Uncovering ecological state dynamics with hidden Markov models

Ecological systems can often be characterised by changes among a finite set of underlying states pertaining to individuals, populations, communities or entire ecosystems through time. Owing to the inherent difficulty of empirical field studies, ecological state dynamics operating at any level of this hierarchy can often be unobservable or 'hidden'. Ecologists must therefore often contend with incomplete or indirect observations that are somehow related to these underlying processes. By formally disentangling state and observation processes based on simple yet powerful mathematical properties that can be used to describe many ecological phenomena, hidden Markov models (HMMs) can facilitate inferences about complex system state dynamics that might otherwise be intractable. However, HMMs have only recently begun to gain traction within the broader ecological community. We provide a gentle introduction to HMMs, establish some common terminology, review the immense scope of HMMs for applied ecological research and provide a tutorial on implementation and interpretation. By illustrating how practitioners can use a simple conceptual template to customise HMMs for their specific systems of interest, revealing methodological links between existing applications, and highlighting some practical considerations and limitations of these approaches, our goal is to help establish HMMs as a fundamental inferential tool for ecologists.

Generalized model-based solutions to false-positive error in species detection/nondetection data

Detection/nondetection data are widely collected by ecologists interested in estimating species distributions, abundances, and phenology, and are often imperfect. Recent model development has focused on accounting for both false-positive and false-negative errors given evidence that misclassification is common across many sampling protocols. To date, however, model-based solutions to false-positive error have largely addressed occupancy estimation. We describe a generalized model structure that allows investigators to account for false-positive error in detection/nondetection data across a broad range of ecological parameters and model classes, and demonstrate that previously developed model-based solutions are special cases of the generalized model. Simulation results demonstrate that estimators for abundance and migratory arrival time ignoring false-positive error exhibit severe (20-70%) relative bias even when only 5-10% of detections are false positives. Bias increased when false-positive detections were more likely to occur at sites or within occasions in which true positive detections were unlikely to occur. Models accounting for false-positive error following the site-confirmation or observation-confirmation designs generally reduced bias substantially, even when few detections were confirmed as true or false positives or when the process model for false-positive error was misspecified. Results from an empirical example focusing on gray fox (Urocyon cinereoargenteus) abundance in Wisconsin, USA reinforce concerns that biases induced by false-positive error can also distort spatial predictions often used to guide decision making. Model sensitivity to false-positive error extends well beyond occupancy estimation, but encouragingly, model-based solutions developed for occupancy estimators are generalizable and effective across a range of models widely used in ecological research.

Identifying error and accurately interpreting environmental DNA metabarcoding results: A case study to detect vertebrates at arid zone waterholes

Environmental DNA (eDNA) metabarcoding surveys enable rapid, noninvasive identification of taxa from trace samples with wide-ranging applications from characterizing local biodiversity to identifying food-web interactions. However, the technique is prone to error from two major sources: (a) contamination through foreign DNA entering the workflow, and (b) misidentification of DNA within the workflow. Both types of error have the potential to obscure true taxon presence or to increase taxonomic richness by incorrectly identifying taxa as present at sample sites, but multiple error sources can remain unaccounted for in metabarcoding studies. Here, we use data from an eDNA metabarcoding study designed to detect vertebrate species at waterholes in Australia's arid zone to illustrate where and how in the workflow errors can arise, and how to mitigate those errors. We detected the DNA of 36 taxa spanning 34 families, 19 orders and five vertebrate classes in water samples from waterholes, demonstrating the potential for eDNA metabarcoding surveys to provide rapid, noninvasive detection in remote locations, and to widely sample taxonomic diversity from aquatic through to terrestrial taxa. However, we initially identified 152 taxa in the samples, meaning there were many false positive detections. We identified the sources of these errors, allowing us to design a stepwise process to detect and remove error, and provide a template to minimize similar errors that are likely to arise in other metabarcoding studies. Our findings suggest eDNA metabarcoding surveys need to be carefully conducted and screened for errors to ensure their accuracy.

Determining multi-species site use outside the protected areas of the Maasai Mara, Kenya, using false positive site-occupancy modelling

Although protected areas are the basis for many conservation efforts they are rarely of an adequate size for the long-term survival of populations of large, wide-roaming mammals. In the Maasai Mara, Kenya, communally owned wildlife conservancies have been developed to expand the area available for wildlife. As these continue to develop it is important to ensure that the areas chosen are beneficial to wildlife. Using presence data for cheetahs Acinonyx jubatus, elephants Loxodonta africana, spotted hyaenas Crocuta crocuta, leopards Panthera pardus, lions Panthera leo and wild dogs Lycaon pictus, collected through interviews with 648 people living outside protected areas, we identify key wildlife areas using false positive site-occupancy modelling. The probabilities of site use were first determined per species based on habitat, distance to protected area, human presence and rivers, and these probabilities were then combined to create a map to highlight key wildlife areas. All species, except hyaenas, preferred sites closer to the protected areas but site use varied by species depending on habitat type. All six species avoided human presence. Leopards, elephants, lions and wild dogs preferred sites closer to rivers. The resulting combined map highlights areas that could potentially benefit from conservation efforts, including the expansion of wildlife areas, and areas where human development, such as a newly tarmacked road, could have an impact on wildlife.

Estimating occupancy of Chinese pangolin (Manis pentadactyla) in a protected and non-protected area of Nepal

Chinese pangolin is the world's most heavily trafficked small mammal for luxury food and traditional medicine. Although their populations are declining worldwide, it is difficult to monitor their population status because of its rarity and nocturnal behavior. We used site occupancy (presence/absence) sampling of pangolin sign (i.e., active burrows) in a protected (Gaurishankar Conservation Area) and non-protected area (Ramechhap District) of central Nepal with multiple environmental covariates to understand factors that may influence occupancy of Chinese pangolin. The average Chinese pangolin occupancy and detection probabilities were Ψ ^  ± SE = 0.77 ± 0.08; p ^  ± SE = 0.27 ± 0.05, respectively. The detection probabilities of Chinese pangolin were higher in PA ( p ^  ± SE = 0.33 ± 0.03) than compared to non-PA ( p ^  ± SE = 0.25 ± 0.04). The most important covariates for Chinese pangolin detectability were red soil (97%), food source (97.6%), distance to road (97.9%), and protected area (97%) and with respect to occupancy was elevation (97.9%). We recommended use of remote cameras and potentially GPS collar surveys to further investigate habitat use and site occupancy at regular intervals to provide more reliable conservation assessments.

A Bayesian hierarchical approach to quantifying stakeholder attitudes toward conservation in the presence of reporting error

Stakeholder support is vital for achieving conservation success, yet there are few reliable mechanisms to monitor stakeholder attitudes toward conservation. Approaches used to assess attitudes rarely account for bias arising from reporting error, which can lead to falsely reporting a positive attitude toward conservation (false-positive error) or not reporting a positive attitude when the respondent has a positive attitude toward conservation (false-negative error). Borrowing from developments in applied conservation science, we used a Bayesian hierarchical model to quantify stakeholder attitudes as the probability of having a positive attitude toward wildlife notionally (or in abstract terms) and at localized scales while accounting for reporting error. We compared estimates from our model, Likert scores, and naïve estimates (i.e., proportion of respondents reporting a positive attitude in at least 1 question that was only susceptible to false-negative error) with true stakeholder attitudes through simulations. We then applied the model in a survey of tea estate staff on their attitudes toward Asian elephants (Elephas maximus) in the Kaziranga-Karbi Anglong landscape of northeast India. In simulations, Bayesian model estimates of stakeholder attitudes toward wildlife were less biased than naïve estimates or Likert scores. After accounting for reporting errors, we estimated the probability of having a positive attitude toward elephants notionally as 0.85 in the Kaziranga landscape, whereas the proportion of respondents who had positive attitudes toward elephants at a localized scale was 0.50. In comparison, without accounting for reporting errors, naïve estimates of proportions of respondents with positive attitudes toward elephants were 0.69 and 0.23 notionally and at local scales, respectively. False (positive and negative) reporting probabilities were consistently not 0 (0.22-0.68). Regular and reliable assessment of stakeholder attitudes-combined with inference on drivers of positive attitudes-can help assess the success of initiatives aimed at facilitating human behavioral change and inform conservation decision making.

Dependent double-observer method reduces false-positive errors in auditory avian survey data

Bias introduced by detection errors is a well-documented issue for abundance and occupancy estimates of wildlife. Detection errors bias estimates of detection and abundance or occupancy in positive and negative directions, which can produce misleading results. There have been considerable design- and model-based methods to address false-negative errors, or missed detections. However, false-positive errors, or detections of individuals that are absent but counted as present because of misidentifications or double counts, are often assumed to not occur in ecological studies. The dependent double-observer survey method is a design-based approach speculated to reduce false positives because observations have the ability to be confirmed by two observers. However, whether this method reduces false positives compared to single-observer methods has not been empirically tested. We used prairie songbirds as a model system to test if a dependent double-observer method reduced false positives compared to a single-observer method. We used vocalizations of ten species to create auditory simulations and used naive and expert observers to survey these simulations using single-observer and dependent double-observer methods. False-positive rates were significantly lower using the dependent double-observer survey method in both observer groups. Expert observers reported a 3.2% false-positive rate in dependent double-observer surveys and a 9.5% false-positive rate in single-observer surveys, while naive observers reported a 39.1% false-positive rate in dependent double-observer surveys and a 49.1% false-positive rate in single-observer surveys. Misidentification errors arose in all survey scenarios and almost all species combinations. However, expert observers using the dependent double-observer method performed significantly better than other survey scenarios. Given the use of double-observer methods and the accumulating evidence that false positives occur in many survey methods across different taxa, this study is an important step forward in acknowledging and addressing false positives.

Native bees of high Andes of Central Chile (Hymenoptera: Apoidea): biodiversity, phenology and the description of a new species of Xeromelissa Cockerell (Hymenoptera: Colletidae: Xeromelissinae)

High-altitude ecosystems are found in mountain chains and plateaus worldwide. These areas tend to be underrepresented in insect biodiversity assessments because of the challenges related to systematic survey at these elevations, such as extreme climatic and geographic conditions. Nonetheless, high-altitude ecosystems are of paramount importance because they have been seen to be species pumps for other geographic areas, such as adjacent locations, functioning as buffers for population declines. Moreover, these ecosystems and their biodiversity have been proposed to be fast-responding indicators of the impacts caused by global climate change. Bees have been highlighted among the insect groups that have been affected by these problems. This work used bees as a proxy to demonstrate and reinforce the importance of systematic surveys of high-altitude ecosystems. Here, field collections were undertaken and an updated review was conducted for the native bee biodiversity of the high-altitude ecosystem found at the Andes system of central Chile, including the phenological trends of these insects during the flowering season. Of the 58 species that have been described for this location, we were able to confirm the occurrence of 46 of these species as a result of our sampling. In addition, thanks to these recent collections, a new species of Xeromelissa Cockerell is described in the present work. These findings highlight the need for further high-altitude insect surveys of this biome, which include both temporal and spatial complexity in their design, to allow for accurate assessment of bee species diversity and compositional changes in these mountain regions.

Improved detection of house infestations with triatomines using sticky traps: a paired-comparison trial in the Argentine Chaco

Background

We conducted a matched-pairs trial of three methods for detecting house infestation with triatominae bugs in a well-defined endemic rural area in the Argentine Chaco.

Methods

The three methods included a simple double-sided adhesive tape (ST) installed near host resting sites; timed-manual collections with a dislodging aerosol (TMC, the reference method used by vector control programmes), and householders’ bug notifications (HN). Triatomine infestations were evaluated in 103 sites of 54 houses, including domiciles, kitchens and storerooms.

Results

In domiciles where Triatoma infestans was collected, sensitivity of each single method decreased from 79% by ST and 77% by HN, to 57% by TMC, and increased to 92% when ST was combined with HN. In peridomestic kitchens and storerooms, TMC was relatively as sensitive as ST and significantly more sensitive than HN. On average, the number of bugs recovered by ST was 0.94 times that collected by TMC. The ST mainly collected early-instar nymphs whereas TMC yielded late (larger) stages. Triatomines caught by ST had significantly lower mean weight-to-length ratios and lower blood-feeding rates than those caught by TMC, suggesting the ST intercepted and trapped vectors seeking a blood meal host.

Conclusions

The ST may effectively replace TMC for detecting T. infestans in domiciles, and is especially apt for early detection of low-density domestic infestations in the frame of community-based surveillance or elimination programmes; decision making on whether an area should be targeted for full-coverage insecticide spraying, and to corroborate that extant conditions are compatible with the interruption of vector-borne transmission.

Megafire effects on spotted owls: elucidation of a growing threat and a response to Hanson et al. (2018)

The extent to which wildfire adversely affects spotted owls (Strix occidentalis) is a key consideration for ecosystem restoration efforts in seasonally dry forests of the western United States. Recently, Jones et al. (2016) demonstrated that the 2014 King Fire (a “megafire”) adversely affected a population of individually-marked California spotted owls (S. o. occidentalis) monitored as part of a long-term demographic study in the Sierra Nevada, California, USA because territory occupancy declined substantially at territories burned at high-severity and GPS-tagged spotted owls avoided large patches of high-severity fire. Hanson et al. (2018) attempted to reassess changes in territory occupancy of the Jones et al. (2016) study population and claimed that occupancy declined as a result of post-fire salvage logging not fire per se and suggested that the avoidance of GPS-marked owls from areas that burned at high-severity was due to post-fire logging rather than a response to high-severity fire. Here, we demonstrate that Hanson et al. (2018) used erroneous data, inadequate statistical analyses and faulty inferences to reach their conclusion that the King Fire did not affect spotted owls and, more broadly, that large, high-severity fires do not pose risks to spotted owls in western North American dry forest ecosystems. We also provide further evidence indicating that the King Fire exerted a clear and significant negative effect on our marked study population of spotted owls. Collectively, the additional evidence presented here and in Jones et al. (2016) suggests that large, high-severity fires can pose a threat to spotted owls and that restoration of natural low- to mixed-severity frequent fire regimes would likely benefit both old-forest species and dry forest ecosystems in this era of climate change. Meeting these dual objectives of species conservation and forest restoration will be complex but it is made more challenging by faulty science that does not acknowledge the full range of wildfire effects on spotted owls.

Traditional trapping methods outperform eDNA sampling for introduced semi-aquatic snakes

Given limited resources for managing invasive species, traditional survey methods may not be feasible to implement at a regional scale. Environmental DNA (eDNA) sampling has proven to be an effective method for detecting some invasive species, but comparisons between the detection probability of eDNA and traditional survey methods using modern occupancy modeling methods are rare. We developed a qPCR assay to detect two species of watersnake (Nerodia fasciata and Nerodia sipedon) introduced to California, USA, and we compared the efficacy of eDNA and aquatic trapping. We tested 3–9 water samples each from 30 sites near the known range of N. fasciata, and 61 sites near the known range of N. sipedon. We also deployed aquatic funnel traps at a subset of sites for each species. We detected N. fasciata eDNA in three of nine water samples from just one site, but captured N. fasciata in traps at three of ten sites. We detected N. sipedon eDNA in five of six water samples from one site, which was also the only site of nine at which this species was captured in traps. Traditional trapping surveys had a higher probability of detecting watersnakes than eDNA surveys, and both survey methods had higher detection probability for N. sipedon than N. fasciata. Occupancy models that integrated both trapping and eDNA surveys estimated that 5 sites (95% Credible Interval: 4–10) of 91 were occupied by watersnakes (both species combined), although snakes were only detected at four sites (three for N. fasciata, one for N. sipedon). Our study shows that despite the many successes of eDNA surveys, traditional sampling methods can have higher detection probability for some species. We recommend those tasked with managing species invasions explicitly compare eDNA and traditional survey methods in an occupancy framework to inform their choice of the best method for detecting nascent populations.

Invasive Plant Species Establishment and Range Dynamics in Sri Lanka under Climate Change

Plant invasion has been widely recognized as an agent of global change that has the potential to have severe impacts under climate change. The challenges posed by invasive alien plant species (IAPS) on biodiversity and ecosystem stability is growing and not adequately studied, especially in developing countries. Defining climate suitability for multiple invasive plants establishment is important for early and strategic interventions to control and manage plant invasions. We modeled priority IAPS in Sri Lanka to identify the areas of greatest climatic suitability for their establishment and observed how these areas could be altered under projected climate change. We used Maximum Entropy method to model 14 nationally significant IAPS under representative concentration pathways 4.5 and 8.5 for 2050 and 2070. The combined climate suitability map produced by summing up climatic suitability of 14 IAPS was further classified into five classes in ArcMap as very high, high, moderate, low, and very low. South and west parts of Sri Lanka are projected to have potentially higher climatic suitability for a larger number of IAPS. We observed suitable area changes (gains and losses) in all five classes of which two were significant enough to make an overall negative impact i.e., (i) contraction of the very low class and (ii) expansion of the moderate class. Both these changes trigger the potential risk from IAPS in Sri Lanka in the future.

Evaluating Cost-Effective Methods for Rapid and Repeatable National Scale Detection and Mapping of Invasive Species Spread

Invasive species can spread rapidly at local and national scales, creating significant environmental and economic impacts. A central problem in mitigation efforts is identifying methods that can rapidly detect invasive species in a cost-effective and repeatable manner. This challenge is particularly acute for species that can spread over large areas (>1 million km²). Wild pigs (Sus scrofa) are one of the most prolific invasive mammals on Earth and cause extensive damage to agricultural crops, native ecosystems, and livestock, and are reservoirs of disease. They have spread from their native range in Eurasia and North Africa into large areas of Australia, Africa, South America, and North America. We show that the range of invasive wild pigs has increased exponentially in Canada over the last 27 years following initial and ongoing releases and escapes from domestic wild boar farms. The cumulative range of wild pigs across Canada is 777,783 km², with the majority of wild pig distribution occurring in the Prairie Provinces. We evaluate eight different data collection and evaluation/validation methods for mapping invasive species over large areas, and assess their benefits and limitations. Our findings effectively map the spread of a highly invasive large mammal and demonstrate that management efforts should ideally rely on a set of complementary independent monitoring methods. Mapping and evaluating resulting species occurrences provide baseline maps against which future changes can be rapidly evaluated.

A practical guide for combining data to model species distributions

Understanding and accurately modeling species distributions lies at the heart of many problems in ecology, evolution, and conservation. Multiple sources of data are increasingly available for modeling species distributions, such as data from citizen science programs, atlases, museums, and planned surveys. Yet reliably combining data sources can be challenging because data sources can vary considerably in their design, gradients covered, and potential sampling biases. We review, synthesize, and illustrate recent developments in combining multiple sources of data for species distribution modeling. We identify five ways in which multiple sources of data are typically combined for modeling species distributions. These approaches vary in their ability to accommodate sampling design, bias, and uncertainty when quantifying environmental relationships in species distribution models. Many of the challenges for combining data are solved through the prudent use of integrated species distribution models: models that simultaneously combine different data sources on species locations to quantify environmental relationships for explaining species distribution. We illustrate these approaches using planned survey data on 24 species of birds coupled with opportunistically collected eBird data in the southeastern United States. This example illustrates some of the benefits of data integration, such as increased precision in environmental relationships, greater predictive accuracy, and accounting for sample bias. Yet it also illustrates challenges of combining data sources with vastly different sampling methodologies and amounts of data. We provide one solution to this challenge through the use of weighted joint likelihoods. Weighted joint likelihoods provide a means to emphasize data sources based on different criteria (e.g., sample size), and we find that weighting improves predictions for all species considered. We conclude by providing practical guidance on combining multiple sources of data for modeling species distributions.

Examining human-carnivore interactions using a socio-ecological framework: sympatric wild canids in India as a case study

Many carnivores inhabit human-dominated landscapes outside protected reserves. Spatially explicit assessments of carnivore distributions and livestock depredation patterns in human-use landscapes are crucial for minimizing negative interactions and fostering coexistence between people and predators. India harbours 23% of the world's carnivore species that share space with 1.3 billion people in approximately 2.3% of the global land area. We examined carnivore distributions and human-carnivore interactions in a multi-use forest landscape in central India. We focused on five sympatric carnivore species: Indian grey wolf Canis lupus pallipes, dhole Cuon alpinus, Indian jackal Canis aureus indicus, Indian fox Vulpes bengalensis and striped hyena Hyaena hyaena. Carnivore occupancy ranged from 12% for dholes to 86% for jackals, mostly influenced by forests, open scrublands and terrain ruggedness. Livestock/poultry depredation probability in the landscape ranged from 21% for dholes to greater than 95% for jackals, influenced by land cover and livestock- or poultry-holding. The five species also showed high spatial overlap with free-ranging dogs, suggesting potential competitive interactions and disease risks, with consequences for human health and safety. Our study provides insights on factors that facilitate and impede co-occurrence between people and predators. Spatial prioritization of carnivore-rich areas and conflict-prone locations could facilitate human-carnivore coexistence in shared habitats. Our framework is ideally suited for making socio-ecological assessments of human-carnivore interactions in other multi-use landscapes and regions, worldwide.

Bats in a changing landscape: Linking occupancy and traits of a diverse montane bat community to fire regime

Wildfires are increasing in incidence and severity across coniferous forests of the western United States, leading to changes in forest structure and wildlife habitats. Knowledge of how species respond to fire-driven habitat changes in these landscapes is limited and generally disconnected from our understanding of adaptations that underpin responses to fire.We aimed to investigate drivers of occupancy of a diverse bat community in a fire-altered landscape, while identifying functional traits that underpinned these relationships.We recorded bats acoustically at 83 sites (n = 249 recording nights) across the Plumas National Forest in the northern Sierra Nevada over 3 summers (2015-2017). We investigated relationships between fire regime, physiographic variables, forest structure and probability of bat occupancy for nine frequently detected species. We used fourth-corner regression and RLQ analysis to identify ecomorphological traits driving species-environment relationships across 17 bat species. Traits included body mass; call frequency, bandwidth, and duration; and foraging strategy based on vegetation structure (open, edge, or clutter).Relationships between bat traits and fire regime were underpinned by adaptations to diverse forest structure. Bats with traits adapting them to foraging in open habitats, including emitting longer duration and narrow bandwidth calls, were associated with higher severity and more frequent fires, whereas bats with traits consistent with clutter tolerance were negatively associated with fire frequency and burn severity. Relationships between edge-adapted bat species and fire were variable and may be influenced by prey preference or habitat configuration at a landscape scale.Predicted increases in fire frequency and severity in western US coniferous forests are likely to shift dominance in the bat community to open-adapted species and those able to exploit postfire resource pulses (aquatic insects, beetles, and snags). Managing for pyrodiversity within the western United States is likely important for maintaining bat community diversity, as well as diversity of other biotic communities.

Megafire effects on spotted owls: elucidation of a growing threat and a response to Hanson et al. (2018)

The extent to which wildfire adversely affects spotted owls (Strixoccidentalis) is a key consideration for ecosystem restoration efforts in seasonally dry forests of the western United States. Recently, Jones et al. (2016) demonstrated that the 2014 King Fire (a “megafire”) adversely affected a population of individually-marked California spotted owls (S.o.occidentalis) monitored as part of a long-term demographic study in the Sierra Nevada, California, USA because territory occupancy declined substantially at territories burned at high-severity and GPS-tagged spotted owls avoided large patches of high-severity fire. Hanson et al. (2018) attempted to reassess changes in territory occupancy of the Jones et al. (2016) study population and claimed that occupancy declined as a result of post-fire salvage logging not fire per se and suggested that the avoidance of GPS-marked owls from areas that burned at high-severity was due to post-fire logging rather than a response to high-severity fire. Here, we demonstrate that Hanson et al. (2018) used erroneous data, inadequate statistical analyses and faulty inferences to reach their conclusion that the King Fire did not affect spotted owls and, more broadly, that large, high-severity fires do not pose risks to spotted owls in western North American dry forest ecosystems. We also provide further evidence indicating that the King Fire exerted a clear and significant negative effect on our marked study population of spotted owls. Collectively, the additional evidence presented here and in Jones et al. (2016) suggests that large, high-severity fires can pose a threat to spotted owls and that restoration of natural low- to mixed-severity frequent fire regimes would likely benefit both old-forest species and dry forest ecosystems in this era of climate change. Meeting these dual objectives of species conservation and forest restoration will be complex but it is made more challenging by faulty science that does not acknowledge the full range of wildfire effects on spotted owls.

Dynamic wildlife occupancy models using automated acoustic monitoring data

Automated acoustic monitoring of wildlife has been used to characterize populations of sound-producing species across large spatial scales. However, false negatives and false positives produced by automated detection systems can compromise the utility of these data for researchers and land managers, particularly for research programs endeavoring to describe colonization and extinction dynamics that inform land use decision-making. To investigate the suitability of automated acoustic monitoring for dynamic occurrence models, we simulated underlying occurrence dynamics, calling patterns, and the automated acoustic detection process for a hypothetical species under a range of scenarios. We investigated an automated species detection aggregation method that considered a suite of options for creating encounter histories. From these encounter histories, we generated parameter estimates and computed bias for occurrence, colonization, and extinction rates using a dynamic occupancy modeling framework that accounts for false positives via small amounts of manual confirmation. We were able to achieve relatively unbiased estimates for all three state parameters under all scenarios, even when the automated detection system was simulated to be poor, given particular encounter history aggregation choices. However, some encounter history aggregation choices resulted in unreliable estimates; we provide caveats for avoiding these scenarios. Given specific choices during the detection aggregation process, automated acoustic monitoring data may provide an effective means for tracking species occurrence, colonization, and extinction patterns through time, with the potential to inform adaptive management at multiple spatial scales.

A hierarchical Bayesian approach for handling missing classification data

Ecologists use classifications of individuals in categories to understand composition of populations and communities. These categories might be defined by demographics, functional traits, or species. Assignment of categories is often imperfect, but frequently treated as observations without error. When individuals are observed but not classified, these "partial" observations must be modified to include the missing data mechanism to avoid spurious inference.We developed two hierarchical Bayesian models to overcome the assumption of perfect assignment to mutually exclusive categories in the multinomial distribution of categorical counts, when classifications are missing. These models incorporate auxiliary information to adjust the posterior distributions of the proportions of membership in categories. In one model, we use an empirical Bayes approach, where a subset of data from one year serves as a prior for the missing data the next. In the other approach, we use a small random sample of data within a year to inform the distribution of the missing data.We performed a simulation to show the bias that occurs when partial observations were ignored and demonstrated the altered inference for the estimation of demographic ratios. We applied our models to demographic classifications of elk (Cervus elaphus nelsoni) to demonstrate improved inference for the proportions of sex and stage classes.We developed multiple modeling approaches using a generalizable nested multinomial structure to account for partially observed data that were missing not at random for classification counts. Accounting for classification uncertainty is important to accurately understand the composition of populations and communities in ecological studies.

Making inference with messy (citizen science) data: when are data accurate enough and how can they be improved?

Measurement or observation error is common in ecological data: as citizen scientists and automated algorithms play larger roles processing growing volumes of data to address problems at large scales, concerns about data quality and strategies for improving it have received greater focus. However, practical guidance pertaining to fundamental data quality questions for data users or managers-how accurate do data need to be and what is the best or most efficient way to improve it?-remains limited. We present a generalizable framework for evaluating data quality and identifying remediation practices, and demonstrate the framework using trail camera images classified using crowdsourcing to determine acceptable rates of misclassification and identify optimal remediation strategies for analysis using occupancy models. We used expert validation to estimate baseline classification accuracy and simulation to determine the sensitivity of two occupancy estimators (standard and false-positive extensions) to different empirical misclassification rates. We used regression techniques to identify important predictors of misclassification and prioritize remediation strategies. More than 93% of images were accurately classified, but simulation results suggested that most species were not identified accurately enough to permit distribution estimation at our predefined threshold for accuracy (<5% absolute bias). A model developed to screen incorrect classifications predicted misclassified images with >97% accuracy: enough to meet our accuracy threshold. Occupancy models that accounted for false-positive error provided even more accurate inference even at high rates of misclassification (30%). As simulation suggested occupancy models were less sensitive to additional false-negative error, screening models or fitting occupancy models accounting for false-positive error emerged as efficient data remediation solutions. Combining simulation-based sensitivity analysis with empirical estimation of baseline error and its variability allows users and managers of potentially error-prone data to identify and fix problematic data more efficiently. It may be particularly helpful for "big data" efforts dependent upon citizen scientists or automated classification algorithms with many downstream users, but given the ubiquity of observation or measurement error, even conventional studies may benefit from focusing more attention upon data quality.

Use of ambiguous detections to improve estimates from species distribution models

As large carnivores recover throughout Europe, their distribution needs to be studied to determine their conservation status and assess the potential for human-carnivore conflicts. However, efficient monitoring of many large carnivore species is challenging due to their rarity, elusive behavior, and large home ranges. Their monitoring can include opportunistic sightings from citizens in addition to designed surveys. Two types of detection errors may occur in such monitoring schemes: false negatives and false positives. False-negative detections can be accounted for in species distribution models (SDMs) that deal with imperfect detection. False-positive detections, due to species misidentification, have rarely been accounted for in SDMs. Generally, researchers use ad hoc data-filtering methods to discard ambiguous observations prior to analysis. These practices may discard valuable ecological information on the distribution of a species. We investigated the costs and benefits of including data types that may include false positives rather than discarding them for SDMs of large carnivores. We used a dynamic occupancy model that simultaneously accounts for false negatives and positives to jointly analyze data that included both unambiguous detections and ambiguous detections. We used simulations to compare the performances of our model with a model fitted on unambiguous data only. We tested the 2 models in 4 scenarios in which parameters that control false-positive detections and true detections varied. We applied our model to data from the monitoring of the Eurasian lynx (Lynx lynx) in the European Alps. The addition of ambiguous detections increased the precision of parameter estimates. For the Eurasian lynx, incorporating ambiguous detections produced more precise estimates of the ecological parameters and revealed additional occupied sites in areas where the species is likely expanding. Overall, we found that ambiguous data should be considered when studying the distribution of large carnivores through the use of dynamic occupancy models that account for misidentification.

Verifying bilby presence and the systematic sampling of wild populations using sign-based protocols – with notes on aerial and ground survey techniques and asserting absence

The recognition of sign such as tracks, scats, diggings or burrows is widely used to detect rare or elusive species. We describe the type of sign that can be used to confirm the presence of the greater bilby (Macrotis lagotis) in comparison with sign that should be used only to flag potential presence. Clear track imprints of the front and hind feet, diggings at the base of plants to extract root-dwelling larvae, and scats commonly found at diggings can be used individually, or in combination, to verify presence, whereas track gait pattern, diggings in the open, and burrows should be used to flag potential bilby activity but not to verify presence. A protocol to assess potential activity and verify bilby presence is provided. We provide advice on the application of a plot-based technique to systematically search for sign and produce data for the estimation of regional occupancy. Digging and burrow activity can be readily detected from the air but systematic ground-based assessment to determine the rate of false-presence and false-absence needs to accompany aerial survey. The approach to estimate survey effort to assert bilby absence is also described.

Disease-structured N-mixture models: A practical guide to model disease dynamics using count data

Obtaining inferences on disease dynamics (e.g., host population size, pathogen prevalence, transmission rate, host survival probability) typically requires marking and tracking individuals over time. While multistate mark-recapture models can produce high-quality inference, these techniques are difficult to employ at large spatial and long temporal scales or in small remnant host populations decimated by virulent pathogens, where low recapture rates may preclude the use of mark-recapture techniques. Recently developed N-mixture models offer a statistical framework for estimating wildlife disease dynamics from count data. N-mixture models are a type of state-space model in which observation error is attributed to failing to detect some individuals when they are present (i.e., false negatives). The analysis approach uses repeated surveys of sites over a period of population closure to estimate detection probability. We review the challenges of modeling disease dynamics and describe how N-mixture models can be used to estimate common metrics, including pathogen prevalence, transmission, and recovery rates while accounting for imperfect host and pathogen detection. We also offer a perspective on future research directions at the intersection of quantitative and disease ecology, including the estimation of false positives in pathogen presence, spatially explicit disease-structured N-mixture models, and the integration of other data types with count data to inform disease dynamics. Managers rely on accurate and precise estimates of disease dynamics to develop strategies to mitigate pathogen impacts on host populations. At a time when pathogens pose one of the greatest threats to biodiversity, statistical methods that lead to robust inferences on host populations are critically needed for rapid, rather than incremental, assessments of the impacts of emerging infectious diseases.

ESTIMATING OCCURRENCE, PREVALENCE, AND DETECTION OF AMPHIBIAN PATHOGENS: INSIGHTS FROM OCCUPANCY MODELS

Understanding the distribution of pathogens across landscapes and their prevalence within host populations is a common aim of wildlife managers. Despite the need for unbiased estimates of pathogen occurrence and prevalence for planning effective management interventions, many researchers fail to account for imperfect pathogen detection. Instead raw data are often reported, which may lead to ineffective, or even detrimental, management actions. We illustrate the utility of occupancy models for generating unbiased estimates of disease parameters by 1) providing a written tutorial describing how to fit these models in Program PRESENCE and 2) presenting a case study with the pathogen ranavirus. We analyzed ranavirus detection data from a wildlife refuge (Maryland, US) using occupancy modeling, which yields unbiased estimates of pathogen occurrence and prevalence. We found ranavirus prevalence was underestimated by up to 30% if imperfect pathogen detection was ignored. The unbiased estimate of ranavirus prevalence in larval wood frog (Lithobates sylvaticus; 0.73) populations was higher than in larval spotted salamander (Ambystoma maculatum; 0.56) populations. In addition, the odds of detecting ranavirus in tail samples were 6.7 times higher than detecting ranavirus in liver samples. Therefore, tail samples presented a nonlethal sampling method for ranavirus that may be able to detect early (nonsystemic) infections.