Book

Bayesian Population Analysis Using WinBUGS

Authors:

Abstract

Bayesian statistics has exploded into biology and its sub-disciplines, such as ecology, over the past decade. The free software program WinBUGS and its open-source sister OpenBugs is currently the only flexible and general-purpose program available with which the average ecologist can conduct standard and non-standard Bayesian statistics. Bayesian Population Analysis Using WinBUGS goes right to the heart of the matter by providing ecologists with a comprehensive, yet concise, guide to applying WinBUGS to the types of models that they use most often: linear (LM), generalized linear (GLM), linear mixed (LMM) and generalized linear mixed models (GLMM). Comprehensive and richly-commented examples illustrate a wide range of models that are most relevant to the research of a modern population ecologist. All WinBUGS/OpenBUGS analyses are completely integrated in software R. Includes complete documentation of all R and WinBUGS code required to conduct analyses and shows all the necessary steps from having the data in a text file out of Excel to interpreting and processing the output from WinBUGS in R.
... Hierarchical modelling. We used an integrated hierarchical model 73 , which is a complex stochastic system partitioned into a dependent sequential set of simpler sub-models dynamically affecting the performance of the main system of interest 73 (Supplementary Methods, Supplementary Table S1). ...
... (4) is specified with Gaussian errors and a priori assumed environmental disturbances (wind) as: y (t) |n (t) ∼ normal n (t) + γ x (t) , τ 2 , where γ is a parameter for east-west aspect winds x (t) , and τ is the standard deviation of a random observation-error process (Supplementary Methods). The joint-likelihood 73 of the integrated hierarchical model combining Eqs. (1), (2) and (4) is summarized with Supplementary Table S1. ...
... This integrated hierarchical model for the long-tailed duck population investigates a complex stochastic system affected dynamically by sub-systems that are also stochastic and regulated by independent variables 73,79 . This model allowed us first, to correct for population estimate errors due to variation in wind velocity and direction at Söderskär, Gulf of Finland, during spring migration, which accounted for 51% of the total observation error variance. ...
Article
Full-text available
Migratory animals experience very different environmental conditions at different times of the year, i.e ., at the breeding grounds, during migration, and in winter. The long-tailed duck Clangula hyemalis breeds in the Arctic regions of the northern hemisphere and migrates to temperate climate zones, where it winters in marine environments. The breeding success of the long-tailed duck is affected by the abundances of predators and their main prey species, lemmings Lemmus sibiricus and Dicrostonyx torquatus , whose population fluctuation is subject to climate change. In the winter quarters, long-tailed ducks mainly eat the blue mussel Mytilus edulis . We examined how North-west Siberian lemming dynamics, assumed as a proxy for predation pressure, affect long-tailed duck breeding success and how nutrient availability in the Baltic Sea influences long-tailed duck population size via mussel biomass and quality. Evidence suggests that the long-tailed duck population dynamics was predator-driven on the breeding grounds and resource-driven on the wintering grounds. Nutrients from fertilizer runoff from farmland stimulate mussel stocks and quality, supporting high long-tailed duck population sizes. The applied hierarchical analysis combining several trophic levels can be used for evaluating large-scale environmental factors that affect the population dynamics and abundance of migrants from one environment to another.
... Using the mark-recapture encounter histories, we estimated annual survival probability between 1979 and 2022 within the integrated population model using an age-structured Cormack-Jolly-Seber model for two age classes. We chose a multinomial likelihood formulation because an individual formulation was computationally prohibitive (Kéry & Schaub, 2012;Lebreton et al., 1992;Seward et al., 2019). The probability of being observed and recorded on Gough Island was assumed to be different for birds returning to Gough Island for the first time, because immature loafers generally visit the breeding areas later in the year and are therefore less likely to be detected than established birds that spend more time on breeding grounds. ...
... To relate our estimates of breeding success and survival to the annual census data of breeding pairs on Gough Island, we used a hierarchical state-space model (Clark & Bjørnstad, 2004;de Valpine, 2003;Kéry & Schaub, 2012) for the annual census data. ...
... Correlations between demographic parameters and the annual population growth rate were calculated using a Pearson correlation coefficient and are presented in the Supplementary Information. We used the integrated population model to project the population size 30 years into the future while accounting for the uncertainty in demographic parameters (Kéry & Schaub, 2012;Oppel et al., 2014;Zipkin & Saunders, 2018). We included survival probabilities for the two age classes, mean breeding success, and incorporated the three scenarios of plausible changes in breeding success or survival. ...
Article
Invasive species are one of the greatest drivers of biodiversity loss worldwide, and the eradication of invasive species from islands is a highly efficient management strategy. Because eradication operations require large financial investments, uncertainty over the magnitude of impacts of both invasive species and their removal can impede the willingness of decision makers to invest in eradication. Such uncertainty is prevalent for long‐lived species that display an inherent lag between life stages affected by invasive species and those used for population status assessments. Albatrosses are amongst the longest‐living bird species and are threatened on land by invasive species and at sea by industrial fisheries. As in many seabird species, usually only a segment of the population (breeding adults) is used for status assessments, making it difficult to assess albatross population trends and the potential benefit of conservation action, such as the management of predatory invasive species. We used population monitoring and mark‐recapture data to estimate the past population trajectory of the critically endangered Tristan albatross Diomedea dabbenena by accounting for unobservable birds at sea in an integrated population model. We then projected the future population trajectory of Tristan albatrosses for scenarios with or without predation by invasive house mice Mus musculus on their main breeding site, Gough Island. The adult breeding population remained stable between 2004 and 2021, but breeding success was low (31%) and our model indicated that the total population (including unobservable immature birds) decreased from a median estimate of 9,795 to 7,752 birds. Eradicating invasive mice leading to a two‐fold increase in breeding success would result in a 1.8–7.6 times higher albatross population by 2050 (median estimate 10,352 individuals) than without this intervention. Low reproductive output for long‐lived species may lead to a cryptic population decrease, which can be obscured from readily available counts of breeding pairs by changes in the population structure. Mouse eradication is necessary to halt the ongoing population decrease of the Tristan albatross, even if this decrease is not yet apparent in the breeding population size. Low reproductive output for long‐lived species may lead to a cryptic population decrease, which can be obscured from readily available counts of breeding pairs by changes in the population structure. Mouse eradication is necessary to halt the ongoing population decrease of the Tristan albatross, even if this decrease is not yet apparent in the breeding population size.
... The data encompassed more severe cold stuns (>1000 turtles) from 7 to 14 January 2010 and from 14 to 21 January 2018 (hereafter occasion 1 and occasion 4) and less severe cold stuns (100-200 turtles) from 13 to 15 January 2011 and from 7 to 9 January 2014 (hereafter occasion 2 and occasion 3; Table 1). We only considered recaptures during occasion 4 for modeling purposes because newly encountered individuals do not provide information about survival (Kéry & Schaub, 2012;Lebreton et al., 1992). ...
... The CJS model is fundamentally a hierarchical model that distinguishes between a binary latent state ("alive" or "dead") and a binary observed state (captured or not captured; Kéry & Schaub, 2012;Lebreton et al., 1992). Observed capture history data y it are recorded as 1 if individual i is captured at occasion t and zero if not captured. ...
... Estimates are computationally similar using either a fixed or a random approach with few levels of a factor (i.e., shrinkage is minimal). However, treating occasion as random in a CJS model eliminates the identifiability problem when estimating the last period for ϕ and p (Brooks et al., 2000;Kéry & Schaub, 2012). Also, with >2 levels of the occasion factor, some degree of robustness was achieved by treating it as random versus fixed (Gelman & Hill, 2007;Wagner et al., 2006). ...
Article
Full-text available
Understanding the effects of extreme weather on animal populations is fundamental to ecological and conservation sciences and species management. Climate change has resulted in both warm and cold temperature extremes, including an increased frequency of severe cold snaps at middle latitudes in North America. These unusually cold air masses cause rapid declines in near-shore ocean temperatures in coastal areas, with detrimental effects on marine organisms. Acute cold-stun events (hereafter cold stuns) occur when hundreds to thousands of resident juvenile sea turtles fail to escape shallow water during cold snaps. Human intervention through rescue and recovery largely mitigates direct juvenile sea turtle mortality, but delayed effects of cold stuns on rescued individuals are not well understood. Our objective was to examine long-term juvenile green turtle (Chelonia mydas) survival across four cold stuns of varying severity in St. Joseph Bay, Florida, between 2010 and 2018. We used the classic Cormack-Jolly-Seber model in a hierarchical Bayesian framework to estimate apparent survival (i.e., emigration and mortality) of rescued turtles at different time intervals. Our results indicated about half of a cohort rescued during a severe cold stun in January 2010 likely remained in the population 1 year later, with 10%-20% remaining 4 years later, and as few as 5% by 2018. The results also suggested higher apparent survival for cohorts rescued during two subsequent milder cold stuns. Emigration was a more plausible ecological explanation for low apparent survival than delayed mortality. Potential ecological mechanisms underlying emigration include a reduction in food availability and a behavioral response to either the severe weather event or handling during rescue (or both). However, the typical annual turnover of juvenile green turtles, though assumed low, is not well known in St. Joseph Bay. Thus, our apparent survival estimates may be reflective of higher-than-expected emigration in the broader population. Our study provides important baseline information about long-term juvenile sea turtle survival after cold stuns in temperate regions. We also highlight the importance of strategic monitoring between cold stuns to examine additional ecological questions.
... These models combine both live recapture and dead recovery data into a single model, which provides a more precise estimate for survival (Gilroy et al., 2012). Mark-recapturerecovery models also provide an estimate for fidelity by accounting for permanent emigrations from the study area, where fidelity is calculated as 1the probability of permanently emigrating the study area (Burnham, 1993;Keŕy and Schaub, 2011). Emigration was included in the model to account for individuals within the population who disappear from the sighting history, though information is lacking about if the individual has died, or if they have moved to a poorly or unstudied area and are missed during survey efforts (Kraus et al., 1986;Right Whale Consortium, 2020). ...
... We used a multi-event Jolly-Seber model (George, 1992;Schwarz and Arnason, 1996;Brooks et al., 2002;King, 2012) for the joint analysis of mark-recapture-recovery data (Lebreton et al., 1999;Keŕy and Schaub, 2011). In this model, we considered ten true biological states: not yet entered (NE), calf alive within study site (CAI), pre-breeder alive within study site (PAI), breeder alive within study site (BAI), pre-breeder alive outside study site (PAO), breeder alive outside study site (BAO), calf recently dead (CRD), pre-breeder recently dead (PRD), breeder recently dead (BRD), and dead ( †). ...
... where Z i,t specifies the state of individual i at time t, and W z i,t ,1… s,i,t describes state membership over time, where s is the number of true states (Keŕy and Schaub, 2011). ...
Article
Full-text available
Abundance and population trends of Critically Endangered North Atlantic right whales ( Eubalaena glacialis , NARW) have been estimated using mark-recapture analyses where an individual’s state is based upon set delineations of age, using historical estimates of age at first reproduction. Here we assigned individual females to states based upon their reproductive experience, rather than age. We developed a Bayesian mark-recapture-recovery model to investigate how survival, recapture, site-fidelity and dead-recovery probabilities vary for female NARW in different states, using data collected from 1977-2018. States were assigned as calves for individuals in their first year; pre-breeder for individuals greater than one year of age who had yet to produce a calf, or breeder if an individual had reproduced. A decline in abundance of female NARW was seen starting in 2014, with 185 females declining yearly to 142 by 2018. The largest decline was seen in breeding females, with 72 estimated to be alive at the beginning of 2018, while female pre-breeder abundance plateaued at around 70 between 2011- 2018. Females born from 2000 onwards had an average 4% (95% CI:0.03-0.06) chance of transitioning from pre-breeder to breeder, compared to 8% (95%CI:0.06-0.1) for females born prior. This reduction in transition rate from non-breeder to breeder for the current cohort resulted in breeding females declining to 51% of the female population by 2018. We show that a collapse in fecundity of breeding females, and the failure of pre-breeders to start breeding, is an important factor in understanding the current decline in abundance of the NARW.
... To describe the timing of downstream movements of Lake Sturgeon for objective 3, we used a state-space multistate model (Kéry and Schaub 2012) to evaluate monthly apparent survival and transition (movement) and detection probabilities in USF, LSF, and the segment of river downstream of Menominee Dam. For this analysis, we used all Lake Sturgeon that were detected ≥3 d after release into USF, regardless of sex or stage (N = 133). ...
... We fitted the multistate model using a Bayesian framework for analyses in the open-source software JAGS (Plummer 2003(Plummer , 2017 and R (R Development Core Team 2018) via the package rjags (Plummer 2019). We constrained priors for p and phi by using a uniform distribution (0, 1), while priors for psi were dependent on specific transitions, as different equations were used (Kéry and Schaub 2012). We used a normal distribution (0, 0.001) for the lpsi parameters in the exponential equations for USF (remain in USF, transition to LSF) and a uniform distribution (0, 1) for remaining in LSF (Kéry and Schaub 2012). ...
... We constrained priors for p and phi by using a uniform distribution (0, 1), while priors for psi were dependent on specific transitions, as different equations were used (Kéry and Schaub 2012). We used a normal distribution (0, 0.001) for the lpsi parameters in the exponential equations for USF (remain in USF, transition to LSF) and a uniform distribution (0, 1) for remaining in LSF (Kéry and Schaub 2012). Transitions from a location must sum to 1, so transitions from USF or LSF to downstream of Menominee Dam did not need priors since they were dependent on the other transition estimates (e.g., psi for transitioning from LSF to downstream of Menominee Dam = 1 − psi for remaining in LSF). ...
Article
Fish behavior after passage or transfer around dams is a critical component in determining whether the goals of these efforts are achieved, but these behaviors are often poorly understood. An elevator was constructed in the lowermost hydroelectric dam on the Menominee River, Wisconsin–Michigan; it is the first elevator specifically designed to capture Lake Sturgeon Acipenser fulvescens for upstream transfer above two dams, providing access to high‐quality spawning and early life habitat. Our objectives were to determine whether (1) Lake Sturgeon transferred upstream remained upstream for at least one spawning opportunity; (2) spawning opportunity, time to reach the next dam upstream, and residency in different segments of the river were related to sex, capture method (elevator versus electrofishing), and season of transfer; and (3) the probability of fish transitioning back downstream of the two dams varied among months. We evaluated posttransfer behaviors of 139 Lake Sturgeon that were captured in the elevator or by electrofishing, implanted with acoustic transmitters, transferred upstream (in spring or fall) from fall 2014 to spring 2017, and monitored until fall 2018 using 20–23 stationary acoustic receivers deployed throughout the river. Most Lake Sturgeon (91%) remained upstream for at least one spawning opportunity. The probability of remaining for one spawning opportunity was not related to sex, fish capture method, or season of transfer. Residency times within the two impoundments and time to reach the next dam upstream varied among individual fish. A multistate model indicated that monthly survival after upstream transfer was high and that Lake Sturgeon typically remained above both dams in late fall to early spring, with most downstream movements occurring in April and May. Our results indicate that Lake Sturgeon transferred upstream have the potential to contribute offspring that may help to bolster the Lake Sturgeon population in Lake Michigan, but additional research may help in determining whether these contributions occur. Impact Statement Lake Sturgeon that were transferred upstream of two dams remained upstream of the dams for at least one spawning opportunity, representing the first step in determining whether transfers can be used to increase Lake Sturgeon abundance in the Great Lakes and beyond.
... Entry could result from on-site recruitment (individuals born locally) or immigration (individuals born outside the study area entering at some point). To ensure that all individuals (N) enter the population at some point during the study, the sum of all entry probabilities must be equal to 1 (Kéry & Schaub, 2012). ...
... The number of individuals entering the population at each sampling occasion is modeled using a multinomial distribution (Kéry & Schaub, 2012), which is a generalization of the binomial distribution used in simulated populations for residency studies based on C-R data by Morteo et al. (2012). The algorithm provides as output, the capture histories of the N individuals, along with the number of individuals that entered the population for the first time and the actual size of the population on each occasion, according to the input parameters provided by the user (details in the next section). ...
... In each scenario, the simulations included the fixed parameters (size of the overpopulation, number of occasions and probability of entry) and different combinations of probabilities of recapture and survival, using an adapted version (available on GitHub) of the "simul.js" algorithm(Kéry & Schaub, 2012) in the Rproject programming language. Therefore, recapture probability (p) between sampling occasions was set at values of 0.01 and subsequently in increments of 0.1 up to 1.0, to represent naturally variable scenarios, considering individuals present both once in the study period (i.e., 0.01) and present on different occasions up to animals sighted in all samplings (i.e., p=1). ...
Thesis
Full-text available
Residency and site fidelity are important parameters in the population ecology of many species, as they indicate temporal and spatial use of individuals. Cetaceans with welldefined patterns of residency and site fidelity may be more vulnerable to anthropogenic disturbances and the study of these parameters makes it possible to assess such impacts. However, due to interspecific differences and logistical restrictions for the study of cetaceans, there is no consensus on the definitions and ways to evaluate these parameters. This project investigated the factors that influence measurements of residency and site fidelity in dolphins through a retrospective, ~20 year (1990-2019), literature review. Simulations of capture-recapture histories based on real scenarios of wild populations of dolphins were used to evaluate residency metrics by varying the parameters in a JollySeber model. We found 30 indicators of residency, where 20 were based on capturerecapture data, and 8 contained categories. These main differences were associated with patterns of temporal use and the thresholds for classifying individuals. Simulations showed sensitivity to different survival estimates, in addition to those of already known catchability. Indicators of site fidelity were more consistent with each other, and the best performing indicators of residency combined monthly, seasonal, and annual data. These findings show the need for paradigm shifts to create comprehensive, standardized, and generalized definitions of residency, which can be compared across different populations of marine mammals around the world.
... When attempting to monitor or manage wild animal populations, particularly elusive or difficult to detect animals, accurate estimations of their abundance and population dynamics are essential (Fuller et al. 2016). Often, this is estimated through capture-mark-recapture (CMR) modelling whereby data obtained from live trapping or other methods of recording the presence of known animals is used to estimate population densities (Kéry and Schaub 2012). A key principle of CMR models is the estimation of the detection probability, which provides the link between the observations of animal presence and true population parameter such as the survival and recapture probability. ...
... However, capture mark recapture studies are inherently spatial: traps are laid out in space, and therefore there is spatial structuring in the ecological processes that give rise to the capture data (Royle and Chandler 2014). This is likely to invalidate the assumption of traditional CMR modelling that detection probabilities are assumed to be equal over all individuals (Kéry and Schaub 2012). In practice, the detection probability is heterogeneous due to the spatial organization of individuals relative to traps: the detection probability depends on the number of traps present inside the home range of an individual (Royle and Chandler 2014) and will vary in space as a result of both individual space use and the configuration of detection devices (Dupont et al. 2019a). ...
... However, it should be recognised that the increased flexibility comes at the expense of being highly computationally demanding (Glennie and Foster 2019), and, due to the inclusion of prior knowledge, can be criticised as being inherently biased. The Bayesian framework allows the user to incorporate the distribution of the data and the distribution of prior knowledge in a natural way (Mariucci, Ray, and Szab 2017), meaning that the posterior distribution is always the result of the chosen prior distribution and the information contained in the data (Kéry and Schaub 2012). The ability to incorporate prior knowledge can result in improved posterior distributions, but if prior knowledge is not truthful, posterior distributions can be deceitful (Kéry and Royle 2016). ...
Preprint
Full-text available
Capture mark recapture (CMR) models allow the estimation of various components of animal populations, such as survival and recapture probabilities. In recent years, incorporating the spatial distribution of the devices used to detect an animals presence has become possible. By incorporating spatial information, we explicitly acknowledge the fact that there will be spatial structuring in the ecological processes which give rise to the capture data. Individual detection probability is not heterogeneous for a range of different reasons, for example the location of traps within an individuals home range, the environmental context around the trap or the individual characteristics of an animal such as its age. Spatial capture recapture models incorporate this heterogeneity by including the spatial coordinates of traps, data which is often already collected in standard CMR approaches. Here, we compared how the inclusion of spatial data changed estimations of survival, detection probability, and to some extent the probability of seroconversion to a common arenavirus, using the multimammate mouse as our model system. We used a Bayesian framework to develop non spatial, partially spatial and fully spatial models alongside multievent CMR models. First, we used simulations to test whether certain parameters were sensitive to starting parameters, and whether models were able to return the expected values. Then we applied the non-spatial, partially spatial and fully spatial models to a real dataset. We found that bias and precision were similar for the three different model types, with simulations always returning estimates within the 95% credible intervals. When applying our models to the real data set, we found that the non-spatial model predicted a lower survival of individuals exposed to Morogoro virus (MORV) compared to unexposed individuals, yet in the spatial model survival between exposed and non-exposed individuals was the same. This suggests that the non-spatial model underestimated the survival of seropositive individuals, most likely due to an age effect. We suggest that spatial coordinates of traps should always be recorded when carrying out CMR and spatially explicit methods of analysis should be used whenever possible, particularly as incorporating spatial variation may more easily capture ecological processes without the need for additional data collection that can be challenging to acquire with wild animals.
... We estimated the total number of breeder males by summing the following: the number of first-year males that survived and recruited to breeders (N 0 t ϕ 0 breeders, 1 − ψ BA t ) during the following year. Next, we multiplied the number of breeding males by fecundity halved ( Ft+1 2 ) because we specified a male-based IPM and assumed an equal sex ratio of births (Kéry and Schaub, 2012). We assigned a Poisson distribution to constrain the number of wild-reared first-years to discrete integers greater than or equal to zero. ...
... This CJS submodel estimated true survival rather than apparent survival, which can be confounded with emigration (Kéry and Schaub, 2012). We included data from both males and females in the CJS submodel to obtain survival estimates for both sexes using a single analysis. ...
... We used Pearson's correlation coefficients to estimate correlations between population growth rates and some vital rates (i.e. survival and fecundity) from the IPM (Kéry and Schaub, 2012). We assigned importance to correlation coefficients when 90 % HDIs did not intersect zero, and the probability of the correlation coefficient being greater than zero was ≥0.90. ...
Article
We examined long-term demography of an endangered subspecies, the Northern Aplomado Falcon (Falco femoralis septentrionalis), in South Texas, USA. The population has been managed and monitored since reintroductions began in 1993. Data spanning 1993–2018 enabled us to build an integrated population model (IPM) and a Cormack-Jolly-Seber model to estimate survival for three life stages (first-year, non-breeders, and breeders) and both sexes, abundance of males, fecundity, immigration, and emigration. Male falcons survived at lower rates than females during their first year; Hurricane Harvey caused a decline in survival rates of first-years and breeders; and fecundity increased after 2011 coinciding with changes in management focused on improving nest platforms and habitat quality. Both immigration of non-breeders and emigration were likely negligible for this population suggesting a potentially isolated population. The IPM likely overestimated immigration of breeders warranting further research. Population growth rates were greatest during years having more released captive-reared young and greater probabilities of breeder survival. Importantly, an apparent decrease in breeder survival of unknown cause occurred during 2006 to 2009 when breeder survival declined and remained low for several years. Our ability to identify the cause for reduced survival is now greatly hampered by the extended time that has passed, limiting the usefulness of our recent awareness of reduced survival for informing management and further highlighting the importance of real-time monitoring for proactive decision making processes. Our study greatly improves knowledge of demographics for a reintroduced, isolated, and intensively managed population of Aplomado Falcons. Applying this IPM to new data each year will enable adaptive management of the South Texas population by providing annual evaluations of vital rates along with revised assessments of monitoring and management.
... (Plummer, 2003) and R 3.1.0 (R Team, 2014), following the basic approach of Kéry and Schaub (2012). We used uninformative priors for all parameters; however, we did limit p to values ≥−3 on the logit scale to avoid boundary issues. ...
... We ran each model for 80,000 iterations with 20,000 discarded as burn-in, retaining the remaining 60,000 for inference. We assessed model fit using the Bayesian p-value described by Kéry and Schaub (2012), and based inference of fixed effects on the precision (95% Bayesian credible intervals) of regression coefficients. If credible intervals of estimated parameters included zero, effects associated with these parameters were considered to be uninformative. ...
Article
Ongoing environmental change across the Arctic is affecting many freshwater ecosystems, including small thaw ponds that support macroinvertebrates, thus potentially affecting important forage for fish and bird species. To accurately predict how fish and wildlife that depend on these macroinvertebrates will be affected by ecosystem change at high latitudes, understanding proximate factors that influence macroinvertebrate abundance is critical. To better understand factors that affect spatial and seasonal (i.e. phenology) patterns in abundance, we collected macroinvertebrates throughout the growing season of a single year from 33 thaw ponds on the Arctic Coastal Plain in northern Alaska. We used hierarchical N‐mixture models to provide detection‐corrected estimates of abundance (of the population exposed to sampling) in relation to pond type and seasonal patterns in environmental variables (i.e., cumulative water temperature, nutrient levels) for five taxonomic groups representing key food items for birds and fish—Anostraca (Arthropoda: Branchiopoda), Chironomidae (Insecta: Diptera), Cladocera (Arthropoda: Branchiopoda), Limnephilidae (Insecta: Trichoptera), and Physidae (Mollusca: Gastropoda). For three of five taxa (Anostraca, Cladocera, Limnephilidae), abundance varied across pond types and was lower in pond types where water temperatures increased more rapidly. Further, seasonal temperature profiles in ponds affected phenology, suggesting that seasonal patterns in abundance were influenced by changes in water temperature. These findings suggest that increases in water temperature in northern areas could alter macroinvertebrate phenology, possibly with consequences for upper level predators if availability of macroinvertebrate prey is reduced or shifted seasonally. Our results will facilitate improved predictions of how changing abiotic conditions could affect inland waters in northern areas, a critical need for conservation of Arctic wildlife and ecosystems.
... Teal hunting primarily occurs in September and October (Devink et al., 2013), while natural mortality of adult female ducks primarily occurs during the breeding season (Arnold et al., 2012;Dufour & Clark, 2002;Hoekman et al., 2002;. Thus, we modelled natural mortality probability as conditional on having survived previous hunting mortality, To model the band-recovery data, we built modified bandrecovery models following Brownie and Pollock (1985), Kéry and Schaub (2012), Ergon et al. (2018), Nater et al. (2020) and , where recoveries of previously marked adult females occurred as a function of annual survival (s t ) and bandrecovery (f t ) probabilities. We estimated annual band-recovery probability as a function of hunting mortality probability ( ), bandreporting probability ( ), and the probability that birds killed by hunting are not retrieved, that is, crippling loss probability (c). ...
... We formatted the capture-recovery data (M) in a multinomial array (Brownie & Pollock, 1985), where the rows (i) represent the year of release, the columns ( j) represented the year of recovery, and T is the number of years where individuals were released. Thus, the cell values in the data represent the number of individuals released in year i and shot, retrieved and reported in year j, and the last column in the m-array is the number of individuals that were never reencountered (Brownie & Pollock, 1985;Kéry & Schaub, 2012). ...
Article
Full-text available
Harvest of wild organisms is an important component of human culture, economy, and recreation, but can also put species at risk of extinction. Decisions that guide successful management actions therefore rely on the ability of researchers to link changes in demographic processes to the anthropogenic actions or environmental changes that underlie variation in demographic parameters. Ecologists often use population models or maximum sustained yield curves to estimate the impacts of harvest on wildlife and fish populations. Applications of these models usually focus exclusively on the impact of harvest and often fail to consider adequately other potential, often collinear, mechanistic drivers of the observed relationships between harvest and demographic rates. In this study, we used an integrated population model and long‐term data (1973‐2016) to examine the relationships among hunting and natural mortality, the number of hunters, habitat conditions, and population size of blue‐winged teal (Spatula discors), an abundant North American dabbling duck with a relatively fast‐paced life history strategy. Over the last two and a half decades of the study, teal abundance tripled, hunting mortality probability increased slightly (<), and natural mortality probability increased substantially (>) at greater population densities. We demonstrate strong density‐dependent effects on natural mortality and fecundity as population density increased, indicative of compensatory harvest mortality and compensatory natality. Critically, an analysis that only assessed the relationship between survival and hunting mortality would spuriously indicate depensatory mortality due to multicollinearity between abundance, natural mortality, and hunting mortality. Our findings demonstrate that models that only consider the direct effect of hunting on survival or natural mortality can fail to accurately assess the mechanistic impact of hunting on population dynamics due to multicollinearity among demographic drivers. This multicollinearity limits inference and may have strong impacts on applied management actions globally.
... We considered four CMR model configurations that differ in how they treat sighting probability [76]. Model M 0 is the null model for which sighting probability is assumed constant over time and across individual dolphins. ...
... Although we included models with heterogeneous sighting probabilities (M h and M th ) for completeness, we note that with only three sampling occasions, such models can produce unreliable estimates [29]. For each of the four configurations, we used data augmentation to estimate abundance from the observed sighting histories [76,77]. Data augmentation attempts to account for individuals that were present in the system but never observed [54]. ...
Article
Full-text available
The social structure of estuarine-resident bottlenose dolphins is complex and varied. Residing in habitats often utilized for resource exploitation, dolphins are at risk due to anthropogenic pressures while still federally protected. Effective conservation is predicated upon accurate abundance estimates. In North Carolina, two estuarine-resident stocks (demographically independent groups) of common bottlenose dolphin have been designated using spatiotemporal criteria. Both stocks are subjected to bycatch in fishing gear. The southern North Carolina estuarine stock was estimated at
... We assessed convergence by visual assessment of MCMC chains and by using the Gelman-Rubin statistic ('Rhat'), with values <1.1 considered acceptable (Gelman & Hill, 2007). An assessment of model fit to the actual dataset was performed using a Bayesian p-value approach and an estimation of a 'lack-of-fit' statistics (Kéry & Schaub, 2012;Broms et al., 2016) (Suppl.S3.3). In contrast to the frequentist models for which the use of model selection criteria is widespread, there is currently no consensus regarding the appropriate methods for covariates selection or model comparison in the framework of complex hierarchical models Broms et al., 2016). ...
... Values larger than 0.95 or smaller than 0.05 indicate that the model did not fit well the data. We also calculated a "lackof -fit" statistics / which is expected to be equal to 1 if the model fits the data perfectly (Kéry & Schaub, 2012). ...
Thesis
Full-text available
Insects are still poorly considered in biodiversity conservation programs, despite their major role in the functioning of terrestrial ecosystems. Because of their great diversity, and a great lack of knowledge of community structuring patterns, the conservation of these organisms faces major challenges related to the evaluation and prioritization of issues as well as the estimation of the species richness of communities. Because they constitute two groups of insects particularly impacted by environmental changes, butterflies and dragonflies are an important part of protected insects in Europe. In this context of low consideration of entomological diversity in conservation biology, this thesis aims to (1) measure the recent dynamics (decline, stability or expansion) of these two taxa in France, and to identify the climatic and ecological factors likely to condition them, (2) for butterflies, to use a fine description of the interactions network between the larvae and their hostplants to question the conservation status according to a systemic approach (3) and finally, with a clearly operational objective, to evaluate the links between the information gains and the efforts / sampling costs mobilized during environmental impact assessments. The comparative study of the temporal dynamics of the populations of these two taxa was carried out at the scale of three countries of Western Europe through a diachronic analysis of the species distribution patterns on a time step of 35 years. This work has highlighted a strong correlation between the artificialisation of landscapes (urbanization, agricultural intensification, regression of wetlands) and the decline of many species, mainly characterized by strict ecological requirements and currently unprotected. On this same scale, the analysis of the architecture of the interactions network between butterfly larvae and their food plants revealed a modular structure in relation to the taxonomy of the partners, as well as a link between the degree specialization and species vulnerability (although the similarity of the species' diet did not seem to condition their conservation status). A final part was focused on the evaluation of the completeness of entomological surveys carried out during environmental impact assessments. A systematic and calibrated return procedure, on sites sampled in Mediterranean scrublands, allowed highlighting the limits of the current protocols in the estimation of species richness and the detection of protected or vulnerable species. This thesis contributes to the articulation between fundamental issues and operational needs, by allowing both a better understanding of the structuring mechanisms of insect communities and the formulation of recommendations for a better local application of conservation policies
... The occupancy model made use of repeated observations in the sites to disentangle the observational component (i.e., detection/nondetection) from the state variable of interest (i.e., "true" occupancy [38,55]). The model had four central assumptions: (i) independence between sampling sites, (ii) independence between repeated observation occasions, (iii) absence of misidentification of the focal species (no false positive error), and (iv) no colonizations or extinctions during the study period; that is, the sampling sites were expected to be "closed" to the occupancy state of species during the study period [55,60,61]. Assumption of a closed occupancy state may be relaxed if the changes in the occupancy state are random, and therefore the occupancy parameter should be interpreted as the probability that the species "uses" the habitat [62]. ...
... The parameters of both detection and occupancy were modeled as a logit function [55,59]. We normalized all continuous explanatory variables to have a mean of 0 and a standard deviation of 1 [61]. ...
Article
Full-text available
Invasive alien species (IAS) are a threat to biodiversity and ecosystem function worldwide. Unfortunately, researchers, agencies, and other management groups face the unresolved challenge of effectively detecting and monitoring IAS at large spatial and temporal scales. To improve the detection of soniferous IAS, we introduced a pipeline for large-scale passive acoustic monitoring (PAM). Our main goal was to illustrate how PAM can be used to rapidly provide baseline information on soniferous IAS. To that aim, we collected acoustic data across Puerto Rico from March to June 2021 and used single-species occupancy models to investigate species distribution of species in the archipelago and to assess the peak of vocal activity. Overall, we detected 16 IAS (10 birds, 3 mammals, and 3 frogs) and 79 native species in an extensive data set with 1,773,287 1-minute recordings. Avian activity peaked early in the morning (between 5 a.m. and 7 a.m.), while amphibians peaked between 1 a.m. and 5 a.m. Occupancy probability for IAS in Puerto Rico ranged from 0.002 to 0.67. In general, elevation and forest cover older than 54 years were negatively associated with IAS occupancy, corroborating our expectation that IAS occurrence is related to high levels of human disturbance and present higher occupancy probabilities in places characterized by more intense human activities. The work presented here demonstrates that PAM is a workable solution for monitoring vocally active IAS over a large area and provides a reproducible workflow that can be extended to allow for continued monitoring over longer timeframes.
... The detection histories were produced using the results of the two surveys for each tissue type, kidney (Kj=1; Kj=2) or liver (L j=1; L j=2), resulting in 16 possible detection histories (Table 2.1). For instance, the following detection history indicates that an individual fish tested negative on both surveys in each tissue (with two observations each for kidney and liver) Prior to running the multi-state occupancy models, I used a multinomial logistic regression to evaluate if fish length, weight, and/or sex affected the state or observation process in kidney and liver tissues (Kéry & Schaub 2011). Only individual effect models were included because length, weight, and sex were known to be correlated since female fish were larger than male fish. ...
... Probabilities of infection are calculated for each of the four states: Ψ ! is the probability of infection for state 1, Ψ " is the probability of infection for state 2, Ψ $ is the probability of infection for state 3, and Ψ ( is the probability of infection for state 4. I assigned a Dirichlet prior(Kéry & Schaub 2011) for the probabilities of infection. To constrain the four probabilities, the hyperparameters (1 &,# ) within the Dirichlet prior were sampled from a non-informative gamma distribution (gamma ...
Thesis
Full-text available
Renibacterium salmoninarum, the causative agent of bacterial kidney disease (BKD), is known to cause high mortality in both wild and cultured salmonids, causing concern for many of the salmonid populations. Bacterial kidney disease caused up to 80% mortality in cultured Pacific salmonids and 40% in Atlantic salmonids. Due to high mortality among salmonid species, the American Fisheries Society has defined R. salmoninarum as a regulated pathogen. Due to its regulated status, research efforts have focused on advancing fish health diagnostics and understanding the transmission of the bacteria. However, many of these studies focus on Pacific northwest salmonids and the understanding of R. salmoninarum dynamics is not well known among inland salmonids. Aquaculture propagation of Greenback Cutthroat Trout (Oncorhynchus clarkii) is a necessary component of their management. Since their protection under the Endangered Species Act, broodstock of Greenback Cutthroat Trout have been established at Colorado Parks and Wildlife (CPW) hatcheries to allow more rapid reintroduction through stocking. In 2017, hatcheries rearing isolated strains of the Greenback Cutthroat Trout contributed 1.5 million eggs during the spawning season. However, one major constraint to maintaining spawning production of the Greenback is the spread of disease within a facility. Increased contact rates between fish in raceways may influence the transmission of a pathogen. To ensure fish health and promote best practices in fish culture, fish health inspections have served as a critical step in identifying prohibitive and regulated pathogens entering or exiting the hatchery systems. Various diagnostic methods have been established to detect R. salmoninarum in salmonids. Culturing the bacteria is the most accurate and reliable assay for detection; however, it is a slow process and not suited for rapid assessment. Other methods used to detect R. salmoninarum include Direct Fluorescent Antibody Tests (DFAT), Polymerase Chain Reaction (PCR), and Enzyme-Linked Immunosorbent Assays (ELISA) and are typically performed using lethally collected kidney tissue. Currently, kidney tissues are used to screen for the presence of the bacteria using DFAT as the initial test and PCR as a confirmatory test, following the American Fisheries Fish (AFS) Health Blue Book protocol. The protocol was developed using highly susceptible Pacific northwest salmonids and it is unknown if the protocol is appropriate for testing inland salmonids which may be less susceptible. In addition, the current protocol requires sacrificing fish, which is undesirable in situations with valuable and sometimes irreplaceable broodstocks. Therefore, I examined the efficacy of the current AFS detection protocol and compared it to other potential approaches (Chapter 2). I also assessed several non- lethal approaches to detecting the bacteria (Chapter 1 and 2). In chapter 1, I compared non-lethal sampling methods with standardized lethal kidney tissue sampling that is used to detect R. salmoninarum infections in salmonids. I collected anal, buccal, and mucus swabs (non-lethal qPCR) and kidney tissue samples (lethal DFAT) from 72 adult Brook Trout (Salvelinus fontinalis) reared at the Colorado Parks and Wildlife Pitkin Brood Unit and tested each sample to assess R. salmoninarum infections. Brook Trout were used as a model species for Cutthroat Trout because they are described as highly susceptible species. Standard kidney tissue detected R. salmoninarum 1.59 times more often than mucus swabs, compared to 10.43 and 13.16 times more often than buccal or anal swabs, respectively, indicating mucus swabs were the most effective and may be a useful non-lethal method. My study highlights the potential of non-lethal mucus swabs to sample for R. salmoninarum and suggests future studies are needed to refine this technique for use in aquaculture facilities and wild populations of inland salmonids. In chapter 2, I assessed the probability of detecting the bacteria in several tissues using standard diagnostic tests. I collected three lethal tissue (kidney, liver, and spleen) and three nonlethal serum (blood, ovarian fluid, and mucus swabs) samples from 781 adult Greenback Cutthroat Trout at the Colorado Parks and Wildlife Poudre Rearing Unit. All tissues were tested for R. salmoninarum via DFAT and qPCR. The overall prevalence (all tissue types) of R. salmoninarum among the fish was 22.7% with DFAT and 81.8% with qPCR. Kidney and liver tissues resulted in the greatest number of detections using either assay. To calculate the probability of infection among kidney and liver tissues and probability of detection between assays, I developed a hierarchical occupancy model. The liver had the highest probability of infection among all fish (0.69) and the probability of detection within the liver was highest with qPCR (0.79). DFAT produced a high probability of false negative detections (0.58). Thus, I suggest that testing a combination of both kidney and liver tissues with qPCR may yield a higher detection rate that better predicts the probability of infection when performing fish health inspections. Management of R. salmoninarum is particularly difficult because the bacterium utilizes both vertical and horizontal transmission. Vertical transmission occurs when infected brood fish transmit the bacterium to their eggs and ultimate their progeny. Previous studies suggest the bacterium cannot be paternally transmitted due to limited success of bacterial entry into the egg from the spermatozoa. Thus, vertical transmission is suggested to be maternal. Horizontal transmission occurs among individuals through the ingestion of contaminated fecal matter or through direct contact with infected fish or water. In previous studies, horizontal transmission has been suggested to contribute more toward infection persistence than vertical transmission in wild and hatchery fish populations. However, the relative importance of horizontal transmission in hatcheries, where flow-through systems may expose multiple fish lots, has received little attention. I conducted experiments to determine rates of vertical and horizontal transmission. In chapter 3, I examined the potential for horizontal transmission among hatchery-reared brood fish at an R. salmoninarum-positive hatchery facility. Juvenile Cutthroat Trout were placed in sentinel cages near positive adult Rainbow Trout and Cutthroat Trout for three, 30-day periods during optimal temperatures for infection. After exposure, the caged Cutthroat Trout were euthanized, and kidney tissue was tested for R. salmoninarum with qPCR. Only one out of 360 potentially exposed fish tested positive. My data suggest that horizontal transmission may play a small role in maintaining infection in hatchery-reared inland trout. However, I also show that horizontal transmission can occur in a short time, an important consideration when moving fish both within a hatchery or from one unit to another. In chapter 4, I assessed whether the bacterium is vertically transmitted in Cutthroat Trout from the Poudre Rearing Unit in Colorado and the rate of transmission from paternal and maternal brood fish. Adult brood fish were lethally tested for R. salmoninarum and stripped of gametes to create 32 families among four R. salmoninarum infection treatments (MNFN, MNFP, MPFN, MPFP; M: male, F: female, P: positive, N: negative). Progeny from each spawning treatment were sampled at 6- and 12-month post swim-up to test for the presence of R. salmoninarum with an enzyme-linked immunosorbent assay (ELISA) and quantitative polymerase chain reaction (qPCR). My study indicates that vertical transmission occurs in inland Cutthroat trout and transmission is high when examined at the family level but is low within a family. These results suggest that hatcheries should limit vertical transmission through practices such as lethal culling, but also that adopting other methods such as testing eggs for R. salmoninarum should be considered in the future.
... We developed an IPM to evaluate site-level management effects on seasonal demography and annual population growth of bobwhite populations. We linked spring whistle counts with nest monitoring and telemetry datasets to estimate abundance, fecundity, and survival to explicitly account for processes underlying population change (Besbeas et al., 2002;Schaub et al., 2007;Schaub and Abadi 2011;Kéry and Schaub 2012;Zipkin and Saunders 2018;Fig. 3). ...
... While we had count and demographic data for 2016 through 2018, we projected Bobwhite abundance and population change for 2019 by sampling from the posterior distribution of our 2018 demographic rates and estimated site-specific pre-breeding season abundance (Kéry and Schaub 2012;Oppel et al., 2014). ...
Article
Understanding the effects of landscape management on northern bobwhite (Colinus virginianus) population growth requires information on seasonal- and stage-specific demographic parameters linked across the full annual cycle. We evaluated site-level seasonal dynamics and population growth of bobwhites in southwest Missouri and compared differences between three extensively and two intensively managed sites from 2016 to 2019. Extensively managed sites were continuous tracts of native prairie. Intensively managed sites were composed of smaller native and non-native grassland units interspersed with strip crops, food plots, and woody field borders. We radio-marked adults and broods to estimate survival and productivity, conducted spring whistle counts to estimate abundance and developed a two-season, two-stage, two-sex integrated population model to estimate population dynamics. The number of young hatched per female-incubated nest was greater on the three extensively managed sites compared to the two intensively managed sites. Six-month period survival of adults during the breeding season was also greater on the three extensively managed sites compared to the two intensively managed sites. One hundred-day juvenile breeding season survival varied among sites and was highest on Talbot, compared to juvenile breeding season survival on the other four study sites. Six-month, non-breeding season period survival was lowest on the two smallest extensively managed sites, Stony Point Prairie and Shelton Conservation Area compared to non-breeding season survival on the other three study sites. Annual changes in bobwhite abundance were weakly correlated with female fecundity, though this positive relationship was stronger on extensively managed sites. Intensively managed sites exhibited low mean fecundity and breeding season adult survival relative to those that resulted in a stable population. Populations across sites declined from 2016 to 2019 and estimates of annual population growth rates overlapped across sites. Differences between observed changes in bobwhite abundance and estimates of observed demographic rates at some sites suggest unmeasured processes such as movement or bias associated with data and model assumptions influenced estimated vital rates. Overall, our integrated population model was an effective tool for understanding site-level seasonal dynamics and population growth of bobwhites. Based on comparisons of demography between extensively and intensively managed sites, we suggest increased native grassland cover managed with prescribed fire, low intensity grazing, and high mowing may increase bobwhite nesting success and breeding season adult survival, however, achieving stable or increasing rates of population growth may also require increased juvenile breeding season and non-breeding season survival.
... We used multi-season occupancy modelling (MacKenzie et al. 2003) to investigate Black-winged Trumpeter occupancy trends. This approach uses detection/non-detection data to estimate occupancy rates (the proportion of sites occupied by the species) and to model temporal changes in occupancy as a function of local survival and colonization processes, while accounting for imperfect detection (Royle and Dorazio 2008, Kéry and Schaub 2012, MacKenzie et al. 2017. The model requires sampling at two temporal scales, namely primary and secondary periods. ...
... We fitted the model in a Bayesian framework, adapting the specifications provided by (Kéry and Schaub 2012). We implemented the model in JAGS (Plummer 2015) using the R2jags package (Su and Yajima 2012). ...
Article
Full-text available
The ‘Critically Endangered’ Black-winged Trumpeter Psophia obscura is endemic to the Belém Centre of Endemism in extreme eastern Amazonia and has a declining population and range. Here we report on a five-year (2016–2020) systematic camera-trap (n = 61) study of the species in Gurupi Biological Reserve, one of its most important conservation areas. We used a multi-season occupancy model to identify factors affecting occupancy rates and to assess occupancy trends in the reserve. Occupancy was negatively related to elevation and site-level tree density, and was positively related to post-logging recovery times. Average annual occupancy rates remained stable throughout the study period (ranging between 0.57 and 0.67) and this stability was largely driven by high between-year survival rates. Results confirm that the Black-winged Trumpeter is an interior-forest specialist that is highly sensitive to forest disturbance, which underlines the importance of the Gurupi Biological Reserve as a core site for the conservation of the species. However, the species is long-lived, so continuous monitoring is needed to further clarify population trends. We also recommended that the status of the species in other forest remnants, most of which remain unprotected, is assessed.
... The last 5000 iterations of each chain were thinned by half, combined, and summarized as posterior probability distributions for the parameters of the model. The model was implemented in Stan (Stan Development Team 2021; Gelman et al. 2013) with the likelihood coded using the forward algorithm to improve computational efficiency (Kery and Schaub 2011;McClintock et al. 2020;Barraquand and Nielsen 2021). We verified chain convergence using the Gelman-Rubin R statistic (Gelman et al. 2013). ...
Article
Full-text available
Ecological traps can be caused when partial restoration leads organisms to make maladaptive habitat choices. One example of this is fishways (e.g., fish ladders) that provide upstream passage at dams, but are not paired with adequate downstream passage. We tested the hypothesis that attracting anadromous fishes to spawn above a dam, but blocking downstream passage of their offspring leads to an ecological trap. Using passive integrated transponder (PIT) tags, we monitored the movements of steelhead (Oncorhynchus mykiss) at a dam and fishway on the central California coast. We found that downstream passage for juveniles and kelts was limited by four factors: migration delay, loss in the reservoir, avoidance of the downstream bypass, and water depths on the spillway. Based on the spillway-passage depth-thresholds, we estimated that the ability for fish to pass downstream was limited to only half of the migration season in 55% of the past 20 years (2002-2021). Our results support the ecological trap hypothesis, which may explain why restoration using fishways has failed to produce recovery gains in this population and elsewhere.
... Overlooking individuals during counts at a site can result in biased estimation of species occurrence and abundance, as well as to misleading inferences on the key parameters of populations or of the relationships between species and habitats. For instance, not considering detection probability can lead to strongly biased estimates of the effect of environmental features on a species' presence or abundance [6]. Furthermore, overlooking an imperfect detection process causes a systematic underestimation of species distribution and abundance across a study area [1]. ...
Article
Full-text available
Most animal species are detected imperfectly and overlooking individuals can result in a biased inference of the abundance patterns and underlying processes. Several techniques can incorporate the imperfect detection process for a more accurate estimation of abundance, but most of them require repeated surveys, i.e., more sampling effort compared to single counts. In this study, we used the dependent double-observer approach to estimate the detection probability of the egg clutches of two brown frog species, Rana dalmatina and R. latastei. We then simulated the data of a declining population at different levels of detection probability in order to assess under which conditions the double counts provided better estimates of population trends compared to naïve egg counts, given the detectability of frog clutches. Both species showed a very high detection probability, with average values of 93% for Rana dalmatina and 97% for R. latastei. Simulations showed that not considering imperfect detection reduces the power of detecting population trends if detection probability is low. However, at high detection probability (>80%), ignoring the imperfect detection does not bias the estimates of population trends. This suggests that, for species laying large and easily identifiable egg clutches, a single count can provide useful estimates if surveys are correctly timed.
... 1. Detection is imperfect and sampling effort heterogeneous in space and time: not all individuals present in the study area are detected (Kéry and Schaub, 2012). 2. Individuals that reside primarily outside the surveyed area may be detected within it. ...
Technical Report
Full-text available
The Scandinavian wolf (Canis lupus) population is being monitored annually using non-invasive genetic sampling (NGS) and recovery of dead individuals. DNA extracted from faeces, urine, hair, and tissue is used to identify the species, sex, and individual from which each sample originated. These data have been compiled in the Scandinavian large carnivore database Rovbase 3.0. Using the Bayesian open-population spatial capture-recapture (OPSCR) model developed by RovQuant, we estimated annual density and vital rates of the Scandinavian wolf population for nine consecutive seasons from 2013/2014 to 2021/2022. We adjusted the OPSCR model, originally used in previous abundance estimations, to increase the size of the study area and the degree of model realism. We generated annual density maps and estimated total and jurisdiction-specific population sizes for wolf from the winter 2013/2014 to 2021/2022. Based on the OPSCR model, the size of the Scandinavian wolf population was likely (95% credible interval) between 472 and 509 individuals in 2021/2022, with 381 to 417 individuals attributed to Sweden and 83 to 101 to Norway. In addition to annual density and jurisdiction-specific abundance estimates, we report annual estimates of cause-specific mortalities, recruitment, and detection probabilities.
... Individual age-and habitat-specific survival probability was estimated using standard multinomial models of mark-recovery data 53 . Within the same model, annual survival (φ t ) was estimated for first-winter birds, from fledging to 30th June the following calendar year ( ϕ fw,h,t ) and, for adults, from 1st July to the following 30th June (ϕ ad,h,t ) in each of the two habitats (h). ...
Article
Full-text available
The influence of supplementary feeding of wildlife on disease transmission and its consequent impacts on population dynamics are underappreciated. In Great Britain, supplementary feeding is hypothesised to have enabled the spread of the protozoan parasite, Trichomonas gallinae, from columbids to finches, leading to epidemic finch trichomonosis and a rapid population decline of greenfinch (Chloris chloris). More recently, chaffinch (Fringilla coelebs), has also declined markedly from the second to fifth commonest bird in Britain. Using citizen science data, we show that both declines were driven primarily by reduced adult survival, with the greatest reductions occurring in peri-domestic habitats, where supplementary food provision is common. Post-mortem examinations showed a proportional increase in chaffinch trichomonosis cases, near-contemporaneous with its population decline. Like greenfinches, chaffinches often use supplementary food, but are less associated with human habitation. Our results support the hypothesis that supplementary feeding can increase parasite transmission frequency within and between common species. However, the dynamics behind resultant population change can vary markedly, highlighting the need for integrating disease surveillance with demographic monitoring. Other species susceptible to T. gallinae infection may also be at risk. Supplementary feeding guidelines for wildlife should include disease mitigation strategies to ensure that benefits to target species outweigh risks.
... We used a Bayesian hierarchical Cormack-Jolly-Seber (CJS) model to test for differences in apparent overwinter survival among itraconazole-treated frogs, experimental controls, and field controls using code developed by Kéry and Schaub (2011). Capture histories included the initial capture at the time of treatment in 2018 and the four surveys conducted in 2019; we did not include the late fall 2018 surveys as our focus was on apparent overwinter survival, and some frogs had likely begun overwintering by the time these surveys were conducted. ...
Article
Full-text available
Abstract Effective management of wildlife populations threatened by disease requires accurate predictions about the consequences of intervention. However, generating such predictions is challenging, especially for organisms with complex life histories that are also threatened by climate change, such as montane amphibians. Cascades frogs (Rana cascadae) in northern California have experienced dramatic declines associated with the fungal pathogen Batrachochytrium dendrobatidis (Bd), and remnant populations are also threatened by changing climate conditions. We evaluated the population‐level impacts of treating Cascades frog metamorphs with the antifungal chemical itraconazole using a field experiment and population simulations. We explored the influence of larval habitat on these treatment effects by including metamorphs from different larval habitat types. We found that frogs treated with itraconazole were more than four times more likely to survive their first winter than untreated controls and had reduced Bd infection intensity compared to other surviving frogs from the same cohort in the following year. We also found an effect of larval habitat type on Bd infection in recently metamorphosed frogs, with the lowest levels of infection occurring in frogs emerging from larval habitats that tend to be intermediate in temperature and drying rate. Applying the differential apparent overwinter survival of treated and untreated metamorphs to population projections suggests that intermittent antifungal treatment of metamorphs has the potential to restore population viability. Our results indicate that in situ treatment of individual hosts may be a useful component of a comprehensive management strategy to reduce the risk of pathogen‐mediated population declines and extirpations.
... Because the study length was relatively short (21 years), we chose to include only one covariate at a time (see covariate correlations in Appendix S4 & S5), fitting as many IPMs as covariates. We computed the probability (hereafter, p) that regression coefficients of explanatory covariates were below zero and examined the size of the effect (Kéry and Schaub 2012). For quadratic polynomials, the probabilities associated to both coefficients are reported below as pX and pX2. ...
Article
There is growing evidence that the Earth's climate is undergoing profound changes that are affecting biodiversity worldwide. This gives rise to the pressing need to develop robust predictions on how species will respond in order to inform conservation strategies and allow managers to adapt mitigation measures accordingly. While predictions have begun to emerge on how species at the extremes of the so‐called slow‐fast continuum might respond to climate change, empirical studies for species for which all demographic traits contribute relatively equally to population dynamics are lacking. Yet, climate change is expected to strongly affect them throughout their entire lifecycle. We built a 21‐year integrated population model to characterize the population dynamics of the rock partridge (Alectoris graeca) in France, and tested the influence of nine weather covariates on demographic parameters. As predicted both annual survival and breeding success were affected by weather covariates. Thick snow cover during winter was associated with low survival and small brood size the following breeding season. Brood size was higher with intermediate winter temperatures and snowmelt timing, positively correlated to breeding period temperature, but negatively correlated to temperature during the coldest fortnight and precipitation during the breeding period. Survival was positively correlated to winter temperature, but negatively to breeding period precipitation. Large‐scale indices indicated that cold and wet winters were associated with small brood size the following breeding season but with high survival. Expected changes of weather conditions due to climate change are likely to impact demographic traits of the rock partridge both positively and negatively depending on the traits and on the affected weather variables. Future population dynamics will thus depend on the magnitude of these different impacts. Our study illustrates the difficulty to make strong predictions about how species with a population dynamic influenced by both survival and fecundity will respond to climate change. This article is protected by copyright. All rights reserved.
... Our occupancy response variable (ψ) can be considered "use" (sensu MacKenzie, 2005;Mackenzie et al., 2006) because birds may be temporarily, but not permanently, absent from a given survey point at random times. In this context, our estimate of occupancy describes the proportion of survey points ever occupied, rather than the survey points that are permanently occupied (Kéry & Schaub, 2012). We retain the term "occupancy" to maintain the terminology used in this modeling approach (e.g., Glisson et al., 2017). ...
Article
Full-text available
Wetland birds are undergoing severe population declines in North America, with habitat degradation and wetland loss considered two of the primary causes. Due to the cryptic nature of many wetland bird species, the ecological conditions (e.g., matrix composition) that influence bird occupancy, and the relevant spatial scales at which to measure bird responses, remain unclear but may affect inference about wetland use and suitability. We conducted wetland bird surveys at 477 points across northeastern Illinois and northwestern Indiana within the highly urbanized landscape surrounding Chicago. Using remotely sensed land cover data, we built occupancy models for 10 wetland bird species (American Coot Fulica americana, Black‐crowned Night‐Heron Nycticorax nycticorax, Blue‐winged Teal Anas discors, Common Gallinule Gallinula galeata, Least Bittern Ixobrychus exilis, Marsh Wren Cistothorus palustris, Pied‐billed Grebe Podilymbus podiceps, Sora Porzana carolina, Swamp Sparrow Melospiza georgiana, and Virginia Rail Rallus limicola) to quantify their responses to wetland cover types (emergent wetland, forested wetland, riverine wetland, and freshwater pond) and urbanization at four spatial scales (200‐, 400‐, 800‐, and 2000‐m radial distances). We also included the distance to Lake Michigan as a covariate in occupancy models to account for ecological differences between coastal and inland wetlands. We found that relationships between land cover types and occupancy differed by species, as did the spatial scale of support. Generally, the presence of emergent wetlands or ponds at immediate (200 m) and local (400 m) spatial scales within the surrounding matrix was positively associated with wetland bird occupancy. Contrary to expectations, we did not find support for a negative relationship between urbanization and occupancy for most focal species, indicating that birds are using available wetland habitats despite surrounding development. While future research should evaluate management strategies at the watershed scale, our findings suggest that wetland conservation planning at immediate and local scales is likely to promote bird habitat use within highly modified landscapes of the Upper Midwestern United States.
... State-space models are hierarchical models with two components; a process model that describes the natural stochasticity in ecological processes, and an observation model that describes the error associated with sampling or observing animals, given their current state (de Valpine and Hastings, 2002;Auger Méthe et al., 2021;Newman et al., 2022). In a state-space framework, multistate mark-recapture models allow for probabilistic transition of individuals between a set of latent states (Gimenez et al., 2007;Kéry and Schaub, 2012). Separation of demographic parameters from nuisance parameters such as detection probabilities through the use of separate state and observation equations permits fitting of complex models in this framework (Clark, 2005;Pradel, 2005;Pedersen and Weng, 2013;Auger-Méthé et al., 2021). ...
Article
Full-text available
Sustainable management of exploited marine fish and wildlife populations requires knowledge about their productivity. Survival from natural causes of mortality is a key component of population productivity, but is notoriously difficult to estimate. We evaluate the potential for acoustic telemetry as a long-term monitoring tool to estimate rates of natural mortality. We present a Bayesian multistate mark-recapture model for telemetry data collected over a decade from 188 Atlantic bluefin tuna ( Thunnus thynnus) and apply it to estimate the rate of natural mortality using only acoustic tag detections for all animals, or using acoustic tag detections for 96 single-tagged tuna plus acoustic tag detections combined with estimated positions from pop-up satellite archival tags for 92 double-tagged animals. We support the model for bluefin tuna with a simulation study to quantify bias in estimates of population dynamics parameters and investigate the effect of auxiliary information from satellite tagging on mortality rate estimates for different acoustic tag detection probability scenarios. We obtained posterior estimates of the instantaneous annual natural mortality (survival) rate across a decade of tagging for Atlantic bluefin tuna of 0.17 yr ⁻¹ (0.84 yr ⁻¹ ) both using only acoustic tagging data, and using a combination of acoustic and satellite tagging data. Use of a prior implying a higher rate of fishing mortality yielded an instantaneous annual natural mortality (survival) estimate of 0.10 yr ⁻¹ (0.90 yr ⁻¹ ), with combined acoustic and satellite tag data. Results from the simulation study indicate that the use of satellite tags can improve the precision and accuracy of estimates of detection probabilities, area-specific movement probabilities and mortality rates, where the extent of the improvement depends on true underlying acoustic tag detection probabilities. Our work demonstrates that long-term acoustic tagging data sets have strong potential for monitoring of highly migratory marine fish and wildlife populations, providing information on a number of key parameters, including survival and movement rates. However, improved information on tag reporting rates or fishing mortality is needed to better separate natural and fisheries mortality for Atlantic bluefin tuna.
... Similar to the findings of several previous studies, density estimates from ML-SECR and Bayesian SCR models were very similar (Noss et al., 2012, Gerber et al., 2011, Gopalaswamy et al., 2012and Ivan et al., 2012. In general, it was observed that the posterior mean summaries for Bayesian estimators were very similar to the maximum likelihood estimates when the priors in the Bayesian models were non-informative (Kery and Schaub 2012). We defaulted to using uniform priors for our Bayesian analysis in program SPACECAP. ...
... We skipped species during closed sessions with <10 individuals captured at a geographic site and those site/month/years with <3 trapping arrays running, for a total of 9,226 models. We assessed model fit for each species closed session with posterior predictive checks (Kéry and Schaub 2011;Gelman et al. 2014) by calculating the sum of squared Pearson residuals. We calculated a Bayesian P-value from posterior simulations and assumed adequate fit if 0.1 < P < 0.9. ...
Article
Small mammals are important to the functioning of ecological communities with changes to their abundances used to track impacts of environmental change. While capture–recapture estimates of absolute abundance are preferred, indices of abundance continue to be used in cases of limited sampling, rare species with little data, or unmarked individuals. Improvement to indices can be achieved by calibrating them to absolute abundance but their reliability across years, sites, or species is unclear. To evaluate this, we used the US National Ecological Observatory Network capture–recapture data for 63 small mammal species over 46 sites from 2013 to 2019. We generated 17,155 absolute abundance estimates using capture–recapture analyses and compared these to two standard abundance indices, and three types of calibrated indices. We found that neither raw abundance indices nor index calibrations were reliable approximations of absolute abundance, with raw indices less correlated with absolute abundance than index calibrations (raw indices overall R2 < 0.5, index calibration overall R2 > 0.6). Performance of indices and index calibrations varied by species, with those having higher and less variable capture probabilities performing best. We conclude that indices and index calibration methods should be used with caution with a count of individuals being the best index to use, especially if it can be calibrated with capture probability. None of the indices we tested should be used for comparing different species due to high variation in capture probabilities. Hierarchical models that allow for sharing of capture probabilities over species or plots (i.e., joint-likelihood models) may offer a better solution to mitigate the cost and effort of large-scale small mammal sampling while still providing robust estimates of abundance.
... We estimated survival probability between sampling occasions using Cormack-Jolly-Seber (CJS) models (Lebreton et al. 1992) in the Bayesian hierarchical approach (Kéry and Schaub 2012). We assumed that individual i of species s survived from occasion t to occasion t + 1 with a species-specific, interval-specific survival probability, ϕ, for species s and occasion t: ...
Article
Full-text available
Animals experience seasonally changing conditions in temperate regions, thus population vital rates change seasonally. However, knowledge is lacking on patterns of seasonal correlation between growth and survival in sympatric ectotherms, and this knowledge gap limits our understanding of environmental change impacts on animal populations and communities. Here, we investigated sub‐seasonal (two‐month intervals) correlation between growth and survival in three stream fishes (bluehead chub Nocomis leptocephalus, creek chub Semotilus atromaculatus and mottled sculpin Cottus bairdii) in South Carolina, USA, via a mark–recapture survey over 28 months. We found that patterns of temporal correlation between the population vital rates differed among the sympatric species. Growth increased and survival decreased with water temperature in two eurythermal species, resulting in negative correlation between growth and survival. Growth peaked in sub‐seasons with an intermediate water temperature range in a third stenothermal species, while survival decreased with water temperature for this species too. Consequently, there was not significant negative or positive correlation between sub‐seasonal growth and survival in the stenothermal species. Body condition (weight at given length) decreased from May through November in all three species, providing a potential physiological explanation for why survival rates were lower during this period. Negative correlation among population vital rates stabilizes population size over time and buffers animal populations from environmental change because the vital rates are not affected simultaneously in the same direction, indicating some degree of resiliency in the face of climate changes in the two eurythermal species. However, such a demographic mechanism of resiliency could be maintained so long as climate warming does not exceed optimal growth temperature, above which negative correlation between growth and survival may no longer be maintained.
... We first plotted predictions of the numbers of breeders and breeding immigrants, and several measures of reproductive output, against relevant observational data to ensure that predictions were not substantially biased (Gelman et al. 2013). Second, we checked for major discrepancies among datasets and between datasets and the population model by comparing posterior distributions of vital rate parameters obtained from the IPM to those obtained from models estimating each vital rate independently (Kéry & Schaub 2012;Gelman et al. 2013). ...
Article
Many migratory species are in decline across their geographical ranges. Single‐population studies can provide important insights into drivers at a local scale, but effective conservation requires multi‐population perspectives. This is challenging because relevant data are often hard to consolidate, and state‐of‐the‐art analytical tools are typically tailored to specific datasets. We capitalized on a recent data harmonization initiative (SPI‐Birds) and linked it to a generalized modeling framework to identify the demographic and environmental drivers of large‐scale population decline in migratory pied flycatchers (Ficedula hypoleuca) breeding across Britain. We implemented a generalized integrated population model (IPM) to estimate age‐specific vital rates, including their dependency on environmental conditions, and total and breeding population size of pied flycatchers using long‐term (34‐64 years) monitoring data from seven locations representative of the British breeding range. We then quantified the relative contributions of different vital rates and population structure to changes in short‐ and long‐term population growth rate using transient life table response experiments (LTREs). Substantial covariation in population sizes across breeding locations suggested that change was the result of large‐scale drivers. This was supported by LTRE analyses, which attributed past changes in short‐term population growth rates and long‐term population trends primarily to variation in annual survival and dispersal dynamics, which largely act during migration and/or non‐breeding season. Contributions of variation in local reproductive parameters were small in comparison, despite sensitivity to local temperature and rainfall within the breeding period. We show that both short‐ and longer‐term population changes of British‐breeding pied flycatchers are likely linked to factors acting during migration and in non‐breeding areas, where future research should be prioritized. We illustrate the potential of multi‐population analyses for informing management at (inter)national scales and highlight the importance of data standardization, generalized and accessible analytical tools, and reproducible workflows to achieve them.
... Random slopes for covariates may also be considered. This model is very similar to the basic Bayesian model for time-series analysis at multiple sites presented by Kéry and Schaub (2012), which has been widely applied in ecological analyses. It can be fitted with common non-Bayesian statistical packages or by Bayesian methods. ...
Article
Full-text available
Time-series data offer wide-ranging opportunities to test hypotheses about the physical and biological factors that influence species abundances. Although sophisticated models have been developed and applied to analyze abundance time series, they require information about species detectability that is often unavailable. We propose that in many cases, simpler models are adequate for testing hypotheses. We consider three relatively simple regression models for time series, using simulated and empirical (fish and mammal) datasets. Model A is a conventional generalized linear model of abundance, model B adds a temporal autoregressive term, and model C uses an estimate of population growth rate as a response variable, with the option of including a term for density dependence. All models can be fit using Bayesian and non-Bayesian methods. Simulation results demonstrated that model C tended to have greater support for long-lived, lower-fecundity organisms (K life-history strategists), while model A, the simplest, tended to be supported for shorter-lived, high-fecundity organisms (r life-history strategists). Analysis of real-world fish and mammal datasets found that models A, B, and C each enjoyed support for at least some species, but sometimes yielded different insights. In particular, model C indicated effects of predictor variables that were not evident in analyses with models A and B. Bayesian and frequentist models yielded similar parameter estimates and performance. We conclude that relatively simple models are useful for testing hypotheses about the factors that influence abundance in time-series data, and can be appropriate choices for datasets that lack the information needed to fit more complicated models. When feasible, we advise fitting datasets with multiple models because they can provide complementary information.
... Integrated population model.--We developed an integrated population model (IPM) that linked spring whistle counts with nest monitoring and telemetry data to estimate abundance, fecundity, and survival, in order to model population change for each site (Schaub et al. 2007, Schaub and Abadi 2011, Kéry and Schaub 2012, Zipkin and Saunders 2018. We used a 2-stage, 2-sex periodic matrix design comprising the breeding season (1 May-31 Oct) and non-breeding season (1 Nov-30 Apr) to account for differences in survival and productivity among adults and juveniles, and males and females among our sites (Burger et al. 1995a, Sandercock et al. 2008. ...
... Integrated population model.--We developed an integrated population model (IPM) that linked spring whistle counts with nest monitoring and telemetry data to estimate abundance, fecundity, and survival, in order to model population change for each site (Schaub et al. 2007, Schaub and Abadi 2011, Kéry and Schaub 2012, Zipkin and Saunders 2018. We used a 2-stage, 2-sex periodic matrix design comprising the breeding season and non-breeding season to account for differences in survival and productivity among adults and juveniles, and males and females among our sites (Burger et al. 1995a. ...
... Finally, we assessed model fit using a Bayesian p-value and a Ĉ score calculated from both simulated and observed data (Allen et al., 2022;Suraci et al., 2021), where Bayesian p-values between 0.25 and 0.75 indicate a good fit and a Ĉ score less than 1.1 suggests that minimal overdispersion remains unaccounted for (Conn et al., 2018;Gelman et al., 1996;Mazerolle, 2017). Both values were calculated using chi-squared discrepancy statistics (Allen et al., 2022;Kéry and Schaub, 2011;Suraci et al., 2021). We compared covariates with significant relationships based on their β coefficient values and Bayesian Credible Intervals (CIs). ...
Article
Full-text available
Non-charismatic species are often understudied, despite having unique ecological roles that are important to understand and value for ecosystem integrity, function, and health. Striped skunks (Mephitis mephitis) are one such species, as they are stigmatized and feared due to their noxious smell and potential for spreading rabies. Yet this species often co-occurs with humans, occupying a distinct but poorly understood ecological niche. To better understand this understudied species and its unique niche, we used a Bayesian N-mixture model to estimate factors driving the local abundance of striped skunks at a continent-wide scale using data from the Snapshot USA programs in 2019 and 2020. In our modelling approach, we included factors that we a-priori hypothesized would affect striped skunk local abundance, including habitat, agriculture, primary productivity, and anthropogenic factors. We found that skunks were most abundant in more productive and anthropogenically affected areas. The factors driving abundance (in order of effect) were primary productivity, impervious surface, cultivated land, and cultivated-wetland edge cover. While our results underscore striped skunks’ role as a synanthropic species that co-occurs with humans, they are more of a synanthropic misanthrope, as their close relationship with people often causes them to suffer persecution and death. But their relationship with highly productive areas also underscores how striped skunks are habitat generalists that exploit resources in many different types of habitats. Our analysis highlights the importance of programs like Snapshot USA that collect standardized data across large geographic areas and allow broad-scale studies for evaluating the local abundance of understudied species. Using these data, we were able to provide foundational information on the relative importance of the major factors affecting the local abundance of striped skunks, with implications for the management and conservation of this unique and understudied mesocarnivore across its range.
... McCaffery and Lukacs (2016) provide an example involving the use of informative priors for survival in birds. Nonetheless, the ideal approach would be to use the encounter histories as data in the IPM joint likelihood when possible and practical (Kéry & Schaub, 2012). ...
Article
Informed conservation and management of wildlife require sufficient monitoring to understand population dynamics and to direct conservation actions. Because resources available for monitoring are limited, conservation practitioners must strive to make monitoring as cost‐effective as possible. Our focus was on assessing the value of monitoring to the adaptive harvest management (AHM) program for pink‐footed geese (Anser brachyrhynchus). We conducted a retrospective analysis to assess the costs and benefits of a capture‐mark‐resight (CMR) program, a productivity survey, and biannual population censuses. Using all available data, we fit an integrated population model (IPM) and assumed that inference derived from it represented the benchmark against which reduced monitoring was to be judged. We then fit IPMs to reduced sets of monitoring data and compared their estimates of demographic parameters and expected management performance against the benchmark IPM. Costs and the precision and accuracy of key demographic parameters decreased with the elimination of monitoring data. Eliminating the CMR program, while maintaining other monitoring instruments, resulted in the greatest cost savings, usually with small effects on inferential reliability. Productivity surveys were also expensive and some reduction in survey effort may be warranted. The biannual censuses were inexpensive and generally increased inferential reliability. The expected performance of AHM strategies was surprisingly robust to a loss of monitoring data. We attribute this result to explicit consideration of parametric uncertainty in harvest‐strategy optimization and the fact that a broad range of population sizes is acceptable to stakeholders. Synthesis and applications: Our study suggests that existing or potential monitoring instruments for wildlife populations should be scrutinized as to their cost‐effectiveness for improving biological inference and management performance. Using Svalbard pink‐footed geese as a case study, we show that the loss of some existing monitoring instruments may not be as adverse as commonly assumed if data are jointly analyzed in an integrated population model. Finally, regardless of the monitoring data available, we suggest that conservation strategies that explicitly account for uncertainty in demography are more likely to be successful than those that do not.
... trap-dependent recapture rates) in the CJS model. In Eq. 2, α 0 and δ k are the intercept terms for adult survival and recapture probabilities, respectively, and recapture is modelled as a function of whether or not an individual was captured in the preceding year (δ 1 = yes, δ 2 = no) (Kéry and Schaub 2012). We also included a random effect term (ε *,t ) in each function to represent annual environmental stochasticity for survival and variation in sampling effort for recapture. ...
Article
Full-text available
The source–sink paradigm predicts that populations in poorer‐quality habitats (‘sinks') persist due to continued immigration from more‐productive areas (‘sources'). However, this categorisation of populations assumes that habitat quality is fixed through time. Globally, we are in an era of wide‐spread habitat degradation, and consequently there is a pressing need to examine dispersal dynamics in relation to local population change. We used an integrated population model to quantify immigration dynamics in a long‐lived colonial seabird, the black‐legged kittiwake Rissa tridactyla, that is classified as globally ‘Vulnerable'. We then used a transient life table response experiment to evaluate the contribution of temporal variation in vital rates, immigration rates and population structure to realised population growth. Finally, we used a simulation analysis to examine the importance of immigration to population dynamics. We show that the contribution of immigration changed as the population declined. This study demonstrates that immigration is unlikely to maintain vulnerable sink populations indefinitely, emphasising the need for temporal analyses of dispersal to identify shifts that may have dramatic consequences for population viability.
Article
Full-text available
Human‐shared landscapes cover much of Earth, yet their conservation value is contested. This controversy may persist because previous studies have examined species diversity, rather than the processes through which such diversity is maintained. For example, a site exhibiting high diversity may not actually bolster populations if the diversity is only maintained through net immigration. Recent research has begun to isolate the processes that maintain metacommunities and develop functional trait methods to identify these processes. However, the processes underlying bird communities remain obscure. Here, we leverage metacommunity theory, functional trait partitioning and a Bayesian multispecies abundance model to assess whether a shared landscape – woody perennial polyculture farms – bolsters bird diversity. Such farms grow multiple species of food‐producing woody perennials together with vegetative groundcover. We surveyed birds and their in situ functional traits across the US Midwest in traditional agriculture, woody perennial polyculture, prairie and woods. We found that woody perennial polycultures exhibited the highest bird diversity and were the most preferred by many species (including threatened ones). Moreover, our functional trait analysis suggests that this diversity is maintained through habitat filtering and competition, rather than merely immigration. Thus, shared landscapes can likely conserve birds by providing a distinct habitat. These results suggest that woody perennial polyculture farms offer substantial potential to support bird populations in the US Midwest. Our study demonstrates the utility of in situ functional trait partitioning within a Bayesian framework to unmask ecological processes and help assess the conservation value of landscapes.
Conference Paper
Full-text available
Northern bobwhite (Colinus virginianus; hereafter, bobwhite) in the Mid-Atlantic United States have been experiencing precipitous population declines due to a combination of habitat deterioration, urban and suburban sprawl, change in forest management regimes, and farming practices. In recent years, restocking of bobwhite through translocation efforts has gained interest to rebuild local populations. However, empirical studies are warranted to understand the limitations of translocation as it relates to its potential use for long-term population recovery and persistence in this region. Further, few studies nation-wide have evaluated resource use and survival during the non-breeding season on translocated sites. As such, we translocated 360 bobwhites from source populations from southern latitudes during March–April 2015–2017 into 2 different landscape types similar to the source population vegetative communities (agricultural cropland dominated in Maryland, USA versus pine forest dominated in New Jersey, USA) and tested the effects of habitat fragmentation on survival and habitat use during the subsequent non-breeding season. We found habitat fragmentation negatively affecting survival and resource use among translocated bobwhite on fragmented cropland-dominated sites as compared to larger unfragmented forested sites. Survival was lower on cropland sites compared to forested sites such that bobwhite in cropland-dominated landscapes were >125 times less likely to survive the winter than those on forested sites. In our examination of resource use, bobwhite in structurally complex forested sites used cut pine, early-successional woody, early-successional herbaceous, and thinned pine more than what was available on the landscape. On the cropland sites bobwhite used food plots, early-successional woody, and mixed woods more than what was available on the landscape and only food plots at the home range scale. While larger unfragmented forested bobwhite habitat ultimately provided a more successful translocation landscape, birds still had large home ranges and relatively low survival. Therefore, proper pine management may be necessary to optimize habitat availability during the non-breeding season. Our findings provide rare information on demographic resiliency and resource use for translocated bobwhite during the non-breeding season. Further, this research provides valuable information to improve future translocation efforts in the Mid-Atlantic.
Chapter
Although a harvest-based model (HBM) is a useful model to estimate wildlife abundance under hunting pressure and is wildly applied in the actual population management of sika deer (Cervus nippon) in Japan, the parameter-identifiable condition of HBM has rarely been clarified. In this study, I examined this condition in relation to the amount of direct abundance index, which is used to calculate absolute abundance from the data only and hunting pressure by data simulation. Furthermore, I applied HBM to the real monitoring data of sika deer that were collected in the Yamanashi Prefecture. Data simulation showed that the parameters of HBM were identifiable if the direct abundance index was available in 5% of the spatial units for abundance estimation, even if it was collected only in 1 year during the survey period. Furthermore, strong hunting pressure, which is comparable to population growth rate, increased the precision of parameter estimation. When HBM was applied to real data, sika deer abundance at 25.3 km2 cells (5.5 km × 4.6 km) could be estimated without any informative prior distribution, and the median of intrinsic population growth rate was 0.146. In conclusion, HBM is a useful model to estimate wildlife abundance, but the estimable conditions should be carefully checked by scientific experts.
Article
This paper describes the development of estimators for the contemporary migration number and rate of adults between two populations in iteroparous species. The proposed estimators are based on known half‐sibling (HS) and/or parent–offspring (PO) relationships observed between populations across breeding seasons. The rationale is that HS and PO pairs exhibit information about the occurrence frequency of parental movements during the breeding interval. The proposed method allows for variance in the average number of offspring per parent within and between populations. In addition, coupled with the PO pairs found within the population, the estimators can be obtained using only genetic data. Generally, a sample size representing the square root of the population size is required to obtain meaningful migration information. We describe a detailed evaluation of the performance of the proposed estimators by running an individual‐based model, and the results provide guidance regarding sample sizes to ensure the required accuracy and precision. In addition, given that there are few effective methods to estimate adult movement (especially when populations cannot be genetically distinct), we discuss the usefulness of the proposed kinship assignment method in terms of conservation biology and wildlife management.
Article
Climate change has altered annual hydrologic cycles globally. Where stream flows are dominated by snow melt in western North America, climate change has resulted in peak flows that occur earlier in the runoff season and reduced flows during the non‐runoff season. Such changes can negatively affect the reproduction success and first‐year survival of stream‐dependent, autumn‐spawning salmonid fishes. We annually sampled bull trout (Salvelinus confluentus Suckley) populations in the St. Mary River catchment in north‐central Montana, U.S.A. Bull trout (≥ age 3; n = 3,502) electrofished from six creeks during 1998–2018 were given passive integrated transponder tags and 739 (21.1%) of those fish were recaptured through 2019 (tag loss c. 4.5%). We used tag–recapture data to (1) determine the spatial extent of the bull trout metapopulation; (2) fit a Bayesian Jolly–Seber model to identify the temporal trends in the sizes, finite growth rates (λ), and apparent survival and detection probabilities of the three main populations; and (3) assess the statistical associations between λ and flow. Bull trout movements among creeks defined the metapopulation's almost catchment‐wide spatial extent. Statistical correlations between λ and flow were clumped in the month × flow variable space at flow lags of 3–7 years, which corresponded with production of the dominant age classes in our tagged populations. Positive correlations occurred during June–September (the bull trout spawning season) and negative correlations during October–May (the embryo‐incubation period). Declining spawning‐season flow could have degraded bull trout spawning habitat, whereas increasing incubation‐period flow could have physically disturbed embryos affecting their survival. This study revealed associations between bull trout population dynamics and climate change as indexed by stream flow. Although the spawning habitats of these bull trout are protected within Glacier National Park and the occurrence of multiple local populations supports metapopulation persistence, that persistence is diminished because flow‐regime change has a strong negative influence on each population. Given the region's apparent climate trajectory, further decline in the bull trout metapopulation is anticipated.
Article
Biologists routinely fit novel and complex statistical models to push the limits of our understanding. Examples include, but are not limited to, flexible Bayesian approaches (e.g. BUGS, stan), frequentist and likelihood‐based approaches (e.g. packages lme4) and machine learning methods. These software and programs afford the user greater control and flexibility in tailoring complex hierarchical models. However, this level of control and flexibility places a higher degree of responsibility on the user to evaluate the robustness of their statistical inference. To determine how often biologists are running model diagnostics on hierarchical models, we reviewed 50 recently published papers in 2021 in the journal Nature Ecology & Evolution, and we found that the majority of published papers did not report any validation of their hierarchical models, making it difficult for the reader to assess the robustness of their inference. This lack of reporting likely stems from a lack of standardized guidance for best practices and standard methods. Here, we provide a guide to understanding and validating complex models using data simulations. To determine how often biologists use data simulation techniques, we also reviewed 50 recently published papers in 2021 in the journal Methods Ecology & Evolution. We found that 78% of the papers that proposed a new estimation technique, package or model used simulations or generated data in some capacity (18 of 23 papers); but very few of those papers (5 of 23 papers) included either a demonstration that the code could recover realistic estimates for a dataset with known parameters or a demonstration of the statistical properties of the approach. To distil the variety of simulations techniques and their uses, we provide a taxonomy of simulation studies based on the intended inference. We also encourage authors to include a basic validation study whenever novel statistical models are used, which in general, is easy to implement. Simulating data helps a researcher gain a deeper understanding of the models and their assumptions and establish the reliability of their estimation approaches. Wider adoption of data simulations by biologists can improve statistical inference, reliability and open science practices.
Article
Natal survival and dispersal have important consequences for populations through the movement of genes and individuals. Metapopulation theory predicts either balanced natal dispersal among regions or source–sink dynamics, which can dramatically change population structure. For species reliant on dynamic, early‐successional habitats, availability and location of habitat will shift from year to year, requiring primiparous individuals to locate an appropriate breeding habitat. We estimated hatch‐year survival to adulthood and natal dispersal rates between two breeding groups of Northern Great Plains piping plovers (Charadrius melodus) from four cohorts (n = 2669 total individuals; 2014–2017). Hatch‐year survival to adulthood was slightly higher for individuals hatched on the Missouri River than on the US Alkali Wetlands but declined over time. Individuals hatched on the US Alkali Wetlands were more likely to disperse to breed on the Missouri River (0.33 [0.20, 0.48]) than vice versa (0.17 [0.11, 0.24]). When more habitat was available at the natal site than in the prior year, natal dispersal rates increased. However, despite higher recruitment rates as a result of higher natal fidelity, the Missouri River showed lower total recruitment with a declining trend in the number of recruits, largely due to differences in abundance between breeding groups. Overall, unbalanced, high natal dispersal rates within the Northern Great Plains indicate high connectivity among distinct regions with different water regimes on the Missouri River and on the US Alkali Wetlands driven by fluctuating availability of habitat. Our results suggest that plovers in the Northern Great Plains take advantage of dynamic habitats where they are available in a broad geographic area, which is consistent with a spatially structured panmictic population rather than a true metapopulation, but further research on adult breeding dispersal is needed to clarify population structure.
Article
Full-text available
Hidden Markov models (HMMs) are broadly applicable hierarchical models that derive their utility from separating state processes from observation processes yielding the data. Multistate models such as mark–recapture and dynamic multistate occupancy models are HMMs frequently used in ecology. In their early formulations, states, such as pathogen infection status, were assumed to be perfectly observed without ambiguity. However, state uncertainty is a pervasive feature of many ecological studies, and multievent models were developed to explicitly account for it. We developed a novel extended multievent mark–recapture model that incorporates state uncertainty at multiple levels of detection. Using a disease‐structured example, both false negative and false positive state assignment errors are modelled at two levels of state assignment—the pathogen sampling process and the diagnostic process that samples are subjected to. We additionally describe methods to jointly model infection intensity to integrate heterogeneity in ecological parameters, such as mortality and infection dynamics, and the pathogen detection processes. We provide code to simulate and analyse datasets with various underlying ecological processes and fit our model to a mark–recapture dataset of Mixophyes fleayi (Fleay's barred frog) infected with the amphibian chytrid fungus (Batrachochytrium dendrobatidis, Bd). In our case study, we found evidence for various state assignment errors: the sampling protocol performed poorly in detecting Bd, pathogen detection was highly dependent on infection intensity and false positives were non‐negligible. Incorporating state uncertainty yielded significantly higher estimates of infection prevalence and 4–5 times lower rates of infection state transitions compared to those obtained from a traditional multistate model. Our results highlight that incorporating state assignment errors improves inference on the ecological process, especially when sensitivity and specificity of the state assignment processes are low. The general model structure can be applied to other HMMs, providing a foundation for modelling state uncertainty in related models. For disease‐structured multistate models, we recommend conducting robust design surveys and collecting samples during each capture event to facilitate incorporating pathogen detection errors.
Article
Full-text available
Background Pine flatwoods of the southeastern United States were shaped by frequent fires. Land managers use prescribed fires to control fuels but also to restore historical fire dynamics. Broad outcomes of this practice are well-understood, but impacts on many organisms are still being explored. Frogs, for example, have upland and wetland requirements, limited mobility, and skin susceptible to desiccation. Treefrogs spend most of their lives in uplands away from water. When fire approaches, animals may escape to an unburned area, shelter in place, or be killed by the fire. We examined which of these mechanisms is the prevailing short-term response for a specialist treefrog in a pyrogenic flatwood system. Results We assessed the short-term impacts of prescribed fire on the dynamics of an upland flatwood specialist, the pinewoods treefrog Dryophytes femoralis , using a replicated before-after-control-impact field experiment. We set pipes as treefrog refugia at 3 m, 6 m, 9 m, and 9 ⁺ m in 12 pine trees spread evenly across two treatments: reference trees in units burned in 2020 and trees in units with 2021 prescribed fire. Prescribed fires occurred on 16 April and 21 July 2021. Every 2 weeks between 5 March and 5 September, we checked pipes for frogs and assigned them unique color marks. We observed 78 individuals with 199 additional recaptures. We modeled abundance (as raw counts), survival, and vertical movement using mark-recapture methods, multi-state, and mixed linear models with a Bayesian framework. Survival and recapture were comparable among prescribed fire treatments, but abundances and movement probability varied. Frogs in trees in areas burned during the study were more likely to stay in place and less likely to descend to lower heights. We observed more frogs in trees after a 2021 fire compared to reference trees. Conclusions The prevailing mechanism for resiliency to fire for pinewoods treefrogs was migration up large pines, then likely recolonization to lower vegetation layers when plants regreen post-fire. This substantiates conclusions from other works that the integrity of mature pines is key to sustaining native biodiversity. Future work and management should consider the three-dimensional structure of habitat when developing burn prescriptions and study designs.
Article
The Northwestern Pond Turtle (Actinemys marmorata; WPT) was once widespread throughout the Sacramento Valley and the Sacramento-San Joaquin River Delta. Much of its historical range has been converted into agricultural land, reducing and altering aquatic habitat and surrounding uplands. Red-eared Sliders (Trachemys scripta elegans; RES) have been introduced throughout much of the existing WPT range, particularly near urban centers, potentially competing with WPT for resources. Previous surveys for turtles in central California have primarily focused on rivers, lakes, and protected wetlands. Little is known about where WPT and RES occur in the vast expanses of agricultural land across the Sacramento Valley and Sacramento-San Joaquin River Delta. Using aquatic hoop nets, we surveyed 142 locations (102 irrigation canal sites, 39 wetlands, 1 tidally influenced slough) across 8 counties during the summers of 2018 and 2019. Both species were detected in agricultural habitats. Using occupancy modeling, we estimated that WPT occur at 44 (95% CRI = 3853) of our trapping sites and RES occur at 51 (4166) sampled sites. Co-occurrence of these 2 species was rare; the species were found together at only 6 sites. RES were primarily found in restored wetlands near major roads and the Sacramento metropolitan area, whereas WPT were more commonly found farther from urban areas in wider canals. Our work provides a picture of how WPT and RES occupy this modified agroecosystem that can inform future conservation efforts.
Preprint
Full-text available
Freshwaters represent the most threatened environments with regard to biodiversity loss and therefore there is a need for national monitoring programs to effectively document species distribution and evaluate potential risks for vulnerable species. The monitoring of species for effective management practices is, however, challenged by insufficient data acquisition when using traditional methods. Here we present the application of environmental DNA (eDNA) metabarcoding of amphibians in combination with quantitative PCR assays for an invasive pathogenic chytrid species (Batrachochytrium dendrobatidis -Bd), a potential threat to endemic and endangered amphibian species. Statistical comparison of amphibian species detection using either traditional or eDNA-based approaches showed weak correspondence. By tracking the distribution of Bd over three years, we concluded that the risk for amphibian extinction is low since Bd was only detected at five sites where multiple amphibians were present over the sampled years. Our results show that eDNA-based detection can be used for simultaneous monitoring of amphibian diversity and the presence of amphibian pathogens at the national level in order to assess potential species extinction risks and establish effective management practices. As such our study represents suggestions for a national monitoring program based on eDNA.
Article
Full-text available
Globally, habitat loss has been deemed a major threat to wetland bird populations. However, the underlying mechanism of population declines and variations in the birds’ vulnerability throughout their annual cycle is challenging to determine, yet critical for development of targeted conservation strategies. Over seven years, landscape water availability explained occupancy of breeding territories best when breeding performance, migratory performance, and annual survival of the White-naped Crane (Grus vipio) population in eastern Mongolia were studied. Also, the hatching success of eggs was positively correlated with water availability in addition to plant productivity. High ambient temperatures and large numbers of herder families (and hence more livestock) negatively affected hatching success. High water availability at Luan, a major stopover site increased migration speed during the cranes’ northbound migration to their breeding grounds. In contrast, when water conditions were favorable, the birds stayed longer at the stopover site during southbound migration. Increased human density reduced the use of the stopover site during northbound migration. Finally, cranes arrived early at the breeding grounds when ambient temperature was high in northeast Mongolia. Combining these findings with historical trends in key environmental factors on their breeding grounds explains the general decline observed in this population of cranes in recent decades. Extrapolating our findings with future climate predictions, the outlook seems poor unless urgent action is taken. Knowledge of the mechanisms underlying White-naped Crane population decline in eastern Mongolia identified in this paper should improve the effectiveness of these actions.
Thesis
Full-text available
Estuaries are commonly touted as nurseries for outmigrating salmonids, providing higher prey availability than streams, a physiological transition zone, and refugia from marine predators. Yet the diversity of estuaries makes it difficult to generalize the effect they have on salmonid recruitment. In bar-built estuaries, sandbars form at the mouth of rivers during periods of low flow, closing access to the ocean and disrupting outmigration. In this thesis, I evaluated how residency in a bar-built estuary affects the growth, survival, and ultimately recruitment of Chinook salmon (Oncorhynchus tshawytscha) in Redwood Creek, California. I conducted a mark-recapture experiment on outmigrating juveniles during the summer of 2018 to determine estuary abundance, growth, and survival. I used scales and sagittal otoliths collected from spawning adult carcasses to quantify the contribution of different juvenile life histories to the adult population. I then integrated these data and monitoring data collected from spawning ground surveys, rotary screw traps, and estuary seines to create a stage-structured matrix model. iii Juveniles that remained in the estuary after the mouth closed were larger at ocean entry than ocean rearing juveniles that entered the ocean earlier in the spring. However, estuary rearing juveniles grew less and ultimately were smaller than ocean rearing juveniles prior to winter. Despite having a larger ocean entry size, estuary rearing juveniles had lower survival from river outmigration to adult return than ocean rearing juveniles and contributed disproportionately less to the spawning population. Lack of marine influence and low river flow are common attributes of bar-built estuaries that may lower food availability and deteriorate conditions in these estuaries. Levees constructed in lower Redwood Creek prevent flooding and establishment of marsh and floodplain habitat, potentially majorly limiting the productivity of the estuary and salmonid growth. Restoration efforts designed to address limitations to growth in the estuary such as low food availability and high temperatures are needed to increase the ocean survival and ultimately contribution of estuary juveniles to the population. iv ACKNOWLEDGEMENTS
Article
The planned removal of four dams on the Klamath River (anticipated 2024) will be the largest river restoration effort ever undertaken on the planet. Dam removal will restore access to >50 km of the Klamath River mainstem for coho salmon, but mainstem habitat may not be suitable for rearing juvenile coho salmon. Instead, small tributaries may provide most rearing habitat for reestablishing coho salmon. We used four approaches to evaluate six Klamath River tributaries above existing dams to assess their potential to support juvenile coho salmon: (1) We measured summer temperature regimes and evaluated thermal suitability. (2) We applied an Intrinsic Potential (IP) model to evaluate large‐scale geomorphological constraints on coho salmon habitat. (3) We used the Habitat Limiting Factors Model (HLFM) to estimate rearing capacity for juveniles given current habitat conditions. (4) We developed an occupancy model using data from reference tributaries to predict coho salmon rearing distribution. All six streams had summer temperatures cooler than the mainstem Klamath River. However, five of the streams have barriers that will restrict coho salmon to within 5 km of the confluence with the Klamath River and two were disconnected mid‐summer. Despite these constraints, the tributaries will likely produce coho salmon. Most streams had high IP in their lower reaches, the HLFM model estimated a total capacity of 105,000 juvenile coho salmon, and the occupancy model predicted juvenile coho salmon will rear throughout the accessible reaches. Protection and habitat enhancement for these tributaries will be important for coho salmon reestablishment post‐dam removal.
Article
Full-text available
Context - Disentangling the effect of environment and biological interaction on community composition with observational data, within the environmental filtering framework, is challenging because the two processes produce non independent results. Objectives - Adopting community N-mixture models with symmetric interactions, we aimed at estimating differential effects of landscape structure and biotic interactions on the local abundance of a Mediterranean reptile community including four lizards (Lacerta bilineata; Podarcis siculus; P. muralis; Chalcides chalcides) and two snakes (Hierophis viridiflavus; Natrix Helvetica). Methods - We sampled reptiles for three consecutive years (2019–2021; 4 surveys/year) on 52 linear transects on a Mediterranean coastal landscape. We analyzed count data by means of a multi-species N-mixture model with symmetric interactions. Interactions within pair of species were estimated from the residual correlation of their realized abundances, after accounting for four landscape features: landscape heterogeneity calculated from land cover data, edge density of woody vegetation patches, tree cover density, net primary productivity. Results - Most species displayed very low detection probability (p ~ 0.10). All species responded with different intensity and sensitivity to landscape predictors. Two biological interactions resulted significant: L. bilineata and P. siculus showed a positive interaction, while P. muralis and C. chalcides displayed a negative interaction. Conclusions - Using community N-mixture models we demonstrated that, also with observational data obtained from a realized community, partitioning the filtering process of the landscape from the one of biotic interactions is possible.
Technical Report
Full-text available
he kuaka / Whenua Hou diving petrel (Pelecanoides whenuahouensis) is a taoka (treasure) to Kāi Tahu, and the wider community. With only ~210 adults remaining it is one of the rarest birds of Aotearoa New Zealand and is listed as Threatened – Nationally Critical in Aotearoa and as Critically Endangered globally. Recent research suggested that, despite the removal of invasive predators from its sole remaining colony on Codfish Island / Whenua Hou, the species has not recovered, and ongoing complex pressures continue to put the species at risk of extinction. To identify the best conservation management options for kuaka recovery, the New Zealand Department of Conservation (DOC) initiated a structured decision-making (SDM) process. During this SDM process, a rōpū (working group; consisting of representatives of the Papatipu Rūnaka, Te Rūnanga o Ngāi Tahu, DOC, Fisheries New Zealand, the fishing industry, and Environment Southland) was formed. This rōpū articulated seven values (objectives) fundamental to kuaka recovery and identified 11 potential conservation strategies (alternatives). Fundamental objectives included: maximising the number of kuaka and kuaka populations, Kāi Tahu to express rangatiratanga (sovereignty) and exercise kaitiakitanga (guardianship), minimising ecosystem impacts, maximising public appreciation, minimising costs to conservation management, and minimising impacts on fisheries. Potential alternatives included: status quo, increasing best practice mitigation of vessel-based light pollution, implementing light curfews, improving plant control, initiating competition management, and intra- or inter-island conservation translocations. The consequences for each alternative across each fundamental objective were then predicted using a variety of modelling techniques and expert elicitations, while explicitly accounting for uncertainty. Once consequences were predicted, the rōpū was able to navigate this decision-landscape explicitly and transparently using a variety of qualitative and quantitative decision-analytical tools to identify the preferred option for kuaka conservation. The kuaka rōpū identified that a combination of improved mitigation of vessel-based light pollution, increased weed and native plant control, competition managment, and inter-island translocations provided the best outcome for kuaka recovery across the multiple objectives. These components should be subjected to a stepwise implementation with the different representatives of the rōpū once adequate funding has been secured. The kuaka SDM process provided an inclusive environment among participants and facilitated a rational and transparent recommendation for the recovery of this taoka in the face of challenges, common for marine species recovery, including competing objectives, differing value judgements, and uncertainty.
Article
A trap-happy response can produce serious underestimates of Ni in the Jolly-Seber model, whereas a trap-shy response can produce substantial positive bias. Relative biases are most severe when the studied population exhibits substantial turnover between sampling periods.-from Authors