Complexity is costly: a meta‐analysis of parametric and non‐parametric methods for short‐term population forecasting

Short‐term forecasts based on time series of counts or survey data are widely used in population biology to provide advice concerning the management, harvest and conservation of natural populations. A common approach to produce these forecasts uses time‐series models, of different types, fit to time series of counts. Similar time‐series models are used in many other disciplines, however relative to the data available in these other disciplines, population data are often unusually short and noisy and models that perform well for data from other disciplines may not be appropriate for population data. In order to study the performance of time‐series forecasting models for natural animal population data, we assembled 2379 time series of vertebrate population indices from actual surveys. Our data were comprised of three vastly different types: highly variable (marine fish productivity), strongly cyclic (adult salmon counts), and small variance but long‐memory (bird and mammal counts). We tested the predictive performance of 49 different forecasting models grouped into three broad classes: autoregressive time‐series models, non‐linear regression‐type models and non‐parametric time‐series models. Low‐dimensional parametric autoregressive models gave the most accurate forecasts across a wide range of taxa; the most accurate model was one that simply treated the most recent observation as the forecast. More complex parametric and non‐parametric models performed worse, except when applied to highly cyclic species. Across taxa, certain life history characteristics were correlated with lower forecast error; specifically, we found that better forecasts were correlated with attributes of slow growing species: large maximum age and size for fishes and high trophic level for birds. Synthesis Evaluating the data support for multiple plausible models has been an integral focus of many ecological analyses. However, the most commonly used tools to quantify support have weighted models’ hindcasting and forecasting abilities. For many applications, predicting the past may be of little interest. Concentrating only on the future predictive performance of time series models, we performed a forecasting competition among many different kinds of statistical models, applying each to many different kinds of vertebrate time series of population abundance. Low‐dimensional (simple) models performed well overall, but more complex models did slightly better when applied to time series of cyclic species (e.g. salmon).     

Using MODIS satellite imagery to predict hantavirus risk

Aims Sin Nombre virus (SNV), a strain of hantavirus, causes hantavirus pulmonary syndrome (HPS) in humans, a deadly disease with high mortality rate (> 50%). The primary virus host is the deer mouse, and greater abundance of deer mice has been shown to increase the human risk of HPS. Our aim is to identify and compare vegetation indices and associated time lags for predicting hantavirus risk using remotely sensed imagery. Location Utah, USA. Methods A 5‐year time‐series of moderate‐resolution imaging spectroradiometer (MODIS) satellite imagery and corresponding field data was utilized to compare various vegetation indices that measure productivity with the goal of indirectly estimating mouse abundance and SNV prevalence. Relationships between the vegetation indices and deer mouse density, SNV prevalence and the number of infected deer mice at various time lags were examined to assess which indices and associated time lags might be valuable in predicting SNV outbreaks. Results The results reveal varying levels of positive correlation between the vegetation indices and deer mouse density as well as the number of infected deer mice. Among the vegetation indices, the normalized difference vegetation index (NDVI) and the enhanced vegetation index (EVI) produced the highest correlations with deer mouse density and the number of infected deer mice using a time lag of 1.0 to 1.3 years for May and June imagery. Main conclusions This study demonstrates the potential for using MODIS time‐series satellite imagery in estimating deer mouse abundance and predicting hantavirus risk. The 1‐year time lag provides a great opportunity to apply satellite imagery to predict upcoming SNV outbreaks, allowing preventive strategies to be adopted. Analysis of different predictive indices and lags could also be valuable in identifying the time windows for data collection for practical uses in monitoring rodent abundance and subsequent disease risk to humans.     

Estimating density dependence from population time series using demographic theory and life-history data

For populations with a density-dependent life history reproducing at discrete annual intervals, we analyze small or moderate fluctuations in population size around a stable equilibrium, which is applicable to many vertebrate populations. Using a life history having age at maturity alpha, with stochasticity and density dependence in adult recruitment and mortality, we derive a linearized autoregressive equation with time lags from 1 to alpha yr. Contrary to current interpretations, the coefficients corresponding to different time lags in the autoregressive dynamics are not simply measures of delayed density dependence but also depend on life-history parameters. The theory indicates that the total density dependence in a life history, D, should be defined as the negative elasticity of population growth rate per generation with respect to change in population size, [Formula: see text], where lambda is the asymptotic multiplicative growth rate per year, T is the generation time, and N is adult population size. The total density dependence in the life history, D, can be estimated from the sum of the autoregression coefficients. We estimate D in populations of seven vertebrate species for which life-history studies and unusually long time series of complete population censuses are available. Estimates of D were statistically significant and large, on the order of 1 or higher, indicating strong density dependence in five of the seven species. We also show that life history can explain the qualitative features of population autocorrelation functions and power spectra and observations of increasing empirical variance in population size with increasing length of time series.     

Comparative population dynamics of large and small mammals in the Northern Hemisphere: deterministic and stochastic forces

Deterministic feedbacks within populations interact with extrinsic, stochastic processes to generate complex patterns of animal abundance over time and space. Animals inherently differ in their responses to fluctuating environments due to differences in body sizes and life history traits. However, controversy remains about the relative importance of deterministic and stochastic forces in shaping population dynamics of large and small mammals. We hypothesized that effects of environmental stochasticity and density dependence are stronger in small mammal populations relative to their effects in large mammal populations and thus differentiate the patterns of population dynamics between them. We conducted an extensive, comparative analysis of population dynamics in large and small mammals to test our hypothesis, using seven population parameters to describe general dynamic patterns for 23 (14 species) time series of observations of abundance of large mammals and 38 (21 species) time series for small mammals. We used state‐space models to estimate the strength of direct and delayed density dependence as well as the strength of environmental stochasticity. We further used phylogenetic comparative analysis to detect differences in population dynamic patterns and individual population parameters, respectively, between large and small mammals. General population dynamic patterns differed between large and small mammals. However, the strength of direct and delayed density dependence was comparable between large and small mammals. Moreover, the variances of population growth rates and environmental stochasticity were greater in small mammals than in large mammals. Therefore, differences in population response to stochastic forces and strength of environmental stochasticity are the primary factor that differentiates population dynamic patterns between large and small mammal species.     

Density‐dependent and density‐independent drivers of population change in Barton Springs salamanders

Understanding population change is essential for conservation of imperiled species, such as amphibians. Worldwide amphibian declines have provided an impetus for investigating their population dynamics, which can involve both extrinsic (density‐independent) and intrinsic (density‐dependent) drivers acting differentially across multiple life stages or age classes. In this study, we examined the population dynamics of the endangered Barton Springs Salamander (Eurycea sosorum) using data from a long‐term monitoring program. We were interested in understanding both the potential environmental drivers (density‐independent factors) and demographic factors (interactions among size classes, negative density dependence) to better inform conservation and management activities. We used data from three different monitoring regimes and multivariate autoregressive state‐space models to quantify environmental effects (seasonality, discharge, algae, and sediment cover), intraspecific interactions among three size classes, and intra‐class density dependence. Results from our primary data set revealed similar patterns among sites and size classes and were corroborated by our out‐of‐sample data. Cross‐correlation analysis showed juvenile abundance was most strongly correlated with a 9‐month lag in aquifer discharge, which we suspect is related to inputs of organic carbon into the aquifer. However, sedimentation limited juvenile abundance at the surface, emphasizing the importance of continued sediment management. Recruitment from juveniles to the sub‐adult size class was evident, but negative density‐dependent feedback ultimately regulated each size class. Negative density dependence may be an encouraging sign for the conservation of E. sosorum because populations that can reach carrying capacity are less likely to go extinct compared to unregulated populations far below their carrying capacity. However, periodic population declines coupled with apparent migration into the aquifer complicate assessments of species status. Although both density‐dependent and density‐independent drivers of population change are not always apparent in time series of animal populations, both have important implications for conservation and management of E. sosorum.     

Short-term effects of reduced white-tailed deer density on insect communities in a strongly overbrowsed boreal forest ecosystem

Browsing by overabundant deer modifies plant communities and alters forest regeneration, which can indirectly impact associated insect fauna. We tested the hypothesis that the response of insect communities to changes in deer abundance should depend on the strength of their association with plants, which we considered as a key functional trait. Seven years after a deer density control experiment was established in partly harvested forests on Anticosti Island (Quebec, Canada), we evaluated the effects of reducing white-tailed deer (Odocoileus virginianus) density from >20 down to 15, 7.5 and 0 deer km^?2, on four insect taxa representing different levels of dependence on plants. As predicted by our hypothesis, the sensitivity of insect taxa to deer density decreased along a gradient representing their degree of association with plants. Carabidae remained unaffected, while Apoidea and Syrphidae communities differed between uncontrolled and reduced deer densities, but not as clearly as for Lepidoptera. As expected, insect communities responded faster in harvested than in forested areas because vegetation changes more rapidly in open habitats. For most insect taxa, dominant species were the most strongly affected by deer density reduction, but it was clearly stronger for predator taxa (Syrphidae and Carabidae). A fast recovery of rare species was observed for macro Lepidoptera. Reducing deer density down to 15 deer km^?2 is sufficient to restore insect diversity on Anticosti Island, but it is unlikely to be efficient in all situations, particularly when competing tree regeneration is firmly established.     

Absence of founder effect and evidence for adaptive divergence in a recently introduced insular population of white‐tailed deer (Odocoileus virginianus)

Islands are generally colonized by few individuals which could lead to a founder effect causing loss of genetic diversity and rapid divergence by strong genetic drift. Insular conditions can also induce new selective pressures on populations. Here, we investigated the extent of genetic differentiation within a white‐tailed deer (Odocoileus virginianus) population introduced on an island and its differentiation with its source mainland population. In response to their novel environmental conditions, introduced deer changed phenotypically from mainland individuals, therefore we investigated the genetic bases of the morphological differentiation. The study was conducted on Anticosti Island (Québec, Canada) where 220 individuals were introduced 120 years ago, resulting in a population size over 160,000 individuals. We used genotyping‐by‐sequencing (GBS) to generate 8,518 filtered high‐quality SNPs and compared patterns of genetic diversity and differentiation between the continental and Anticosti Island populations. Clustering analyses indicated a single panmictic island population and no sign of isolation by distance. Our results revealed a weak, albeit highly significant, genetic differentiation between the Anticosti Island population and its source population (mean F_ST = 0.005), which allowed a population assignment success of 93%. Also, the high genetic diversity maintained in the introduced population supports the absence of a strong founder effect due to the large number of founders followed by rapid population growth. We further used a polygenic approach to assess the genetic bases of the divergent phenotypical traits between insular and continental populations. We found loci related to muscular function and lipid metabolism, which suggested that these could be involved in local adaptation on Anticosti Island. We discuss these results in a harvest management context.     

Direct negative density‐dependence in a pond‐breeding frog population

Understanding population dynamics is critical for the management of animal populations. Comparatively little is known about the relative importance of endogenous (i.e. density‐dependent) and exogenous (i.e. density‐independent) factors on the population dynamics of amphibians with complex life cycles. We examined the potential effects of density‐dependent and ‐independent (i.e. climatic) factors on population dynamics by analyzing a 15‐yr time series data of the agile frog Rana dalmatina population from Târnava Mare Valley, Romania. We used two statistical models: 1) the partial rate correlation function to identify the feedback structure and the potential time lags in the time series data and 2) a Gompertz state‐space model to simultaneously investigate direct and delayed density dependence as well as climatic effects on population growth rate. We found evidence for direct negative density dependence, whereas delayed density dependence and climate did not show a strong influence on population growth rate. Here we demonstrated that direct density dependence rather than delayed density dependence or climate determined the dynamics of our study population. Our results confirm the findings of many experimental studies and suggest that density dependence may buffer amphibian populations against environmental stress. Consequently, it may not be easy to scale up from individual‐level effects to population‐level effects.     

Modeling the Relationship Between Fecal Pellet Indices and Deer Density

Abstract: Although fecal pellet counts have been widely used to index changes in deer abundance in forests, few studies have modeled the relationship between the indices and deer density. We examined the relationships between 3 fecal pellet indices (total pellets, pellet groups, and pellet frequency) and the density of deer (primarily red deer [Cervus elaphus scoticus]) in 20 enclosures in the North and South islands of New Zealand. In each enclosure we estimated the 3 indices on 30 randomly located 150-m transects, with each transect having 30 circular plots of 3.14 m². We developed 4 candidate models (1 linear and 3 nonlinear) to describe the relationship between the indices and deer density. We used a Bayesian analysis to account for uncertainty in the estimates of deer abundance and to facilitate fitting models that included random transect effects. The 4 models explained the relationship between the 3 indices and deer density similarly well. The slopes of the linear relationships between the 3 indices and deer density were positive. Our results suggest that fecal pellet counts may be useful indices of deer abundance.     

Disentangling demographic effects of red deer on chamois population dynamics

Investigating the impact of ecological factors on sex‐ and age‐specific vital rates is essential to understand animal population dynamics and detect the potential for interactions between sympatric species. We used block count data and autoregressive linear models to investigate variation in birth rate, kid survival, female survival, and male survival in a population of Alpine chamois Rupicapra rupicapra rupicapra monitored over 27 years within the Stelvio National Park, Central Italian Alps, as function of climatic variables, density dependence, and interspecific competition with red deer Cervus elaphus. We also used path analysis to assess the indirect effect of deer abundance on chamois growth rate mediated by each demographic parameter. Based on previous findings, we predicted that birth rate at [t] would negatively relate to red deer abundance at year [t − 1]; survival rates between [t] and [t + 1] would negatively relate to red deer abundance at year [t − 1] and to the interactive effect of winter precipitation at [t + 1] and chamois density at [t]. Our results showed that birth rate was positively related to spring–summer precipitation in the previous year, but this effect was hampered by increasing red deer abundance. Kid and female survival rates were negatively related to the combined effect of chamois abundance and winter precipitation. Male and female survival rates were negatively related to lagged red deer abundance. The path analysis supported a negative indirect effect of red deer abundance on chamois growth rate mediated by birth rate and female survival. Our results suggest that chamois population dynamics was largely explained by the synergistic effect of density dependence and winter harshness, as well as by interspecific competition with red deer, whose effects were seemingly stronger on the kid–female segment of the population.     

Incorporating capture heterogeneity in the estimation of autoregressive coefficients of animal population dynamics using capture–recapture data

Population dynamic models combine density dependence and environmental effects. Ignoring sampling uncertainty might lead to biased estimation of the strength of density dependence. This is typically addressed using state‐space model approaches, which integrate sampling error and population process estimates. Such models seldom include an explicit link between the sampling procedures and the true abundance, which is common in capture–recapture settings. However, many of the models proposed to estimate abundance in the presence of capture heterogeneity lead to incomplete likelihood functions and cannot be straightforwardly included in state‐space models. We assessed the importance of estimating sampling error explicitly by taking an intermediate approach between ignoring uncertainty in abundance estimates and fully specified state‐space models for density‐dependence estimation based on autoregressive processes. First, we estimated individual capture probabilities based on a heterogeneity model for a closed population, using a conditional multinomial likelihood, followed by a Horvitz–Thompson estimate for abundance. Second, we estimated coefficients of autoregressive models for the log abundance. Inference was performed using the methodology of integrated nested Laplace approximation (INLA). We performed an extensive simulation study to compare our approach with estimates disregarding capture history information, and using R‐package VGAM, for different parameter specifications. The methods were then applied to a real data set of gray‐sided voles Myodes rufocanus from Northern Norway. We found that density‐dependence estimation was improved when explicitly modeling sampling error in scenarios with low process variances, in which differences in coverage reached up to 8% in estimating the coefficients of the autoregressive processes. In this case, the bias also increased assuming a Poisson distribution in the observational model. For high process variances, the differences between methods were small and it appeared less important to model heterogeneity.     

Parasite abundance contributes to condition‐dependent dispersal in a wild population of large herbivore

Parasite abundance has been shown to have major consequences for host fitness components such as survival and reproduction. However, although natal dispersal is a key life history trait, whether an individual's decision to disperse or not is influenced by the abundance of parasites it carries remains mostly unknown. Current and opposing hypotheses suggest that infected individuals should either be philopatric to avoid the energetic costs of dispersal (condition dependence) or disperse to escape from heavily parasitised habitats. From intensive monitoring of a roe deer population inhabiting a multi‐use and spatially heterogeneous agricultural landscape, we evaluated the link between an individual's parasite abundance and its propensity to disperse, while accounting for confounding effects of body mass. Dispersal propensity generally decreased with both increasing nematode abundance and with decreasing body mass. Within the dispersing segment of the population, individuals with high nematode abundance left their natal home range later in the season than less parasitised deer. These results clearly show that parasite abundance is an important component of condition‐dependent dispersal in large herbivores. However, unexpectedly, three individuals that were both heavily parasitised and of low body mass dispersed. We suggest that this ‘leave it’ response to high parasite levels in the natal habitat could represent a last ditch attempt to improve reproductive prospects, constituting a form of emergency life history strategy.     

Dynamics of an introduced population of mouflon Ovis aries on the sub‐Antarctic archipelago of Kerguelen

A commonly reported pattern in large herbivores is their propensity to irrupt and crash when colonizing new areas. However, the relative role of density‐dependence, climate, and cohort effects on demographic rates in accounting for the irruptive dynamics of large herbivores remains unclear. Using a 37‐yr time series of abundance in a mouflon Ovis aries population located on Haute Island, a sub‐Antarctic island of Kerguelen, 1) we investigated if irruptive dynamics occurred and 2) we quantified the relative effects of density and climate on mouflon population dynamics. Being released in a new environment, we expected mouflon to show rapid growth and marked over‐compensation. In support of this prediction, we found a two‐phase dynamics, the first phase being characterised by an irruptive pattern best described by the θ‐Caughley model. Parameter estimates were r_m=0.29±0.005(maximum growth rate), K=473±45 (carrying capacity) and S=2903±396 (surplus) mouflon. With a θ=3.18±0.69 our model also supported the hypothesis that density dependence is strongest at high density in large herbivores. The second phase was characterised by an unstable dynamics where growth rate was negatively affected by population abundance and winter precipitation. Climate, however, did not trigger population crashes and our model suggested that lagged density‐dependence and over‐grazing were the probable causes of mouflon irruptive dynamics. We compare our results with those of Soay sheep and discuss the possibility of a reversible alteration of the island carrying capacity after the initial over‐grazing period.     

Estimating density dependence in time-series of age-structured populations

For a life history with age at maturity alpha, and stochasticity and density dependence in adult recruitment and mortality, we derive a linearized autoregressive equation with time-lags of from 1 to alpha years. Contrary to current interpretations, the coefficients for different time-lags in the autoregressive dynamics do not simply measure delayed density dependence, but also depend on life-history parameters. We define a new measure of total density dependence in a life history, D, as the negative elasticity of population growth rate per generation with respect to change in population size, D = - partial differential lnlambda(T)/partial differential lnN, where lambda is the asymptotic multiplicative growth rate per year, T is the generation time and N is adult population size. We show that D can be estimated from the sum of the autoregression coefficients. We estimated D in populations of six avian species for which life-history data and unusually long time-series of complete population censuses were available. Estimates of D were in the order of 1 or higher, indicating strong, statistically significant density dependence in four of the six species.     

Sex- and age-dependent effects of population density on life history traits of red deer Cervus elaphus in a temperate forest

We studied both the short‐ and long‐term effects of density on three life history traits of a red deer population inhabiting a temperate forest. Both male and female body mass increased when population density decreased, but male mass changed to a greater extent than female mass. Density did not influence female survival irrespective of age, however, survival of males was lower at high density for all age classes except the prime‐age class. Pregnancy rates of primiparous females increased markedly with decreasing density, whereas those of adult hinds were fairly constant and unrelated to density. For both sexes, of the studied life history traits we detected a long‐term effect of density at birth (cohort effect) only on body mass. These results suggest that density influences life history traits in the same way as factors of environmental variation such as climate. In this population we did not find any evidence for an influence of climatic conditions on life history traits of red deer. Both mild winters and the absence of summer droughts during the study period could account for such an absence of climatic effects. We interpreted our results to show that 1) as expected for a highly dimorphic and polygynous species such as red deer, male traits showed consistently higher sensitivity to variation in density than female traits, illustrating possible costs caused by sexual selection in males, 2) the female‐based Eberhardt's model according to which increasing density should sequentially affect juvenile survival, reproductive rates of primiparous females, reproductive rates of adults and lastly adult survival was only partly supported because we found that pregnancy rate of primiparous females rather than juvenile survival was the most sensitive trait to variation in density. We propose that including variation in male traits would improve the accuracy of models of population dynamics of large mammals, at least for highly dimorphic species. Because the population we studied was not fenced, we only measured apparent survival. We discuss how dispersal, in relation to the phenotypic quality of young deer, might be a potential regulating factor under such conditions.     

Modelling geographic patterns of population density of the white‐tailed deer in central Mexico by implementing ecological niche theory

Conservation and management of species require basic knowledge on their geographic distribution and abundance. Here, we propose a novel approach, based on the theory of the ecological niche, to model the spatial patterns of the white‐tailed deer Odocoileus virginianus population density in two regions of central Mexico (Balsas Basin and Tehuacán‐Cuicatlán Valley). We used an ecological niche model to generate binary geographic distribution maps of the white‐tailed deer in each region based on occurrence data and a set of environmental variables. Then, the centroid of the distributions was calculated in ecological space (niche centroid) and the multidimensional Euclidian ecological distance of each pixel to the niche centroid was estimated. Finally, for each region the distance to the niche centroid (DNC) was regressed against 14 independent occurrence points in each site containing white‐tailed deer density information to determine the function describing the DNC‐density relationship, which was used to generate maps describing the distribution of white‐tailed deer density. Our results indicated an inverse DNC‐density relationship in both regions (Balsas Basin: r² = 0.90 and Tehuacán‐Cuicatlán: r² = 0.76) that was validated via bootstrapping resulting in a predicting capacity of near 62% for Balsas Basin and 65% for Tehuacán‐Cuicatlán Valley. Our results suggest that the distance to the niche centroid method is a robust, science‐based correlative approach that resulted useful to predict the population density of the white‐tailed deer in a spatially explicit fashion. The proposed approach is suitable for predicting the distribution of density for white‐tailed deer for which occurrence data with accompanying density information exists, but relative abundance can also be estimated when no abundance data are available.     

The effects of asymmetric competition on the life history of Trinidadian guppies

The effects of asymmetric interactions on population dynamics has been widely investigated, but there has been little work aimed at understanding how life history parameters like generation time, life expectancy and the variance in lifetime reproductive success are impacted by different types of competition. We develop a new framework for incorporating trait‐mediated density‐dependence into size‐structured models and use Trinidadian guppies to show how different types of competitive interactions impact life history parameters. Our results show the degree of symmetry in competitive interactions can have dramatic effects on the speed of the life history. For some vital rates, shifting the competitive superiority from small to large individuals resulted in a doubling of the generation time. Such large influences of competitive symmetry on the timescale of demographic processes, and hence evolution, highlights the interwoven nature of ecological and evolutionary processes and the importance of density‐dependence in understanding eco‐evolutionary dynamics.     

Are patterns of density dependence in the Global Population Dynamics Database driven by uncertainty about population abundance?

Density dependence in population growth rates is of immense importance to ecological theory and application, but is difficult to estimate. The Global Population Dynamics Database (GPDD), one of the largest collections of population time series available, has been extensively used to study cross-taxa patterns in density dependence. A major difficulty with assessing density dependence from time series is that uncertainty in population abundance estimates can cause strong bias in both tests and estimates of strength. We analyse 627 data sets in the GPDD using Gompertz population models and account for uncertainty via the Kalman filter. Results suggest that at least 45% of the time series display density dependence, but that it is weak and difficult to detect for a large fraction. When uncertainty is ignored, magnitude of and evidence for density dependence is strong, illustrating that uncertainty in abundance estimates qualitatively changes conclusions about density dependence drawn from the GPDD.     

Applicability of a Bayesian state-space model for evaluating the effects of localized culling on subsequent density changes: sika deer as a case study

At the landscape scale, localised culling is often conducted to achieve various deer management aims. However, few studies have assessed the effects of localised culling on deer population dynamics, owing to the spatially and temporally insufficient datasets of deer abundance that are derived from limited survey efforts. In this study, we estimated the population dynamics of a sika deer (Cervus nippon) population in the Tanzawa Mountains, central Japan, by Bayesian state-space modelling using spatiotemporally insufficient abundance indices and evaluated the effects of unit-specific culls on subsequent density changes in 56 units. The responses of deer density to unit-specific culls differed greatly among units, and a very weak correlation was observed between the intensities of unit-specific culls and the reduction in density. Deer populations in some units tended to resist density decreases despite high culling pressure, whereas those in other units were susceptible to density decreases with little to no culling pressure. Because the spatial scales of each unit were relatively small, annual density changes in each unit were largely influenced by deer movement in this estimation. The obscured effects of unit-specific culls, which were probably derived from deer movement among units in this case study, re-emphasized that deer migration should be incorporated into the planning of localised culling and that deer management should be coordinated over a wide area beyond landscape components and landownerships. Thus, we conclude that Bayesian state-space modelling is valuable for practical deer management programs at a large spatial scale even if different abundance indices are used.     

Using spotlight observations to predict resource selection and abundance for white-tailed deer

Spotlight surveys for white-tailed deer (Odocoileus virginianus) can yield large presence-only datasets applicable to a variety of resource selection modeling procedures. By understanding how populations distribute according to a given resource for a reference area, density and abundance can be predicted across new areas assuming the relationship between habitat quality (measured by an index of selection) and species distribution are equivalent. Habitat-based density estimators have been applied to wildlife species and are useful for addressing conservation and management concerns. Although achieving reliable population estimates is a primary goal for spotlighting studies, presence-only models have yet to be applied to spotlight data for estimating habitat selection and abundance for deer. From 2012 to 2017, we conducted spring spotlight surveys in each of 99 counties in Iowa, USA, and collected spatial locations for 20,149 groups of deer (n = 71,323 individuals). We used a resource selection function (RSF) based on deer locations to predict the relative probability of use for deer at the population level and to estimate statewide abundance. The number of deer observed statewide increased significantly with increasing RSF value for all years and the mean RSF value along survey transects explained 59% of the variability in county-level deer counts, indicating that a functional response between habitat quality and deer distribution existed at landscape scales. We applied our RSF to a habitat-based density estimator (extrapolation) and zero-inflated Poisson (ZIP) and negative binomial (ZINB) count models to predict statewide abundance from spotlight counts. Population estimates for 2012 were variable, indicating that atypical weather conditions may affect spotlight counts and population estimates in some years. For 2013–2017, we predicted a mean population of 439,129 (95% CI ∼ ± 55,926), 440,360 (∼ ± 43,676), and 465,959 (∼ ± 51,242) deer across years for extrapolation, ZIP, and ZINB models, respectively. Estimates from all models were not significantly different than estimates from an existing deer population accounting model in Iowa for 2013 and 2016, and differed by <76,000 deer for all models from 2013–2017. Extrapolation and ZIP models performed similarly and differed by <2,897 deer across all years, whereas ZINB models showed inconsistencies in model convergence and precision of estimates. Our results indicate that presence-only models are capable of producing reliable and precise estimates of resource selection and abundance for deer at broad landscape scales in Iowa and provide a tool for estimating deer abundance in a spatially explicit manner. © 2019 The Wildlife Society.