Interpreting habitat distribution models of an elusive species,the marbled murrelets: a response to Burger and Page

Burger and Page (this volume) evaluated our models of habitat preferences and breeding success of a threatened seabird, the marbled murrelet (Brachyramphus marmoratus), based on the largest available set of confirmed nest-sites found in coastal old-growth forest of the Pacific North-West. We believe our study documented novel and unexpected patterns of landscape-level distribution of marbled murrelets in both heavily logged and relatively intact old-growth landscapes and provided insights into how these patterns influence their reproduction, and, eventually, management. Considering the importance of the issue and to ensure appropriate and responsible use of the information we welcome discussion, detailed scrutiny and evaluation of our original results. Burger and Page claim to have identified flaws with model interpretation, data quality, statistical approaches, presentation and interpretation of our results that would invalidate our conclusions. We respond that most of their critique is irrelevant and/or misdirected with respect to our study and the interpretation of GIS data models, and that valid aspects of their claims do not critically affect our conclusions.     

Interaction of food resources and landscape structure in determining the probability of patch use by carnivores in fragmented landscapes

Studies on the distribution of mammalian carnivores in fragmented landscapes have focused mainly on structural aspects such as patch and landscape features; similarly, habitat connectivity is usually associated with landscape structure. The influence of food resources on carnivore patch use and the important effect on habitat connectivity have been overlooked. The aim of this study is to evaluate the relative importance of food resources on patch use patterns and to test if food availability can overcome structural constraints on patch use. We carried out a patch-use survey of two carnivores: the beech marten (Martes foina) and the badger (Meles meles) in a sample of 39 woodland patches in a fragmented landscape in central Italy. We used the logistic model to investigate the relative effects on carnivore distribution of patch, patch neighbourhood and landscape scale variables as well as the relative abundance of food resources. Our results show how carnivore movements in fragmented landscapes are determined not only by patch/landscape structure but also by the relative abundance of food resources. The important take-home message of our research is that, within certain structural limits (e.g. within certain limits of patch isolation), by modifying the relative amount of resources and their distribution, it is possible to increase suitability in smaller/relatively isolated patches. Conversely, however, there are certain thresholds above which an increase in resources will not achieve high probability of presence. Our findings have important and generalizable consequences for highly fragmented landscapes in areas where it may not be possible to increase patch sizes and/or reduce isolation so, for instance, forest regimes that will increase resource availability could be implemented. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Historic landscape change and habitat loss: the case of black grouse in Lower Saxony,Germany

The declines of many specialist bird species in the agricultural landscapes of Central Europe have resulted in small and isolated populations. In the case of the black grouse, a ground-nesting bird species with large spatial requirements, empiric evidence about underlying landscape changes is scarce. In this study, we examined land cover and land cover changes in a farmland-forest mosaic in eastern Lower Saxony, Germany and how they affect occurrence and persistence of black grouse. Spatial information came from historic topographic maps from 1958 to 1975. The results show profound conversions of habitat to forest and farmland but also an increase in settlement area. Habitat conversions and suburbanization were negative correlates of black grouse persistence. Habitat models from before and after a decline period differed in some of the predictors and suggest black grouse habitat to be more diverse before the land cover changes. Our study confirms that land use factors at a landscape scale extent contribute to explain black grouse occurrence and thus can complement important small scale factors like the quality and size of individual habitat patches. Results also show that landscape factors affect black grouse distribution predominantly from an area much greater than an individual black grouse home range. Our models may be further evaluated on present-day landscapes and might be used to evaluate large-scale habitat availability for black grouse.     

Testing assumptions of cost surface analysis—a tool for invasive species management

Applied ecology could benefit from new tools that identify potential movement pathways of invasive species, particularly where data are sparse. Cost surface analysis (CSA) estimates the permeability (friction) across a landscape and can be applied to dispersal modelling. Increasingly used in a diversity of applications, several fundamental assumptions that might influence the outputs of CSA (cost surfaces and least-cost pathways) have yet to be systematically examined. Thus, we explore two issues: the presumed relationship between habitat preferences and dispersal behaviour as well as the degree of landscape fragmentation through which an organism moves by modelling a total of 18 sensitivity and dispersal scenarios. We explored the effect of fragmentation by altering the friction values (generally assigned using expert opinion) associated with patch and linear features. We compared these sensitivity scenarios in two sites that differed in fragmentation. We also used eastern grey squirrels (Sciurus carolinensis) as an example invading species and compared diffusion models and two contrasting cost surface dispersal scenarios. The diffusion model underestimated spread because squirrels did not move randomly through the landscape. Despite contrasting assumptions regarding dispersal behaviour, the two cost surfaces were strikingly similar while the least-cost paths differed. Furthermore, while the cost surfaces were insensitive to changes in friction values for linear features, they were sensitive to assumptions made for patch features. Our results suggest that movement in fragmented landscapes may be more sensitive to assumptions regarding friction values than contiguous landscapes. Thus, the reliability of CSA may depend not only on the range of friction values used for patches but also the degree of contiguity in the landscape.     

Effects of landscape composition on spread of an herbicide-resistant weed

Widespread adoption of genetically modified glyphosate-resistant (GR) crops in the US has dramatically changed the agricultural landscape to one that selects for establishment and spread of weedy species resistant to glyphosate, a commonly applied herbicide. Weed species that possess the means to readily spread across the landscape will be contained by weed management strategies that limit weed establishment and prevent seed set. An empirically-derived simulation model was developed to explore GR Conyza canadensis spread in relation to characteristics of the agricultural landscape. C. canadensis seeds are carried in the wind and move among fields and therefore, access high quality habitat (GR crops) at long distances. The baseline scenario was the current GR adoption levels in many US agricultural landscapes with corn and soybean rotated annually. Alternate scenarios examined the interacting effects of management uniformity (GR crop adoption) and increased landscape richness (three crops: corn, soybean, alfalfa, instead of two), over a 10 year simulation period. When landscape uniformity increased (increased GR corn adoption), 3× more fields would be infested with the resistant biotype and a specific field would have up to 24% greater likelihood of being infested compared to the current GR crop adoption levels. Increased landscape richness (adding alfalfa as a third crop) slightly decreased GR C. canadensis abundance. Reduced GR management uniformity by way of reducing GR soybeans to half their current adoption levels had the greatest impact on spread and prevented GR C. canadensis from reaching high abundance. Large-scale reliance on glyphosate for weed management has increased high-fitness habitat and will result in rapid spread of glyphosate-resistant weeds. Without significant reductions of glyphosate use and without spatial coordination of weed and crop management practices, GR weeds will continue to spread rapidly and impact agricultural practices in areas reliant on glyphosate.     

Importance of agricultural landscapes to nesting burrowing owls in the Northern Great Plains, USA

Anthropogenic habitat loss and fragmentation are the principle factors causing declines of grassland birds. Declines in burrowing owl (Athene cunicularia) populations have been extensive and have been linked to habitat loss, primarily the decline of black-tailed prairie dog (Cynomys ludovicianus) colonies. Development of habitat use models is a research priority and will aid conservation of owls inhabiting human-altered landscapes. From 2001 to 2004 we located 160 burrowing owl nests on prairie dog colonies on the Little Missouri National Grassland in North Dakota. We used multiple linear regression and Akaike’s Information Criterion to estimate the relationship between cover type characteristics surrounding prairie dog colonies and (1) number of owl pairs per colony and (2) reproductive success. Models were developed for two spatial scales, within 600 m and 2,000 m radii of nests for cropland, crested wheatgrass (Agropyron cristatum), grassland, and prairie dog colonies. We also included number of patches as a metric of landscape fragmentation. Annually, fewer than 30% of prairie dog colonies were occupied by owls. None of the models at the 600 m scale explained variation in number of owl pairs or reproductive success. However, models at the 2,000 m scale did explain number of owl pairs and reproductive success. Models included cropland, crested wheatgrass, and prairie dog colonies. Grasslands were not included in any of the models and had low importance values, although percentage grassland surrounding colonies was high. Management that protects prairie dog colonies bordering cropland and crested wheatgrass should be implemented to maintain nesting habitat of burrowing owls.     

Nest site use by crested ibis: dependence of a multifactor model on spatial scale

The crested ibis (Nipponia nippon), a species at the brink of extinction in 1981, remain restricted to a small (25 km radius) area of temperate forests in central China. To improve the chances of successful reintroduction into new areas we developed a multifactor logistic regression model of habitat association at multiple scales. Using habitat variables, i.e. vegetation, human impact, elevation, and wetland, we compared occupied and unoccupied sites at grain sizes ranging from 1 to 6400 ha. The goodness-of-fit of the habitat suitability model depended on grain size, with the best fit (most information) at a grain size of 2 ha. Semivariograms showed the habitat variables at control sites have a gradient pattern, yet the crested ibis had their specific habitat preferences, and only selected a narrow range from the available gradient. Our results indicated that spatial scale needs to be considered in developing habitat models for applications such as conservation planning.     

The Importance of Spatial Scale in Habitat Models: Capercaillie in the Swiss Alps

The role of scale in ecology is widely recognized as being of vital importance for understanding ecological patterns and processes. The capercaillie (Tetrao urogallus) is a forest grouse species with large spatial requirements and highly specialized habitat preferences. Habitat models at the forest stand scale can only partly explain capercaillie occurrence, and some studies at the landscape scale have emphasized the role of large-scale effects. We hypothesized that both the ability of single variables and multivariate models to explain capercaillie occurrence would vary with the spatial scale of the analysis. To test this hypothesis, we varied the grain size of our analysis from 1 to just over 1100 hectares and built univariate and multivariate habitat suitability models for capercaillie in the Swiss Alps. The variance explained by the univariate models was found to vary among the predictors and with spatial scale. Within the multivariate models, the best single-scale model (using all predictor variables at the same scale) worked at a scale equivalent to a small annual home range. The multi-scale model, in which each predictor variable was entered at the scale at which it had performed best in the univariate model, did slightly better than the best single-scale model. Our results confirm that habitat variables should be included at different spatial scales when species-habitat relationships are investigated.     

Effect of matrix habitat on the spread of flea beetle introductions for biological control of leafy spurge

Matrix habitats are known to influence the movement patterns of a variety of species but it is less well known whether these effects have strong implications for spatial population dynamics, including the spread of biological introductions. Using a spatially explicit simulation model parameterized with empirical data, we examine how grass and shrub matrix habitats, each offering different resistance to dispersal, influence the spread and impact of a biocontrol agent, Aphthona lacertosa, on the invasive weed, leafy spurge. Model predictions indicate that differential responses to matrix habitat have little effect on the agent’s spread over the study landscape and this is supported by statistical models fit to observed A. lacertosa incidence on the same landscape. Subsequent experimentation with the simulation model suggested that A. lacertosa colonization rates were largely unaffected by increases in amount of the more restrictive shrub matrix. However, simulations of an hypothetical species with greater overall dispersal ability but reduced dispersal rate through shrub matrix showed that colonization rates were noticeably reduced when the percentage of shrub matrix on the landscape approached 50%. Combined these results suggest that some tailoring of release strategies may be required to accommodate the unique dispersal capabilities of different biocontrol agents on particular release landscapes, but for A. lacertosa there appears to be little effect of matrix habitat structure on rates of spread.     

Movements of cactus bugs: Patch transfers,matrix resistance,and edge permeability

Individual movement is a key process affecting the distribution of animals in heterogeneous landscapes. For specialist species in patchy habitat, a central issue is how dispersal distances are related to landscape structure. We compared dispersal distances for cactus bugs (Chelinidea vittiger) on two naturally fragmented landscapes (≤ 4% suitable habitat) with different matrix structures (i.e., vegetation height of nonsuitable habitat between suitable patches). Using mark-release-recapture studies, we determined that most transfers between cactus patches occurred during the mating season. Dispersal distances were reduced by > 50% on the landscape that had reduced structural connectivity due to relatively high matrix structure and low patch density. An experiment with detailed movement pathways demonstrated that greater matrix structure decreased mean step lengths, reduced directionality, and thus decreased net displacement by > 60%. However, habitat edges between two matrix elements that differed substantially in resistance to movement were completely permeable. Therefore, the difference in distributions of dispersal distances between the two landscapes mainly reflected the average resistance of matrix habitat and not the level of matrix heterogeneity per se. Our study highlights the merits of combining estimates of dispersal distances with insights on mechanisms from detailed movement pathways, and emphasizes the difficulty of treating dispersal distances of species as fixed traits independent of landscape structure.     

How far do birds disperse seeds in the degraded tropical landscape of Hong Kong,China?

Information on seed dispersal distances is critical for understanding plant species persistence in habitat fragments and vegetation recovery when disturbance is reduced. In the degraded upland landscape of Hong Kong (22°N), the bulbuls Pycnonotus sinensis and P. jocosus are responsible for a large proportion of seed movements. Dispersal distances were estimated from gut passage times (GPTs) and movement patterns determined by radio-telemetry. Estimates were also made for the hwamei, Garrulax canorus. Seven adult P. sinensis and four adult G. canorus were tracked in winter, and six juvenile P. sinensis, three juvenile P. jocosus and two juvenile G. canorus in summer. GPTs were 5–122 min in the bulbuls and 18–61 min in the hwamei. Most 10-min movements were <100 m for the bulbuls and <50 m for the hwamei, but the largest were >1300 m for both bulbuls and 940 m for the hwamei. Displacement-time graphs generally levelled off rapidly, with median displacements <60 m after an hour, except with summer juvenile bulbuls. Median displacements during median gut passage times for seeds from single-seeded fruits were 40, 116 and 131 m, respectively, for winter adult P. sinensis and summer juvenile P. sinensis and P. jocosus. Maximum observed displacements during the maximum measured GPTs were >1 km for all bulbuls. Estimated dispersal distances were shorter for hwameis. The radio-telemetry results were supplemented by 49 h of visual observations, during which 1,510 bird movements across open areas were observed, 64% by P. sinensis, 13% by P. jocosus, and 0.5% by G. canorus. The bulbuls, therefore, connect habitat fragments in upland Hong Kong for plants with fruits within their maximum gape width.     

Importance of Wetland Landscape Structure to Shorebirds Wintering in an Agricultural Valley

Only recently has the influence of landscape structure on habitat use been a research focus in wetland systems. During non-breeding periods when food can be locally limited, wetland spatial pattern across a landscape may be of great importance in determining wetland use. We studied the influence of landscape structure on abundances of wintering Dunlin (Calidris alpina) and Killdeer (Charadrius vociferus) observed on wetlands in the agricultural Willamette Valley of Oregon, USA, during two winters (1999–2000, 2000–2001) of differing rainfall. We examined (1) shorebird use within a sample of 100 km² regions differing in landscape structure (hectares of shorebird habitat [wet, unvegetated]) and (2) use of sites differing in landscape context (area of shorebird habitat within a species-defined radius). For use of sites, we also assessed the influence of two local characteristics: percent of soil exposed and area of wet habitat. We analyzed data using linear regression and information-theoretic modeling. During the dry winter (2000–2001), Dunlin were attracted to regions with more wetland habitat and their abundances at sites increased with greater area of shorebird habitat within both the site and the surrounding landscape. In contrast, Dunlin abundances at sites were related to availability of habitat at only a local scale during the wet winter (1999–2000). Regional habitat availability was of little importance in predicting Killdeer distributions, and Killdeer site use appeared unrelated to habitat distributions at both landscape and local scales. Results suggest prioritizing sites for conservation that are located in areas with high wetland coverage.     

The importance of functional connectivity in the conservation of a ground-dwelling mammal in an urban Australian landscape

The distribution of the northern brown bandicoot (Isoodon macrourus), a medium-sized ground-dwelling marsupial, was examined in habitat fragments within the urban landscape of the city of Brisbane, Australia. From surveys conducted in 68 fragments, bandicoots were found to be present in 33 (49%) despite widespread habitat loss and fragmentation. Logistic regression analysis revealed that of 13 measured independent variables, functional connectivity was the only factor that significantly predicted the presence of bandicoots within fragments, with connectivity positively correlated with the likelihood of occupation. Functional connectivity was equated to the likelihood of bandicoot immigration into the focal fragment from the nearest occupied fragment, based on the estimated resistance to movement offered by the intervening matrix. Within Brisbane, riparian habitat fragments typically have a relatively high level of functional connectivity, as thin strips of vegetation fringing waterways serve as corridors between larger riparian areas and facilitate the movement of bandicoots between patches. Analyses based on the Akaike Information Criterion revealed that the optimal model based on landscape context variables was convincingly better supported by the data than the optimal model produced from fragment characteristics. However, it is important to examine both internal attributes of habitat fragments and external features of the surrounding landscape when modelling the distribution of ground-dwelling fauna in urban environments, or other landscapes with a highly variable matrix. As urban centres throughout the world expand, it is crucial that the ecology of local wildlife be considered to ensure functional connection is maintained between habitat patches, especially for the conservation of species that are highly susceptible to fragmentation.     

Contribution of Topographically Based Landslide Hazard Modelling to the Analysis of the Spatial Distribution and Ecology of Kauri (Agathis australis)

In this paper the use of topographical attributes for the analysis of the spatial distribution and ecological cycle of kauri (Agathis australis), a canopy emergent conifer tree from northern New Zealand, is studied. Several primary and secondary topographical attributes are derived from a Digital Elevation Model (DEM) for a study area in the Waitakere Ranges. The contribution of these variables in explaining presence or absence of mature kauri is assessed with logistic regression and Receiver Operating Characteristic (ROC) plots. A topographically based landslide hazard index, calculated by combining a steady state hydrologic model with the infinite slope stability equation, appears to be very useful in explaining the occurrence and ecological dynamics of kauri. It is shown that the combination of topographical, soil physical and hydrological parameters in the calculation of this single landslide hazard index, performs better in explaining presence of mature kauri than using topographical attributes calculated from the DEM alone. Moreover, this study demonstrates the possibilities of using terrain attributes for representing geomorphological processes and disturbance mechanisms, often indispensable in explaining a species’ ecological cycle. The results of this analysis support the ‘temporal stand replacement model’, involving disturbance as a dominant ecological process in forest regeneration, as an interpretation of the community dynamics of kauri. Furthermore a threshold maturity stage, in which trees become able to stabilize landslide prone sites and postpone a possible disturbance, together with great longevity are seen as major factors making kauri a ‘landscape engineer’.     

Landscape attributes and life history variability shape genetic structure of trout populations in a stream network

Spatial and temporal landscape patterns have long been recognized to influence biological processes, but these processes often operate at scales that are difficult to study by conventional means. Inferences from genetic markers can overcome some of these limitations. We used a landscape genetics approach to test hypotheses concerning landscape processes influencing the demography of Lahontan cutthroat trout in a complex stream network in the Great Basin desert of the western US. Predictions were tested with population- and individual-based analyses of microsatellite DNA variation, reflecting patterns of dispersal, population stability, and local effective population sizes. Complementary genetic inferences suggested samples from migratory corridors housed a mixture of fish from tributaries, as predicted based on assumed migratory life histories in those habitats. Also as predicted, populations presumed to have greater proportions of migratory fish or from physically connected, large, or high quality habitats had higher genetic variability and reduced genetic differentiation from other populations. Populations thought to contain largely non-migratory individuals generally showed the opposite pattern, suggesting behavioral isolation. Estimated effective sizes were small, and we identified significant and severe genetic bottlenecks in several populations that were isolated, recently founded, or that inhabit streams that desiccate frequently. Overall, this work suggested that Lahontan cutthroat trout populations in stream networks are affected by a combination of landscape and metapopulation processes. Results also demonstrated that genetic patterns can reveal unexpected processes, even within a system that is well studied from a conventional ecological perspective.     

Multi-scale use of lands providing anthropogenic resources by American Crows in an urbanizing landscape

The conversion of forests and farmlands to human settlements has negative impacts on many native species, but also provides resources that some species are able to exploit. American Crows (Corvus brachyrhynchos), one such exploiter, create concern due to their impact as nest predators, disease hosts, and cultural harbingers of evil. We used various measures of crow abundance and resource use to determine crows’ response to features of anthropogenic landscapes in the Puget Sound region of the United States. We examined land cover and land use composition at three spatial scales: study sites (up to 208 ha), crow home ranges within sites (18.1 ha), and local land cover (400 m²). At the study site and within-site scales crow abundance was strongly correlated with land cover providing anthropogenic resources. In particular, crows were associated with the amount of ‘maintained forest’ cover, and were more likely to use grass and shrub cover than forest or bare soil cover. Although crows did not show a generalized response to an edge variable, they exhibited greater use of patchy habitat created by human settlements than of native forests. Radio-tagged territorial adults used resources within their home ranges relatively evenly, suggesting resource selection had occurred at a larger spatial scale. The land conversion pattern of new suburban and exurban settlements creates the mix of impervious surfaces and maintained vegetation that crows use, and in our study area crow populations are expected to continue to increase. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Estimation of the proportions of offspring having soft fruit and juicy fruit as evaluated by sensory tests in Japanese persimmon

Firmness and juiciness are important fruit attributes targeted in the breeding of Japanese persimmon (Diospyros kaki Thunb.). To facilitate the efficient production of cultivars with soft and juicy fruit, we successfully used the mid-parental values (MP) of sensory evaluations as independent variables in logistic regression models to estimate the proportions of offspring from a given cross that would have soft fruit and juicy fruit. The best model for estimating firmness contained only the MP for firmness (MPf) as an explanatory variable, whereas the best model for juiciness contained both the MP for juiciness (MPj) and the MPf as explanatory variables. The regression lines from these models suggested that crosses with a low MPf and high MPj can increase the proportion of offspring that have soft fruit and juicy fruit, respectively. However, the slopes of the regression lines for both firmness and juiciness were gentle, indicating that genetic segregation in offspring is wide for both attributes. We infer from these data that softer and juicier fruit can be bred from parents having only moderate ratings for these traits.     

Predicting the spatial distribution of an invasive plant species (<Emphasis Type="Italic">Eupatorium adenophorum</Emphasis>) in China

Invasive alien species can pose a severe threat to biodiversity and stability of the ecosystems they invade. Predicting distribution patterns of invasive species in regions outside their native range is a fundamental component of early warning systems. Crofton weed (Eupatorium adenophorum Spreng.) was first discovered in the Yunnan Province of China around the 1940s. The well-documented invasion history of this plant species provided the opportunity for us to examine the spatiotemporal patterns of biological invasion by crofton weed. Using the datasets documenting 441 known localities invaded by crofton weed in China over the past 50 years and 23 environmental variables generated by the genetic algorithm for rule-set production (GARP) model, we tested the predictability of crofton weed distribution with a high degree of accuracy. Both the Kappa statistics and the receiver–operator characteristic (ROC) analysis indicated that it is possible to predict the geographical spread of crofton weed in China. Precipitation in the coldest quarter of the year, extremely low air temperature, and maximum annual air temperature strongly influenced the predictions. Our results indicate that crofton weed may break out in Yungui Plateau, Sichuan Basin, southeastern Coastlands, Hainan Island, and Taiwan although currently it is either absent or has only recently been recorded in these regions. Redundancy analysis (RDA) ordination results demonstrated that temperature and precipitation play an important role in confining the spread of crofton weed. Over the past 20 years, crofton weed has spread from subtropical areas with higher annual mean temperature and lower climatic fluctuations to much cooler and dryer areas at higher altitudes. The distribution of crofton weed was restricted mainly to regions with mean annual air temperature ranging from 10 to 22°C and annual precipitation from 800 to 2000 mm. Our results could help in developing and implementing early detection measures to minimize the ecological impacts of crofton weed invasion in China.     

The need for biological realism in habitat modeling: a reinterpretation of Zharikov et al. (2006)

Zharikov et al. (2006: Landscape Ecology 21:107–120) modeled the nest-site habitat use of marbled murrelets (Brachyramphus marmoratus) in Desolation Sound (DS) and Clayoquot Sound (CS), British Columbia. They compared known nest sites, located with radio-telemetry, with randomly-located points within the same areas. Their conclusions suggest that murrelets tended to nest in disproportionately smaller fragments within the more disturbed DS landscape; streams, steeper slopes, and lower elevations were selected in both landscapes; murrelets nested closer to recent clearcuts than would be expected in the DS landscape; and survivorship of nestlings was greater in areas with recent clearcuts and was positively correlated with recent habitat fragmentation. These conclusions are contrary to current management guidelines in British Columbia, and therefore require close scrutiny. Our detailed examination reveals flaws in their use of data, application of modeling, and most seriously, interpretation of the results. Problems include: conceptual errors in the interpretation of models; inappropriate spatial resolution; confusing use and interpretation of fragmentation and patch size data; overemphasis of statistically significant but biologically trivial results; and ignoring some contradictory studies. We conclude that it would be risky to apply the results from Zharikov et al. in the selection of murrelet nesting habitat for management purposes in British Columbia. Our review identifies issues that may arise in other ecological modeling studies and stresses the need for biological realism in addition to statistical rigour.     

Effects of habitat loss on populations of white-footed mice: testing matrix model predictions with landscape-scale perturbation experiments

Habitat loss is the leading cause of decline in wildlife diversity and abundance throughout the world, but its effects on wildlife are not always predictable. Matrix population modeling is an increasingly common tool used to predict the effects of habitat loss. In spite of the growing number of studies using this approach, and its wide use in conservation practice, the predictions generated by matrix population models are rarely explicitly tested in the field. We compared the ability of a suite of spatially explicit demographic matrix models to predict the response of white-footed mice to loss of high quality habitat at mosaic sites in northeast Connecticut, USA. We tested short-term model predictions with landscape scale habitat perturbation experiments, including clear-cut logging or prescribed burning of high quality habitat at two study sites. Comparison of each model’s predictions with the observed responses at both sites qualitatively supported predictions that perturbation of high quality habitat would have negative effects extending into the surrounding landscape. The best-supported model assumed that evicted residents of the perturbed habitat would successfully resettle in nearby intact habitats, and allowed for gradual population recovery in the perturbed habitat. Similarly, long-term simulations (20 years) revealed how loss of a single habitat could trigger population declines throughout a mosaic site. This study shows that careful consideration of model assumptions such as those pertaining to resettlement behavior is crucial if predictions are to be reliable, and highlights the role of experiments for comparing alternative model predictions.