islets and seem to these factors is made, it is a matter of determining how to
favour specific areas of each islets used (Oremus and Mattei qualify them from direct observations (field data) and from
2017). This study will focus on identifying how the mor- indirect sources such as satellite images, aerial photographs
phological characteristics of the islets influence the nests’ or any other useful sources (geo-referenced databases). The
repartition. objective is to provide an indicator value for each factor
and then to carry out a multi-criteria analysis to determine a
Sites unique indicator of nesting conditions. Relevant indicators
defining their favourable, unfavourable or critical charac-
The Grande-Terre is New Caledonia’s main island (South teristics are selected. Methods for evaluating and mapping
Pacific, between 18° and 23°S and 164° and 167° E, inset these indicators are proposed.
Fig. 1). It is a long, narrow island 400 km in length and For each homogeneous islets’ coastline segment, a sim-
50 km in width (about 16,000 km2 in area) located on the ple multicriteria analysis of indicators is realized. This is
margin of the Coral Sea’s margin, halfway between Aus- comprised of five possible values from “very favourable”
tralia and Fiji. The Grande-Terre island is surrounded by to “very unfavourable”. The annual turtle nest inventories
a barrier reef (linear extent of about 1,600 km) enclosing made by biologists are put into perspective using the turtle
a large lagoon covering about 24,000 km2 (Bonnet et al. nesting indicator maps for each islet in order to validate the
2014). Fifteen thousand square kilometres of New Caledo- correlations and causality links.
nia’s lagoon have been listed as UNESCO World Heritage
Site since 2008 and New Caledonia’s entire maritime area
(1,300,000 km²) became the Coral Sea Nature Park in April
2014. The Western barrier reef is sub-continuous with few
passes, while the Eastern barrier is irregular with numerous
passes. The lagoon is relatively deep on the Eastern side
with depths greater than 30 m, while the Western lagoon is
1
shallow with depths under 15 m (Coutures, 2000; Flamand French Coral Reef Preservation Initiative.
2
Service Hydrographique et Océanographique de la Marine: the
French Navy’s hydrographic and oceanographic service.

13
Assessment method of the sea turtle-nesting habitat of small reef islands Page 3 of 13 24

Fig. 1 Location map

Identification and characteristics of turtles’ nesting Introduction

criteria
Turtles nest in areas above high tide level in areas covered

13
24 Page 4 of 13 M. Garcin et al.

by thin sands, without vegetation or with low-density herba- Beachrocks

ceous vegetation (Hays et al. 1995; Wood & Bjorndal, 2000;
Kamel and Mrosovsky 2004; Read et al. 2020). Sometimes Description.
some nests are observed behind shrubbery vegetation’s Numerous beachrocks are present around New Cale-
limit when its density is low. Nests are marked by an ellipti- donian islets. According to their morphology and intrinsic
cal shaped surface that was first dug and then covered and characteristics, beachrocks can constitute a constraint or an
stirred. Failed nesting attempts are indicated by a bowl or obstacle to turtle nesting. Based on our knowledge of New
funnel morphology. This shape corresponds to the begin- Caledonian islets (Garcin et al., 2016) we have identified
ning of nest excavation, which did not succeed and was eight beachrock types/characteristics useful to determine
therefore not plugged. their influence on turtle nesting (Table 1, Online Resource
The stopping of digging by the turtle can be linked to: (i) 1):
the sediment’s granulometry if the sediment is too coarse
in depth to be favourable for nesting; (ii) the presence of ● Slope in line with that of the beach and at the same topo-
roots in depth; (iii) an exogenous disturbance at the time of graphic level than the beach. In this case the beachrock
digging. does not constitute an obstacle to nesting;
In the event of failure, the turtle will try to return quickly ● Slope in line with that of the beach but with a higher
to lay eggs near or on another site of the same island or elevation than the beach level; the beach rock consti-
even on other islet (Beyneto & Delcroix, 2005; Murphy and tutes an obstacle to nesting depending on the difference
Hopkins-Murphy 1990). in elevation between the two;
Thus, criteria can be grouped into two families: ● Beachrock beds highly oblique relative to the actual
beach orientation. If this latter is not in relief it doesn’t
● Criteria defining the nesting site’s availability and constitute a constraint for nesting;
quality, ● The beachrock slope is inverted in relation to the beach
● Criteria defining the nesting site’s accessibility. slope: beachrock can constitutes an obstacle for nesting;
● The beachrock is multi-beds and stair-like: this case is
not necessarily an insurmountable obstacle for adults
A suitable surface area for turtle nesting beaching for nesting at high tide. Conversely, it consti-
tutes a trap for juveniles going to sea after hatching. This
Description. obstacle highly increases their mortality rate. In this spe-
To establish their nests, turtles need an available surface cific case, the beachrock does not prevent nesting but
area on the backshore of either bare sand or sand with a low drastically lowers its reproductive success ;
vegetation cover. The greater the area available, the greater ● Raised beachrock (relative to actual sea level): in this
the number of nests may be. If the surface area is too limited case, the beachrock is an insurmountable obstacle,
and the number of turtles is high, there is a risk that the nest ● Chaotic dismantled beachrock constituting an obstacle
digging will be done at the expense of a pre-existing one, for turtle nesting ;
resulting in the destruction of the latter. In this case, the rate ● Highly weathered and/or holed: is unfavourable for tur-
of hatchlings reaching maturity per number of nests will be tle nesting.
reduced dramatically.
The upper beach and backshore areas may decrease Table 1 Assessment of the impact of beachrock types on turtle nesting
if they are affected by erosion. On the coastal segments site accessibility
affected by erosion, one notes their disappearance and the Beachrock characteristics Nesting site Indi-
appearance and development of an erosion scarp (Garcin et accessibility cator
Value
al., 2016). When the erosion affects the entire islet coast,
No beachrock Very favourable 2
these areas can be completely destroyed. In contrast, some
Very oblique in relation to the shore Very favourable 2
islets have stable or accretionary segments. A gently sloping Inline slope without protruding Favourable 1
beach develops while the backshore area increases over one Inline slope with protruding Unfavourable -1
or more years. These new areas are favourable for nesting. Inverted slope Unfavourable -1
Finally, it should be pointed out that certain islet segments Stair-like multi-beds Unfavourable -1
are affected, over several years or even decades, by alternat- Weathered and holed Unfavourable -1
ing phases of erosion and accretion. In this case, the attrac- Dismantled Highly -2
tiveness in terms of nesting will vary according to these unfavourable
phases. Raised Highly unfavourable -2
(inaccessible)

13
Assessment method of the sea turtle-nesting habitat of small reef islands Page 5 of 13 24

The beachrocks have various ages from millennia to several Table 2 Assessment of the impact of erosion scarp on turtle nesting
site accessibility
centuries. Due to the islets shape‘s changes in relation to
Erosion scarp height Nesting site accessibility Indicator value
erosion and accretion processes, beachrocks can appear dur- (m) Green Logger- Green Logger-
ing erosion phases, and conversely, they can be covered by turtle head turtle turtle head
sand during accretion phases. These changes can occur rap- turtle
idly (few hours to few days) during a single energetic hydro- No erosion scarp Easily Easily 2 2
meteorological event (storm, cyclone, southern swell…). accessible accessible
Erosion scarp up to Accessible Accessible 1 0
0.4 m
Erosion scarps Erosion scarp up to Poorly Hardly 0 -1
0.8 m accessible accessible
Some islets are affected by significant erosion of their shore- Erosion scarp up Hardly Not -1 -2
lines. This erosion creates an erosion scarp whose height is to 1 m accessible crossable
variable and linked to the topography and morphology of Erosion scarp higher Not crossable Not crossable -2 -2
than 1 m
the islet (Online Resource 1). The scarp height at which it
becomes an obstacle for turtle nesting depends on the spe-
cies concerned. Green turtles (average size of adult around Tree stumps, roots and plant debris
115 cm) have higher overcoming capacities than loggerhead
turtles (average size of adult around 70–95 cm). Green tur- In eroded areas, seaward and on the scarp, it is quite com-
tles have an overcoming capacity around 0.8 to 1 m, while mon to observe roots, tree stumps and other woody debris
loggerhead turtles have an overcoming capacity around 0.6 (Online Resource 1). These constitute an unfavourable fac-
to 0.8 m. The islets within the Grande-Terre‘s lagoon are tor to turtle nesting for two reasons.
mainly frequented by loggerhead turtles, while the islets First, the wood stumps, roots and woody debris consti-
of Entrecasteaux and Chesterfield archipelagos are mainly tute some obstacles preventing turtles from accessing the
frequented by green turtles (Read et al. 2015). The erosion nesting sites. The second is, if the turtle finally accesses the
scarp height to be taken into account is therefore dependant site, presence of woody-debris can prevent them from creat-
on the species breeding on the islet (Read et al. 2015). In ing the nest. In this case, unsuccessful nesting attempts are
any case, the presence of an erosion scarp has an impact on observed. The presence of roots and woody debris in the
the hatchlings mortality rate during their journey towards sand (surface and in depth) are unfavourable factors that are
the sea. often the cause of abandoned nesting attempts.
Erosion scarps evolve according to the phases of erosion, If a significant accretion/deposition phase occurs after
stability or sedimentation that affect the islet. If erosion is the erosional phase (and consecutive apparition of woody
chronic and almost continuous, the scarp will persist over debris, roots etc.), previous debris are buried and accessibil-
time while moving towards the centre of the island (Gar- ity to nesting area becomes easier. Nevertheless, these new
cin et al., 2016). The scarp’s height depends on the islet’s sandy areas hide, in depth, numerous debris that constitute
topography. If the erosion affects a relatively low-lying
island, the height will remain moderate to low (decimetre
to few decimetres). Conversely, if the island is older (i.e. Table 3 Assessment of the impact of tree stumps, roots and plant
established during the higher Holocene sea level; Yamano debris on turtle nesting site accessibility
et al. 2014; Garcin et al., 2016) or if there are some sandy Vegetal debris Nesting site accessibility, Indi-
ridges or small dunes in the backshore, its height will be Nesting excavation cator
Value
higher. If a period of coastline stability occurs after the ero-
Lack of vegetal and Good accessibility, herbaceous 2
sion phase, a beach profile regularization will occur. The wood debris cover of the backshore or lack of
profile then becomes less steep and therefore can be more vegetation
easily overcome by turtles. If a phase of accretion succeeds Rare vegetal debris, Fairly good accessibility, disrup- 1
to the erosion phase, the sand deposition at scarp’s foot will some isolated tree tions during nest excavation,
reduce its height (thus making it easier to overcome). In branches herbaceous backshore with scarce
bushes
some cases, the sedimentation could be sufficient to bury
Backshore with bushes Accessibility sometimes difficult, -1
the scarp morphology. or low density stumps difficulties to excavate
and roots
Numerous wood tree Difficult to impossible access, -2
roots exposed excavation sometimes impossible
Numerous wood or Difficult to impossible access, exca- -2
bushes stumps vation sometimes impossible

13
24 Page 6 of 13 M. Garcin et al.

an unfavourable context for nest digging and egg survival the granulometry of a given beach (or of backshore) can
(Read et al., 2019). change quickly (few hours to few days) during periods of
erosion or accretion. A very rapid change of the granulom-
Sediment granulometry etry of the surficial sediments can occur due to the supply
of coarse debris during storms or high-energy events (e.g.
Sediments of islets beaches have various granulometry: southern swells, cyclone, tsunami…). Access to the nesting
from very fine sand to blocks and pebbles (Online Resource site will not be as easy and if the backshore is also covered
1). A sediment with a fine granulometry is a favourable by coarse sediments, the site will be inappropriate for turtle
factor for turtles’ nesting. The zones where sediments are nesting. Conversely, new fine sand layer can be deposited
homogeneous and fine are preferred to those with a hetero- during accretion or recovery phase; in this case, the beach
metric grain size or those constituted of coarser element becomes friendlier for turtles. Not all the islets are affected
(gravels and pebbles, coralline debris etc.). If the sediment by these granulometry changes but a good part of them are;
is a fine sand but with a low proportion of coarse elements it mainly depends on each islet’s dynamics of changes, in
(coral debris), nesting is still observed. On the other hand, relation with the forcing factors acting upon it.
during the digging of the nest, if the turtle finds in depth
coarser beds (pebbles, coral debris, bioclastic debris, etc.) Reef-flat wideness
it will give up.
The beach and backshore sediment granulometry is dif- A narrow reef-flat (small distance between the reef front
fers from one islet to another and is variable at the islet scale. and the nesting area) is a favourable factor because it makes
The granulometry variability is related to each local context access to the nesting area easier, less risky and faster than in
(exposure to waves, currents, distance to reef passes, exis- the case of a wider shelf.
tence/non-existence of a seaward barrier reef …). Moreover, The reef-flat widths surrounding the islets of the Cale-
donian lagoon are highly variable (Online Resource 1). The
Table 4 Assessment of the sites attractivness for turtle in relation with values range from a few metres to several hundred metres.
the sediment granulometry (unpublished data) Moreover, for some islets, the reef-flat width is also highly
Sediment granulometry at turtle nest- Nesting site Value
variable depending on the segment of shoreline considered.
ing site attractiveness assign to
indicator Moreover, for a given width of reef-flat, the presence of a
Fine sand Very attractive 2 pass leads to a local decrease in the width of the reef-flat and
Medium sand Attractive 1 thus allows turtles to access the beach more easily (Online
Sand (fine to medium) with rare Attractive 1 Resource 1).
debris Based on the location of 140 nests (data acquired by
Coarse sand Moderately 0 WWF during 2018–2019 campaign) observed on eleven
attractive
islets spread across Southern Lagoon we have analysed the
Sand (fine to medium) with numerous Low attractive -1
debris (coral, bioclasts, …) reef-flat width in front of these nests. The reef-flat widths
Coarse coral debris Not adapted -2 crossed by turtle have an average value of 129 m, a median
Beachrock bed or blocks Not adapted -2 of 126 m with a 45 m standard deviation. For each islet,
an analysis of nest locations in relation to reef-flat width

Table 5 Distance from nests to the reef-crest and reef-flat wideness for some islets of the Southern lagoon
Islet name Nest number Median of reef-crest to Minimum reef-crest Maximum reef-crest Minimum reef-flat Maximum
nest distance (m) to nest distance (m) to nest distance (m) wideness (m) reef-flat
wideness
(m)
Ua 1 91 91 91 50 317
Léroué 2 142 141 142 140 413
N’Da 4 111 51 128 41 202
Rédika 8 96 73 120 90 220
Uaterembi 8 186 133 212 130 227
Uatio 12 99 82 213 82 274
Kouaré 13 119 55 131 55 172
Vua 13 108 99 221 99 257
Atiré 25 179 118 209 118 221
Gi 26 138 45 224 45 260
N’Gé 28 85 42 196 42 326

13
Assessment method of the sea turtle-nesting habitat of small reef islands Page 7 of 13 24

Fig. 2 Distance between the

turtle nests (2018–2019 cam-
paign) and the reef crest for
eleven islets. For each islet: the
bottom of the bar is the lower
distance, the top of the bar is
the higher distance, the black
dot represents the median of the
distance nest to reef crest for
each islet

crossed by turtles was realized. Table 5 shows: (i) the very but that once the island is chosen, the turtles will tend to
great variability of the number of nests per islet (1 to 28) in show preference to the sectors with the narrowest reef-flat
the Southern Lagoon; (ii) the strong variability of reef-flat’s (Online resource 1).
widths crossed by turtles to make their nests (from 42 to Consequently, the reef-flat wideness indicator must be
224 m, Fig. 2). evaluated at the scale of each islet. The value assigned to
The very wide dispersion of the lengths crossed on the this indicator is completely dependant of the reef-flat char-
most nested islands also demonstrates that the number of acteristics of each islet. Then, the reef-flat wideness range is
nests on an island is not only controlled by the width of the computed (maximum wideness – minimum wideness) and
platform to be crossed (Fig. 2). split into five classes of equal interval (Delta). The indicator
For example, on islands Gi and N’Gé where respectively value decrease from the smaller reef-flat wideness classes
26 and 28 nests were observed, the reef-flat widths crossed (value + 2) to the larger ones (value − 2; Table 6).
by turtles are very variable (from about 42 to 224 m). Sim- The reef-flat width can be evolutionary even in the short
ilarly, the median crossed is 179 m at Atiré and 85 m at term (one year to few years) by the islet migration on the
N’Gé, while the number of nests recorded is very close (25 reef-flat over time. The distance can also change locally for
and 28 respectively). That implies that the reef-flat width to a particular shoreline segment due to erosion (in this case
cross is not the primary island choice criterion for nesting the distance between the reef front and the islet coastline
increase and consequently also the reef-flat) or after an
Table 6 Assessment of the impact of reef-flat wideness on turtle nest- accretion phase (in this case the width lower). At multi-
ing site accessibility for a given islet ( MinW = Minimum Reef-flat
wideness)
yearly and multi-decadal time scales, this variation of the
Reef-flat wideness Value reef-flat width can become significant.
Lower bound Upper bound assigned
to The selected indicators
indicator
MinW MinW + Delta 2 Five geomorphological indicators were considered as
MinW + Delta MinW + (2*Delta) 1
the most relevant and were selected. They are namely
MinW + (2*Delta) MinW + (3*Delta) 0
beachrock, Erosion scarp, reef-flat width (which integrates
MinW + (3*Delta) MinW + (4*Delta) -1
the distance to reef-passes if present), granulometry, stumps
MinW + (4*Delta) MinW + (5*Delta) -2

13
24 Page 8 of 13 M. Garcin et al.

Table 7 Indicators used for the assessment of turtle nesting potential for a coastal segment with their characteristics
Indicator Nesting Site Nest- Data sources Evolutivity Retrospective Link with
Accessibility ing Site analysis processes
Quality
Beachrock (presence & x Remote sensing, field High Partially possible Erosion / accretion
typology)
Erosion scarp height x Field High Partially possible Erosion / accretion
Stumps and woody debris x x Field, partially remote High Partially Erosion / accretion
sensing
Sediment granulometry x x Field Medium Impossible Erosion / accre-
tion, exposure
to waves energy,
extreme events
Reef-flat width x Remote sensing, field Medium Possible Erosion / accretion

and woody debris. A large majority of the parameters can Assessment method of the turtle-nesting potential
evolve at short term during morphogenetic events (storms, index
cyclones, Southern or cyclonic swells). The parameters’
evolutions are related to erosion and accretion processes For each homogeneous coastal segment, the five indicators
acting on the islets (Table 7). These processes constitute the are aggregated in a unique and mappable single turtle nest-
principal forcing factor leading to an evolution in the turtle ing potential index.
nesting habitats. The turtle nesting potential index assessment is a three-
The indicators can be grouped into three categories: step process:

i) one dealing with the accessibility of the nesting site, ● (i) the average value of the five single indicators is
ii) one dealing with the quality of the nesting sites, computed (real number). Result range from − 2 to + 2
iii) one dealing with the accessibility and quality of the (i.e. from Very unfavourable to Very favourable) as for
sites. single indicators.
● (ii) if one (or more) of the following single indicators
Indicators related to the accessibility of nesting sites are (“Beachrock”, “Erosion scarp”, “Sediment Granulom-
the most important. Indeed, if turtles have difficulty or etry”, “Stumps and wood debris”) is -2, this value is
cannot reach the top of the beach nor the backshore, there assigned to the segment, whatever the average obtains.
will be few or no nests even if the sectors behind hold very This is justified by the fact that even if all other indica-
favourable characteristics. The analysis of the parameters tors are favourable, the “Very unfavourable” value of
(Table 7) shows that, in detail, those that affect the nesting these is critical. For the specific case of the “Reef-flat
site quality criterion always affect the accessibility criterion wideness” indicator, we consider that the “Very unfa-
as well (e.g. Stumps and wood debris). Consequently, no vourable” value is not critical because even though the
parameters are only related to the quality of the nesting site. nesting site is harder to reach, it remains possible.
On the other hand, some parameters only affect accessibility ● (iii) The numeric value obtained is then converted into
(e.g. beachrock, reef-flat width). one of five classes giving the value of the turtle nesting
Each homogeneous shoreline’s segment is assigned a potential index (Table 8).
representative value of the state of each geomorphologi-
cal parameter. Five values can be assigned according to the
characteristics’ favourability to turtles’ nesting: Very favour-
able, Favourable, Moderately favourable, Unfavourable
Table 8 Correspondance between numerical value and the turtle nest-
and Very unfavourable. ing potential index
It is then possible to map the favourability value of the Numerical value Turtle nesting Index
segment for each parameter. Min (>) Max (<=) Value Label
1.5 2 2 Very favourable
0.5 1.5 1 Favourable
-0.5 0.5 0 Moderately
favourable
-1.5 -0.5 -1 Unfavourable
-2 -1.5 -2 Very
unfavourable

13
Assessment method of the sea turtle-nesting habitat of small reef islands Page 9 of 13 24

This integrate index permits an inter-comparison of the female population abundance. Although it is not as accurate
nesting potential of the shoreline segments of one islet and as the count of uniquely identified females, it requires less
an inter-comparison of the nesting potential of several islets. effort and presence to adapt to the logistical constraints of
Due to the geomorphological evolutivity of the islets, the the nesting site (SWOT 2011). All tracks of nesting activi-
value of this indicator is only representative at a given date. ties more recent than the previous survey are counted,
This justifies the fact that the geomorphological analysis of assessed, geolocated and erased to prevent recounting on
the islets must be done synchronously with the nests’ count- the next patrol.
ing campaigns. Geomorphological and sedimentological descriptors
and indicators were recorded directly in the field using the
method presented previously. Field observations were sup-
Results and discussion: application of the plemented by a GIS analysis (coastline mapping, computa-
method to 13 islets of the southern lagoon tion of the “Reef-flat wideness” indicator). Islets coastlines
were then segmented according to their characteristics and
Our study focuses on the breeding ground of the Great the indicators aggregated and associated with each coastal
South Lagoon, in the south part of the Grande Terre of New segment. Islands’ nesting potential were mapped according
Caledonia (Fig. 1, Online Ressource 2). It is an extensive
shallow reef area of about 2,600 km2, accounting more
than 30 islets widely spread, all of them possibly hosting
marine turtle nesting activities. The first step in assessing
the conservation status of a given marine turtle population
is determining how many reproductive females exist in the
population (SWOT 2011). For a long-lived and late mature
species like the loggerhead turtle, population monitoring
must persist for several years or decades to detect a signifi-
cant population trend (Chaloupka et al. 2008). It is relatively
easy to identify a marine turtle nest, because females leave
wide, deep tracks on the open sandy beaches they come to
nest on (Girondot 2017). We therefore chose to conduct a
count of the number of clutches detected (110 nests) in 13
of the islets (Fig. 3) during the nesting season (from late
November to late March, 2019–2020), with a two to three Fig. 4 Example of coast segmentation using the turtle nesting potential
weeks delay between each count for a total of six counts index and location of turtle nests (Feb. 2020) of the N’Gé islet (South-
ern Lagoon, New-Caledonia)
per season. This count can be considered a proxy for total

Fig. 3 Location map of the 13

studied islets

13
24 Page 10 of 13 M. Garcin et al.

Table 9 The turtle nesting potential index of the coastlines of the 13 (cumulative length 3 162 m) and 24.3% are “Very favour-
islets
able” (2,604 m). The “Unfavourable” category is not repre-
Turtle nesting potential Cumu- Length Turtle Turtle
lative percent- nests nests sented in this islets set.
length age (%) number percent- The analysis of the nests’ locations in relation with our
(m) age (%) coastline classification shows that in (Table 9; Fig. 5):
Very favourable 2 833 24.3 62 56.4 − 56.4% of the observed turtle nests concern coastal seg-
Favourable 2 934 25.2 34 30.9 ments characterized as “Very favourable”,
Moderately favourable 1 025 8.8 6 5.5 − 30.9% concern sectors characterized as “Favourable”,
Unfavourable 0 0.0 0 0.0
− 5.5% concern sectors characterized as “moderately
Very Unfavourable 4 874 41.8 8 7.3
favourable”,
− 7.3% of the observed turtle are located in segments
characterized as “Very unfavourable”.
This cross analysis between the index value and the num-
ber of nests observed in the field indicates that our index is
discriminant enough to identify the nesting habitat attrac-
tiveness of coastal segments for sea-turtle nesting. Never-
theless, few nests have been observed in Very unfavourable
coastal segments for unknown reasons. This could be linked
to the fact that turtles try to nest in the same place as pre-
vious years even if the nesting conditions have become
unfavourable due to the geomorphological evolution of the
islets.
The average number of nests by coast length (km) by
nesting potential index is computed (Table 11). Even if
Fig. 5 Percentage of observed turtle nests relative to the turtle nesting
potential index
these preliminary results are obtained on a still limited set
of data, the “Very Favourable” coasts have a sea turtle nest
to the five categories (from most to least favourable) in order number per km around 13.7 fold higher than the less favour-
to evaluate their attractiveness (Fig. 4). The number of nests able coasts. The “Favourable” coasts have a ratio of around
is associated with each coastal segment. The detailed results 7.25 and “Moderately favourable” of around 3.7.
obtained (maps) are presented in detail in the supplementary In order to sort out the attractiveness of each islet towards
materials. others, we have computed for each islet the number of nests
The cumulative coastline length investigated (sum of using the average nest-by-kilometre ratio for each types of
coastline of the 13 islets) is 11,665 m. 41.8% of the cumu- coasts. We obtain a theoretical number of nests for each islet
lative coastline length are “Very unfavourable” (cumula- as a function of their coastal characteristics. For a given islet,
tive length 4,874 m), 8.8% are “Moderately favourable” if the number of computed nests is higher than the observed
(cumulative length 1 025 m), 25.2% are “Favourable” number, we conclude that the islet is under-nested compared

Table 10 Sorting of islets as a function of their cumulative length of favourable coastline for turtle nesting
Length of coast- 1) Very 2) Favourable More attractive 3) Mod- 4) Unfavourable 5) Very Less attractive #
line (m) / Nesting favourable coast length erately Unfavourable coast length Nests
potential (1 + 2) favourable (3 + 4 + 5)
Uatérembi 0 778 778 0 0 259 259 5
Leroué 118 482 600 0 0 334 334 1
N’Do 575 0 575 0 0 723 723 11
Ua 429 105 534 0 0 489 489 8
Kouaré 81 378 459 281 0 334 334 4
Noé 301 134 435 0 0 275 275 3
Rédika 125 309 434 258 0 479 479 14
Uié 226 169 395 0 0 248 248 8
N’Gé 289 97 386 31 0 456 456 18
Petit Mato 117 215 332 0 0 127 127 2
Uatio 331 0 331 455 0 131 131 12
Atiré 0 267 267 0 0 243 243 18
Puemba 241 0 241 0 0 776 776 6

13
Assessment method of the sea turtle-nesting habitat of small reef islands Page 11 of 13 24

Fig. 6 Left : difference between

the number of observed turtle
nests and theoretical number of
nests; right : percentage of the
difference between the number
of turtle nests observed and the
theoretical number of nests;
over nested islets in green,
under nested islets in red

sedimentological parameters. Comparison of the classifi-

Table 11 Average number of sea-turtle nest by km of each category of cation of the coastal segments with the turtle nest counts
coast attractiveness for all the islets investigated
shows a good agreement between the classification obtained
Coast attractiveness Number
of sea and the nests density. This method will be useful to biolo-
turtle gists monitoring turtle nesting sites and to decision-makers
nest/ km in charge of marine area management.
“Very Favourable” 21.9 Assessing the influence of morphological and environ-
“Favourable” 11.6 mental parameters (and their evolution) on turtle nesting
“Moderately Favourable” 5.9
is useful to (1) identify the most sensitive areas for turtles
“Unfavourable” -
in the Southern Lagoon (at the islet scale), (2) provide
“Very Unfavourable” 1.6
elements for prioritising turtle conservation and manage-
ment measures, (3) identify priority conservation areas at
to its potential (e.g. Kouaré, Leroué, Noé). Conversely, if the island level (restricted areas), but also at the southern
computed nests number is lower than the observed number, lagoon level (island closure during the nesting period).
we conclude that the islet is over-nested compared to its It enables a relationship to be established between mor-
potential (e.g. Atiré, N’Gé, Rédika; Fig. 6). phological and sedimentological states (and changes) of
There are probably several factors leading the over-nest- beaches and backshore and the nesting potential of marine
ing or under nesting of an islet. These factors may include turtles. This method is easily transposable and applicable to
the distance of the islet from the barrier-reef passes and the all nesting sites on coral islands around the world but also
ease of access from the open ocean to the islet, channels and transposable to others sea turtle nesting contexts. Future
currents in the lagoon, and human frequentation. Human morphological changes in coral islets due to climate change
frequentation is a particularly important factor for turtle and sea level rise will affect sea turtle nesting habitats. The
nesting. On the New Caledonian islets, this frequentation impact of future changes can be assessed using this method.
is very variable from one islet to another according to its
accessibility and its distance from either ports or launching Supplementary information The online version contains
supplementary material available at https://doi.org/10.1007/s11852-
docks. Human frequentation of the islets is mainly recre- 022-00870-7.
ational (fishing, swimming, camping, kite surfing…). The
most critical point is night light pollution around the camp- Acknowledgements This study was carried out as part of the OBLIC
sites (Witherington and Martin 2000). This can make turtles project (Observatoire du littoral de Nouvelle-Calédonie) supported by
give up on accessing the nesting sites. However, despite the the New Caledonia Government (Direction de l’Industrie, des Mines et
de l’Environnement de Nouvelle-Calédonie) and the BRGM (French
importance of this factor, human frequentation is not taken Geological Survey: Bureau de Recherche Géologique et Minière), by
into account in our approach because it is not a parameter the BRGM’s Resilience project and by the World Wide Fund for Nature
directly assignable to the characterization of the nesting (WWF). Thanks to Cécile Capderrey and Audrey Baills (BRGM), Julie
habitat. It can be considered as an additional exogenous fac- Mounier and Julie-Anne Kerandel (Direction des Affaires Maritimes
de Nouvelle-Calédonie) for their helpful discussions and comments,
tor that will have to be taken into account once the nesting to Claudine Jego and Jack Challis for improving the English. Finally,
habitat’s qualification has been carried out. we would like to thank the reviewers for their insightful comments and
suggestions for improving this document.

Conclusions Authors’ contributions Methodology development: MG and MVL;

fieldwork and data acquisition: MG, MVL, TR, MO, HB; GIS, data-
bases and computing: MG and MVL; writing the original paper: MG;
This method makes it possible to define and map the level Improving the final paper: MG, MVL, TR, MO, HB.
of attractiveness of coral islands coasts for turtle nesting.
It uses several easily accessible geomorphological and Funding This study was carried out as part of the OBLIC project (Ob-
servatoire du littoral de Nouvelle-Calédonie) supported by the New

13
24 Page 12 of 13 M. Garcin et al.

Caledonia Government (DIMENC) and the BRGM (French Geologi- Couture E (2000) Distribution of phyllosoma larvae of Scyllaridae and
cal Survey), by the BRGM’s Resilience project and by the WWF. Palinuridae (Decapoda: Palinuridea) in the south-western lagoon
of New Caledonia. Mar Freshw Res 51(4):363–369
Availability of data and material Not applicable. Debenay JP, Fernandez JM (2009) Benthic foraminifera records of
complex anthropogenic environmental changes combined with
geochemical data in a tropical bay of New Caledonia (SW Pacific.
Declarations Mar Poll Bull 59:311–322
Dickinson WR (1999) Holocene sea-level record on Funafuti and
Conflict of interest The authors declare that they have no conflict of potential impact of global warming on Central Pacific atolls. Qua-
interest. tern Res 51:124–132
Faure V, Pinazo C, Torréton JP, Douillet P (2010) Modelling the spatial
Conflicts of interest/competing interests We wish to confirm that and temporal variability of the SW lagoon of New Caledonia II:
there are no known conflicts of interest associated with this publica- Realistic 3D simulations compared with in situ data. Mar Poll
tion “Assessment method of the Sea turtle nesting habitat of small reef Bull 61:480–502
islands” and there has been no significant financial support for this Fish MR, Côté IM, Gill JA, Jones AP, Renshoff S, Watkinson AR
work that could have influenced its outcome. (2005) Predicting the impact of sea-level rise on Caribbean Sea
turtle nesting habitat. Conserv Biol 19(2):482–491
Ethics approval Not applicable. Flamand B (2006) Les pentes externes du récif barrière de la Grande-
Terre de Nouvelle-Calédonie: morphologie, lithologie, contrôle
de la tectonique et de l’eustatisme. Ph.D. thesis, 221 p. France:
Consent to participate Not applicable.
Université de Bretagne Occidentale
Ford M (2012) Shoreline changes on an urban atoll in the Central
Consent for publication Not applicable. Pacific Ocean: Majuro Atoll, Marshall Islands. J Coast Res
28(1):11–22
Open Access This article is licensed under a Creative Commons Ford M (2013) Shoreline changes interpreted from multi-temporal
Attribution 4.0 International License, which permits use, sharing, aerial photographs and high resolution satellite images: Wotje
adaptation, distribution and reproduction in any medium or format, atoll, marshall islands. Remote Sens Environ 135:130–140.
as long as you give appropriate credit to the original author(s) and the doi:https://doi.org/10.1016/j.rse.2013.03.027. ISSN 0034-4257
source, provide a link to the Creative Commons licence, and indicate ;1879 – 0704
if changes were made. The images or other third party material in this Ford M, Kench P (2014) Formation and adjustment of typhoon-
article are included in the article’s Creative Commons licence, unless impacted reef islands interpreted from remote imagery: Nadikdik
indicated otherwise in a credit line to the material. If material is not Atoll, Marshall Islands. Geomorphology 214:216–222
included in the article’s Creative Commons licence and your intended Garcin M, Vendé-Leclerc M, Robineau B, Maurizot P, Le Cozannet
use is not permitted by statutory regulation or exceeds the permitted G, Nicolae-Lerma A, 2016 - Lagoon islets as indicators of recent
use, you will need to obtain permission directly from the copyright environmental changes in the South Pacific - The New Caledo-
holder. To view a copy of this licence, visit http://creativecommons. nian example.Continental Shelf Research, 122,120–140
org/licenses/by/4.0/. Girondot M (2017) Optimizing sampling design to infer the number of
marine turtles nesting on low and high density sea turtle rooker-
ies using convolution of negative binomial distribution. Ecol Ind
81:83–89
References Hays GC, Mackay A, Adams CR, Mortimer JA, Speakman JR, Boer-
ema M (1995) Nest site selection by sea turtles. J Mar Biol Asso-
Ballu V, Bouin MN, Siméoni P, Crawford WC, Calmant S, Boré JM, ciation United Kingd 75(3):667–674
Kanas T, Pelletier B (2011) Comparing the role of absolute sea- Jimenez H, Dumas P, Léopold M, Ferraris J (2011) Invertebrate har-
level rise and vertical tectonic motions in coastal flooding, Tor- vesting on tropical urban areas: Trends and impact on natural pop-
res Islands (Vanuatu). Proceedings of the National Academy of ulations (New Caledonia, South Pacific). Fish Res 108:195–204
Sciences, USA, 108, 13019–13022. doi: https://doi.org/10.1073/ Kamel SJ, Mrosovsky N (2004) Nest site selection in leatherbacks,
pnas.1102842108 Dermochelys coriacea: individual patterns and their conse-
Beyneto S, Delcroix E (2005) Underwater Oviposition by a Hawks- quences. Anim Behav 68(2):357–366
bill turtle in Guadeloupe, French West Indies. Mar Turt Newletter Long TM, Angelo J, Weishampel JF (2011) LiDAR-derived measures
107:14 of hurricane-and restoration-generated beach morphodynam-
Bonnet X, Briand M, Brischoux F, Letourneur Y, Fauvel T, Busta- ics in relation to sea turtle nesting behaviour. Int J Remote Sens
mante P (2014) Anguilliform fish reveal large scale contamina- 32(1):231–241
tion by mine trace elements in the coral reefs of New Caledonia. Mounier S (2007) Bilan préliminaire des volets survol et mission ter-
Science of the Total Environment, 2014, 470–471, 876–882 rain de l’opération tortues (2006/2007). WWF Nouméa
Bozec Y, Dolédec S, Kulbicki M (2005) An analysis of fish-habitat Murphy TM, Hopkins-Murphy SR (1990) Homing of translocated
associations on disturbed coral reefs: chaetodontid fishes in New gravid loggerhead turtles. In: T.H. Richardson, J.I. Richardson &
Caledonia. Journ Fish Bio 66:966–982 M. Donnelly (Compilers) Proceedings of the Tenth Annual Work-
Brischoux F, Bonnet X (2009) Life history of sea kraits in New Cale- shop on Sea Turtle Biology and Conservation. NOAA Technical
donia, in Grandcolas P. (ed.), Zoologia neo caledonica 7. Biodi- Memorandum NMFS-SEFC-278, Miami, Florida, pp. 123–124
versity studies in New Caledonia. Mémoires du Muséum national Nicholls RJ, Cazenave A (2010) Sea-level rise and its impact on
d’Histoire naturelle, 198: 133–147. Paris coastal zones. Science 328(1517). doi:https://doi.org/10.1126/
Chaloupka M, Bjorndal KA, Balazs GH, Bolten AB, Ehrhart LM, Lim- science.1185782
pus CJ, Suganuma H, Troeng S, Yamaguchi M (2008) Encour- Nunn PD (2013) The end of the pacific? Effects of sea level rise on
aging outlook for recovery of a once severely exploited marine pacific island livelihoods. Singap J Trop Geogr 34:143–171.
megaherbivore. Glob Ecol Biogeogr 17:297–304 doi:https://doi.org/10.1111/sjtg.12021

13
Assessment method of the sea turtle-nesting habitat of small reef islands Page 13 of 13 24

Nurse L, McLean R, Agard J, Briguglio L, Duvat-Magnan V, Pele- Roy P et J. Connell (1991) «Climatic-change and the future of Atoll
sikoti N, Tompkins E, Webb. A (2014) Small islands, Climate states». J Coastal Res 7(4):1057–1075 ISSN 0749 – 0208
Change 2014: Impacts, Adaptation, and Vulnerability. Part B : Stoddard DR (1990) Coral reefs and islands and predicted sea-level
Regional Aspects, Contribution of Working Group II to the Fifth rise. Prog Phys Geogr 14(4):521–536
Assessment Report of the Intergovernmental Panel on Climate SWOT Scientific Advisory Board (2011) The State of the World’s Sea
Change, Cambridge University Press, Cambridge, United King- Turtles (SWOT) Minimum Data Standards for Nesting Beach
dom and New York, NY, USA, p. 1613–1654 Monitoring. Technical Report, 24 pp
Oremus M, Mattei J (2017) Tortues «grosse tête» du Grand Lagon Testut L, Duvat V, Ballu V, Fernandes RMS, Pouget F, Salmon C,
Sud. Inventaire des sites de ponte et réflexion sur la mise en place Dyment J (2015) Shoreline changes in a rising sea-level con-
d’un protocole de suivi de la population. Rapport de projet WWF text: the example of Grande Glorieuse, Scattered Islands, West-
France, p 26 ern Indian Ocean. Acta Oecol 1–10. https://doi.org/10.1016/j.
Ouillon S, Douillet P, Lefebvre JP, Le Gendre R, Jouon A, Bonneton actao.2015.10.002
P, Fernandez JM, Chevillon C, Magand O, Lefèvre J, Le Hir P, Webb AP, Kench PS (2010) The dynamic response of reef islands to
Dumas F, Marchesiello P, Madani B, Andréfouet A, Panché S, sea-level rise: evidence from multi-decadal analysis of island
Fichez JY, R (2010) Circulation and suspended sediment trans- change in the Central Pacific. Glob Planet Change 72:234–246
port in a coral reef lagoon: The south-west lagoon of New Cale- Witherington BE, Martin RE(2000) Understanding, Assessing, and
donia. Mar Poll Bull 61:269–296 Resolving Light-Pollution Problems on Sea Turtle Nesting
Pandolfi Benoit M, Bretagnolle V (2002) Seabirds of the Southern Beaches. Florida Marine Research Institute. Technical Report
Lagoon of New Caledonia; Distribution, Abundance and Threats Wood DW, Bjorndal KA(2000) Relation of temperature, moisture,
Waterbirds2002 25 (2), 202–213 salinity, and slope to nest site selection in loggerhead sea turtles.
Ramsay D (2011) Coastal erosion and inundation due to climate Copeia, 2000 (1), pp.119–119
change in the Pacific and East Timor. NIWA report 76p. ISBN Woodroffe CD (2008) Reef-island topography and the vulnerability of
978-925006-03-2 atolls to sea-level rise. Glob Planet Change 62:77–96
Read TC, FitzSimmons NN, Wantiez L, Jensen MP, Keller F, Chateau Xue C(1997) Coastal erosion and management of Vaitupu Island,
O, Limpus CJ (2015) Mixed stock analysis of a resident green Tuvalu. Technical report 243, SOPAC
turtle, Chelonia mydas, population in New Caledonia links rook- Yamano H, Cabioch G, Join CCh, J.-L (2014) Late Holocene sea-level
eries in the South Pacific. Wildl Res 42(6):488–499 change and reef island evolution in New Caledonia. Geomorphol-
Read TC, Petit M, Magnan M, Booth D (2020) Going back to the roots: ogy 222:39–45
finding a strategy for the management of nesting loggerhead sea Yates M, Le Cozannet G, Garcin M, Salaï E, Walker P (2013) Multi-
turtles in New Caledonia. Australian J Zool 66(6):394–400 decadal atoll shoreline change on Manihi and Manuae, French
Read T, Booth DT, Limpus CJ (2013) Effect of nest temperature on Polynesia. Journ Coast Res 29:4, 870–882
hatchling phenotype of loggerhead turtles (Caretta caretta) from
two South Pacific rookeries, Mon Repos and La Roche Percée. Publisher’s note Springer Nature remains neutral with regard to juris-
Australian J Zool 60(6):402–411 dictional claims in published maps and institutional affiliations.

13