Landscape Extents Research Articles

Towards a mechanistic understanding of animal-ecosystem interactions.

Research Highlight: Ferraro, K. M., Welker, L., Ward, E. B., Schmitz, O. J., & Bradford, M. A. (2023). Plant mycorrhizal associations mediate the zoogeochemical effects of calving subsidies by a forest ungulate. Journal of Animal Ecology, https://doi.org/10.1111/1365-2656.14002. Animals play large roles in ecosystem elemental cycling but predicting effects in diverse contexts remains a substantial challenge. Fundamental to progress is (1) identifying mechanisms by which animals impact nutrient distribution and cycling, and (2) disentangling how environmental context mediates the operation of alternative mechanisms. In an elegant field experiment, Ferraro etal. (2023) provide the first detailed exploration of the impact of nutrient inputs from mammalian parturition on soil functioning and the stoichiometry of plant tissues. The authors find that nitrogen from experimental additions of ungulate parturition material (natal fluids) is rapidly incorporated into microsite soil organic pools and plant tissues. They also find that soil processes (soil microbial biomass, rates of carbon mineralization, nitrogen mineralization and nitrification) and the nitrogen content of plant tissues above- and belowground are increased by addition of parturition material. Notably, the authors identify that increases in some soil processes and plant tissue nitrogen are weaker in microsites dominated by ericoid mycorrhizal plants than those dominated by ectomycorrhizal plants. These findings demonstrate that parturition depositions, a ubiquitous but overlooked mechanism of mammalian impacts on ecosystems, impact ecosystem processes and plant tissue stoichiometry. Furthermore, plant-fungal associations are a predictive axis of context dependency mediating zoogeochemical effects at fine scales. Ferraro etal.'s (2023) novel approach simultaneously advances mechanistic understanding of animal-ecosystem interactions at fine scales and facilitates prediction of ungulate effects on nutrient availability at landscape extents.

A general meta‐ecosystem model to predict ecosystem functions at landscape extents

The integration of ecosystem processes over large spatial extents is critical to predicting whether and how local and global changes may impact biodiversity and ecosystem functions. Yet, there remains an important gap in meta‐ecosystem models to predict multiple functions (e.g. carbon sequestration, elemental cycling, trophic efficiency) across ecosystem types (e.g. terrestrial‐aquatic, benthic‐pelagic). We derive a flexible meta‐ecosystem model to predict ecosystem functions at landscape extents by integrating the spatial dimension of natural systems as spatial networks of different habitat types connected by cross‐ecosystem flows of materials and organisms. We partition the physical connectedness of ecosystems from the spatial flow rates of materials and organisms, allowing the representation of all types of connectivity across ecosystem boundaries. Through simulating a forest‐lake‐stream meta‐ecosystem, our model illustrates that even if spatial flows induced significant local losses of nutrients, differences in local ecosystem efficiencies could lead to increased secondary production at regional scale. This emergent result, which we dub the ‘cross‐ecosystem efficiency hypothesis', emphasizes the importance of integrating ecosystem diversity and complementarity in meta‐ecosystem models to generate empirically testable hypotheses for ecosystem functions.

Predicting ecosystem pattern across landscapes

Predictive modelling is fundamental to ecology and essential for objective biodiversity assessment. However, while predictive biodiversity models are generally well‐developed, models for predicting patterns within (e.g. composition, structure) and among ecosystems (e.g. ecosystem identity, spatial properties) have not been adequately operationalized. We contend the scarcity of such models marks a concerning gap in the scientific community's ability to make ecosystem predictions across landscapes, and more broadly for supporting the conservation of biodiversity and ecosystem functions. We propose ecosystem spatial pattern models (ESPM) to fill this gap in modelling capacity. Under our approach to ESPM, spatial patterns of ecosystem properties are the basis for resolving ecosystem organization at local and landscape extents. Our integrative modelling framework differs from others in that it accords biotic and abiotic constituents equally, consistent with their joint mechanistic influence on ecosystem establishment and dynamics. Development of ESPM is especially timely for ecosystem assessment is undergoing a groundswell, as scientists and conservation groups propose ambitious targets for ecosystem conservation and restoration.

Frugivore zoogeochemistry in tropical forest ecosystems

Frugivore zoogeochemistry in tropical forest ecosystems

How do forest landscapes respond to elevated CO2 and ozone? Scaling Aspen‐FACE plot‐scale experimental results

AbstractThe Aspen‐FACE (Free‐Air Carbon Enrichment) experiment was an 11‐yr study of the effect of elevated CO2 and ozone (alone and in combination) on the growth productivity of model aspen communities (pure aspen, aspen‐birch, and aspen‐maple) in the field in northern Wisconsin, USA. Uncertainty remains about how these short‐term plot‐level responses might play out at landscape scales where climate change, competition, succession, and disturbances interact with tree‐level responses. In this study, we used a recent physiology‐based approach (PnET‐Succession v3.1) within the forest landscape model LANDIS‐II to scale the site‐scale FACE results to landscape extents by mechanistically accounting for the globally changing drivers of CO2, ozone, temperature, and precipitation. We conducted a factorial simulation experiment to test five hypotheses about the effects of three treatments (CO2 concentration, cumulative ozone exposure, and disturbance). CO2 was clearly the dominant driver of landscape response, with disturbance also having a large effect. Ozone was not a dominant driver of landscape dynamics or total landscape biomass, but its negative effect on mean landscape biomass was nevertheless significant. We found that CO2 mitigation of water stress may not have a major effect on species composition or biomass accumulation. We found that species diversity was somewhat decreased by elevated CO2 as expected, but somewhat increased by O3, contrary to expectations. The spatial pattern of the landscape was minimally affected by the treatments. While rising CO2 concentrations have some mitigating effect on the negative O3 effect on the species studied, additional research is needed to confirm whether researchers and managers can be justified in disregarding O3 as a primary driver of forest dynamics in other ecosystems. Our results also add more support to the growing consensus that projections of climate change effects must include robust, direct links between CO2 and tree growth and competition; temperature effects (as demonstrated elsewhere) appear to be less by comparison.

The Currency of Conservation: How Is Landscape Extent Applied in Conservation Planning?

We asked how the concept of landscape extent is used in conservation planning. To narrow the scope of our review, we focused on four iconic species depicted on Canadian currency: the beaver (Castor canadensis), the caribou (Rangifer tarandus), the common loon (Gavia immer), and the polar bear (Ursus maritimus). Landscape ecology research suggests that a landscape is a species-specific construct arising from the spatial scales over which ecological processes occur. Species traits can be correlated with the scales of important effects and thus can identify relevant landscape extents for research and conservation. It is unclear, however, how the theoretical concept of landscape extent is applied in conservation practice. For the four iconic species in our study, we found an inconsistent use of species-specific traits in definitions of landscape extent for conservation planning. For two species, the caribou and the polar bear, species-specific traits were used to define conservation units, for defining either Designatable Units for assessment of at-risk status or for defining herds or subpopulations for management actions. No species-specific traits were identified in spatial frameworks used in loon or beaver conservation planning, as both were included in multi-species management approaches. Our review suggests a type of research implementation gap. Conservation units are often defined based on political or administrative boundaries rather than by species-specific processes. We suggest that landscape extent be implemented as a hypothesis in an adaptive management framework when possible. Additionally, we encourage research approaches that explicitly incorporate policy considerations into research design.

Mapping upland peat depth using airborne radiometric and lidar survey data

A method to estimate peat depth and extent is vital for accurate estimation of carbon stocks and to facilitate appropriate peatland management. Current methods for direct measurement (e.g. ground penetrating radar, probing) are labour intensive making them unfeasible for capturing spatial information at landscape extents. Attempts to model peat depths using remotely sensed data such as elevation and slope have shown promise but assume a functional relationship between current conditions and gradually accrued peat depth. Herein we combine LiDAR-derived metrics known to influence peat accumulation (elevation, slope, topographic wetness index (TWI)) with passive gamma-ray spectrometric survey data, shown to correlate with peat occurrence, to develop a novel peat depth model for Dartmoor.Total air absorbed dose rates of Thorium, Uranium and Potassium were calculated, referred to as radiometric dose. Relationships between peat depth, radiometric dose, elevation, slope and TWI were trained using 1334 peat depth measurements, a further 445 measurements were used for testing. All variables showed significant relationships with peat depth. Linear stepwise regression of natural log-transformed variables indicated that a radiometric dose and slope model had an r2 = 0.72/0.73 and RMSE 0.31/0.31 m for training/testing respectively. This model estimated an area of 158 ±101 km2 of peaty soil >0.4 m deep across the study area. Much of this area (60 km2) is overlain by grassland and therefore may have been missed if vegetation cover was used to map peat extent. Using published bulk density and carbon content values we estimated 13.1 Mt. C (8.1–21.9 Mt. C) are stored in the peaty soils within the study area. This is an increase on previous estimates due to greater modelled peat depth. The combined use of airborne gamma-ray spectrometric survey and LiDAR data provide a novel, practical and repeatable means to estimate peat depth with no a priori knowledge, at an appropriate resolution (10 m) and extent (406 km2) to facilitate management of entire peatland complexes.

Determining the size of a complete disturbance landscape: multi-scale, continental analysis of forest change.

The scale of investigation for disturbance-influenced processes plays a critical role in theoretical assumptions about stability, variance, and equilibrium, as well as conservation reserve and long-term monitoring program design. Critical consideration of scale is required for robust planning designs, especially when anticipating future disturbances whose exact locations are unknown. This research quantified disturbance proportion and pattern (as contagion) at multiple scales across North America. This pattern of scale-associated variability can guide selection of study and management extents, for example, to minimize variance (measured as standard deviation) between any landscapes within an ecoregion. We identified the proportion and pattern of forest disturbance (30m grain size) across multiple landscape extents up to 180km2. We explored the variance in proportion of disturbed area and the pattern of that disturbance between landscapes (within an ecoregion) as a function of the landscape extent. In many ecoregions, variance between landscapes within an ecoregion was minimal at broad landscape extents (low standard deviation). Gap-dominated regions showed the least variance, while fire-dominated showed the largest. Intensively managed ecoregions displayed unique patterns. A majority of the ecoregions showed low variance between landscapes at some scale, indicating an appropriate extent for incorporating natural regimes and unknown future disturbances was identified. The quantification of the scales of disturbance at the ecoregion level provides guidance for individuals interested in anticipating future disturbances which will occur in unknown spatial locations. Information on the extents required to incorporate disturbance patterns into planning is crucial for that process.

Tree senescence as a direct measure of habitat quality: Linking red-edge Vegetation Indices to space use by Magellanic woodpeckers

Accurate estimates of the quality and quantity of remnant habitats is critical for planning management activities for the conservation of threatened species. Although habitat quality usually is understood from a multidimensional niche space approach, the availability of foraging substrates can be a suitable and more proximate index of habitat quality for species with specialized habitat requirements, like woodpeckers that feed almost exclusively on larvae of wood-boring beetles in the trunks and branches of infested trees. Recent approaches use simple mathematical algorithms on spectral bands called Vegetation Indices (VI) to identify infestations, providing a new opportunity to assess habitat quality for woodpeckers. In this paper, we tested the ability of 102 VI to estimate tree attributes explaining habitat quality for Magellanic woodpeckers for its usage as a reliable foraging habitat quality estimator. We hypothesized that space use of Magellanic woodpeckers is positively associated with the spatial distribution of decayed trees in the landscape. We developed a methodological framework based on high-resolution, multispectral imagery with three basic steps. First, we mapped individual Nothofagus trees based on estimates of species composition from a supervised classification procedure, VI estimates and image segmentation. Second, we selected the best VI predicting the tree quality for Magellanic woodpeckers. Third, we tested these habitat quality predictors, the species composition and tree age, by using two Synoptic Models of Space Use (SMSU) of Magellanic woodpeckers based on very high-frequency (VHF) radio-telemetry and global positioning system (GPS) telemetry data.Generalized Linear Mixed Models (GLMM) showed that the VI that best predicted habitat quality at the tree-scale was the Plant Senescence Reflectance Index (PSRI, computed as [Red-Blue]/Red-edge), included in almost all the most parsimonious models. The most parsimonious SMSU included only PSRI as an independent covariate, with a strong positive relation. Although coefficient differences were found between telemetry data (VHF vs. GPS data) both showed a positive overall response. Consequently, Red-edge based PSRI can be considered a reliable estimator of tree-scale foraging habitat quality at landscape extents for future research and management activities including Magellanic woodpeckers living on heterogeneous Nothofagus forests.

Evolution of Canada's Boreal Forest Spatial Patterns as Seen from Space.

Understanding the development of landscape patterns over broad spatial and temporal scales is a major contribution to ecological sciences and is a critical area of research for forested land management. Boreal forests represent an excellent case study for such research because these forests have undergone significant changes over recent decades. We analyzed the temporal trends of four widely-used landscape pattern indices for boreal forests of Canada: forest cover, largest forest patch index, forest edge density, and core (interior) forest cover. The indices were computed over landscape extents ranging from 5,000 ha (n = 18,185) to 50,000 ha (n = 1,662) and across nine major ecozones of Canada. We used 26 years of Landsat satellite imagery to derive annualized trends of the landscape pattern indices. The largest declines in forest cover, largest forest patch index, and core forest cover were observed in the Boreal Shield, Boreal Plain, and Boreal Cordillera ecozones. Forest edge density increased at all landscape extents for all ecozones. Rapidly changing landscapes, defined as the 90th percentile of forest cover change, were among the most forested initially and were characterized by four times greater decrease in largest forest patch index, three times greater increase in forest edge density, and four times greater decrease in core forest cover compared with all 50,000 ha landscapes. Moreover, approximately 18% of all 50,000 ha landscapes did not change due to a lack of disturbance. The pattern database results provide important context for forest management agencies committed to implementing ecosystem-based management strategies.

Citizen science and smartphones take roadkill monitoring to the next level

Road networks, even in industrialized countries, become denser year after year and traffic volumes continue to increase at a steady pace. It is imperative that we monitor the impact of this trend on wildlife, but monitoring roads for flattened fauna is a time consuming effort and roadkill monitoring projects conducted up till now have been relatively small scale both in terms of time and space. This hampers the progress of road ecology analyses at the population level and at larger landscape extents. We demonstrate that citizen science projects in combination with smartphones and other new technologies allow analysis at this level and extent, and simultaneously offer more complete data for safer transportation and mitigation of roadkill hotspots. Monitoring roadkill with citizen scientists poses certain challenges regarding data quality and people management, but we show that these challenges can be addressed, which allows researchers to benefit from the many other advantages and possible applications of monitoring roadkill with citizen scientists, including raising public awareness on the matter.

Informing Disease Models with Temporal and Spatial Contact Structure among GPS-Collared Individuals in Wild Populations

Contacts between hosts are essential for transmission of many infectious agents. Understanding how contacts, and thus transmission rates, occur in space and time is critical to effectively responding to disease outbreaks in free-ranging animal populations. Contacts between animals in the wild are often difficult to observe or measure directly. Instead, one must infer contacts from metrics such as proximity in space and time. Our objective was to examine how contacts between white-tailed deer (Odocoileus virginianus) vary in space and among seasons. We used GPS movement data from 71 deer in central New York State to quantify potential direct contacts between deer and indirect overlap in space use across time and space. Daily probabilities of direct contact decreased from winter (0.05–0.14), to low levels post-parturition through summer (0.00–0.02), and increased during the rut to winter levels. The cumulative distribution for the spatial structure of direct and indirect contact probabilities around a hypothetical point of occurrence increased rapidly with distance for deer pairs separated by 1,000 m – 7,000 m. Ninety-five percent of the probabilities of direct contact occurred among deer pairs within 8,500 m of one another, and 99% within 10,900 m. Probabilities of indirect contact accumulated across greater spatial extents: 95% at 11,900 m and 99% at 49,000 m. Contacts were spatially consistent across seasons, indicating that although contact rates differ seasonally, they occur proportionally across similar landscape extents. Distributions of contact probabilities across space can inform management decisions for assessing risk and allocating resources in response.

Body size and spatial scales in avian response to landscapes: a meta‐analysis

Response of species to their environment is dependent on the scales that individuals interact with landscapes. Several life‐history traits have been suggested to influence scales of response to landscapes, but little empirical work has addressed this issue, limiting our ability to predict the relevant scale(s) at which species respond to their environment. Body size is frequently hypothesized to be relevant for predicting the effects of landscape structure on species given correlations between body size and various measures of mobility or perceptual range, yet evidence of this relationship remains equivocal. We conducted a meta‐analysis of 22 studies to address the question of whether body size is related to the characteristic scale at which bird species respond to landscape structure. On the basis of correlation coefficients and linear mixed models, we show that body size is positively related to the characteristic scale of response to landscapes. The strength of this relationship is less than body size – dispersal allometries, suggesting that response to landscapes may be influenced, in part, by factors other than dispersal (e.g. territory size). However, the strength of the observed body size – scale relationship may have been affected by several factors, including limitations in the range of body sizes and landscape extents tested. Our findings suggest that life‐history traits mediate aspects of the scales at which species respond to the landscape and contribute to developing a more predictive framework for investigating scalar relationships in nature.

Threshold Responses of Forest Birds to Landscape Changes around Exurban Development

Low-density residential development (i.e., exurban development) is often embedded within a matrix of protected areas and natural amenities, raising concern about its ecological consequences. Forest-dependent species are particularly susceptible to human settlement even at low housing densities typical of exurban areas. However, few studies have examined the response of forest birds to this increasingly common form of land conversion. The aim of this study was to assess whether, how, and at what scale forest birds respond to changes in habitat due to exurban growth. We evaluated changes in habitat composition (amount) and configuration (arrangement) for forest and forest-edge species around North America Breeding Bird Survey (BBS) stops between 1986 and 2009. We used Threshold Indicator Taxa Analysis to detect change points in species occurrence at two spatial extents (400-m and 1-km radius buffer). Our results show that exurban development reduced forest cover and increased habitat fragmentation around BBS stops. Forest birds responded nonlinearly to most measures of habitat loss and fragmentation at both the local and landscape extents. However, the strength and even direction of the response changed with the extent for several of the metrics. The majority of forest birds’ responses could be predicted by their habitat preferences indicating that management practices in exurban areas might target the maintenance of forested habitats, for example through easements or more focused management for birds within existing or new protected areas.

The accuracy of land cover-based wetland assessments is influenced by landscape extent

Widespread degradation of wetlands has motivated the development of tools to evaluate wetland condition. The application of field-based tools over large regions can be prohibitively expensive; however, land cover data may provide a surrogate for intensive assessments, enabling rapid and cost-effective evaluation of wetlands throughout whole regions. Our goal was to determine if land cover data could be used to estimate the biotic integrity of wetlands in Alberta’s Beaverhills watershed. Biotic integrity was measured using both plant- and bird-based indices of biotic integrity (IBIs) in 45 wetlands. Land cover data were extracted from seven nested landscape extents (100–3,000 m radii) and used to model IBI scores. Strong, significant predictions of IBI scores were achieved using land cover data from every spatial extent, even after factoring out the influence of location to address the spatial autocorrelation of land cover classes. Plant-based IBI scores were best predicted using data from 100 m buffers and bird-based IBI scores were best predicted using data extracted from 500 m buffers. Road cover or density and measures of the proportion of disturbed land were consistent predictors of IBI score, suggesting their universal importance to plant and bird communities. Simplified models using the proportion of undisturbed land were less accurate than more detailed models (reductions in r 2 of 0.31–0.32). Regardless of the level of detail in land cover classification, our results emphasize the need to optimize landscape extent for the taxonomic group of interest: an issue that is typically poorly articulated in studies reporting on the development of GIS-based assessment methods. Our results also highlight the need to calibrate models in test areas before scaling up, to ensure predictive accuracy.

PREDICTING LANDSCAPE STRUCTURAL METRICS USING ASTER SATELLITE DATA / KRAŠTOVAIZDŽIO STRUKTŪRINIŲ METRIKŲ NUSTATYMAS REMIANTIS ASTER PALYDOVINIAIS DUOMENIMIS

The aim of this study was to predict landscape structural metrics using the features extracted from the ASTER multispectral satellite imagery with 15 m spatial resolution. The landscape structural metrics were calculated on the basis of forest map polygons generated from 1:15000 scaled aerial photos by photo-interpretation technique. The landscape metrics and corresponding image features that are texture parameters and segmentation polygons were determined for four different landscape extents. A stepwise multiple linear regression analysis was carried out to identify the most significant image-derived predictors of landscape metrics for each extent. The regression models established for the landscape metrics including the Number of Patches (NUMP), Edge Density (ED), Shannon's Diversity Index (SDI) and Patch Richness (PR) performed moderately with adjusted R2 values of 0.50 and 0.53 (P<0.01), indicating that 50–53% of the total variation in these landscape metrics can be explained by image-derived features. By contrast, the regression analyses showed that there were weak relationships between the image features and Interspersion Juxtaposition Index (IJI), and Shannon's Evenness Index (SEI) (adj. R2 is varied from 0.12 to 0.30, P<0.01). According to the results of model evaluation, the Entropy measures based on Grey Level Co-occurrence Matrix (GLCM) calculated from the infrared and red bands of ASTER were found as the most correlated parameters with the landscape metrics. Besides, the window size (extent) of 81–144 ha (between 900×900 and 1200×1200 m) might be recommended in estimating the landscape metrics using the ASTER or similar satellite imagery. It can be concluded that the 15 m resolution satellite data used for estimating landscape spatial structure cannot replace aerial photos or very high resolution satellite imagery for local-level inventories. However, it might have potential for predicting landscape heterogeneity for large-scale inventories. Santrauka Tyrimo tikslas buvo prognozuoti kraštovaizdžio struktūrines metrikas, gaunamas iš Aster daugiaspektrių 15 m rezoliucijos palydovinių vaizdų. Atsako kraštovaizdžio struktūrinės metrikos, taikant fotonuotraukų interpretavimo techniką, buvo apskaičiuojamos remiantis miškų žemėlapių, gautų iš 1:15 000 mastelio aerofotonuotraukų, poligonais. Nustatytos keturių skirtingų kraštovaizdžio lygių metrikų ir atitinkamos vaizdų savybės (reljefo parametrai ir poligonų padalijimas). Palaipsnė sudėtinė tiesinė regresinė analizė buvo atliekama identifikuojant reikšmingiausią kiekvieno kraštovaizdžio lygmens atvaizdą. Pagal modelio įvertinimo rezultatus, entropijos matavimo duomenys, pagrįsti Grey lygio tų pačių bendrų įvykių matrica, apskaičiuota remiantis infraraudonųjų spindulių ir ASTER raudonosiomis horizontaliomis juostomis, buvo labiausiai su kraštovaizdžio metrikomis koreliuojantys parametrai. Be to, vertinant kraštovaizdžio metrikas pagal ASTERarpanašius palydovinius duomenis, gali būti rekomenduojama 81–144 ha (tarp 900 ir 900×1200×1200 m) lango dydis (dydis). Galima daryti išvadą, kad 15 m rezoliucijos palydoviniai duomenys, naudojami vertinant kraštovaizdžio erdvinę struktūrą, lokaliajam lygiui inventorizuoti negali pakeisti aerofotonuotraukų ar labai didelės skiriamosios gebos palydovinių vaizdų.

Spatial, temporal, and life history assumptions influence consistency of landscape effects on species distributions

Models describing relationships between landscape features and species distribution patterns often display inter-study inconsistencies. Identifying factors contributing to these inconsistencies is a vital step in clarifying the ecological importance of landscape features and synthesizing an effective knowledge base for use in conservation contexts. We examined the influence of several spatial, temporal, and life history assumptions on the outcomes of distribution versus landscape models (DLMs) relating wetland bird communities at 29 Massachusetts (USA) sites to independent urbanization, wetland, forest, and agricultural landscape gradients. We considered a bird specialization index as well as obligate and facultative species richness as response variables. Landscape gradients were quantified at 10 landscape extents (0–1000 m in 100 m increments) and three time periods (1971, 1985, 2005). Univariate models indicated that our specialization index showed: (1) the strongest response to landscape gradients at small extents (200 m); (2) a negative, threshold response to urbanization was superior to a linear fit; and (3) no evidence of time-lagged effects of landscape change. Interestingly, the form of our model (i.e. linear versus threshold) influenced the extent at which strongest effects were detected. Multivariate models relating the specialization index as well as obligate and facultative species richness to landscape gradients showed evidence of annual variability (i.e. composition, parameter estimates, and variability explained) that did not depend upon an organism’s degree of specialization. Our results provide evidence that violations of common assumptions (e.g. selection of appropriate extent, lack of time-lagged effects, etc.) can impact the outcome of DLMs, which could lead to inter-study inconsistencies.

Regional variation in habitat–occupancy thresholds: a warning for conservation planning

Summary An important target for conservation planning is the minimum amount of habitat needed in a landscape to ensure the persistence of a species. Appropriate targets can be determined by identifying thresholds in the amount of habitat, below which persistence, abundance or occupancy declines rapidly. Although some studies have identified habitat thresholds, we currently have little understanding of the extent to which thresholds vary spatially. This is important for establishing whether we can apply the same planning targets across broad geographical regions. We quantified habitat–occupancy relationships for the koala Phascolarctos cinereus (Goldfuss) in three study regions that span much of its geographical range. Standard and piecewise (broken‐stick/segmented) logistic regression were used to model linear and threshold habitat–occupancy relationships. We then used an information‐theoretic approach to test: (1) whether habitat–occupancy relationships were described better by threshold or linear models and (2) where threshold models were better, whether, and to what extent, threshold points varied among study regions. There was substantially greater support for the threshold than the linear models across a range of habitat qualities and landscape extents. The threshold models generally predicted a rapid decline in occupancy below the threshold points. Estimated threshold points varied, sometimes substantially, among study regions. This may relate to cross‐regional differences in habitat quality, demographic rates, and land‐use patterns. The role of habitat fragmentation is unclear. Synthesis and applications. Variation in threshold points among study regions suggests that we should be wary of using thresholds derived in one region for setting conservation planning targets in another. Rather, we should aim to set specific targets for individual locations (and species), while acknowledging the inherent uncertainties in these targets. This has implications for our ability to make general conservation prescriptions for widely distributed species. Future research should aim to develop generic models capable of predicting threshold responses across different landscapes and life‐history characteristics.

Multi-scale models of grassland passerine abundance in a fragmented system in Wisconsin

Fragmentation of grasslands has been implicated in grassland bird population declines. Multi-scale models are being increasingly used to assess potential factors that influence grassland bird presence, abundance, and productivity. However, studies rarely assess fragmentation metrics, and seldom evaluate more than two scales or interactions among scales. We evaluated the relative importance of characteristics at multiple scales to patterns in relative abundance of Savannah Sparrow (Passerculus sandwichensis), Grasshopper Sparrow (Ammodramus savannarum), Eastern Meadowlark (Sturnella magna), and Bobolink (Dolichonyx oryzivorus). We surveyed birds in 74 southwestern Wisconsin pastures from 1997 to 1999 and compared models with explanatory variables from multiple scales: within-patch vegetation structure (microhabitat), patch (macrohabitat), and three landscape extents. We also examined interactions between macrohabitat and landscape factors. Core area of pastures was an important predictor of relative abundance, and composition of the landscape was more important than configuration. Relative abundance was frequently higher in pastures with more core area and in landscapes with more grassland and less wooded area. The direction and strength of the effect of core pasture size on relative abundance changed depending on amount of wooded area in the landscape. Relative abundance of grassland birds was associated with landscape variables more frequently at the 1200-m scale than at smaller scales. To develop better predictive models, parameters at multiple scales and their interactive effects should be included, and results should be evaluated in the context of microhabitat variability, landscape composition, and fragmentation in the study area.

Thresholds in Songbird Occurrence in Relation to Landscape Structure

Theory predicts the occurrence of threshold levels of habitat in landscapes, below which ecological processes change abruptly. Simulation models indicate that below critical thresholds, fragmentation of habitat influences patch occupancy by decreasing colonization rates and increasing rates of local extinction. Uncovering such putative relationships is important for understanding the demography of species and in developing sound conservation strategies. Using segmented logistic regression, we tested for thresholds in occurrence of 15 bird species as a function of the amount of suitable habitat at multiple scales (150-2000-m radii). Suitable habitat was defined quantitatively based on previously derived, spatially explicit distribution models for each species. The occurrence of 10 out of 15 species was influenced by the amount of habitat at a landscape scale (>or=500-m radius). Of these species all but one were best predicted by threshold models. Six out of nine species exhibited asymptotic thresholds; the effects of habitat loss intensified at low amounts of habitat in a landscape. Landscape thresholds ranged from 8.6% habitat to 28.7% (x= 18.5 +/- 2.6%[95% CI]). For two species landscape thresholds coincided with sensitivity to fragmentation; both species were more likely to occur in large patches, but only when the amount of habitat in a landscape was low. This supports the fragmentation threshold hypothesis. Nevertheless, the occurrence of most species appeared to be unaffected by fragmentation, regardless of the amount of habitat present at landscape extents. The thresholds we identified may be useful to managers in establishing conservation targets. Our results indicate that findings of landscape-scale studies conducted in regions with relatively high proportions of habitat and low fragmentation may not be applicable in regions with low habitat proportions and high fragmentation.