Landsc Ecol (2023) 38:2717–2730

https://doi.org/10.1007/s10980-023-01708-9

EDITORIAL

Toward conciliation in the habitat fragmentation

and biodiversity debate
Jonathon J. Valente · Dustin G. Gannon · Jessica Hightower · Hankyu Kim · Kara G. Leimberger ·
Rossana Macedo · Josée S. Rousseau · Matthew J. Weldy · Rachel A. Zitomer · Lenore Fahrig ·
Robert J. Fletcher Jr. · Jianguo Wu · Matthew G. Betts

Published online: 12 September 2023

This is a U.S. Government work and not under copyright protection in the US; foreign copyright protection may apply 2023

Landscape-scale conservation planning is urgent advance research that leads to consensus rather than
given the extent of anthropogenic land-use change the perpetuation of disagreement. Explicit efforts
and its pervasive impacts on Earth’s biodiversity. to develop and test multiple competing hypotheses,
However, such efforts are hindered by disagree- inter-laboratory collaborations, and acknowledgement
ments over the effects of habitat fragmentation on of multiple interacting effects will be vital for mov-
biodiversity that have persisted since the mid-1970s. ing the fragmentation debate forward. We argue that
We contend that nearly 50 years later, these disa- we in the ecology community should be responsible
greements have become a locked-in debate charac- for helping to reconcile different views across scales,
terized by polarized, unproductive discourse and a systems, and methodological approaches to advance
lack of consistent guidance for landscape managers conservation planning within a landscape ecology
and policy makers. Here, we highlight the need for framework.
a unified set of principles regarding conservation in
fragmented landscapes, identify potential reasons
for disparate conclusions in fragmentation research,
and suggest ways for the ecological community to

J. J. Valente (*) H. Kim

U.S. Geological Survey, Alabama Cooperative Fish Department of Forest and Wildlife Ecology, University
and Wildlife Research Unit, College of Forestry, Wildlife of Wisconsin–Madison, Madison, WI, USA
and Environment, Auburn University, Auburn, AL, USA
e-mail: jvalente@usgs.gov J. S. Rousseau
Cornell Lab of Ornithology, Ithaca, NY, USA
D. G. Gannon
Department of Botany and Plant Pathology, Oregon State L. Fahrig
University, Corvallis, OR, USA Department of Biology, Carleton University, Ottawa, ON,
Canada
D. G. Gannon
Department of Statistics, Oregon State University, R. J. Fletcher Jr.
Corvallis, OR, USA Department of Wildlife Ecology and Conservation,
University of Florida, Gainesville, FL, USA
J. Hightower · K. G. Leimberger · R. Macedo ·
M. J. Weldy · R. A. Zitomer · M. G. Betts J. Wu
Department of Forest Ecosystems and Society, Oregon School of Life Sciences and School of Sustainability,
State University, Corvallis, OR, USA Arizona State University, Tempe, AZ, USA

Vol.: (0123456789)
13
 2718 Landsc Ecol (2023) 38:2717–2730

The fragmentation debate ecology, including disagreements about the relation-

ship between productivity and diversity (Adler et al.
Disagreement is fundamental to the scientific process. 2011; Fridley et al. 2012; Grace et al. 2012; Pan et al.
Rigorous scientific inquiry encourages researchers to 2012) and local biodiversity dynamics under global
be at war with their own ideas, to design studies that change (Cardinale et al. 2018; Gonzalez et al. 2016;
challenge their hypotheses, and to ensure that those Vellend 2017; Vellend et al. 2017). Unfortunately,
hypotheses can withstand scrutiny (Popper 1959). locked-in debates tend to impede scientific progress
More broadly, scientific progress happens in fits and by hindering productive discourse and reducing com-
starts with periods of normal, incremental knowledge plex ecological processes to polarized arguments.
growth punctuated by moments in which established More concerning is that this affords decisionmakers
theories are challenged, tested, and debated (Kuhn the freedom to single out results that support their
1962). It is during these latter paradigm-shifting agenda or even dismiss scientific findings outright
moments that the potential for disagreement and divi- due to a perceived lack of credibility (Norberg et al.
sion among scientists is highest, but also when the 2022). Thus, locked-in debates can be detrimental to
greatest advances in knowledge are likely to occur. scientific progress and conservation efforts.
When researchers become entrenched in their dif- For half a century, ecologists have wrestled with
fering perspectives, however, these debates often a locked-in debate regarding the effects of habitat
become unproductive and locked in (sensu Norberg fragmentation per se (conditional on a given amount
et al. 2022). Recently, Norberg et al. (2022) provided of habitat; Fig. 1) on biodiversity. Early disparity
several examples of locked-in debates in the field of over the application of island biogeography theory

Fig. 1  Landscape (a) is

composed of 100% habitat
(green). Habitat loss can
occur without fragmenta-
tion (b) while the process
of fragmentation involves
habitat loss that results in
multiple smaller patches
(c). Increasing the number
of patches without changing
the total amount of habitat
in a landscape is referred to
as fragmentation per se (b
to d and c to d)

Vol:. (1234567890)
13
Landsc Ecol (2023) 38:2717–2730 2719

(MacArthur and Wilson 1967) to terrestrial systems takeaway from these reviews is that accomplished
fueled the SLOSS controversy in the 1970s and 1980s scientists come to seemingly mutually exclusive,
regarding whether a single large reserve would sup- opposite conclusions regarding fragmentation
port greater species richness than several smaller effects on biodiversity and ecosystem processes.
reserves (Terborgh 1974; Diamond 1975; May 1975; Critically, the consequences of this locked-in
Wilson and Willis 1975; Simberloff and Abele 1976; debate extend far beyond its influences on scien-
Quammen 1997). This controversy led to a discussion tific progress. Although many factors come into
about the effects of fragmentation per se on species play when developing conservation strategies (e.g.,
distributions (Wilcox and Murphy 1985), which in location, cost, threatened species), land managers,
turn has fractured into numerous interrelated disa- conservation practitioners, and policymakers tasked
greements. These include back-and-forths in the lit- with conserving at-risk species and their habitats
erature regarding the relative effects of habitat loss across the globe are increasingly incorporating the
and fragmentation per se on species distributions and composition and configuration of protected areas
abundances (Betts et al. 2006; Laurance 2008; Fahrig into the planning process. Broad-scale initiatives
2003, 2013, 2015; Hanski 2015; Haddad et al. 2017; such as the Mesoamerican Biological Corridor
Watling et al. 2020), the expected effects of habi- (Miller et al. 2001; Independent Evaluation Group
tat amount on species diversity (Fahrig 2013, 2021; 2011), the Yellowstone to Yukon Conservation Ini-
Saura 2020, 2021), the general pattern of biodiversity tiative (Aenǵst 1999), the Northwest Forest Plan
response to fragmentation (Haddad et al. 2015; Fahrig (Spies et al. 2019), and the America the Beautiful
2017; Fletcher et al. 2018a; Fahrig et al. 2019), and Challenge (U.S. Depts. of Interior, Agriculture, and
whether habitat patches are relevant units for study- Commerce 2021) have all emphasized the need for
ing biodiversity in terrestrial systems (McIntyre and science-based guidance for improving biodiversity
Barrett 1992; Lindenmayer and Fischer and 2007; protection through regional planning efforts. Given
Fahrig 2013; Mendenhall et al. 2014). Most concern- that decisions regarding resource allocation at large
ing is that while natural systems are being lost and spatial expanses are typically irreversible in the
fragmented at an increasing rate globally (Haddad short term, we need to ensure this polarized debate
et al. 2015; Taubert et al. 2018), these debates have is not undermining effective decision-making.
become more frequent and polarized (Fig. 2). Our goal here is neither to take sides, nor to crit-
To appreciate how locked-in and confusing this icize the valuable progress that has been made by
debate has become, one needs to look no further fragmentation researchers to date, but to provide a
than two relatively recent reviews by some of the framework for moving forward. To do this, we have
leading thinkers in landscape ecology. Haddad et al. generated several hypotheses to explain how differ-
(2015) reviewed 76 studies across five long-running ent scientists can come to discordant conclusions
experiments in which habitat fragments were cre- regarding the effects of fragmentation on biodiver-
ated. They concluded that “… habitat fragmenta- sity, and we encourage a shift towards exploring
tion reduces biodiversity by 13 to 75% and impairs these potential mechanisms rather than defending
key ecosystem functions by decreasing biomass and a specific position in the debate. Our hope is that
altering nutrient cycles.” Shortly thereafter, Fahrig developing an empirically verified understanding
(2017) reviewed 118 studies that measured frag- of the biological processes and study design fac-
mentation across whole landscapes and concluded tors driving the debate will lead to more unified and
that across a range of variables related to biodi- management-relevant knowledge of how fragmen-
versity and ecosystem services, “… the significant tation affects species distributions and biodiversity
responses to habitat fragmentation independent of across scales.
habitat amount are rare and mostly positive.” While
there are some important differences in the ques-
tions addressed by these researchers and the crite- Why are we locked in?
ria used for including studies in their reviews, these
differences are likely to be lost on many students, The heart of many locked-in scientific debates is a
land managers, and policymakers. The most salient failure to progress from studies focused on testing

Vol.: (0123456789)
13
 2720 Landsc Ecol (2023) 38:2717–2730

Fig. 2  This figure, adapted from Norberg et al. (2022), high- 26 published journal articles spanning that time period (panel
lights the four phases of adaptive theory development (con- a). Unlocking this debate will require collaboration among
ception, demonstration, investigation, and consolidation). The researchers with different perspectives to develop a common
hypothesis that fragmentation reduces biodiversity has been conceptual framework for habitat fragmentation and to use that
locked in a demonstration and counter-demonstration phase framework to develop and test hypotheses to explain disparity
since the mid-1970s, as shown by a non-random sample of in previous findings (panel b)

Vol:. (1234567890)
13
Landsc Ecol (2023) 38:2717–2730 2721

support for a hypothesis or theory (theory demonstra- landscapes by altering the size, shape, and isolation
tion) to those that explore the adequacy and limita- of individual patches, which theory predicts should
tions of the theory (theory investigation; Fig. 2; Nor- affect species distributions and dispersal rates (Mac-
berg et al. 2022). In the case of the fragmentation Arthur and Wilson 1967; Levins 1969). For this rea-
debate, studies focused on theory demonstration do son, other researchers have argued that fragmentation
indeed still abound (Evju and Sverdrup-Thygeson effects on biodiversity are best measured as a function
2016; Haddad et al. 2017; Lindgren and Cousins of the size, shape, or isolation of individual patches
2017; Watling et al. 2020; With and Payne 2021). (Thornton et al. 2011; Hanski 2015; Fletcher et al.
However, conflicting findings from such studies can- 2018a, 2023).
not fully explain the sustained division given that Interestingly, there is some evidence that study
reviews examining the mechanistic effects of frag- conclusions can be confounded by the study design.
mentation on biodiversity have also yielded mixed Reviews that rely exclusively on landscape-scale stud-
results (Debinski and Holt 2000; Haddad et al. 2015; ies designed to measure the effects of fragmentation
Fahrig 2017). For example, Fahrig’s (2017) review per se seem more likely to conclude that fragmen-
found no evidence that tropical species are more tation positively affects biodiversity (Fahrig 2003,
negatively affected by fragmentation than temperate 2017; Riva and Fahrig 2022). In contrast, reviews
species, while another global analysis using the BIO- relying on patch-scale studies seem more likely to
FRAG database (Pfeifer et al. 2014) revealed that spe- come to the opposite conclusion (Gilbert-Norton
cies near the equator are 6 times more likely to show et al. 2010; Thornton et al. 2011; Haddad et al. 2015;
negative responses to fragmentation than those at Chase et al. 2020). We suspect this difference has to
higher latitudes (Betts et al. 2019). In addition, while do with disparate mechanisms dominating the under-
several studies have identified trait-based predictors lying biodiversity patterns in such studies. When
of species response to fragmentation, these predictors comparing landscapes with similar amounts of habi-
are inconsistent across studies (Hatfield et al. 2018). tat, increased fragmentation per se (Fig. 1) will likely
Notably, there are several key biological and meth- result in increased inter- and intra-patch heterogeneity
odological factors that may cause variability among leading to greater diversity within whole landscapes
fragmentation studies and ultimately contribute to (gamma diversity). Not only does this increase the
such inconsistencies. pool of species in the landscape available to colonize
patches, but under this model, fragmentation reduces
Patch‑scale vs. landscape‑scale study design inter-patch nearest neighbor distances making coloni-
zation events potentially more likely, thereby increas-
A carefully designed study is critical to testing any ing local (alpha) diversity. Therefore, the sum of these
hypothesis and drawing reliable inferences. Yet what processes may have a net positive effect on some bio-
constitutes the best, or even a reasonable approach, to diversity metrics (see ‘Measurements of biodiver-
testing fragmentation effects on species or communi- sity’ below). Conversely, the process of fragmenta-
ties remains one of the most divisive issues in frag- tion (which typically stems from habitat loss; Fig. 1)
mentation research. We suggest that study design may results in patches of habitat that are smaller and more
be the most important cause of opposing conclusions isolated from one another. Populations of species that
about fragmentation effects. Some researchers have rely on that habitat are smaller in remnant patches and
argued that because fragmentation is a landscape- thus more likely to go extinct, while isolated patches
scale process, and because patch size and isolation are less likely to be recolonized due to dispersal limi-
can be confounded with habitat amount, effects of tation. Moreover, the quality of individual patches
fragmentation per se can only be studied by measur- may be reduced for individual species due to altered
ing biodiversity responses to fragmentation metrics biophysical properties stemming from edge effects
within whole landscapes (e.g., mean patch size, num- (Ries et al. 2004). Thus, comparisons among patches
ber of patches, mean interpatch distance; McGari- with different emergent properties stemming from
gal and Cushman 2002; Fahrig 2017; Fahrig et al. fragmentation may yield negative effects on biodiver-
2019). On the other hand, the process of fragmenta- sity metrics. Further empirical research is needed to
tion (which results from habitat loss; Fig. 1) affects

Vol.: (0123456789)
13
 2722 Landsc Ecol (2023) 38:2717–2730

Fig. 3  The process of fragmenting one land cover type can Bunting (Passerina cyanea) habitat, creating larger, more con-
have very different effects on the distribution of habitat for tiguous patches (top). c To measure habitat amount and frag-
different species. Using species distribution models (SDM) mentation, these SDMs must be dichotomized by selecting a
that linked known occurrences with satellite imagery, we (a) threshold value in the occupancy probability gradient to distin-
examined the effects of forest loss and fragmentation from road guish patches from matrix; the choice of this threshold value
development on breeding bird distributions in southern Indi- (0.5 or 0.7) will affect measurements of habitat amount and
ana (Valente and Betts 2017; base map from Esri and its licen- fragmentation. Unedited Indigo Bunting and Ovenbird images
sors, copyright 2023). b The forest fragmentation process split were provided by Dan Pancamo and Mike’s Birds, respectively,
Ovenbird (Seiurus aurocapilla) habitat into smaller and more under a Creative Commons Attribution-ShareAlike 2.0 license
isolated units (bottom) but had the opposite effect on Indigo (https://​creat​iveco​mmons.​org/​licen​ses/​by-​sa/2.​0/​legal​code)

develop a comprehensive understanding of how study to the physical material on the Earth’s surface in an
design affects conclusions in fragmentation studies. area, such as vegetation communities, water bodies,
and artificial structures. A land cover classification
Fragmentation of habitat vs. land cover (e.g., forest, grassland, or chapparal) can be used as a
habitat proxy, but it is unlikely to be a direct analog of
Quantifying the structure of a landscape is a funda- habitat for all, or even most, individual species (Hal-
mental precursor to measuring fragmentation effects stead et al. 2019).
on a species or community. However, while there is a While fragmentation of a land cover type is likely
difference between measuring fragmentation of habi- to affect the distribution of habitat for many species,
tat and fragmentation of a land cover type, many stud- it will not affect them all equally. For some species,
ies use land cover to represent habitat in their analyses the process of fragmenting a particular land cover
(e.g., Evju and Sverdrup-Thygeson 2016; Lindgren type may be comparable to loss and fragmentation
and Cousins 2017; Valente and Betts 2019; With and of habitat, but for others it could create more habi-
Payne 2021). Habitat is a species-specific concept tat aggregated in larger patches (Fig. 3a, b). That is,
that refers to the biotic and abiotic conditions neces- the fragmentation process makes space for additional
sary for a given species to occupy an area, which may species that have niches more aligned with the eco-
lead to survival and reproduction (Lindenmayer and systems that replace the disturbed land cover type.
Fischer 2007). On the other hand, land cover refers Fragmentation could also add habitat for species that

Vol:. (1234567890)
13
Landsc Ecol (2023) 38:2717–2730 2723

respond positively to the altered biophysical prop- Fragmentation metrics and measurement scales
erties of the remnant patches stemming from edge
effects (Ries et al. 2004; Fletcher et al. 2007). Indeed, After defining patch boundaries, researchers must
this increased heterogeneity could help explain why a also choose among the dozens of metrics that can
collection of small patches of a particular land cover be used to quantify fragmentation (Neel et al. 2004;
can often support more species than large patches of Wang et al. 2014). To some extent, fragmentation
the same area (Wintle et al. 2019). To be clear, under- metrics are confounded with study design as land-
standing the responses of individual species to frag- scape-scale studies, by definition, measure char-
mentation of both habitat and land cover types is use- acteristics describing the whole landscape (e.g.,
ful from a conservation perspective, but the former mean patch size or edge density in a landscape)
is a test of island biogeography and metapopulation while patch-scale studies measure characteristics
theory while the latter is a test of how species with of individual patches (e.g., focal patch size or edge
different habitat requirements respond to landscape density). Although previous research has compared
change. Failure to recognize this subtle, but impor- metrics in their abilities to quantify landscape pat-
tant, difference could help explain how researchers terns (e.g., Wang et al. 2014), empirical evidence
can find seemingly idiosyncratic responses of species linking landscape- or patch-scale fragmentation
to fragmentation (Betts et al. 2014). metrics with the mechanistic processes that could
lead to a fragmentation response in a species (e.g.,
Subjectivity in patch delineation dispersal ability, edge effects) are lacking (Li and
Wu 2004; but see Fletcher et al. 2018b).
Regardless of whether a researcher measures habi- Similarly, researchers must identify a scale at
tat or a land cover type, they must establish rules which to quantify selected fragmentation metrics,
regarding patch boundaries before measuring frag- such as when using a landscape-scale or focal-patch
mentation. Establishing rules is not necessarily study design (Fletcher et al. 2023). The choice
straightforward as it requires drawing hard lines of scale can have strong effects on the value of
in landscapes often characterized by gradual bio- these metrics (Wu 2004; Wu et al. 2002) and their
geophysical gradients (Fig. 3c). Where patches inferred relationships with biological responses
are separated by distinct boundaries, they may be (Holland et al. 2004). For example, Valente et al.
functionally connected depending on the species, (2023) demonstrated that habitat fragmentation had
its dispersal capabilities, and the matrix structure significant positive effects on the distribution of an
(Taylor et al. 1993; Ricketts 2001), which further endangered species when measured at a fine spatial
contributes to ambiguity regarding where one scale, but significant negative effects when meas-
patch stops and another starts. In many ways, patch ured at a broader spatial scale. Although much has
boundaries are thus often based on a series of sub- been said about matching spatial scales of meas-
jective decisions made by the researcher which urement with those at which relevant biological
could affect the values of measured fragmenta- processes operate (e.g., Jackson and Fahrig 2012),
tion metrics and thus the perceived relationship most ecological studies do not provide biological
between fragmentation and species or community justification for the scale at which landscapes are
distribution patterns. While this perception has led defined (Jackson and Fahrig 2014) which could
some researchers to argue that the patch-matrix lead to spurious conclusions regarding fragmenta-
model is of limited use in terrestrial systems (McI- tion effects.
ntyre and Barrett 1992; Fischer and Lindenmayer
2006; Fahrig 2013; Mendenhall et al. 2014), few Context dependency of fragmentation effects
studies have empirically investigated whether deci-
sions regarding how to define the landscape mosaic Evidence indicates that fragmentation effects can
affect study conclusions (but see Moilanen 2002). be highly context dependent which almost certainly
contributes added variance. For instance, Betts et al.
(2019) demonstrated the role of evolutionary context
where species that evolved in regions with frequent

Vol.: (0123456789)
13
 2724 Landsc Ecol (2023) 38:2717–2730

and severe disturbances were less likely to be nega- as familiar as species richness could vary depend-
tively affected by modern processes causing frag- ing on whether the focus is on local (alpha) richness,
mentation. Ecological contexts also play an impor- regional (gamma) richness, or their ratio (beta rich-
tant role, as characteristics of the intervening matrix ness; Valente et al. 2022).
(Prugh et al. 2008; Ricketts 2001) or the amount of
habitat remaining in the landscape (Andrén 1994; Other factors
Herse et al. 2020; Püttker et al. 2020) can moderate
fragmentation effects. For example, there is evidence The list of factors above is not meant to be compre-
that the effects of fragmentation can be amplified in hensive, and other study characteristics might affect a
landscapes with more (Herse et al. 2020; Püttker et al. researcher’s conclusions regarding the effects of frag-
2020) or less (Andrén 1994; Betts et al. 2007) habi- mentation on biodiversity. These include, for exam-
tat. Additionally, intra-species fragmentation effects ple, the amount of time since the habitat loss and
are known to vary across a species’ geographic range fragmentation occurred (Vellend et al. 2006; Haddad
(Banks-Leite et al. 2022; Valente et al. 2023), making et al. 2015; but see Fahrig 2020), how the confound-
it difficult to draw unequivocal conclusions from stud- ing effects of habitat amount are accounted for in
ies conducted at a single location within the range. estimating the effects of fragmentation per se (Koper
et al. 2007), and the variation in quality among rem-
Measurements of biodiversity nant habitat patches (Mortelliti et al. 2010). Given
such complexity, it is no wonder that researchers who
The United Nations defines biodiversity as the varia- work with different species using different measures
bility of life on Earth, including “diversity within spe- and scales of fragmentation in different regions can
cies, between species, and of ecosystems,” (Conven- arrive at contrasting conclusions. Moving forward,
tion on Biological Diversity 2010). In other words, fragmentation researchers could explore these dispar-
the variety of response variables that can be explored ities so that patterns can be elucidated and commu-
under the banner of “biodiversity” is massive nicated to practitioners and policymakers. Identify-
(Haddad et al. 2015; Fahrig 2017). As noted above, ing the patterns that exist and clearly delineating the
responses of individual species to fragmentation can situations in which these patterns do and do not hold
be idiosyncratic and influenced by habitat require- would benefit effective conservation policy (Fahrig
ments, life history, or both. Hence, the emergent et al. 2022).
structure of the biotic community then depends on
the cumulative effects of fragmentation on individual
species, subsequent inter-specific interactions, as well How do we unlock?
as interacting effects of local conditions (i.e., habitat
quality) and fragmentation. As a result, choices of Norberg et al. (2022) highlighted several instances
focal species or species groups (Bender et al. 1998; in which researchers with opposing viewpoints were
Valente and Betts 2019) can affect perception of how able to unite and make substantial advancements in
fragmentation affects biodiversity as can the choice their field through collaboration and productive dis-
to use taxonomic, phylogenetic, or functional diver- course (e.g., Kahneman and Klein 2009). These col-
sity metrics within the resulting community (Devictor laborations can be useful for elucidating the source
et al. 2010). Even within an individual species, frag- and extent of disagreement (Scott-Phillips et al.
mentation could have divergent effects on distribu- 2014), clarifying the study designs suitable for test-
tion, behavior, fitness, or population genetic structure. ing existing hypotheses, generating new testable
For example, fragmentation can simultaneously have hypotheses with buy-in from both sides of a debate
positive effects on the distribution of a species but (Matske et al. 2015), and bringing opposing views
negative effects on its reproductive output (Ries and closer together (Cowan et al. 2020). Indeed, research
Fagan 2003). As a result, the choice of which bio- indicates that collaborations among individuals with
logical process to measure may have strong effects on contrasting perspectives generally lead to higher-
whether positive or negative effects of fragmentation quality products (Shi et al. 2019). Thus, we hope that
on biodiversity are detected. Even effects on a metric a concerted effort by ecologists to come together and

Vol:. (1234567890)
13
Landsc Ecol (2023) 38:2717–2730 2725

unlock the fragmentation debate will be highly pro- et al. 2023) and again measure richness in the focal
ductive for understanding the ecological mechanisms patch (alpha diversity) and in all patches within the
structuring biodiversity and providing consistent, evi- buffer (gamma richness). We would then model alpha
dence-based recommendations to land managers and and gamma richness as a function of landscape-scale
conservation biologists around the planet. fragmentation metrics (e.g., number of patches in the
landscape) or patch-scale fragmentation metrics (e.g.,
Action items for individuals focal patch size) while statistically controlling for
habitat loss by including remnant habitat area within
As a first step in bringing ecologists together, we sug- landscapes or patch buffers as a covariate in our mod-
gest that the discipline would be enhanced by test- els. Such a study design would also allow evaluating
ing the hypothesized ecological and methodological interactions between the two factors (see next para-
mechanisms that could underlie inconsistent results graph). This exercise of explicitly testing multiple
described above. It will be important to start with the hypotheses can reduce intellectual myopia and the
acknowledgment that our colleagues are competent post-hoc tendency to construct compelling narra-
and acting in good faith, which will fuel open-minded tives around results, including just-so stories (Nuzzo
hypothesis development and testing (Loehle 1987). 2015).
That said, science is conducted by humans who are It is also important to consider among the set of
inherently subject to implicit biases (Betini et al. alternative hypotheses explanations that involve two
2017). Further, the need to acquire grants and publi- or more potentially interacting factors rather than
cations for career advancement can foment absolute only those that limit inference to mutually exclusive
thinking (see Norberg et al. 2022 for details). One univariate hypotheses (Hilborn and Stearns 1982).
way for all scientists to avoid these pitfalls is to push Inclusion of multifactor explanations reduces the like-
themselves to consider multiple alternative hypoth- lihood of eliminating potential synergistic causes of a
eses or hierarchies of hypotheses when designing pattern or factors that only act in a subset of scenar-
studies and then evaluate their relative support based ios and is vital to recognizing compatibility among
on empirical results (Betini et al. 2017). The multi- multiple competing hypotheses (Hilborn and Stearns
ple working hypotheses approach was originally sug- 1982; McIntire and Fajardo 2009). For example,
gested by Chamberlin (1890) to circumvent becoming physiological stressors, competition, and bioclimatic
emotionally attached to a favorite hypothesis and pre- suitability experienced by a species vary in space and
maturely dismissing alternative explanations of phe- these factors can interact with landscape disturbances
nomena (Betts et al. 2021). Indeed, this approach was to cause heterogeneity in effects of fragmentation
later championed by Platt (1964) as integral to strong across the species’ range (Banks-Leite et al. 2022;
inference and rapid scientific progress. Valente et al. 2023). As another example, Fahrig et al.
For example, to understand why results from stud- (2022) proposed (and suggested methods for testing)
ies measuring the relationship between fragmenta- the “SLOSS cube hypothesis,” which posits that the
tion of a landcover type and species richness vary effects of fragmentation per se on gamma diversity
so widely, we could simultaneously test the hypoth- can be predicted from the intersection of three fac-
eses that this relationship is driven by (i) the scale tors: between-patch movement, across-habitat hetero-
at which fragmentation is measured (landscape vs. geneity, and the effects of spreading-of-risk on land-
patch) or (ii) the scale at which biodiversity is meas- scape-scale population persistence.
ured (alpha vs. gamma diversity). We could test these Finally, we also suggest researchers consider bio-
two hypotheses by first selecting landscapes across logical null models that address patterns and pro-
a gradient of fragmentation per se (keeping habitat cesses expected in the absence of mechanisms being
amount constant) and then measuring diversity at tested. Indeed, some models for the neutral geometric
multiple sites across each landscape to allow calcula- effects of fragmentation on biodiversity have recently
tion of both alpha and gamma richness. Subsequently, been developed (e.g., May et al. 2019; Deane et al.
we could select one individual patch within each 2022). These null models allow researchers to test
landscape and draw an ecologically relevant buffer whether hypothesized mechanisms are necessary and
around that patch (i.e., a focal-patch design; Fletcher sufficient to explain observed patterns (Pearson and

Vol.: (0123456789)
13
 2726 Landsc Ecol (2023) 38:2717–2730

Gardner 1997). Understanding what processes or situ- Conservation Biology. The goal of these symposia
ations might allow the patterns expected under the would be to highlight and discuss different perspec-
null model to occur can also lead to more biologically tives, or to develop a working group that generates a
realistic alternative hypotheses (Nuzzo 2015). set of consistent principles to communicate to policy
makers and land managers. This approach has prec-
Community‑level action items edence in conservation biology (Lindenmayer et al.
2007). Indeed, straightforward efforts to reach out to
We propose that the way to resolve scientific disa- colleagues with opposing positions to develop col-
greement on the human front is to facilitate con- laborative projects examining the root causes of con-
structive conversation between opposing sides of flicting findings will likely lead to a much stronger
the debate, allowing all perspectives to be heard and understanding of how landscape configuration struc-
understood. We therefore envision a collaborative tures biological communities and will represent an
effort among ecologists with the main goals being important step towards a more productive scientific
to: (1) develop a set of consensus principles to serve discourse.
as best practices for landscape management and bio-
diversity conservation; and (2) identify hypotheses
that may help explain the continued lack of consen- Conclusions
sus regarding fragmentation effects on biodiversity
(building on our incipient effort here) and use these Nearly 50 years after the initial application of island
hypotheses to steer future research into when, where, biogeography theory to terrestrial conservation plan-
why, and how fragmentation affects biodiversity (e.g., ning, we appear to be no closer to a discipline-wide
Fahrig et al. 2022). consensus regarding the effects of fragmentation on
One potential way to achieve both goals is to use biodiversity or the role of landscape configuration in
the Delphi method, a technique developed to help designing effective reserve networks (Fig. 2). We as
groups of experts (hereafter "participants") achieve scientists are responsible for this debate and should
consensus on complex problems (see Mukherjee et al. hold ourselves accountable for helping to settle it
2015 for review). In a Delphi study, communication and ultimately provide clear, scientifically defensible,
between participants is highly structured, anonymous, and timely messaging to managers and policymakers.
and indirect. First, participants respond to a survey, Critically evaluating management and policy recom-
after which a facilitator summarizes the responses mendations indicated by multiple competing hypothe-
and shares the summary with the group; the sum- ses may enable recognition of the degree of common-
mary can also include explanations for any dissenting ality in their practical implications and consolidate
responses. Then, participants take the survey again, points of agreement regarding conservation strategy
providing an opportunity to revise answers based on (Sutherland et al. 2019). With a renewed spirit of col-
group feedback. The process continues until consen- laboration, curiosity, and humility, we hope that ecol-
sus (or an alternate goal) is reached. Although use of ogists can begin to unlock the fragmentation debate
the Delphi method remains relatively rare in ecology and provide sound, unified advice to policymakers
(Mukherjee et al. 2015), this approach has neverthe- and practitioners who desperately need it.
less proven useful in numerous conservation appli-
cations, including evaluating the conservation value Funding This work was supported in part by NSF grant
LTER8 DEB-2025755.
of different forest types (Geneletti 2007), estimating
species-specific connectivity (Scolozzi and Geneletti Declarations
2012), summarizing the effects of forest management
on biodiversity (Filyushkina et al. 2018), and assign- Competing interests The authors have not disclosed any
competing interests.
ing IUCN Red List status (McBride et al. 2012).
Another logical step may be to organize a series
of symposia at international scientific meetings, such
as that of the International Association for Land-
scape Ecology or the International Congress for

Vol:. (1234567890)
13
Landsc Ecol (2023) 38:2717–2730 2727

References Camos V, Logie RH (2020) How do scientific views

change? Notes from an extended adversarial collabo-
Adler PB et al (2011) Productivity is a poor predictor of plant ration. Perspect Psychol Sci. https://​doi.​org/​10.​1177/​
species richness. Science 333:1750–1753 17456​91620​906415
Aenǵst P (1999) The Yellowstone to Yukon initiative: A new Deane DC, Xing D, Hui C, McGeoch M, He F (2022) A null
conservation paradigm to protect the heart of North model for quantifying the geometric effect of habitat
America. Proceedings of a conference on the biology and subdivision on species diversity. Glob Ecol Biogeogr
management of species and habitats at risk Vol. 2. BC, 31:440–453
Kamloops Debinski DM, Holt RH (2000) A survey and overview of habi-
Andrén H (1994) Effects of habitat fragmentation on birds and tat fragmentation experiments. Conserv Biol 14:342–355
mammals in landscapes with different proportions of Devictor V, Mouillot D, Meynard C, Jiguet F, Thuiller W,
suitable habitat: a review. Oikos 71:355–366 Mouquet N (2010) Spatial mismatch and congruence
Banks-Leite C, Betts MG, Ewers RM, Orme CDL, Pigot AL between taxonomic, phylogenetic and functional diver-
(2022) The macroecology of landscape ecology. Trends sity: the need for integrative conservation strategies in a
Ecol Evol 37:480–487 changing world. Ecol Lett 13:1030–1040
Bender DJ, Contreras TA, Fahrig L (1998) Habitat loss and Diamond JM (1975) The island dilemma: lessons of modern
population decline: a meta-analysis of the patch size biogeographic studies for the design of natural reserves.
effect. Ecology 79:517–533 Biol Conserv 7:129–146
Betini GS, Avgar T, Fryxell JM (2017) Why are we not evalu- Evju M, Sverdrup-Thygeson A (2016) Spatial configuration
ating multiple competing hyptheses in ecology and evo- matters: a test of the habitat amount hypothesis for plants
lution? Royal Soc Open Sci. https://​doi.​org/​10.​2307/​ in calcareous grasslands. Landsc Ecol 31:1891–1902
35458​23 Fahrig L (2003) Effects of habitat fragmentation on biodiver-
Betts MG, Forbes GJ, Diamond AW (2007) Thresholds in sity. Annu Rev Ecol Evol Syst 34:487–515
songbird occurrence in relation to landscape structure. Fahrig L (2013) Rethinking patch size and isolation effects: the
Cons Biol 21:1046–1058 habitat amount hypothesis. J Biogeogr 40:1649–1663
Betts MG, Forbes GJ, Diamond AW, Taylor PD (2006) Inde- Fahrig L (2015) Just a hypothesis: a reply to Hanski. J of Bio-
pendent effects of fragmentation on forest songbirds: an geogr 42:993–994
organism-based approach. Ecol Appl 16:1076–1089 Fahrig L (2017) Ecological responses to habitat fragmentation
Betts MG, Fahrig L, Hadley AS, Halstead KE, Bowman J, per se. Annu Rev Ecol Evol Syst 48:1–23
Robinson WD, Wiens JA, Lindenmayer DB (2014) A Fahrig L (2020) Why do several small patches hold more
species-centered approach for uncovering generalities in species than few large patches? Glob Ecol Biogeogr
organism responses to habitat loss and fragmentation. 29:615–628
Ecography 37:517–527 Fahrig L (2021) What the habitat amount hypothesis does
Betts MG, Wolf C, Pfeifer M, Banks-Leite C, Arroyo-Rod- and does not predict: a reply to Saura. J Biogeogr
ríguez V, Ribeiro DB, Barlow J, Eigenbrod F, Faria D, 48:1530–1535
Fletcher RJ, Hadley AS, Hawes JE, Holt RD, Klingbeil Fahrig L, Arroyo-Rodríguez V, Bennett JR, Boucher-Lalonde
B, Kormann U, Lens L, Levi T, Medina-Rangel GF, V, Cazetta E, Currie DJ, Eigenbrod F, Ford AT, Harrison
Melles SL, Ewers RM (2019) Extinction filters mediate SP, Jaeger JAG, Koper N, Martin AE, Martin JL, Metzger
the global effects of habitat fragmentation on animals. JP, Morrison P, Rhodes JR, Saunders DA, Simberloff D,
Science 366:1236–1239 Smith AC, Watling JI (2019) Is habitat fragmentation
Betts MG, Hadley AS, Frey DW, Frey SJK, Gannon D, Har- bad for biodiversity? Biol Conserv 230:179–186
ris SH, Kim H, Kormann UG, Leimberger K, Moriarty Fahrig L, Watling JI, Arnillas CA, Arroyo-Rodríguez V,
K, Northrup JM, Phalan B, Rousseau JS, Stokely TD, Jörger-Hickfang T, Müller J, Pereira H, Riva F, Rösch
Valente JJ, Wolf C, Zárrate-Charry D (2021) When are V, Seibold S, Tscharntke T, May F (2022) Resolving
hypotheses useful in ecology and evolution? Ecol Evol the SLOSS dilemma for biodiversity conservation: a
11:5762–5776 research agenda. Biol Rev 97:99–114
Cardinale BJ, Gonzalez A, Allington GRH, Loreau M (2018) Filyushkina A, Strange N, Löf M, Ezebilo EE, Boman M
Is local biodiversity declining or not? A summary of (2018) Applying the Delphi method to assess impacts of
the debate over analysis of species richness time trends. forest management on biodiversity and habitat preserva-
Biol Conserv 219:175–183 tion. For Ecol Manag 409:179–189
Chamberlin TC (1890) The method of multiple working Fischer J, Lindenmayer D (2006) Beyond fragmentation: the
hypotheses. Science 15:92–96 continuum model for fauna research and conservation in
Chase JM, Blowes SA, Knight TM, Gerstner K, May F human-modified landscapes. Oikos 112:473–480
(2020) Ecosystem decay exacerbates biodiversity loss Fletcher RJ Jr, Ries L, Battin J, Chalfoun AD (2007) The
with habitat loss. Nature 584:238–243 role of habitat area and edge in fragmented landscapes:
Convention on Biological Diversity (2010) Global Biodiver- definitively distinct or inevitably intertwined? Can J Zool
sity Outlook 3. Montréal, 94 pages. http://​www.​cbd.​int/​ 85:1017–1030
gbo/​gbo3/​doc/​GBO3-​final-​en.​pdf Fletcher RJ Jr, Didham RK, Banks-Leite C, Barlow J, Ewers
Cowan N, Belletier C, Doherty JM, Jaroslawska AJ, Rho- RM, Rosindell J, Holt RD, Gonzalez A, Pardini R, Dam-
des S, Forsberg A, Naveh-Benjamin M, Barrouillet P, schen EI, Melo FP (2018) Is habitat fragmentation good
for biodiversity? Biol Conserv 226:9–15

Vol.: (0123456789)
13
 2728 Landsc Ecol (2023) 38:2717–2730

Fletcher RJ Jr, Reichert BE, Holmes K (2018b) The negative Jackson HB, Fahrig L (2014) Are ecologists conducting
effects of habitat fragmentation operate at the scale of research at the optimal scale. Glob Ecol Biogeogr
dispersal. Ecology 99:2176–2186 24:52–63
Fletcher RJ Jr, Smith TAH, Kortessis N, Bruna EM, Holt RD Kahneman D, Klein G (2009) Conditions for intuitive exper-
(2023) Landscape experiments unlock relationships tise: a failure to disagree. Amer Psychol 64:515–526
among habitat loss, fragmentation, and patch-size effects. Koper N, Schmiegelow FKA, Merrill EH (2007) Residuals
Ecology. https://​doi.​org/​10.​1002/​ecy.​4037 cannot distinguish between ecological effects of habitat
Fridley JD, Grime JP, Huston MA, Pierce S, Smart SM, amount and fragmentation : implications for the debate.
Thompson K et al (2012) Comment on “Productivity Landsc Ecol 22:811–820
is a poor predictor of plant species richness.” Science Kuhn T (1962) The structure of Scientific Revolutions. Univer-
335:1441 sity of Chicago Press, Chicago
Geneletti D (2007) Expert panel-based assessment of for- Laurance W (2008) Theory meets reality: how habitat frag-
est landscapes for land use planning. Mt Res Dev mentation research has transcended island biogeographic
27:220–223 theory. Biol Conserv 141:1731–1744
Gilbert-Norton L, Wilson R, Stevens JR, Beard KH (2010) A Levins R (1969) Some demographic and genetic consequences
meta-analytic review of corridor effectiveness. Cons Biol of environmental heterogeneity for biological control.
24:660–668 Bull Entomol Soc Am 15:237–240
Gonzalez A, Cardinale BJ, Allington GRH, Byrnes J, Endsley Li H, Wu J (2004) Use and misuse of landscape indices.
KA, Brown DG, Hooper DU, Isbell F, O’Connor MI, Landsc Ecol 19:389–399
Loreau M (2016) Estimating local biodiversity change: a Lindenmayer DB, Fischer J (2007) Tackling the habitat frag-
critique of papers claiming no net loss of local diversity. mentation panchreston. Trends Ecol Evol 22:127–132
Ecology 97:1949–1960 Lindenmayer DB, Hobbs RJ, Montague-Drake R, Alexandra
Grace JB et al (2012) Response to comments on “Productiv- J, Bennett A, Burgman M, Cale P, Calhoun A, Cramer
ity is a poor predictor of plant species richness.” Science V, Cullen P, Driscoll D, Fahrig L, Fischer J, Frank-
335:1441 lin J, Haila Y, Hunter M, Gibbons P, Lake S, Luck G,
Haddad NM, Brudvig LA, Clobert J, Davies KF, Gonzalez A, MacGregor C, McIntyre S, Mac Nally R, Manning A,
Holt RD, Lovejoy TE, Sexton JO, Austin MP, Collins Miller J, Mooney H, Noss R, Possingham H, Saunders
CD, Cook WM, Damschen EI, Ewers RM, Foster BL, D, Schmiegelow F, Scott M, Simberloff D, Sisk T, Tabor
Jenkins CN, King AJ, Laurance WF, Levey DJ, Mar- G, Walker B, Wiens J, Woinarski J, Zavaleta E (2007) A
gules CR, Townshend JR (2015) Habitat fragmentation checklist for ecological management of landscapes for
and its lasting impact on Earth’s ecosystems. Sci Adv conservation. Ecol Lett 11:78–91
1:e1500052 Lindgren JP, Cousins SAO (2017) Island biogeography theory
Haddad NM, Gonzalez A, Brudvig LA, Burt MA, Levey DJ, outweighs habitat amount hypothesis in predicting plant
Damschen EI (2017) Experimental evidence does not species richness in small grassland remnants. Landsc
support the Habitat amount hypothesis. Ecography Ecol 32:1895–1906
40:48–55 Loehle C (1987) Hypothesis testing in ecology: psychological
Halstead KE, Alexander JD, Hadley AS, Stephens JL, Yang J, aspects and the importance of theory maturation. Q Rev
Betts MG (2019) Using a species-centered approach to Biol 62:397–409
predict bird community responses to habitat fragmenta- MacArthur RH, Wilson EO (1967) The theory of Island Bioge-
tion. Landsc Ecol 34:1919–1935 ography. Princeton University Press, Princeton
Hanski I (2015) Habitat fragmentation and species richness. J Matske D, Nieuwenhuis S, van Rijn H, van der Slagter HA,
Biogeogr 42:989–994 Wagenmak E (2015) The effects of horizontal eye
Hatfield JH, Orme CDL, Tobias JA, Banks-Leite C (2018) movements on free recall: a preregistered adversarial
Trait-based indicators of bird species sensitivity to collaboration. J Exp Psychol Gen. https://​doi.​org/​10.​
habitat loss are effective within but not across data sets. 1037/​xge00​00038
Ecol Appl 28:28–34 May RM (1975) Island biogeography and the design of wild-
Herse MR, With KA, Boyle WA (2020) Grassland fragmen- life preserves. Nature 254:177–178
tation affects declining tallgrass prairie birds most May F, Rosenbaum B, Schurr FM, Chase JM (2019) The
where large amounts of grassland remain. Landsc Ecol geometry of habitat fragmentation: Effects of species
35:2791–2804 distribution patterns on extinction risk due to habitat
Hilborn R, Stearns SC (1982) On inference in ecology and conversion. Ecol Evol 9:2775–2790
evolutionary biology: the problem of multiple causes. McBride MF, Garnett ST, Szabo JK et al (2012) Structured
Acta Biotheor 31:145–164 elicitation of expert judgments for threatened species
Holland JD, Bert DG, Fahrig L (2004) Determining the spa- assessment: a case study on a continental scale using
tial scale of a species’ response to habitat. Bioscience email. Methods Ecol Evol 3:906–920
54:227–233 McGarigal K, Cushman SA (2002) Comparative evaluation
Independent Evaluation Group (2011) The mesoamerican of experimental approaches to the study of habitat frag-
biological corridor. World Bank, Washington mentation effects. Ecol Appl 12:335–345
Jackson HB, Fahrig L (2012) What size is a biologically rel- McIntire EJB, Fajardo A (2009) Beyond description: the
evant landscape? Landsc Ecol 27:929–941 active and effective way to infer processes from spatial
patterns. Ecology 90:46–56

Vol:. (1234567890)
13
Landsc Ecol (2023) 38:2717–2730 2729

Mcintyre S, Barrett GW (1992) Habitat variegation, an alter- Saura S (2020) The habitat amount hypothesis implies nega-
native to fragmentation. Conserv Biol 6:146–147 tive effects of habitat fragmentation on species richness.
Mendenhall CD, Karp DS, Meyer CFJ, Hadly EA, Daily GC J Biogeogr 48:11–22
(2014) Predicting biodiversity change and averting col- Saura S (2021) The habitat amount hypothesis predicts that
lapse in agricultural landscapes. Nature 509:213–217 habitat fragmentation poses a threat to biodiversity. J
Miller K, Chang E, Johnson N (2001) Defining common Biogeogr 48:1536–1540
ground for the Mesoamerican Biological Corridor. Scolozzi R, Geneletti D (2012) Assessing habitat connectiv-
World Resources Institute, Washington ity for land-use planning: a method integrating land-
Moilanen A (2002) Implications of empirical data quality to scape graphs and Delphi survey. J Environ Plan Manag
metapopulation model parameter estimation and appli- 55:813–830
cation. Oikos 96:516–530 Scott-Phillips TC, Laland KN, Shuker DM, Dickins TE, West
Mortelliti A, Amori G, Boitani L (2010) The role of habitat SA (2014) The niche construction perspective: a critical
quality in fragmented landscapes: a conceptual over- appraisal. Evolution 68:1231–1243
view and prospectus for future research. Oecologia Shi F, Teplitskiy M, Duede E, Evans JA (2019) The wisdom of
163:535–547 polarized crowds. Nat Hum Behav 3:329–336
Mukherjee N, Hugé J, Sutherland WJ, McNeill J, Van Opstal Simberloff DS, Abele LG (1976) Island biogeography theory
M, Dahdouh-Guebas F, Koedam N (2015) The Del- and conservation practice. Science 191:285–286
phi technique in ecology and biological conserva- Spies TA, Long JW, Charnley S, Hessburg PF, Marcot BG,
tion: applications and guidelines. Methods Ecol Evol Reeves GH, Lesmeister DB, Reilly MJ, Cerveny LK,
6:1097–1109 Stine PA, Raphael MG (2019) Twenty-five years of the
Neel MC, McGarigal K, Cushman SA (2004) Behavior of Northwest forest plan :what have we learned ? Front Ecol
class-level landscape metrics across gradients of class Environ 17:511–520
aggregation and area. Landsc Ecol 19:435–455 Sutherland et al (2019) Building a tool to overcome barriers in
Norberg J, Blenckner T, Cornell SE, Petchey OL, Hillebrand research-implementation spaces :the conservation evi-
H (2022) Failures to disagree are essential for environ- dence database. Biol Conserv 238:108199. https://​doi.​
mental science to effectively influence policy develop- org/​10.​1016/j.​biocon.​2019.​108199
ment. Ecol Lett 25:1075–1093 Taubert F, Fischer R, Groeneveld J, Lehmann S, Müller, Rödig
Nuzzo R (2015) Fooling ourselves. Nature 526:182–185 E, Wiegand T, Huth A (2018) Global patterns of tropical
Pan X, Liu F, Zhang M (2012) Comment on “Productivity forest fragmentation. Nature 554:519–522
is a poor predictor of plant species richness.” Science Taylor PD, Fahrig L, Henein K, Merriam G (1993) Connec-
335:1441 tivity is a vital element of landscape structure. Oikos
Pearson SM, Gardner RH (1997) Neutral models: useful 68:571–573
tools for understanding landscape patterns. In: Bisson- Terborgh J (1974) Preservation of natural diversity: the prob-
ette JA (ed) Wildlife and Landscape Ecology. Springer, lem of extinction prone species. Bioscience 24:715–722
New York Thornton DH, Branch LC, Sunquist ME (2011) The influence
Pfeifer M, Lefebvre V et al (2014) BIOFRAG—a new data- of landscape, patch, and within-patch factors on species
base for analyzing BIOdiversity responses to forest presence and abundance: a review of focal patch studies.
FRAGmentation. Ecol Evol 4:1524–1537 Landsc Ecol 26:7–18
Platt JR (1964) Strong inference. Science 146:347–353 U.S. Department of the Interior, U.S. Department of Agricul-
Popper KR (1959) The logic of Scientific Discovery. Hutch- ture, U.S. Department of Commerce, Council on Envi-
inson, London ronmental Quality (2021) Conserving and restoring
Prugh LR, Hodges KE, Sinclair ARE, Brashares JS (2008) America the beautiful. https://​www.​doi.​gov/​sites/​doi.​
Effect of habitat area and isolation on fragmented animal gov/​files/​report-​conse​r ving-​and-​resto​r ing-​ameri​ca-​the-​
populations. P Nat A Sci 105:20770–20775 beaut​iful-​2021.​pdf. Accessed 5 Dec 2022
Püttker T, Crouzeilles R, Almeida-Gomes M et al (2020) Indi- Valente JJ, Betts MG (2017) Using dynamic occupancy and
rect effects of habitat loss via habitat fragmentation: a species distribution models to understand species sen-
cross-taxa analysis of forest-dependent species. Biol sitivity to landscape configuration. In: Marra PP, Ryder
Cons 241:1083268 TB, Sillett S, Betts M, Siegel R, Saracco J, Fischer R.
Quammen D (1997) The song of the Dodo; Island Biogeogra- Using a hierarchical approach to model regional source-
phy in an age of extinction. Scribner, New York sink dynamics for Neotropical-Nearctic songbirds to
Ricketts TH (2001) The matrix matters: effective isolation in inform management practices on Department of Defense
fragmented landscapes. Am Nat 158:87–99 installations. Final Report: SERDP Project RC-2121.
Ries L, Fagan WF (2003) Habitat edges as a potential ecologi- https://​serdp-​estcp.​org/​proje​cts/​detai​ls/​5dbdd​303-​fb46-​
cal trap for an insect predator. Ecol Entomol 28:567–572 496b-​bf54-​a8e91​8dc2e​21
Ries L, Fletcher RJ, Battin J, Sisk TD (2004) Ecological Valente JJ, Betts MG (2019) Response to fragmentation by
responses to habitat edges: mechanisms, models, and var- avian communities is mediated by species traits. Divers
iability explained. Annu Rev Ecol Evol Syst 35:491–522 Distrib 25:48–60
Riva F, Fahrig L (2022) The disproportionately high value of Valente JJ, Bennett RE, Gomez C, Bayly NJ, Rice RA, Ryder
small patches for biodiversity conservation. Conserv Lett TB, Sillett TS (2022) Land-sparing and land-sharing
15:e12881 provide complementary benefits for conserving avian

Vol.: (0123456789)
13
 2730 Landsc Ecol (2023) 38:2717–2730

biodiversity in coffee-growing landscapes. Biol Conserv Wilson EO, Willis EO (1975) Applied biogeography. In: Cody
270:109568 ML, Diamond JM (eds) Ecology and Evolution of Com-
Valente JJ, Rivers JW, Yang Z, Nelson SK, Northrup JM, Roby munities. Belknap, Cambridge
DD, Meyer CB, Betts MG (2023) Fragmentation effects Wintle BA, Kujala H, Whitehead A, Cameron A, Veloz S,
on an endangered species across a gradient from the inte- Kukkala A, Moilanen A, Gordon A, Lentini PE, Caden-
rior to edge of its range. Cons. Biol. https://​doi.​org/​10.​ head NCR, Bekessy SA (2019) Global synthesis of con-
1111/​cobi.​14091 servation studies reveals the importance of small habitat
Vellend M (2017) The biodiversity conservation paradox. Am patches for biodiversity. P Nat A Sci 116:909–914
Sci 105:94–101 With KA, Payne AR (2021) An experimental test of the habitat
Vellend M, Verheyen K, Jacquemyn H, Kolb A, Van Calster amount hypothesis reveals little effect of habitat area but
H, Peterken G, Hermy M (2006) Extinction debt of forest transient or indirect effects of fragmentation on local spe-
plants persists for more than a century following habitat cies richness. Landsc Ecol 36:2505–2517
fragmentation. Ecology 87:542–548 Wu J (2004) Effects of changing scale on landscape pattern
Vellend M, Dornelas M, Baeten L et al (2017) Estimates of analysis: scaling relations. Landsc Ecol 19:125–138
local biodiversity change over time stand up to scrutiny. Wu J, Shen W, Sun W, Tueller PT (2002) Empirical patterns
Ecology 98:583–590 of the effects of changing scale on landscape metrics.
Wang X, Blanchet FG, Koper N (2014) Measuring habitat frag- Landsc Ecol 17:761–782
mentation: an evaluation of landscape pattern metrics.
Methods Ecol Evol 5:634–646 Publisher’s Note Springer Nature remains neutral with regard
Watling JI, Arroyo-Rodríguez, Pfeifer M et al (2020) Support to jurisdictional claims in published maps and institutional
for the habitat amount hypothesis from a global synthesis affiliations.
of species density studies. Ecol Lett 23:674–681
Wilcox BA, Murphy DD (1985) Conservation strategy:
the effects of fragmentation on extinction. Am Nat
125:879–887

Vol:. (1234567890)
13