Melinda G. Knutson

ESTIMATING SITE OCCUPANCY, COLONIZATION, AND LOCAL EXTINCTION WHEN A SPECIES IS DETECTED IMPERFECTLY

Few species are likely to be so evident that they will always be detected when present. Failing to allow for the possibility that a target species was present, but undetected, at a site will lead to biased estimates of site occupancy, colonization, and local extinction probabilities. These population vital rates are often of interest in long-term monitoring programs and metapopulation studies. We present a model that enables direct estimation of these parameters when the probability of detecting the species is less than 1. The model does not require any assumptions of process stationarity, as do some previous methods, but does require detection/nondetection data to be collected in a manner similar to Pollocks robust design as used in mark-recapture studies. Via simulation, we show that the model provides good estimates of parameters for most scenarios considered. We illustrate the method with data from monitoring programs of Northern Spotted Owls ( Strix occiden- talis caurina) in northern California and tiger salamanders (Ambystoma tigrinum) in Min- nesota, USA.

AGRICULTURAL PONDS SUPPORT AMPHIBIAN POPULATIONS

In some agricultural regions, natural wetlands are scarce, and constructed agricultural ponds may represent important alternative breeding habitats for amphibians. Properly managed, these agricultural ponds may effectively increase the total amount of breeding habitat and help to sustain populations. We studied small, constructed agricultural ponds in southeastern Minnesota to assess their value as amphibian breeding sites. Our study examined habitat factors associated with amphibian reproduction at two spatial scales: the pond and the landscape surrounding the pond. We found that small agricultural ponds in southeastern Minnesota provided breeding habitat for at least 10 species of amphibians. Species richness and multispecies reproductive success were more closely associated with characteristics of the pond (water quality, vegetation, and predators) compared with char- acteristics of the surrounding landscape, but individual species were associated with both pond and landscape variables. Ponds surrounded by row crops had similar species richness and reproductive success compared with natural wetlands and ponds surrounded by non- grazed pasture. Ponds used for watering livestock had elevated concentrations of phos- phorus, higher turbidity, and a trend toward reduced amphibian reproductive success. Spe- cies richness was highest in small ponds, ponds with lower total nitrogen concentrations, tiger salamanders ( Ambystoma tigrinum) present, and lacking fish. Multispecies reproduc- tive success was best in ponds with lower total nitrogen concentrations, less emergent vegetation, and lacking fish. Habitat factors associated with higher reproductive success varied among individual species. We conclude that small, constructed farm ponds, properly managed, may help sustain amphibian populations in landscapes where natural wetland habitat is rare. We recommend management actions such as limiting livestock access to the pond to improve water quality, reducing nitrogen input, and avoiding the introduction of fish.

A hierarchical spatial model of avian abundance with application to Cerulean Warblers

Surveys collecting count data are the primary means by which abundance is indexed for birds. These counts are confounded, however, by nuisance effects including observer effects and spatial correlation between counts. Current methods poorly accom- modate both observer and spatial effects because modeling these spatially autocorrelated counts within a hierarchical framework is not practical using standard statistical approaches. We propose a Bayesian approach to this problem and provide as an example of its imple- mentation a spatial model of predicted abundance for the Cerulean Warbler (Dendroica cerulea) in the Prairie-Hardwood Transition of the upper midwestern United States. We used an overdispersed Poisson regression with fixed and random effects, fitted by Markov chain Monte Carlo methods. We used 21 years of North American Breeding Bird Survey counts as the response in a loglinear function of explanatory variables describing habitat, spatial relatedness, year effects, and observer effects. The model included a conditional autoregressive term representing potential correlation between adjacent route counts. Cat- egories of explanatory habitat variables in the model included land cover composition and configuration, climate, terrain heterogeneity, and human influence. The inherent hierarchy in the model was from counts occurring, in part, as a function of observers within survey routes within years. We found that the percentage of forested wetlands, an index of wetness potential, and an interaction between mean annual precipitation and deciduous forest patch size best described Cerulean Warbler abundance. Based on a map of relative abundance derived from the posterior parameter estimates, we estimated that only 15% of the species population occurred on federal land, necessitating active engagement of public landowners and state agencies in the conservation of the breeding habitat for this species. Models of this type can be applied to any data in which the response is counts, such as animal counts, activity (e.g., nest) counts, or species richness. The most noteworthy practical application of this spatial modeling approach is the ability to map relative species abundance. The functional relationships that we elucidated for the Cerulean Warbler provide a basis for the development of management programs and may serve to focus management and monitoring on areas and habitat variables important to Cerulean Warblers.

Climate change, uncertainty, and natural resource management†

ABSTRACT n Climate change and its associated uncertainties are of concern to natural resource managers. Although aspects of climate change may be novel (e.g., system change and nonstationarity), natural resource managers have long dealt with uncertainties and have developed corresponding approaches to decision-making. Adaptive resource management is an application of structured decision-making for recurrent decision problems with uncertainty, focusing on management objectives, and the reduction of uncertainty over time. We identified 4 types of uncertainty that characterize problems in natural resource management. We examined ways in which climate change is expected to exacerbate these uncertainties, as well as potential approaches to dealing with them. As a case study, we examined North American waterfowl harvest management and considered problems anticipated to result from climate change and potential solutions. Despite challenges expected to accompany the use of adaptive resource management to address problems associated with climate change, we conclude that adaptive resource management approaches will be the methods of choice for managers trying to deal with the uncertainties of climate change.

Predicting Regional Abundance of Rare Grassland Birds with a Hierarchical Spatial Count Model

Abstract Grassland birds are among the most imperiled groups of birds in North America. Unfortunately, little is known about the location of regional concentrations of these birds, thus regional or statewide conservation efforts may be inappropriately applied, reducing their effectiveness. We identified environmental covariates associated with the abundance of five grassland birds in the upper midwestern United States (Bobolink [Dolichonyx oryzivorus], Grasshopper Sparrow [Ammodramus savannarum], Henslows Sparrow [A. henslowii], Sedge Wren [Cistothorus platensis], and Upland Sandpiper [Bartramia longicauda]) with a hierarchical spatial count model fitted with Markov chain Monte Carlo methods. Markov chain Monte Carlo methods are well suited to this task because they are able to incorporate effects associated with autocorrelated counts and nuisance effects associated with years and observers, and the resulting models can be used to map predicted abundance at a landscape scale. Environmental covariates were derived from five suites of variables: landscape composition, landscape configuration, terrain heterogeneity and physiognomy, climate, and human influence. The final models largely conformed to our a priori expectations. Bobolinks and Henslows Sparrows were strongly sensitive to grassland patch area. All of the species except Henslows Sparrows exhibited substantial negative relations with forest composition, often at multiple spatial scales. Climate was found to be important for all species, and was the most important factor influencing abundance of Grasshopper Sparrows. After mapping predicted abundance, we found no obvious correspondence in the regional patterns of the five species. Thus, no clearly defined areas exist within the upper midwestern United States where management plans can be developed for a whole suite of grassland birds. Instead, a larger, region-wide initiative setting different goals for different species is recommended.

Discontinuities, cross-scale patterns, and the organization of ecosystems

Ecological structures and processes occur at specific spatiotemporal scales, and interactions that occur across multiple scales mediate scale-specific (e.g., individual, community, local, or regional) responses to disturbance. Despite the importance of scale, explicitly incorporating a multi-scale perspective into research and management actions remains a challenge. The discontinuity hypothesis provides a fertile avenue for addressing this problem by linking measureable proxies to inherent scales of structure within ecosystems. Here we outline the conceptual framework underlying discontinuities and review the evidence supporting the discontinuity hypothesis in ecological systems. Next we explore the utility of this approach for understanding cross-scale patterns and the organization of ecosystems by describing recent advances for examining nonlinear responses to disturbance and phenomena such as extinctions, invasions, and resilience. To stimulate new research, we present methods for performing discontinuity analysis, detail outstanding knowledge gaps, and discuss potential approaches for addressing these gaps.

Scaling Local Species-habitat Relations to the Larger Landscape with a Hierarchical Spatial Count Model

Much of what is known about avian species-habitat relations has been derived from studies of birds at local scales. It is entirely unclear whether the relations observed at these scales translate to the larger landscape in a predictable linear fashion. We derived habitat models and mapped predicted abundances for three forest bird species of eastern North America using bird counts, environmental variables, and hierarchical models applied at three spatial scales. Our purpose was to understand habitat associations at multiple spatial scales and create predictive abundance maps for purposes of conservation planning at a landscape scale given the constraint that the variables used in this exercise were derived from local-level studies. Our models indicated a substantial influence of landscape context for all species, many of which were counter to reported associations at finer spatial extents. We found land cover composition provided the greatest contribution to the relative explained variance in counts for all three species; spatial structure was second in importance. No single spatial scale dominated any model, indicating that these species are responding to factors at multiple spatial scales. For purposes of conservation planning, areas of predicted high abundance should be investigated to evaluate the conservation potential of the landscape in their general vicinity. In addition, the models and spatial patterns of abundance among species suggest locations where conservation actions may benefit more than one species.

A cautionary tale regarding use of the National Land Cover Dataset 1992

Abstract Digital land-cover data are among the most popular data sources used in ecological research and natural resource management. However, processes for accurate land-cover classification over large regions are still evolving. We identified inconsistencies in the National Land Cover Dataset 1992, the most current and available representation of land cover for the conterminous United States. We also report means to address these inconsistencies in a bird-habitat model. We used a Geographic Information System (GIS) to position a regular grid (or lattice) over the upper midwestern United States and summarized the proportion of individual land covers in each cell within the lattice. These proportions were then mapped back onto the lattice, and the resultant lattice was compared to satellite paths, state borders, and regional map classification units. We observed mapping inconsistencies at the borders between mapping regions, states, and Thematic Mapper (TM) mapping paths in the upper midwestern United States, particularly related to grassland-herbaceous, emergent-herbaceous wetland, and small-grain land covers. We attributed these discrepancies to differences in image dates between mapping regions, suboptimal image dates for distinguishing certain land-cover types, lack of suitable ancillary data for improving discrimination for rare land covers, and possibly differences among image interpreters. To overcome these inconsistencies for the purpose of modeling regional populations of birds, we combined grassland-herbaceous and pasture-hay land-cover classes and excluded the use of emergent-herbaceous and small-grain land covers. We recommend that users of digital land-cover data conduct similar assessments for other regions before using these data for habitat evaluation. Further, caution is advised in using these data in the analysis of regional land-cover change because it is not likely that future digital land-cover maps will repeat the same problems, thus resulting in biased estimates of change.

Adaptive management in the U.S. National Wildlife Refuge System: Science-management partnerships for conservation delivery

Adaptive management is an approach to recurrent decision making in which uncertainty about the decision is reduced over time through comparison of outcomes predicted by competing models against observed values of those outcomes. The National Wildlife Refuge System (NWRS) of the U.S. Fish and Wildlife Service is a large land management program charged with making natural resource management decisions, which often are made under considerable uncertainty, severe operational constraints, and conditions that limit ability to precisely carry out actions as intended. The NWRS presents outstanding opportunities for the application of adaptive management, but also difficult challenges. We describe two cooperative programs between the Fish and Wildlife Service and the U.S. Geological Survey to implement adaptive management at scales ranging from small, single refuge applications to large, multi-refuge, multi-region projects. Our experience to date suggests three important attributes common to successful implementation: a vigorous multi-partner collaboration, practical and informative decision framework components, and a sustained commitment to the process. Administrators in both agencies should consider these attributes when developing programs to promote the use and acceptance of adaptive management in the NWRS.

EVALUATION OF SPATIAL MODELS TO PREDICT VULNERABILITY OF FOREST BIRDS TO BROOD PARASITISM BY COWBIRDS

We constructed alternative spatial models at two scales to predict Brown- headed Cowbird (Molothrus ater) parasitism rates from land cover maps. The local-scale models tested competing hypotheses about the relationship between cowbird parasitism and distance of host nests from a forest edge (forest-nonforest boundary). The landscape models tested competing hypotheses about how landscape features (e.g., forests, agricultural fields) interact to determine rates of cowbird parasitism. The models incorporate spatial neighbor- hoods with a radius of 2.5 km in their formulation, reflecting the scale of the majority of cowbird commuting activity. Field data on parasitism by cowbirds (parasitism rate and number of cowbird eggs per nest) were collected at 28 sites in the Driftless Area Ecoregion of Wisconsin, Minnesota, and Iowa and were compared to the predictions of the alternative models. At the local scale, there was a significant positive relationship between cowbird parasitism and mean distance of nest sites from the forest edge. At the landscape scale, the best fitting models were the forest-dependent and forest-fragmentation-dependent models, in which more heavily forested and less fragmented landscapes had higher parasitism rates. However, much of the explanatory power of these models results from the inclusion of the local-scale relationship in these models. We found lower rates of cowbird parasitism than did most Midwestern studies, and we identified landscape patterns of cowbird parasitism that are opposite to those reported in several other studies of Midwestern songbirds. We caution that cowbird parasitism patterns can be unpredictable, depending upon ecoregional location and the spatial extent, and that our models should be tested in other ecoregions before they are applied there. Our study confirms that cowbird biology has a strong spatial component, and that improved spatial models applied at multiple spatial scales will be required to predict the effects of landscape and forest management on cowbird parasitism of forest birds.