Conservation of Fragmented Populations

In this paper we argue that landscape spatial structure is of central importance in understanding the effects of fragmentation on population survival. Landscape spatial structure is the spatial relationships among habitat patches and the matrix in which they are embedded. Many general models of subdivided populations make the assumptions that (1) all habitat patches are equivalent in size and quality and (2) all local populations (in the patches) are equally accessible by dispersers. Models that gloss over spatial details of landscape structure can be useful for theoretical developments but will almost always be misleading when applied to real-world conservation problems. We show that local extinctions of fragmented populations are common. From this it follows that recolonization of local extinctions is critical for regional survival of fragmented populations. The probability of recolonization depends on (1) spatial relationships among landscape elements used by the population, including habitat patches for breeding and elements of the inter-patch matrix through which dispersers move, (2) dispersal characteristics of the organism of interest, and (3) temporal changes in the landscape structure. For endangered species, which are typically restricted in their dispersal range and in the kinds of habitat through which they can disperse, these factors are of primary importance and must be explicitly considered in management decisions.     

Effectiveness of Corridors Relative to Enlargement of Habitat Patches

Abstract:  The establishment of biological corridors between two otherwise isolated habitat patches is a common yet contentious strategy for conserving populations in fragmented landscapes. We compared the effectiveness of corridors with the effectiveness of an alternate conservation strategy, the enlargement of existing habitat patches. We used a spatially explicit population model that simulated population size in two kinds of patches. One patch had a corridor that connected it to a larger "source" patch and the other patch was unconnected and enlarged at the periphery by an area the same size as the corridor. Patch isolation, corridor width, patch size, and the probability that individuals would cross the border from habitat to matrix were varied independently. In general, population size was greater in enlarged patches than in connected patches when patches were relatively large and isolated. Corridor width and the probability of crossing the border from habitat to matrix did not affect the relative benefit of corridors versus patch enlargement. Although biological corridors may mitigate potential effects of inbreeding depression at long time scales, our results suggest that they are not always the best method of conserving fragmented populations.     

Corridor Length and Patch Colonization by a Butterfly, Junonia coenia

Abstract: Corridors have been proposed to reduce isolation and increase population persistence in fragmented landscapes, yet little research has evaluated the types of landscapes in which corridors will be most effective. I tested the hypothesis that corridors increase patch colonization by a butterfly, Junonia coenia , regardless of the butterfly's initial distance from a patch. I chose J. coenia because it has been shown to move between patches preferentially through corridors. Individuals were released 16–192 m away from open experimental patches into adjacent open corridors or forest. Neither corridors nor distance had a significant effect on patch colonization, but there was a significant interaction between the presence or absence of corridors and distance. At small distances (16–64 m), J. coenia was more likely to colonize open patches when released within forest than within open corridors, most likely because J. coenia used corridors as habitat. Nevertheless, patch colonization by butterflies released within forest decreased rapidly as distance from patches increased, as predicted by a null model of random movement. Colonization did not change with distance in the corridor, and at long distances (128–192 m), butterflies released in corridors were twice as likely to colonize open patches as those released in forest. These results suggest that one critical factor, interpatch distance, may determine the relative effectiveness of corridors and other landscape configurations, such as stepping stones, in reducing isolation in fragmented landscapes. When distances between patches are short compared to an animal's movement ability, a stepping stone approach may most effectively promote dispersal. Alternatively, the conservation value of corridors is highest relative to other habitat configurations when longer distances separate patches in fragmented landscapes.     

The Effect of Fragment Shape and Species' Sensitivity to Habitat Edges on Animal Population Size

Abstract:  Habitat fragmentation causes extinction of local animal populations by decreasing the amount of viable "core" habitat area and increasing edge effects. It is widely accepted that larger fragments make better nature reserves because core-dwelling species have a larger amount of suitable habitat. Nevertheless, fragments in real landscapes have complex, irregular shapes. We modeled the population sizes of species that have a representative range of preferences for or aversions to habitat edges at five spatial scales (within 10, 32, 100, 320, and 1000 m of an edge) in a nation-wide analysis of forest remnants in New Zealand. We hypothesized that the irregular shapes of fragments in real landscapes should generate statistically significant correlations between population density and fragment area, purely as a "geometric" effect of varying species responses to the distribution of edge habitat. Irregularly shaped fragments consistently reduced the population size of core-dwelling species by 10–100%, depending on the scale over which species responded to habitat edges. Moreover, core populations within individual fragments were spatially discontinuous, containing multiple, disjunct populations that inhabited small spatial areas and had reduced population size. The geometric effect was highly nonlinear and depended on the range of fragment sizes sampled and the scale at which species responded to habitat edges. Fragment shape played a strong role in determining population size in fragmented landscapes; thus, habitat restoration efforts may be more effective if they focus on connecting disjunct cores rather than isolated fragments.     

Geometrical Factors and Metapopulation Dynamics of the Bush Cricket, Metrioptera bicolor Philippi (Orthoptera: Tettigoniidae)

This paper presents a metapopulation study of the bush cricket, Metrioptera bicolor , living in a recently fragmented landscape. The species inhabits grass and heathland patches of varying area and isolation. Analyses are made of how these geometrical factors affect local population size and density, distribution pattern, and the probability of local extinction and colonization. The proportion of available patches occupied varied between 72 and 79% during 1985–1990. Unoccupied patches were smaller and more isolated than those that were occupied. Patches where populations became extinct during this period were smaller than those with persisting populations. Since local population size was well correlated with patch area, it was clear that stochastic extinctions only occurred in small populations. Critical patch size for population extinction was approximately half a hectare. Colonized patches were less isolated than those that had not been colonized. Critical inter-patch distance for colonization was about 100 meters. The turnover was restricted to an identifiable share of the available patches. Only 33% of the patches were so small that extinction due to stochastic causes could be considered highly probable. This metapopulation will therefore most likely persist over a considerable period in its present spatial structure. There are apparent threats of further fragmentation, however, and nothing is known about the likelihood of large-scale extinctions resulting from extremely unfavorable weather conditions. Nevertheless, our results show that it is appropriate to include geometrical factors in metapopulation models.     

Comparison of Two Types of Metapopulation Models in Real and Artificial Landscapes

Abstract: Application of metapopulation models is becoming increasingly widespread in the conservation of species in fragmented landscapes. We provide one of the first detailed comparisons of two of the most common modeling techniques, incidence function models and stage-based matrix models, and test their accuracy in predicting patch occupancy for a real metapopulation. We measured patch occupancies and demographic rates for regional populations of the Florida scrub lizard (   Sceloporus woodi ) and compared the observed occupancies with those predicted by each model. Both modeling strategies predicted patch occupancies with good accuracy ( 77–80%) and gave similar results when we compared hypothetical management scenarios involving removal of key habitat patches and degradation of habitat quality. To compare the two modeling approaches over a broader set of conditions, we simulated metapopulation dynamics for 150 artificial landscapes composed of equal-sized patches (2–1024 ha) spaced at equal distances (50–750 m). Differences in predicted patch occupancy were small to moderate (<20%) for about 74% of all simulations, but 22% of the landscapes had differences openface> 50%. Incidence function models and stage-based matrix models differ in their approaches, assumptions, and requirements for empirical data, and our findings provide evidence that the two models can produce different results. We encourage researchers to use both techniques and further examine potential differences in model output. The feasibility of obtaining data for population modeling varies widely among species and limits the modeling approaches appropriate for each species. Understanding different modeling approaches will become increasingly important as conservation programs undertake the challenge of managing for multiple species in a landscape context.     

Fragments as Islands: a Synthesis of Faunal Responses to Habitat Patchiness

Abstract:  Scientific interest in the impact of habitat fragmentation on biodiversity is increasing, but our understanding of fragmentation is clouded by a lack of appreciation for fundamental similarities and differences across studies representing a wide range of taxa and landscape types. In an effort to synthesize data describing ecological responses of animals to fragmentation across two classes of independent variables (taxonomic group and landscape), we sampled 148 studies of five major faunal groups from the primary literature and analyzed data on 13 variables extracted from those studies. We focused our analyses on three classes of dependent variables (effects of area and isolation on species richness, z values, and nestedness and species composition). Area ranged over more orders of magnitude than isolation and tended to explain more variation in species richness than isolation. There were few matrix or taxon effects on the patterns we investigated, although we did find that sky islands tended to manifest isolation effects on both species richness and nestedness more frequently than other patch types. Sky islands may offer insight into the future of habitat patches fragmented by contemporary habitat loss, and because they show a stronger effect of isolation than other patch types, we suggest that isolation will play an increasing role in the biology of habitat fragments. We use multiple lines of evidence to suggest that our understanding of the role of isolation on community assembly in fragmented landscapes is inadequate. Finally, our observation that consistent taxonomic differences in community patterns were minimal suggests that conservation actions intended to mitigate the negative effects of extinction may have far-reaching effects across taxonomic groups.     

Population Estimates of Fruit Bats (Pteropus mariannus) in the Mariana Islands

Abstract: Populations of Marianas fruit bats, Pteropus mariannus, were surveyed on each of the 15 Mariana Islands in 1983–1984. It is estimated that a minimum of 8,700–9,000 fruit bats occur in the archipelago, with about 8245% of these bats found on the nine northernmost and largely uninhabited islands. The islands of Anatahan, Pagan, and Agrihan had the largest populations, with minimum population estimates of 3,000, 2,500, and 1,000 bats, respectively. Smaller populations of about 400–1,000 fruit bats occurred on Asuncion, Guam, Rota, and Guam. The remaining is-ands in the archipelago were each estimated to have fewer than 125 bats. Population densities of fruit bats were highest on islands with little bunting but were generally much lower on human-inhabited islands where bunting was common. Quantity and quality of existing habitat were other important factors regulating the size of fruit bat populations in the Marianas. Increased enforcement of existing laws protecting bats and a public awareness program are important tasks needed to conserve and manage fmit bats in the Marianu Islands. The census techniques used in this study may be applicable to other fypes of colonial, mobile wildlife that inhabit islands.     

Iberian Lynx in a Fragmented Landscape: Predispersal, Dispersal, and Postdispersal Habitats

Abstract: Applied conservation biology must provide solutions for the conservation of species in modern landscapes, where prime habitats are being continuously fragmented and altered and animals are restricted to small, nonviable populations. We studied habitat selection in a fragmented population of endangered Iberian lynx (   Lynx pardinus ) by examining 14 years of radiotracking data obtained from lynx trapped in two different source areas. Habitat selection was studied independently for predispersal lynx in the source areas, for dispersing individuals through the region, and for postdispersing animals, most of which settled far from their point of origin. A multivariate analysis of variance showed that habitat use differed significantly among these phases and between area of origin, but not between sexes. The habitat type most used, and best represented within home ranges, was the mediterranean scrubland. Pine plantations were also important during and after dispersal. The rest of the habitats were either avoided (open habitats) or used according to availability ( pine and eucalyptus plantations) by dispersing lynx. Differences due to lynx origin were detected only during predispersal and dispersal and were observed because animals from each area had different habitat availability. Lynx with established territories did not use areas at random. They occupied patches of mediterranean scrubland more often than would be expected from scrubland availability during predispersal; the rest of the habitats were included within home ranges less than would be expected from their availability in the landscape. Results indicate that dispersing animals may use habitats of lower quality than habitats used by resident individuals, which suggests that conservation strategies applied across regions might be a viable objective.     

Consequences of ignoring dispersal variation in network models for landscape connectivity

Habitat loss and fragmentation can negatively influence population persistence and biodiversity, but the effects can be mitigated if species successfully disperse between isolated habitat patches. Network models are the primary tool for quantifying landscape connectivity, yet in practice, an overly simplistic view of species dispersal is applied. These models often ignore individual variation in dispersal ability under the assumption that all individuals move the same fixed distance with equal probability. We developed a modeling approach to address this problem. We incorporated dispersal kernels into network models to determine how individual variation in dispersal alters understanding of landscape-level connectivity and implemented our approach on a fragmented grassland landscape in Minnesota. Ignoring dispersal variation consistently overestimated a population's robustness to local extinctions and underestimated its robustness to local habitat loss. Furthermore, a simplified view of dispersal underestimated the amount of habitat substructure for small populations but overestimated habitat substructure for large populations. Our results demonstrate that considering biologically realistic dispersal alters understanding of landscape connectivity in ecological theory and conservation practice.     

Forest Fragmentation Increases Nest Predation in the Eurasian Treecreeper

Abstract:  We used long-term breeding data to monitor the influences of fragmentation and habitat composition at different spatial scales on the reproductive success of Eurasian Treecreepers ( Certhia familiaris ) breeding in nest boxes. We collected data from the same forest patches (2.7–65.1 ha in size) during seven breeding seasons. Nest predation varied considerably over the years and was the primary cause of nesting failure (mean annual rate of 21.6 ± 12.8%). Nest predation explained most of the variation in fledgling production during the study period. Landscape-level fragmentation (radius of 500 m from territory center) affected nest predation more than did fragmentation on the territory scale (radius of 200 m from territory center). In general, nest loss due to predation in fragmented landscapes (32.4%) was almost threefold that of less fragmented (12.0%) landscapes. Of the habitat variables, predation rate correlated positively with the density of edges between forest and open land and with the proportion of sapling stands on the spatial scale of 500 m around a nest. In the core area of a territory (radius of 30 m from territory center), a high density of trees increased the frequency of nest predation. Further, a high proportion of agricultural land close to a nest site increased nest losses of treecreepers, probably because of a high degree of mustelid predation. Our results showed that the spatial scale on which we examined nest predation is important and that even within moderately fragmented landscapes it is possible to detect fragmentation-related nest predation.     

Fish Responses to Experimental Fragmentation of Seagrass Habitat

Abstract: Understanding the consequences of habitat fragmentation has come mostly from comparisons of patchy and continuous habitats. Because fragmentation is a process, it is most accurately studied by actively fragmenting large patches into multiple smaller patches. We fragmented artificial seagrass habitats and evaluated the impacts of fragmentation on fish abundance and species richness over time (1 day, 1 week, 1 month). Fish assemblages were compared among 4 treatments: control (single, continuous 9‐m² patches); fragmented (single, continuous 9‐m² patches fragmented to 4 discrete 1‐m² patches); prefragmented/patchy (4 discrete 1‐m² patches with the same arrangement as fragmented); and disturbance control (fragmented then immediately restored to continuous 9‐m² patches). Patchy seagrass had lower species richness than actively fragmented seagrass (up to 39% fewer species after 1 week), but species richness in fragmented treatments was similar to controls. Total fish abundance did not vary among treatments and therefore was unaffected by fragmentation, patchiness, or disturbance caused during fragmentation. Patterns in species richness and abundance were consistent 1 day, 1 week, and 1 month after fragmentation. The expected decrease in fish abundance from reduced total seagrass area in fragmented and patchy seagrass appeared to be offset by greater fish density per unit area of seagrass. If fish prefer to live at edges, then the effects of seagrass habitat loss on fish abundance may have been offset by the increase (25%) in seagrass perimeter in fragmented and patchy treatments. Possibly there is some threshold of seagrass patch connectivity below which fish abundances cannot be maintained. The immediate responses of fish to experimental habitat fragmentation provided insights beyond those possible from comparisons of continuous and historically patchy habitat.     

Influence of habitat quality, population size, patch size, and connectivity on patch-occupancy dynamics of the middle spotted woodpecker

Despite extensive research on the effects of habitat fragmentation, the ecological mechanisms underlying colonization and extinction processes are poorly known, but knowledge of these mechanisms is essential to understanding the distribution and persistence of populations in fragmented habitats. We examined these mechanisms through multiseason occupancy models that elucidated patch-occupancy dynamics of Middle Spotted Woodpeckers (Dendrocopos medius) in northwestern Spain. The number of occupied patches was relatively stable from 2000 to 2010 (15-24% of 101 patches occupied every year) because extinction was balanced by recolonization. Larger and higher quality patches (i.e., higher density of oaks >37 cm dbh [diameter at breast height]) were more likely to be occupied. Habitat quality (i.e., density of large oaks) explained more variation in patch colonization and extinction than did patch size and connectivity, which were both weakly associated with probabilities of turnover. Patches of higher quality were more likely to be colonized than patches of lower quality. Populations in high-quality patches were less likely to become extinct. In addition, extinction in a patch was strongly associated with local population size but not with patch size, which means the latter may not be a good surrogate of population size in assessments of extinction probability. Our results suggest that habitat quality may be a primary driver of patch-occupancy dynamics and may increase the accuracy of models of population survival. We encourage comparisons of competing models that assess occupancy, colonization, and extinction probabilities in a single analytical framework (e.g., dynamic occupancy models) so as to shed light on the association of habitat quality and patch geometry with colonization and extinction processes in different settings and species.     

Spatial Structure and Population Extinction: A Study with Drosophila Flies

Abstract: The total amount of habitat and also its distribution and subdivision affect the extinction probability of a resident population Two species of Drosophila are studied in spatial configurations of a single large habitat patch, single small habitat patches, and two small but connected habitat patches in which a low rate of migration, roughly one fly per generation, is possible. The single large habitat patch shows the lowest extinction rate lower than the combined rate of two small patches of the same total size. For one of the species, the "corridor" between the pair of small patches seems to produce a "rescue effect" that lowers extinction rates, probably due to a decrease in the coefficient of variation in fluctuations of the population sire in this coupled system. The systems seem to have been influenced by demographic stochasticity, based on the relationship of population size to extinction probability.     

Wood Thrush Populations Are Not All Sinks in the Agricultural Midwestern United States

Abstract: Conservation strategies for forest-breeding, Neotropical migratory birds focus on the identification and preservation of source habitats. In sources, populations maintain relatively high levels of reproductive success, whereas in sinks, reproductive success is consistently below net replacement rate. In agricultural regions of the midwestern United States, forest patches of various sizes appear to act as sinks for Neotropical migrants. In 1994–1996 I monitored Wood Thrush (   Hylocichla mustelina ) nests in four forest patches in an agricultural landscape in northern Indiana. Given the prevailing estimates of adult and juvenile survival, I report productivity data that show substantial annual variation and that are sufficient to suggest the maintenance of temporal sources. Indeed, some Wood Thrush populations may act as source habitats in regions with relatively high levels of Brown-headed Cowbird (   Molothrys ater ) parasitism and nest predation because pairs produce multiple clutches and have the ability to fledge cowbirds and their own young from parasitized nests. My results and those from other studies of Wood Thrushes in landscapes with fragmented forests illustrate the variable ways in which source-sink dynamics are affected by habitat fragmentation. In some agricultural regions, populations in small forest patches may act as sources for Wood Thrushes and may be a legitimate focus of conservation efforts aimed at some migratory songbirds.     

The Quantitative Incidence Function Model and Persistence of an Endangered Butterfly Metapopulation

The incidence function model is derived from a linear first-order Markov chain of the presence or absence of a species in a habitat patch. The model can be parameterized with "snapshot" presence/absence data from a patch network. Using the estimated parameter values the Markov chain can be iterated in the same or in some other patch network to generate quantitative predictions about transient metapopulation dynamics and the stochastic steady state. We tested the ability of the incidence function model to predict patch occupancy using extensive data on an endangered butterfly, the Glanville fritillary ( Melitaea cinxia ) Parameter values were estimated with data collected from a 50-patch network in 1991. In 1993 we surveyed the entire geographic range of the species in Finland, within an area of 50 × 70 km², with 1502 habitat patches (dry meadows) of which 536 were occupied. Model predictions were generated for the 1502 patches and were compared with the observed pattern of occupancy in 1993. The model predicted patch occupancy well in more than half of the study area, but prediction was poor for one quarter of the area, probably because of regional variation in habitat quality and because metapopulations may have been perturbed away from the steady state. The incidence function model provides a practical tool for making quantitative predictions about metapopulation dynamics of species living in fragmented landscapes.     

Extinction Rates in Archipelagoes: Implications for Populations in Fragmented Habitats

To study the effect of habitat fragmentation on population viability, I used extinction rates on islands in archipelagoes and estimated the relative probability of extinction per species on single large islands and sets of smaller islands with the same total area. Data on lizards, birds, and mammals on oceanic islands and mammals on mountaintops and in nature reserves yield similar results. Species are likely to go extinct on all the small islands before they go extinct on the single, large island. In the short term, the analysis indicates that extinction probabilities may be lower on a set of small islands. This is perhaps an artifact due to underestimation of extinction rates on small islands and/or the necessity of pooling species in a focal taxon to obtain estimates of extinction rates (which may obscure area thresholds and underestimate the slope and curvature of extinction rates as a function of area). Ultimately, cumulative extinction probabilities are higher for a set of small islands than for single large islands. Mean and median times to extinction tend to be shorter in the fragmented systems, in some cases much shorter. Thus, to minimize extinction rates in isolated habitat remnants and nature reserve systems, the degree of fragmentation should be minimized     

From population sources to sieves: the matrix alters host-parasitoid source-sink structure

Field experiments that examine the impact of immigration, emigration, or landscape structure (e.g., the composition of the matrix) on the source sink dynamics of fragmented populations are scarce. Here, planthoppers (Prokelisia crocea) and egg parasitoids (Anagrus columbi) were released among host-plant patches that varied in structural (caged, isolated, or in a network of other patches) and functional (mudflat matrix that impedes dispersal vs. brome-grass matrix that facilitates dispersal) connectivity. Planthoppers and parasitoids on caged patches exhibited density-dependent growth rates, achieved high equilibrium densities, and rarely went extinct. Therefore, experimental cordgrass patches were classified as population sources. Because access to immigrants did not result in elevated population densities, source populations were not also pseudosinks, i.e., patches whose densities occur above carrying capacity due to high immigration. Planthoppers and parasitoids in open patches in mudflat had dynamics similar to those in caged patches, but went extinct in 4-5 generations in open patches in brome. Brome-embedded patches leaked emigrants at a rate that exceeded the gains from reproduction and immigration; populations of this sort are known as population sieves. For species whose suitable patches are becoming smaller and more isolated as a result of increased habitat fragmentation, emigration losses are likely to become paramount, a condition favoring the formation of population sieves. An increase in the proportion of patches that are sieves is predicted to destabilize regional population dynamics.     

Using Montane Mammals to Model Extinctions Due to Global Change

We use data on the species-area relationship and the nested subset structure of the boreal mammal faunas inhabiting isolated mountaintops in the Great Basin to develop a simple quantitative model that predicts the number and identity of species that would go extinct under an assumed scenario of changing climate and vegetation. Global warming of 3°C is predicted to cause the loss of 9–62% of the species inhabiting each mountain range and the extinction of three of fourteen species throughout the region. These results suggest (1) that it is possible to make highly plausible predictions about the susceptibility of species to extinction without detailed information about their population biology, and (2) that global and regional environmental changes seriously threaten the survival of species that are restricted in distribution to both natural "habitat islands" and biological reserves.