Search

Context Existing demographic models of California condors have not simultaneously considered individual condor movement paths, the distribution and juxtaposition of release sites, habitat components, or the spatial distribution of threats. Objectives Our objectives were to develop a dynamic spatially explicit and individual-based model (IBM) of California condor demography and to evaluate its ability to replicate empirical data on demography and distribution from California (1995–2019). Methods We built a female-only spatially explicit California condor IBM in HexSim, using a daily timestep that allowed us to simulate the foraging behavior of condors, changes in food distribution and availability, and the ephemeral threat of lead in decaying food resources. Results Simulated population size was highly correlated with annual population census data once the population became established with > 50 females (r2 = 0.99). Mean simulated fecundity and mortality estimates were not significantly different from empirical trends (p > 0.05), although empirical data had higher interannual variability. The geographic distribution of modeled condors was similar to the empirical distribution with an overall accuracy of 79%, a commission error of 27%, and an omission error of 9%. Simulated movement density corresponded moderately well to the density of observed GPS locations (weighted kappa = 0.44). Conclusions We developed a spatially explicit California condor IBM that is well-calibrated to empirical data from California. Given its mechanistic underpinnings and flexibility to incorporate a variety of spatial and demographic inputs, we expect our model to be useful for assessing the relative risks and benefits of future condor reintroduction and management scenarios.

Understanding habitat associations is vital for conservation of at‐risk marsh‐endemic wildlife species, particularly those under threat from sea level rise. We modeled environmental and habitat associations of the marsh‐endemic, Federally endangered salt marsh harvest mouse (Reithrodontomys raviventris, RERA) and co‐occurrence with eight associated small mammal species from annual trap data, 1998–2014, in six estuarine marshes in North San Francisco Bay, California. Covariates included microhabitat metrics of elevation and vegetation species and cover; and landscape metrics of latitude–longitude, distance to anthropogenic features, and habitat patch size. The dominant cover was pickleweed (Salicornia pacifica) with 86% mean cover and 37 cm mean height, and bare ground with about 10% mean cover. We tested 38 variants of Bayesian network (BN) models to determine covariates that best account for presence of RERA and of all nine small mammal species. Best models had lowest complexity and highest classification accuracy. Among RERA presence models, three best BN models used covariates of latitude–longitude, distance to paved roads, and habitat patch size, with 0% error of false presence, 20% error of false nonpresence, and 20% overall error. The all‐species presence models suggested that within the pickleweed marsh environment, RERA are mostly habitat generalists. Accounting for presence of other species did not improve prediction of RERA. Habitat attributes compared between RERA and the next most frequently captured species, California vole (Microtus californicus), suggested substantial habitat overlap, with RERA habitat being somewhat higher in marsh elevation, greater in percent cover of the dominant plant species, closer to urban areas, further from agricultural areas, and, perhaps most significant, larger in continuous size of marsh patch. Findings will inform conservation management of the marsh environment for RERA by identifying best microhabitat elements, landscape attributes, and adverse interspecific interactions., We modeled environmental and habitat associations of federally endangered salt marsh harvest mouse (SMHM) and co‐occurrence with eight associated small mammal species from annual trap data, 1998–2014, in six estuarine marshes in North San Francisco Bay, California. Within the pickleweed marsh environment, SMHM are mostly habitat generalists, with substantial habitat overlap with California voles and other species. SMHM selected for higher marsh elevation, greater plant cover, and large marsh patches.

Classical biological control, the introduction of natural enemies to new environments to control unwanted pests or weeds, is, despite numerous successful examples, associated with rising concerns about unwanted environmental impacts such as population decline of nontarget species. Recognition of these biosafety risks is globally increasing, and prerelease assessments of biological control agents (BCAs) have become more rigorous in many countries. We review the current approaches to risk assessment for BCAs as used in Australasia, Europe, and North America. Traditionally, these assessments focus on providing assurance about the specificity of a proposed BCA, generally via a list of suitable versus nonsuitable hosts determined through laboratory specificity tests (i.e., by determining the BCA's physiological host range). The outcome of interactions of proposed agents in the natural environment can differ from laboratory-based predictions. Potential nontarget host testing may be incomplete, additional ecological barriers under field conditions may limit encounters between BCA and nontargets or reduce attack levels, and BCAs could disperse to habitats beyond those used by the target species and adversely affect nontarget species. We advocate for the adoption of more comprehensive, ecologically-based, probabilistic risk assessment approaches to BCA introductions. An example is provided using a Bayesian network that can integrate information on probabilities and uncertainties of a BCA to spread and establish in new habitats, interact with nontarget species in these habitats, and eventually negatively impact the populations of these nontarget species. Our new model, Biocontrol Adverse Impact Probability Assessment, aims to be incorporated into a structured decision-making framework to support national regulatory authorities.

Alaska has diverse boreal ecosystems across heterogeneous landscapes driven by a wide range of biological and geomorphic processes associated with disturbance and successional patterns under a changing climate. To assess historical patterns and rates of change, we quantified the areal extent of ecotypes and the biophysical factors driving change through photo-interpretation of 2200 points on a time-series (∼1949, ∼1978, ∼2007, ∼2017) of geo-rectified imagery for 22 grids across central Alaska. Overall, 68.6% of the area had changes in ecotypes over ∼68 years. Most of the change resulted from increases in upland and lowland forest types, with an accompanying decrease in upland and lowland scrub types, as post-fire succession led to mid- and late-successional stages. Of 17 drivers of landscape change, fire was by far the largest, affecting 46.5% of the region overall from 1949 to 2017. Fire was notably more extensive in the early 1900s. Thermokarst nearly doubled from 3.9% in 1949 to 6.3% in 2017. Riverine ecotypes covered 7.8% area and showed dynamic changes related to channel migration and succession. Using past rates of ecotype transitions, we developed four state-transition models to project future ecotype extent based on historical rates, increasing temperatures, and driver associations. Ecotype changes from 2017 to 2100, nearly tripled for the driver-adjusted RCP6.0 temperature model (30.6%) compared to the historical rate model (11.5%), and the RCP4.5 (12.4%) and RCP8.0 (14.7%) temperature models. The historical-rate model projected 38 ecotypes will gain area and 24 will lose area by 2100. Overall, disturbance and recovery associated with a wide range of drivers across the patchy mosaic of differing aged ecotypes led to a fairly stable overall composition of most ecotypes over long intervals, although fire caused large temporal fluctuations for many ecotypes. Thermokarst, however, is accelerating and projected to have increasingly transformative effects on future ecotype distributions.

Use and popularity of Bayesian network (BN) modeling has greatly expanded in recent years, but many common problems remain. Here, I summarize key problems in BN model construction and interpretation, along with suggested practical solutions. Problems in BN model construction include parameterizing probability values, variable definition, complex network structures, latent and confounding variables, outlier expert judgments, variable correlation, model peer review, tests of calibration and validation, model overfitting, and modeling wicked problems. Problems in BN model interpretation include objective creep, misconstruing variable influence, conflating correlation with causation, conflating proportion and expectation with probability, and using expert opinion. Solutions are offered for each problem and researchers are urged to innovate and share further solutions.

Environmental decision-making requires an understanding of complex interacting systems across scales of space and time. A range of statistical methods, evaluation frameworks and modeling approaches have been applied for conducting structured environmental decision-making under uncertainty. Bayesian Decision Networks (BDNs) are a useful construct for addressing uncertainties in environmental decision-making. In this paper, we apply a BDN to decisions regarding fire management to evaluate the general efficacy and utility of the approach in resource and environmental decision-making. The study was undertaken in south-eastern Australia to examine decisions about prescribed burning rates and locations based on treatment and impact costs. Least-cost solutions were identified but are unlikely to be socially acceptable or practical within existing resources; however, the statistical approach allowed for the identification of alternative, more practical solutions. BDNs provided a transparent and effective method for a multi-criteria decision analysis of environmental management problems.

The northern spotted owl (Strix occidentalis caurina) was listed as threatened under the U.S. Endangered Species Act (ESA) in 1990. We applied modern spatial conservation theory and models to evaluate several candidate critical habitat networks, and sought an efficient conservation solution that encompassed the highest value lands for spotted owl recovery rather than maximizing the total area of potential critical habitat. We created a map of relative habitat suitability, which served as input to the spatial conservation prioritization program Zonation. We used the spatially-explicit individual-based population model HexSim to estimate and compare simulated spotted owl population outcomes among a suite of candidate critical habitat networks that varied in size and spatial arrangement under alternative scenarios of future habitat suitability and barred owl (S. varia) effects. We evaluated simulated spotted owl population outcomes, including total population size, and extinction and quasi-extinction likelihoods for 108 combinations of candidate critical habitat networks by habitat change by barred owl scenarios, both range-wide and within 11 distinct portions of the owl’s range. Barred owl encounter rates and the amount and suitability of habitat had substantial effects on simulated spotted owl populations. When barred owl encounter rates were high, changes in the amount and suitability of habitat had minimal impacts on population performance. Under lowered barred owl encounter rates, candidate critical habitat networks that included most existing high suitability habitat supported a high likelihood of long-term population persistence. Barred owls are currently the primary driving force behind poor population performance of NSOs; however, our models demonstrated that a sufficient area of high suitability habitat remains essential for recovery when effects of barred owls can be reduced. The modeling approach we employed is sufficiently flexible to incorporate new information about spotted owls as it becomes available and could likely be applied to conservation planning for other species.

The Ecosystem Management Decision Support (EMDS) system has been further enhanced with an analytical engine--BayesFusion's SMILE (Structural Modeling, Inference, and Learning Engine) that comes with the GeNIe (Graphical Network Interface) software--for creating Bayesian network (BN) models. BNs are graphical networks of variables linked by probabilities, that have proven useful in decision-aiding for risk analysis and risk management.

Ecological niche models can be a useful tool to identify candidate reintroduction sites for endangered species but have been infrequently used for this purpose. In this paper, we (1) develop activity-specific ecological niche models (nesting, roosting, and feeding) for the critically endangered California condor (Gymnogyps californianus) to aid in reintroduction planning in California, Oregon, and Washington, USA, (2) test the accuracy of these models using empirical data withheld from model development, and (3) integrate model results with information on condor movement ecology and biology to produce predictive maps of reintroduction site suitability. Our approach, which disentangles niche models into activity-specific components, has applications for other species where it is routinely assumed (often incorrectly) that individuals fulfill all requirements for life within a single environmental space. Ecological niche models conformed to our understanding of California condor ecology, had good predictive performance when tested with data withheld from model development, and aided in the identification of several candidate reintroduction areas outside of the current distribution of the species. Our results suggest there are large unoccupied regions of the California condor’s historical range that have retained ecological features similar to currently occupied habitats, and thus could be considered for future reintroduction efforts. Combining our activity-specific ENMs with ground reconnaissance and information on other threat factors that could not be directly incorporated into empirical ENMs will ultimately improve our ability to select successful reintroduction sites for the California condor.

We project the effects of transitional changes among 60 vegetation and other land cover types (“ecotypes”) in northwest Alaska over the 21st century on habitats of 162 bird and 39 mammal species known or expected to occur regularly in the region. This analysis, encompassing a broad suite of arctic and boreal wildlife species, entailed building wildlife-habitat matrices denoting levels of use of each ecotype by each species, and projecting habitat changes under historic and expected accelerated future rates of change from increasing mean annual air temperature based on the average of 5 global climate models under the A1B emissions scenario, and from potential influence of a set of 23 biophysical drivers. Under historic rates of change, we project that 52 % of the 201 species will experience an increase in medium- and high-use habitats, 3 % no change, and 45 % a decrease, and that a greater proportion of mammal species (62 %) will experience habitat declines than will bird species (50 %). Outcomes become more dire (more species showing habitat loss) under projections made from effects of biophysical drivers and especially from increasing temperature, although species generally associated with increasing shrub and tree ecotypes will likely increase in distribution. Changes in wildlife habitats likely will also affect trophic cascades, ecosystem function, and ecosystem services; of particular significance are the projected declines in habitats of most small mammals that form the prey base for mesocarnivores and raptors, and habitat declines in 25 of the 50 bird and mammal species used for subsistence hunting and trapping.

Sensitivity analysis—determination of how prediction vari- ables affect response variables—of individual-based models (IBMs) are few but important to the interpretation of model output. We present sensitiv- ity analysis of a spatially explicit IBM (HexSim) of a threatened species, the Northern Spotted Owl (NSO; Strix occidentalis caurina) in Washington, USA. We explored sensitivity to HexSim variables representing habitat qual- ity, movement, dispersal, and model architecture; previous NSO studies have well established sensitivity of model output to vital rate variation. We de- veloped "normative" (expected) model settings from field studies, and then varied the values of 1 input parameter at a time by ±10% and ±50% of their normative values to determine influence on response variables of pop- ulation size and trend. We determined time to population equilibration and dynamics of populations above and below carrying capacity. Recovery time from small population size to carrying capacity greatly exceeded decay time from an overpopulated condition, suggesting lag time required to repopu- late newly available habitat. Response variables were most sensitive to input parameters of habitat quality which are well-studied for this species and con- trollable by management. HexSim thus seems useful for evaluating potential NSO population responses to landscape patterns for which good empirical information is available.

Climate warming affects arctic and boreal ecosystems by interacting with numerous biophysical factors across heterogeneous landscapes. To assess potential effects of warming on diverse local-scale ecosystems (ecotypes) across northwest Alaska, we compiled data on historical areal changes over the last 25–50 years. Based on historical rates of change relative to time and temperature, we developed three state-transition models to project future changes in area for 60 ecotypes involving 243 potential transitions during three 30-year periods (ending 2040, 2070, 2100). The time model, assuming changes over the past 30 years continue at the same rate, projected a net change, or directional shift, of 6 % by 2100. The temperature model, using past rates of change relative to the past increase in regional mean annual air temperatures (1 °C/30 year), projected a net change of 17 % in response to expected warming of 2, 4, and 6 °C at the end of the three periods. A rate-adjusted temperature model, which adjusted transition rates (±50 %) based on assigned feedbacks associated with 23 biophysical drivers, estimated a net change of 13 %, with 33 ecotypes gaining and 23 ecotypes losing area. Major drivers included shrub and tree expansion, fire, succession, and thermokarst. Overall, projected changes will be modest over the next century even though climate warming increased transition rates up to 9 fold. The strength of this state-transition modeling is that it used a large dataset of past changes to provide a comprehensive assessment of likely future changes associated with numerous drivers affecting the full diversity of ecosystems across a broad region.

Conservation practitioners are increasingly looking to species translocations as a tool to recover imperiled taxa. Quantitative predictions of where animals are likely to move when released into new areas would allow managers to better address the social, institutional, and ecological dimensions of conservation translocations. Using >5 million California condor (Gymnogyps californianus) occurrence locations from 75 individuals, we developed and tested circuit-based models to predict condor movement away from release sites. We found that circuit-based models of electrical current were well calibrated to the distribution of condor movement data in southern and central California (continuous Boyce Index = 0.86 and 0.98, respectively). Model calibration was improved in southern California when additional nodes were added to the circuit to account for nesting and feeding areas, where condor movement densities were higher (continuous Boyce Index = 0.95). Circuit-based projections of electrical current around a proposed release site in northern California comported with the condor’s historical distribution and revealed that, initially, condor movements would likely be most concentrated in northwestern California and southwest Oregon. Landscape linkage maps, which incorporate information on landscape resistance, complement circuit-based models and aid in the identification of specific avenues for population connectivity or areas where movement between populations may be constrained. We found landscape linkages in the Coast Range and the Sierra Nevada provided the most connectivity to a proposed reintroduction site in northern California. Our methods are applicable to conservation translocations for other species and are flexible, allowing researchers to develop multiple competing hypotheses when there are uncertainties about landscape or social attractants, or uncertainties in the landscape conductance surface.

Daily flight activity patterns of forest insects are influenced by temporal and meteorological conditions. Temperature and time of day are frequently cited as key drivers of activity; however, complex interactions between multiple contributing factors have also been proposed. Here, we report individual Bayesian network models to assess the probability of flight activity of three exotic insects, Hylurgus ligniperda, Hylastes ater, and Arhopalus ferus in a managed plantation forest context. Models were built from 7,144 individual hours of insect sampling, temperature, wind speed, relative humidity, photon flux density, and temporal data. Discretized meteorological and temporal variables were used to build naive Bayes tree augmented networks. Calibration results suggested that the H. ater and A. ferus Bayesian network models had the best fit for low Type I and overall errors, and H. ligniperda had the best fit for low Type II errors. Maximum hourly temperature and time since sunrise had the largest influence on H. ligniperda flight activity predictions, whereas time of day and year had the greatest influence on H. ater and A. ferus activity. Type II model errors for the prediction of no flight activity is improved by increasing the model’s predictive threshold. Improvements in model performance can be made by further sampling, increasing the sensitivity of the flight intercept traps, and replicating sampling in other regions. Predicting insect flight informs an assessment of the potential phytosanitary risks of wood exports. Quantifying this risk allows mitigation treatments to be targeted to prevent the spread of invasive species via international trade pathways.