No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Detecting Long-Term Population Trends for an Elusive Reptile Species
Abstract
Many reptile species are in decline and turtles are especially susceptible. In Massachusetts, eastern box turtle (Terrapene carolina carolina) population densities are critically low, and they are listed as a Species of Special Concern. To aid in the conservation of this species, we developed a statewide population monitoring program to track large-scale population trends. We used GENPRES3 to identify the most efficient sampling design a priori. Using this design, we performed visual surveys in 2010–2012 and used site occupancy models to evaluate baseline occupancy and abundance data. We surveyed 62 4-ha monitoring plots within early successional and forest edge habitat where box turtles congregate in the spring for foraging, mating, nesting, and thermoregulation. We also used radio-telemetry at 2 survey sites to evaluate assumptions and further assess occupancy rates, detection estimates, and population size. The best fit Royle–Nichols model predicted a probability of box turtle occupancy of 0.81 ± 0.10 (mean ± SE) and a mean probability of detection of 0.29 ± 0.18. Roads and vegetation density were important covariates affecting the probability of occurrence. Survey start time, humidity, and surveyor were important covariates affecting detection probability. A power analysis indicated that we could detect a 10% decline in occupancy between 5-year sampling rounds within 15 years. The proportion of radio-tagged turtles inside the survey plots during surveys was relatively constant at each site (0.44–0.63 and 0.36–0.43), mean detection rate was 0.35 ± 0.10, and the total estimated population size of the 2 survey plots (8 ha total) was 13.31 ± 1.53. Our results can be used to track the status of this rare species as well as guide conservation actions and evaluate the effectiveness of site-specific and statewide management plans. Our approach and design can serve as a model for other states developing monitoring programs for the eastern box turtle and other similar, rare and difficult to detect species. Published 2015. This article is a U.S. Government work and is in the public domain in the USA.
... Currently, T. c. carolina is listed as a species of conservation priority in more than half of the states in which it occurs (Erb et al. 2015, Kiester and. However, standardized assessments of population trends and characteristics at multiple study sites over broad geographic areas and diverse environmental contexts are lacking (but see Erb et al. 2015). Due to the cryptic nature of T. c. carolina and their perceived historical commonness, few states have official monitoring programs designed to track population trends that could quantify population declines, identify threats, or assess population responses to conservation and management actions. ...
... From these studies, it is clear that T. carolina populations are in decline throughout much of their geographic range. However, prior to this study, only limited systematic monitoring programs have been conducted on multiple populations to detect longterm and broad-scale population trends in T. c. carolina, with the exception of a state-wide project in Massachusetts in the initial stages of monitoring (Erb et al. 2015). Contrary to our hypotheses informed by declining trends in other areas, we found no evidence of T. c. carolina population declines over as many as 10 yr of continuous monitoring at numerous sites, with population densities (and thus population sizes within stable study areas) fluctuating from year to year but remaining largely temporally stable throughout the study period. ...
... Finally, encounter probability estimates were highly uncertain at some sites, making it difficult to discern methodological artifacts from actual changes in population size. Terrapene carolina carolina activity is highly dependent on temperature, humidity, rainfall, and other environmental conditions that can change over the course of a day or season and vary between years and thus influence turtle detection during surveys (Erb et al. 2015, Boers et al. 2017, Parlin et al. 2019. Even when conditions are optimal, human observers and wildlife detection dogs vary in their effectiveness of detecting turtles depending on experience level (Erb et al. 2015, Boers et al. 2017. ...
Abstract Turtles are in decline worldwide, and the magnitude and recent acceleration of population declines requires immediate action to inform conservation and management plans. Long‐term studies of population trends and characteristics covering multiple populations across a range of environmental contexts are needed to guide the most effective management decisions, yet such studies are uncommon. Eastern box turtle (Terrapene carolina carolina) populations have declined and require conservation action throughout much of their range. We assessed sources of variation and temporal trends in T. c. carolina population characteristics from 39 sites spanning four ecoregions in North Carolina using capture–mark–recapture. Surprisingly, there was no evidence of population decline at any site over a ten‐year period (2008–2017). Population densities ranged from 0.2 to 6.0 turtles per hectare, decreasing with more urban development and increasing with more wetland habitats in the surrounding landscape. Populations varied in demographic structure but had similar sex ratios and more adults than immature turtles in each ecoregion even after adjusting for variation in encounter probability among groups and body sizes. Survivorship was similar among life stages and sexes and only weakly related to body size. After adjusting for emigration, annual survival probability ranged from 90.7% to 96.8% for the various demographic groups and body sizes. Growth rates decreased with increasing body size, with rates ranging from approximately 15‐mm carapace length per year in the smallest individuals to near zero in the largest individuals. No aspect of population demographic structure or vital rates was related to surrounding landscape context. Population trends and characteristics were consistent among ecoregions, suggesting no need for region‐specific conservation or management with respect to the population characteristics examined. However, given the high site‐specific variability in nearly all estimated parameters, managers would benefit from targeting local threats such as urban land development and wetland destruction to ensure local population viability. The population characteristics reported here can serve as a useful baseline to compare population trends at these sites as monitoring continues. Our approach could be a useful model for other regions to follow in developing monitoring programs for this and other imperiled turtles.
... For species of conservation concern, such as those listed in the United States or Canada as threatened or endangered at either the state/ province or federal level, changes in the number and size of populations are the primary metrics used for assessing species status and recovery [3,4]. Thus, much research has been devoted to development and assessment of survey protocols and statistical models to improve reliability of population inferences [e.g., [5][6][7][8]. ...
... While our study focused specifically on investigating assumptions of a survey protocol for the wood turtle, many other species monitoring programs use count data from spatially constrained VES protocols to estimate occurrence or abundance, such as for eastern box turtles (Terrapene carolina) in Massachusetts [7], foothill yellow-legged frogs (Rana boylii) in California [54], and Shenandoah salamanders (Plethodon shenandoah) in Virginia [55]. Every species monitoring program operates with limited resources and the associated survey protocols reflect tradeoffs between intensity of sampling at individual sites and amount of spatial coverage across the monitoring area. ...
Reliable population estimates are important for making informed management decisions about wildlife species. Standardized survey protocols have been developed for monitoring population trends of the wood turtle (Glyptemys insculpta), a semi-aquatic freshwater turtle species of conservation concern throughout its distribution in east-central North America. The protocols use repeated active search surveys of defined areas, allowing for estimation of survey-specific detection probability (p) and site-specific abundance. These protocols assume population closure within the survey area during the survey period, which is unlikely to be met as wood turtles are a highly mobile species. Additionally, current protocols use a single-pass design that does not allow for separation of availability (pa) and detectability (pd). If there are systematic influences on pa or pd that are not accounted for in the survey design or data analysis, then resulting abundance estimates could be biased. The objectives of this study were to determine if pa is a random process and if pa and pd are influenced by demographic characteristics. We modified the wood turtle survey protocol used in the upper Midwest to include a double-pass design, allowing us to estimate pa and pd using a robust design capture-recapture model. The modified protocol was implemented at 14 wood turtle monitoring sites in Minnesota and Wisconsin between 2017 and 2022. Our results indicated that pa was non-random and that pd increased with turtle carapace length. Our study suggests that model assumptions for current wood turtle population models may be violated, likely resulting in an overestimation of abundance. We discuss possible protocol and modeling modifications that could result in more accurate wood turtle abundance estimates.
... If inadequate sampling design or effort result in false negatives, populations may not receive the protection and management needed to persist [4,5]. Consequently, determining the appropriate design and effort required to evaluate status and trends of imperiled and cryptic species is a vital first step for surveying and monitoring programs [6,7]. ...
... Our decline in detection rates illustrates the reduced need for basking during periods when shaded air temperatures are greater than 24 °C. A similar trend was found by Erb et al. [7] for Eastern Box Turtles (Terrapene carolina), which had significantly higher detection rates earlier in the day. However, the pattern does not hold for all reptile species. ...
Population surveying and monitoring are important for identifying conservation needs and tracking trends in populations, communities, and ecosystems over time and laying the groundwork for conservation management and policy decisions. If species or populations go undetected because of inadequate effort or sampling design, protection and management cannot be properly provided. Due to the widespread loss of populations, the Eastern Massasauga (a rattlesnake) was recently listed as a federally threatened species in the United States; it is also listed as threatened in Canada. Given its current conservation status, there is considerable interest at state and federal levels in determining how to best survey for Eastern Massasaugas to aid in management decisions. Using a 16-year dataset, we examined the relationships among environmental, temporal, area, management, and search effort factors on the detection probability of Eastern Massasaugas. We found that four abiotic parameters (solar irradiance, shaded air temperature, three-day maximum air temperature, and humidity) and three search parameters (effort per researcher, search area, and search time of day) influenced detection of Eastern Massasaugas. As the current biodiversity crisis continues, the cost-effective use of resources and scientific expertise will continue to increase in importance. We hope our results stimulate similar analyses in other taxa, which will be critical for designing and implementing regional survey and monitoring programs.
... Although several terrestrial turtle species are of conservation concern worldwide, little is known about survey efficacy. For 8 visual encounter surveys, detection probability was 0.03 for telemetered ornate box turtles (Terrapene ornata) in a sand prairie in northwestern Illinois, USA (Refsnider et al. 2011), and 0.29 (SE ¼ 0.18) for box turtles in early successional and forest edge habitats in Massachusetts, USA (Erb et al. 2015). Detection probability can apparently be improved by using dogs. ...
... During the growing season (defined here as after leaf-out) when box turtles are active, we suggest that visual surveys can be more effective by watching for turtles as they move to refugia during prescribed fires, and focusing postfire searches on refugia like large downed logs, areas of unburned vegetation, and adjacent unburned areas (particularly wetlands). Additionally, conducting surveys during times of the day when turtles are known to be active can increase detection probability (e.g., Erb et al. 2015). When planning a growing season fire, we encourage managers to purposefully retain unburned refugia within the burn area and ensure that turtles can negotiate the flame front to these refugia. ...
Land managers generally assume that postfire surveys conducted by fire personnel are effective at quantifying mortality or injury of box turtles (Terrapene spp.) because individuals should be readily observable in burned areas. Box turtle surveys conducted by humans in unburned habitats can be ineffective, yet little information exists on the efficacy of postfire surveys. We quantified detection probability of eastern box turtles (Terrapene carolina carolina) shortly after a prescribed fire in southwestern Michigan, USA, during May 2015. Immediately prior to a May (growing season) fire, we confirmed that 7 adult box turtles fitted with radiotransmitters occupied the proposed burn area. Two days after the burn we reconfirmed turtle locations and subsequently conducted 6 independent visual‐encounter surveys through 2 burned areas (0.75 and 1.0 ha) that contained telemetered turtles. For these 12 surveys, we found that average detection probability per survey was low (0.11, SE = 0.09) and highly variable among observers (range = 0.00–0.50). We found that individual turtles directly exposed to fire remained buried for up to 12 hr after the fire was extinguished and were thus unavailable for detection immediately after the burn. Further confounding postfire survey results, buried turtles rapidly moved to unburned areas after emerging from their subterranean refugia. Our results suggested that typical visual‐encounter surveys conducted for eastern box turtles after prescribed burning do not accurately reflect occupancy status or fire‐caused mortality. © 2018 The Wildlife Society.
... We used bootstrapping to compute average density for the NCSO range in a way that reflected its multiple sources of error (Efron 1982). We took 10,000 resamples without replacement of our data such that each resample permitted variation due to multiple sources of error simultaneously (e.g., site-level abundance, home range size, seasonal scaling of home range size). ...
... The RN model has not been applied widely to estimate abundance, in part because it is more sensitive to parametric assumptions and performs poorly when abundance is high (K ery and Royle 2015). Based on our literature search of the term, "RN model," we found only four published studies that used this model to estimate abundance of a wildlife species (Stanley and Royle 2005, Ariefiandy et al. 2014, Duquette et al. 2014, Erb et al. 2015. However, Royle and Dorazio (2008) demonstrated use of the model to estimate local abundance of fishers from track plate survey data in the NCSO range. ...
Estimating density and population size is often more challenging than measuring indices of abundance because of uncertainty about the effective area of surveys. We combined hierarchical modeling of detection/non-detection data from camera stations with auxiliary information on home range sizes to address this issue. We used this approach to estimate the total population size of fishers (Pekania pennanti) throughout the largest remaining native range (Northern California and Southern Oregon [NCSO], 48,760 km2) of this species in the Pacific States of the United States. After controlling for various habitat, gender, and survey factors affecting detection probability, local abundance, and home range size, we estimated an average density of 6.6 fishers per 100 km2 (95% confidence interval [CI]: 5.1–8.6) and a total of 3196 fishers (95% CI: 2507–4184). We mapped how fisher density varied throughout the range and demonstrated spatial autocorrelation in density at lag distances up to 40 km. These findings represent the first robust estimate of fisher population size for the range in NCSO. They are important for setting a baseline against which to monitor changes in population status and spatial distribution of fishers which are a species of conservation concern at federal and state levels. However, we note that our estimate of population size is very sensitive to assumptions about the effective area of camera surveys. Our methods could likely be applied to other forest carnivores and highlight the benefits of coordination between researchers to collect and share comparable survey and telemetry data.
... Due to their ecology and life history traits, many amphibians and reptiles are very difficult to detect using classical monitoring methods (Bailey et al. 2004a;Couturier et al. 2013;Daigle and Jutras 2005;Lind et al. 2005;Price et al. 2012). To increase detection probability with practical efforts, surveys and monitoring programs are generally limited to a time window maximizing species detection (Erb et al. 2015; Massachusetts Cooperative Fish and Wildlife Research Unit and the Northeast Wood Turtle Working Group 2013; Petitot et al. 2014). Additionally, the time window optimizing species detection differs from one species to another, therefore, limiting the ability of developing simultaneous monitoring programs for the different species (Bailey et al. 2004b;de Solla et al. 2005;Pellet and Schmidt 2005). ...
... Our results provide an empirical demonstration of the effectiveness of the eDNA method to efficiently characterize the presence of amphibians and reptiles in natural environments and show that eDNA may provide quantitative data comparable to measurements obtained by classical visual surveys. The extension or contraction of distribution ranges is one indicator of a species trend (Erb et al. 2015;Mackenzie et al. 2002;Zylstra et al. 2010) and is a common indicator in recovery plan objectives (e.g., Environment Canada 2014). Environmental DNA can be used to define species range data from different habitats and merge sampling efforts when resources are limited. ...
Among vertebrates, herpetofauna has the highest proportion of declining species. Detection of eDNA is a promising method towards significantly increasing large-scale herpetological conservation efforts. However, the integration of eDNA results within a management framework requires an evaluation of the efficiency of the method in large natural environments and the calibration of eDNA surveys with the quantitative monitoring tools currently used by conservation biologists. Towards this end, we first developed species-specific primers to detect the Wood turtle (Glyptemys insculpta) a species at risk in Canada, by quantitative PCR (qPCR). The rate of eDNA detection obtained by qPCR was also compared to the relative abundance of this species in nine rivers obtained by standardized visual surveys in the Province of Québec (Canada). Secondly, we developed multi-species primers to detect North American amphibian and reptile species using eDNA metabarcoding analysis. An occurrence index based on the distribution range and habitat type was compared with the eDNA metabarcoding dataset from samples collected in seven lakes and five rivers. Our results empirically support the effectiveness of eDNA metabarcoding to characterize herpetological species distributions. Moreover, detection rates provided similar results to standardized visual surveys currently used to develop conservation strategies for the Wood turtle. We conclude that eDNA detection rates may provide an effective semi-quantitative survey tool, provided that assay calibration and standardization is performed.
... Some studies have demonstrated higher abundance together with higher mortality of species in roadside habitats (e.g. Erb et al. 2015 for box tortoises, Meek 2014 for lizards, and Baxter-Gilbert et al. 2015 for pollinating insects). However, such results do not necessarily prove a trap effect. ...
... When surveying for snakes, nondetections can arise from at least 3 factors: the snake may be hidden in refugia, observers may overlook snakes present, or the site may be temporarily unoccupied (Flatt et al. 1997). Thus, surveys for snakes commonly require several site visits before presence can be confirmed or absence inferred (Kéry 2002, Durso et al. 2011, Erb et al. 2015. Although there is a lack of empirical studies assessing snake occupancy, the consensus among herpetologists is snakes are experiencing population declines and range contractions (Mullin and Seigel 2009). ...
Snakes are difficult to study due to their cryptic coloration, minimal movements, and use of inaccessible habitats. Although well‐timed surveys during a species' active season can result in higher detection rates and conserve survey resources (i.e., time and money), survey effort may not ensure the detection of rare and cryptic species. Thus, in such instances, a strategic species‐specific sampling design is needed. The Kirtland's snake ( Clonophis kirtlandii ) is a rare, cryptic species assumed to be experiencing range‐wide declines. Naturalists have noted the disappearance of Kirtland's snakes from various habitats since the early 1970s. The primary objective of our study was to determine detection of Kirtland's snakes and the environmental and temporal factors influencing detection. We calculated the effort needed to detect individuals at sites by estimating detection probabilities of 3 known Kirtland's snake populations in Illinois from 2019 to 2021. Based on 77 Kirtland's snake detections over 226 site visits (34.1%) across 3 study sites, we found that high cloud cover, moderate air temperature, and low relative humidity enhanced the detection probability of this species. The middle of May to the beginning of July was the best time to conduct surveys when detection rates were highest. As our results suggested, it is imperative to establish strategic monitoring programs maximizing conservation resources to document populations for conservation action and range shifts for species of conservation concern, such as Kirtland's snakes.
... Consequently, based on the recorded fatalities, it is likely that the size of the local viper population exceeds considerably the population size formerly estimated. Estimation of the population size of elusive reptiles is often challenging, because of their low detectability and restricted activity (Durso et al. 2011, Erb et al. 2015. Newly collected data suggest that the population size of Hungarian meadow vipers is at least two orders of magnitude larger than previous estimates, as indicated by the SDF (Móré et al. unpubl.). ...
Animal populations are regulated by multiple factors, e.g. resource availability, diseases and predators. However, these factors can be significantly influenced by anthropogenic effects (e.g. biological invasions, mesopredator release) and either subserve populations or push them towards local extinction. The Hungarian meadow viper Vipera ursinii rakosiensis is an endangered reptile, with only a few surviving populations following recent alterations and fragmentation of their habitats. In Hungary, all documented populations are located within nature conservation areas. Significant practical conservation efforts have been made, including habitat reconstruction and development, as well as ex situ breeding and reintroductions. Nevertheless, the estimated overall population size of Hungarian meadow viper remains very low, and the impact of conservation interventions is practically immeasurable according to low densities. It is hypothesized that predation is the main factor interfering with conservation efforts of the vipers. Here, we analysed the diet of two mammalian predators (European badger and red fox) in a Hungarian meadow viper habitat. We found a high prevalence of viper remains in the processed faecal samples, indicating that a high number of individuals vipers are subject to predation even within a short period of time. We conclude that effective predator control might be necessary to increase the impact of standard conservation measures (e.g. habitat development, ex situ breeding and reintroduction) that target the restoration of Hungarian meadow viper populations.
... At a low elevation site in northeastern Virginia, USA (Boucher et al. 2017), Eastern Box Turtles entered into hibernacula in December and emerged in March and were found with almost all carcasses (suggesting recent death), we find this unlikely. In fact, we suspect that fire-related mortality almost certainly exceeded our count of dead turtles given that we did not search all burn units, and that the probability of box turtle detection is low, especially for subadults (Refsnider et al. 2011;Erb et al. 2015;Melvin and Roloff 2018). The fact that we continued to find additional dead, burned turtles when surveying burn units a second time lends credence to this. ...
Fire is a key natural process that has shaped ecosystem distribution and composition for hundreds of millions of years. Prescribed fire has become a common and important technique in habitat management. With its ability to alter physical and vegetative conditions, prescribed fire has the potential to change wildlife habitat characteristics in ways that benefit or harm populations of wildlife, but implications for reptiles are considered only infrequently. In May and June 2018, a prescribed burn was conducted in Rhode Island, USA. After reports of dead Eastern Box Turtles (Terrapene c. carolina), we initiated surveys in the burned area to gauge the extent of mortality and develop guidance to protect Eastern Box Turtle populations where prescribed fire is used. We found 49 dead Eastern Box Turtles across a 31.2 ha area, and based on the available evidence, we suspect that 47 of those were killed as a direct result of the prescribed fire. The post-burn mortality observed during this study is the latest among a growing list of observations of significant mortality of box turtles associated with fire, both wild and prescribed. We urge land managers to carefully consider the location, timing, and frequency of prescribed burns within the range of the Eastern Box Turtle and consider mechanical treatments as an alternative to meet habitat management goals, or prior to burning to reduce fire intensity in sensitive areas. Burning during the growing season is particularly problematic due to the high risk to box turtles active on the surface.
... In New York, box turtles are listed as a species of special concern [40,41]. Population estimates are not available for the species in New York State, likely due to the difficulty of surveying elusive and solitary reptiles, however successful surveys have been performed in neighboring states [42]. The frequency of vehicular trauma in box turtles presented to wildlife rehabilitators suggests that roadway construction and an increasing density of vehicular traffic are directly impacting this species. ...
Wildlife rehabilitation is a publicly popular practice, though not without controversy. State wildlife agencies frequently debate the ecological impact of rehabilitation. By analyzing case records, we can clarify and quantify the causes for rehabilitation, species involved, and treatment outcomes. This data would aid regulatory agencies and rehabilitators in making informed decisions, as well as gaining insight into causes of species mortality. In New York State, the Department of Environmental Conservation (NYSDEC) has licensed rehabilitators since 1980 and annual reporting is required. In this study, we analyzed 58,185 individual wildlife cases that were attended by New York rehabilitators between 2012 and 2014. These encompassed 30,182 (51.9%) birds, 25,447 (43.7%) mammals, 2,421 (4.2%) reptiles, and 75 (0.1%) amphibians. We identified patterns among taxonomic representation, reasons for presentation to a rehabilitation center, and animal disposition. Major causes of presentation were trauma (n = 22,156; 38.1%) and orphaning (n = 21,679; 37.3%), with habitat loss (n = 3,937; 6.8%), infectious disease (n = 1,824; 3.1%), and poisoning or toxin exposure (n = 806; 1.4%) playing lesser roles. The overall release rate for animals receiving care was 50.2% while 45.3% died or were euthanized during the rehabilitation process. A relatively small number (0.3%) were permanently non-releasable and placed in captivity; 4.1% had unknown outcomes. A comparable evaluation in 1989 revealed that wildlife submissions have increased (annual mean 12,583 vs 19,395), and are accompanied by a significant improvement in release (50.2% in the study period vs 44.4% in 1989) (χ²(1) = 90.43, p < 0.0001). In this manuscript, we aim to describe the rehabilitator community in New York State, and present the causes and outcomes for rehabilitation over a three-year period.
... We estimated an average local abundance of 2.75 individuals per camera station (Table 3). This was moderate to high in comparison with published population studies of other species using RN models: higher than some (Furnas et al. 2017, Tobajas et al. 2020) and within the upper range of others (Arefiandy et al. 2014, Erb et al. 2015, O'Brien 2019, Nichols 2019, Duquette 2020. While RBR daily detection probability was also fairly high, we were reassured by the fact that our estimates were stable between surveys, that standard errors were not high, and that tests indicated good model fit. ...
EXECUTIVE SUMMARY We report here on population assessment of riparian brush rabbit (Sylvilagus bachmani riparius, RBR) at the San Joaquin River National Wildlife Refuge (SJRNWR). Our results set a baseline population density at the refuge for RBR which are an endangered species at federal and state levels. We combined hierarchical modeling of detection/non-detection data from camera stations with information on RBR home range sizes from telemetry. Our camera surveys were conducted in July and October 2020 on 852 ha (2,104 acres) of restored riparian habitat in the SJRNWR-West Unit using camera traps baited with alfalfa cubes left out for a 23-day survey. For RBR home range size, we relied on a prior radiotelemetry study conducted at the refuge in 2002-2005 on 101 RBR. After combining these data and controlling for behavioral and survey factors affecting detection probability at camera stations, our summer survey estimated an average of 1.74 RBR/ha (95% CI: 1.29-2.31; 0.71 RBR/acre) for a total of 1,485 RBR (95% CI: 1,097-1,970) in the study area. The fall survey estimate was similar.
[Line break. Start paragraph 2.] This RBR population estimate should be viewed as a conservative estimate of population size at the SJRNWR in two respects. The area included in this model is limited to the described 852 ha (2,104 acres) because this is the area of contiguous restored riparian habitat where the camera traps were placed in our survey, and due to the general homogeneity of the vegetation. However, areas with known riparian brush rabbit occupancy were not included in this model, including the adjacent 676 ha (1,671 acres) of native habitat (Gardner's Cove, Christman Island, Hospital Creek, Peninsula, Colwell River, Lara/Arambel River), the restored and native habitat on 93 ha (229 acres) of the disjunct Buffington Unit on the Stanislaus River, and the dense vegetation Population estimation of riparian brush rabbit 2 planted specifically for riparian brush rabbit flood refugia on eight miles of levee slopes. This limitation is one element that produced a conservative population estimate.
[Line break. Start paragraph 3.] Our findings on RBR density and population size were sensitive to estimates of RBR home range size. We used information based on radio-telemetry from 88 of the first RBR released on the Refuge in 2002-2005 plus 13 native-born RBR. The large home range size calculations from these initial RBR releases may reflect a growing population in its first 3 years of reintroduction and prior to several large riparian habitat restoration projects. RBR home ranges are expected to decrease as habitat quality increases and population density increases (Zoloth 1969, Kelt et al. 2014). If home range size at the refuge has decreased between the 2002-2005 study and our 2020 surveys, then population density at the refuge is proportionately higher than we have reported here.
[Line break. Start paragraph 4.] Management of RBR will require continued monitoring of population trends. In a power analysis we demonstrated our methods may be viable for monitoring long-term changes in population status of RBR. Achieving this would require new telemetry surveys to be established and changes in camera sampling methods to be implemented which we outline in our recommendations. We further recommend consideration of more robust survey methods such as spatial capture-recapture, conventional capture-recapture or hierarchical distance sampling. These methods may provide more robust monitoring at similar cost and can also be combined with the methods of this study at a subset of sites or surveys to validate accuracy.
... Compared to distance sampling, measuring site occupancy has been shown to be more efficient and robust (capable of detecting a 5% annual decline after 10 y of sampling) when used to monitor populations of Sonoran Desert Tortoise (Gopherus morafkai); Zylstra et al. 2010). Similarly, Erb et al. (2015) showed that a 10% decline in occupancy between 5-y sampling rounds could be detected for low-density populations of Eastern Box Turtles (Terrapene carolina carolina), highlighting the use of this approach. Despite its recent popularity and successful application with other chelonian species, occupancy sampling has not been previously applied to Gopher Tortoise populations at a landscape scale. ...
Assessing population trends for imperiled species that occur at low densities across large geographic areas can be challenging. Standard sampling techniques are often designed for small areas where target species can be easily observed across most of the study site. We evaluated the use of an occupancy framework for sampling a low-density Gopher Tortoise (Gopherus polyphemus) population found on Eglin Air Force Base, Florida, USA. We examined the effects of habitat type and proximity to historic Gopher Tortoise observations on site occupancy across the base. We surveyed 469 1-ha sites using two observers to walk 10 m transects across the survey site. We determined occupancy using the presence of active or inactive tortoise burrows as a surrogate for actual tortoise observations. We surveyed approximately 58% of sites twice to estimate the probability of burrow detection. We encountered Gopher Tortoise burrows at 53 survey sites (11%), and the detection probability for burrows was high (p = 0.951). Occupancy probability decreased from 0.42 to 0.01 as the distance from historic tortoise burrow observations increased, regardless of the habitat type. Power analyses indicated that a 3-5% annual decline in Gopher Tortoise occupancy would likely be detected by repeat surveys within 5-10 y. While our approach does not estimate tortoise abundance or density changes over time, it offers natural resource managers a technique to monitor the area occupied by tortoises over a large geographic area and broadly assess the effects of ongoing management actions.
... Turtle populations are threatened by habitat loss and degradation, poaching, introduced diseases, increased meso-predator abundance, and other anthropogenic sources of additive adult mortality (Klemens, 2000;Lovich et al., 2018). Sometimes these declines are directly attributable to habitat loss, road mortality, predation from corvids, or poaching (Garber and Burger, 1995;Dorcas et al., 2006;Walker and Rafeliarisoa, 2012;Loehr, 2017); however, many times the declines are due to synergistic effects from multiple factors that are much more difficult to identify and manage (Lovich, 1989;McCoy et al., 2006;Browne and Hecnar, 2007;Erb et al., 2015;Lovich et al., 2018). ...
Long-term studies on wildlife populations are necessary to track population abundance and shifts in demography over
time, yet such studies are difficult to plan, fund, and conduct and are therefore rarely undertaken. Such studies are
especially important for long-lived species that can persist for long periods of time with little to no reproductive output
or recruitment. We conducted two population studies spanning a 30-year time frame on the globally endangered
Spotted Turtle (Clemmys guttata) on protected land in the center of their range. Spotted Turtles are endangered in
Canada, listed as globally endangered on the IUCN red list, and declining throughout their range. However, there has
only been one previous long-term study tracking their long-term population trajectory. Here, we use mark–recapture
data collected over a 30-year time frame and report that the estimated population size of Spotted Turtles has decreased
by 49% at our study site despite the habitat residing with a protected area. This decline was concurrent with a
significant increase in the proportion of larger individuals within the population, indicating a lack of recruitment into
the sub-adult stage class. These results highlight the value of long-term studies in monitoring population changes of
long-lived species, the importance of active management within protected areas, and the ability of long-lived species to
persist for long periods of time despite having little recruitment and a declining population trajectory.
... Such a lack of statistical power in short-term trend analysis has been identified for other types of monitoring programs and indicator species (Erb et al. 2015;Nielsen et al. 2009;Helander et al. 2008;Meyer et al. 2010). At short timescales, the natural variation in the population parameters of interest tends to mask other, long-term changes that the system is experiencing. ...
Long-term monitoring programs can identify environmental trends or reveal limitations to protocols, as long as their results are analysed appropriately. While monitoring programs are not necessarily hypothesis-driven, their data are important for conservation and can guide improvements to monitoring programs. Here, we present a case study using dynamic occupancy models to guide the optimization of time and effort in a long-term terrestrial salamander monitoring program. To ensure a detailed analysis, we analysed the available long-term data to first identify estimates of occupancy and detection parameters for the salamanders. Using these estimates, we created simulations to identify the optimal number of years for monitoring and the optimal allocation of spatial and temporal survey replicates. Our data support previous claims that monitoring programs should be allowed to run for at least a decade. We also found that in order to obtain accurate estimates of species occupancy, programs should consider appropriate partitioning of monitoring effort across spatial and temporal scales. We show how analyses of long-term monitoring datasets are valuable not only for trend detection but also for the development of templates to guide the design and optimization of similar programs.
... Population monitoring is an integral tool in ecological research and conservation biology that yields essential information on the density, distribution, and persistence of organisms (Amano, Lamming, & Sutherland, 2016). Information on population distribution is critical for evaluating restoration and restocking programmes (Erb, Willey, Johnson, Hines, & Cook, 2015;Noon, Bailey, Sisk, & McKelvey, 2012). The increasing use of predictive habitat suitability models requires population monitoring to ground-truth model predictions, so that outputs can be incorporated into regional planning strategies (Beck & Odaya, 2001;Manley, Zielinski, Schlesinger, & Mori, 2004). ...
• Assessing the conservation status of a species is strongly dependent upon data on species distribution and abundance. With the emergence of novel methods for species monitoring – such as the use of environmental DNA (eDNA) – monitoring success can be improved at reduced expenditure in the field, particularly in remote regions and terrains where access is difficult or dangerous.
• The highly endangered crocodile lizard (Shinisaurus crocodilurus Ahl, 1930) inhabits fragmented sites of the remaining evergreen forest with running water systems in a narrow distribution range in southern China and north‐east Vietnam. Crocodile lizards spend most of the day within or above water bodies, which are commonly remote and inaccessible.
• To monitor recent spatial occurrences, and to confirm the persistence or extinction of previously reported populations (especially in heavily altered habitats), the suitability of using eDNA and quantitative polymerase chain reaction (qPCR) was tested as an alternative method for monitoring this semiaquatic lizard.
• To assess the accuracy and limitations of this method, eDNA results from the field were compared with eDNA data from mesocosms and census data on the actual abundance of this species in the field.
• Environmental DNA of the crocodile lizard was detected in all of the positive controls, and in four of six natural sites; thus, all data collected using traditional field surveys were confirmed with eDNA results.
• eDNA monitoring was found to be a reliable method for assessing the viability of populations; we suggest that it should be developed as a tool for efficient wildlife management, particularly under difficult field and funding conditions.
... Conventional physical methods can also cause destructive impacts on the environment and to biological communities (Wheeler, Raven, & Wilson, 2004), making them difficult to apply in a conservation context. Furthermore, when a species' behaviour or size makes it difficult to survey them (e.g., small-bodied or elusive species), conventional methods can require specialized equipment or species-specific observation times, thus making species richness and relative abundance estimates for entire communities intractable (e.g., many amphibians and reptiles, Erb, Willey, Johnson, Hines, & Cook, 2015;Price, Eskew, Cecala, Browne, & Dorcas, 2012). These reasons highlight the continued need to develop improved ways to survey global biodiversity, and the unique ways eDNA metabarcoding can complement conventional methods. ...
The genomic revolution has fundamentally changed how we survey biodiversity on earth. High-throughput sequencing (‘HTS’) platforms now enable the rapid sequencing of DNA from diverse kinds of environmental samples (termed ‘environmental DNA’ or ‘eDNA’). Coupling HTS with our ability to associate sequences from eDNA with a taxonomic name is called ‘eDNA metabarcoding’ and offers a powerful molecular tool capable of non-invasively surveying species richness from many ecosystems. Here, we review the use of eDNA metabarcoding for surveying animal and plant richness, and the challenges in using eDNA approaches to estimate relative abundance. We highlight eDNA applications in freshwater, marine, and terrestrial environments, and in this broad context, we distill what is known about the ability of different eDNA sample types to approximate richness in space and across time. We provide guiding questions for study design and discuss the eDNA metabarcoding workflow with a focus on primers and library preparation methods. We additionally discuss important criteria for consideration of bioinformatic filtering of data sets, with recommendations for increasing transparency. Finally, looking to the future, we discuss emerging applications of eDNA metabarcoding in ecology, conservation, invasion biology, biomonitoring, and how eDNA metabarcoding can empower citizen science and biodiversity education. This article is protected by copyright. All rights reserved.
... Accurate and precise population estimates are particularly important for species of conservation concern, such as those listed in the United States as federally or state threatened or endangered. Thus, research has been devoted to development and improvement of monitoring programs for focal species (Flint and Harris 2005, Probst et al. 2005, Brown et al. 2013, Erb et al. 2015. However, many species of concern do not have standardized monitoring protocols because of limited funding for wildlife research, or the protocols are not amenable to contemporary analyses because of rapid advancements in statistical approaches for analyzing monitoring data, which have inherent sampling design requirements. ...
Population monitoring is a fundamental component of wildlife management, and is necessary to track site- and regional-level status and recovery of species of conservation concern. The wood turtle (Glyptemys insculpta) is a species of conservation concern for federal and state agencies because of population declines across the species’ range. We developed and tested a survey and analysis design to assist agencies in the Upper Midwest, USA, with establishment of long-term monitoring programs for wood turtle populations. In spring of 2016, we conducted 8 replicate population surveys at 8 candidate long-term monitoring sites in northeastern Minnesota, USA. Using field survey data and simulation models, we assessed the influence of distance from river surveyed, number of survey replications, and number of sites on abundance estimates; we also delineated important survey covariates and compared demographic estimates based on distance from river surveyed. We estimated site-level abundances and compared survey designs using a multinomial N-mixture model that included a removal sampling observation process. Mean abundance estimates were similar when surveying 2 transects (i.e., the river-land interface to ∼25 m inland) or 4 transects (i.e., the river-land interface to ∼55 m inland), but decreasing the survey distance from river reduced the precision of estimates. Mean abundance estimates were similar with ≥6 replications. Air temperature was an important predictor of survey-specific detection probability, with maximum detectability at 19−23°C. Sex ratio and mean carapace length did not differ based on whether we surveyed 2 or 4 transects, and percentage of individuals by size class was nearly identical between the sampling designs. Simulations indicated that 75% of mean abundance estimates were within ±8% of true abundance when ≥15 sites were surveyed. The wood turtle survey and analysis design we developed and tested was effective for estimating abundance of wood turtle populations in northeastern Minnesota, and we encourage its use as a template for wood turtle monitoring programs in the Upper Midwest.
... Methodological comparisons A number of investigators have evaluated the means by which we might increase our understanding of difficult to detect species, such as reclusive reptiles Erb et al. 2015). McKnight et al. (2015) found, similar to our results, that even though they are nonsystematic, incidental encounters often provide a relatively accurate depiction of diversity of snakes occupying a site. ...
Sonoran Desert habitat in southern Arizona is increasingly altered by urban development near metropolitan areas. Understanding how reptiles respond in these impacted habitats is critical to conservation efforts to retain intact biotic communities, especially those with a high diversity of reptile species. We surveyed snakes at one impacted site on the northern edge of the Phoenix metropolitan area in desert/urban interface, and at another site in a desert/rural interface near Florence, Arizona. The site near Phoenix was lower in species richness (15 spp.), and evenness: two snake species accounted for 75 % of all snakes encountered (total = 420). The site near Florence was higher in species richness (19 spp.), and a more even community: no species accounted for more than 20 % of snakes encountered (total = 594). Sampling methodology had a strong influence on species richness and abundance of snakes at the respective sites: road riding, coverboards, and traps each provided evidence of unique species missed by other methods. These results were compared with inventories at three other sites in the central Sonoran Desert of Arizona, and are consistent with the view that the impacted site near Phoenix is uneven, potentially as a result of a single species reacting to shifts in prey availability.
Knowledge of a species’ distribution throughout a landscape is crucial for developing long-term conservation practices and effective management plans. Accurate distributional data are particularly important for threatened or rare species with a broad geographic range and low detectability in their preferred habitats, as surveying and successfully encountering this type of species is oftentimes difficult. One such species is the eastern massasauga (Sistrurus catenatus), a small-bodied, federally threatened rattlesnake with a distribution centered around the Great Lakes region. The current status of many massasauga populations remains uncertain, hindering effective recovery efforts. We used single-season occupancy models to reassess the status of historic massasauga sites (n = 33) distributed throughout Michigan’s Lower Peninsula. We assessed site- and survey-specific covariates at both the local- and landscape-scale to inform occupancy and detection probabilities. Average canopy cover best predicted occupancy probabilities, while search effort best predicted detection probabilities. Our top model which included canopy cover (occupancy) and search effort (detection) estimated an average occupancy probability of 0.30 (CI = 0.15–0.50) and average detection probability of 0.55 (CI = 0.31–0.78). Landscape-scale covariates derived from remotely-sensed data did not improve model fit for occupancy models. Our findings suggest that persistence of eastern massasauga populations relies upon canopy cover of
Having an accurate estimate of population size and density is imperative to the conservation of chelonian species and a central objective of many monitoring programs. Capturerecapture and related methods are widely used to obtain information about population size of chelonians. However, classical capturerecapture methods have strict spatial sampling requirements and do not account for lack of geographic closure caused by movement of individuals in and out of the surveyed landscape. Newly developed spatial capturerecapture (SCR) models address these limitations by specification of explicit models for spatial sampling as well as the spatial distribution of individuals in the population. Spatial capturerecapture models have not yet been applied to the study of chelonian populations. Here we demonstrate their application to a population of box turtles in Maryland that has been studied for 75 yr. Results support dramatic declines in population size of box turtles since the 1940s.
Management decisions for species are often based on estimates of abundance, which can be difficult to obtain for species that are a challenge to survey, as are some reptiles. Information on abundance and population status are lacking for the diamondback terrapin (Malaclemys terrapin), a coastal species that inhabits brackish waters and plays an important trophic role in the saltmarsh ecosystem. Population declines are suspected throughout the species' range, and its population status is unknown in Florida. Of the 5 subspecies that inhabit Florida's coastline, the most understudied subspecies may be the ornate diamondback terrapin (M. t. macrospilota). We conducted a capturemarkrecapture study of M. t. macrospilota during the summer of 2013 on 3 adjacent coastal islands in the eastern panhandle of Florida that provided information on population size and demography. We captured 334 individuals; modeling estimated a population size of 1282 (8671905 95% CI) and a density of 150 terrapins/ha. Population size decreased throughout the study, suggesting that this population is an aggregation that seasonally emigrates from the islands. This emigration trend was more evident for females. Males outnumbered females 4:1, and females were larger than males. Our study is the first to report on M. t. macrospilota populations in the Florida panhandle. We recommend collaborative, long-term population monitoring at our sites to estimate population trends that will be crucial for managing this subspecies.
Prescribed fire is an essential management practice in pyrogenic ecosystems, but fire can also be a significant disturbance and source of mortality for both target and non-target species. Seasonal periods of animal inactivity may provide opportunities to design burn plans that minimize negative impacts to species of conservation concern, but few studies have rigorously examined this possibility. Using radiotelemetry, we studied overwintering behavior and interactions with fire in a forest-dwelling terrestrial turtle, the Eastern Box Turtle (Terrapene carolina carolina), over an eight-year period at two sites that use prescribed fire in forest management. Turtles at both sites selected predominantly hardwood forests and mesic habitats and avoided upland pine forests. Turtles buried deepest (2.9 – 3.2 cm) below the soil-litter interface in late February and then moved gradually shallower until emergence in early April. Emergence timing varied over a 58-day period, but was consistent within individuals from year to year. Turtles also maintained fidelity to refuge locations, but those overwintering in burned areas selected sites over twice as far from refuges used in previous years compared to those in unburned areas. The areas and habitats selected by turtles during winter served as refugia from fire, and those whose refuges did burn remained buffered from lethal temperatures even at shallow burial depths. The only fire-related injury or mortality occurred during seasons of surface activity. Timing burning and other forest management practices during periods of winter dormancy may thus minimize threats to turtle populations, but modifications to prescribed fire regimes must also be balanced with other management objectives.
Sustainable wildlife management is required to guarantee source species viability; however, it is practiced rarely in the tropics. The yellow anaconda (Eunectes notaeus) has a long history of being harvested for its leather. Since 2002 its harvest has operated under a management program in northeastern Argentina, which relies on adaptive management through limiting the minimum anaconda length, number of hunters and restricting the hunting season. We investigated the effects of the anaconda harvest on its biological parameters based on 2002-2016 data. Here we show that the levels of species exploitation are sustainable. The gradual reduction in the annual hunting effort, due to a decrease in number of hunters and hunting season duration, reduced the total number of anacondas harvested. Conversely, captures per unit effort increased across the study time-period. There was no variation in the mean length of anacondas harvested, or in largest anaconda sizes. Though more females than males were caught, the sex ratio did not vary significantly. We also found that a decrease in mean temperature positively influenced anaconda harvest and the captures of giant individuals. Because sustainable use is a powerful tool for conservation, those discoveries are highly applicable to other species and regions.
Wildlife rehabilitation is a publicly popular though highly controversial practice. State wildlife agencies frequently debate the ecological impact of rehabilitation. Analysis of case records could inform that debate by clarifying and quantifying the causes for rehabilitation, species involved, and treatment outcomes. This information could aid in the ability of regulatory agencies and rehabilitators to make informed decisions and gain insight into causes of species decline. In New York, rehabilitators are licensed by the Department of Environmental Conservation (NYSDEC) and thus, are required to submit annual reports. Between 2012-2014, 59,370 individual wildlife cases were seen by licensed rehabilitators comprising 31,229 (52.6%) birds, 25,490 (42.9%) mammals, 2,423 (4.1%) reptiles, and 73 (0.1%) amphibians. We identified patterns among taxonomic representation, reasons for presentation, and disposition. Major causes of presentation were trauma (n = 22,672, 38.2%) and orphaning (n = 21,876, 36.8%), with habitat loss (n =3,746, 6.3%), infectious disease (n = 1,992, 3.4%), and poisoning or toxin exposure (n = 864, 1.5%) playing lesser roles. The overall release rate for animals receiving care was 50.2%; 45.4% were either euthanized or died during the rehabilitation process. A relatively small number (0.3%) were permanently non-releasable and placed in captivity, and 4.1% had unknown outcomes. In comparison to data from 1989, wildlife submissions have increased (annual mean 12,583 vs 19,790), as has the release rate, from 44.4% to 50.2%. Utilizing a large data set allowed us to fill knowledge gaps, which can help inform management by both the rehabilitators and the state agencies that regulate them, deepening understanding of the scope and impacts of wildlife rehabilitation.
The Box Turtle Connection is a 100-year, citizen-science project that aims to build a long-term database to serve as a conservation tool and a model for current and future turtle research. Ten years underway, we've created a revised book publication that features new information on pioneering and maintaining a box turtle study, strategies for quality data collection and standardization of protocols, habitat creation and management, and community engagement.
In the early summer of 1993-1995, a survey was conducted in central Ontario to determine the amount of road kill of Snapping Turtles during their nesting period. Peak nesting occurred during mid June in all three years. The annual mortality of Snapping Turtles during this period was 30.5% of all observed turtles. Only 24% of the dead snapping turtles sampled were mature females. Snapping Turtle road mortality during the nesting period was high and could have negative impacts on the sustainability of local populations.
We examine the degree to which field observers can meet the assumptions underlying line transect sampling to monitor populations of desert tortoises (Gopherus agassizii). We present the results of 2 field trials using artificial tortoise models in 3 size classes. The trials were conducted on 2 occasions on an area south of Las Vegas, Nevada, where the density of the test population was known. In the first trials, conducted largely by experienced biologists who had been involved in tortoise surveys for many years, the density of adult tortoise models was well estimated (-3.9% bias), while the bias was higher (-20%) for subadult tortoise models. The bias for combined data was -12.0%. The bias was largely attributed to the failure to detect all tortoise models on or near the transect centerline. The second trials were conducted with a group of largely inexperienced student volunteers and used somewhat different searching methods, and the results were similar to the first trials. Estimated combined density of subadult and adult tortoise models had a negative bias (-7.3%), again attributable to failure to detect some models on or near the centerline. Experience in desert tortoise biology, either comparing the first and second trials or in the second trial with 2 experienced biologists versus 16 novices, did not have an apparent effect on the quality of the data or the accuracy of the estimates. Observer training, specific to line transect sampling, and field testing are important components of a reliable survey. Line transect sampling represents a viable method for largescale monitoring of populations of desert tortoise; however, field protocol must be improved to assure the key assumptions are met.
This study documents the detrimental effects of human recreation on the North American wood turtle (Clemmys insculpta) in Connecticut. We chronicled the dynamics of two allopatric wood turtle populations in a protected southern New England wildlife reserve for 20 yr (1974-1993). Both wood turtle populations were reproductively isolated from one another, physically separated 100 yr ago when a 1.5 km long human-made pond was constructed. We conducted a mark-and-recapture study on a 1000-ha section of a protected watershed in south-central Connecticut (New Haven County). During this study we monitored 133 different wood turtles, observing them a total of 1176 times. Human and wood turtle demographics were recorded throughout this period. The data support the following conclusions: (1) following a period of apparent stability, two populations of wood turtles declined; (2) the declines were more or less synchronous in both populations; (3) the beginning of each decline corresponded to the opening of the habitat for recreation; (4) an increase in mean turtle age suggests a failure of recruitment; however, (5) a simultaneous reduction in numbers of adult females suggests that the failure of recruitment alone is not sufficient to explain the declines. Throughout our study the size of the forest remained the same, road building was restricted, and the quality of the air and water were constant. The wood turtle populations remained stable when people were denied access to the property. When this area was opened to human recreation (hiking, fishing) the two discrete wood turtle populations declined steadily; the total number of turtles in both populations declined by 100% in 10 yr. As wilderness areas become mixed-use recreation areas, wood turtle populations may suffer. We conclude that without proper management, the increasing recreational use of parks, reservoirs, and wildlife reserves will adversely affect the long-term survival of the North American wood turtle.
Successful monitoring underpins effective wildlife management insofar as monitoring serves to track the response of wildlife resources to management and to identify whether management should be continued or changed. Here we provide both general guidelines and specific examples for the design and implementation of effective monitoring programs for adaptive wildlife management based, in part, on lessons we have learned in the Galapagos Islands, where development of a comprehensive monitoring program for its wildlife is un derway. To be effective, wildlife monitoring programs should (l) be framed by well-articulated objectives that are closely linked to management goals; (2) measure a subset of informative indicators with sampling methods that permit unbiased and statistically powerful results while minimizing costs and logistical problems; (3) ensure program continuity despite the vagaries of change in personnel, technology, and program objectives; and (4) quickly make accessible appropriately analyzed information to a wide audience, particularly policymakers. Only through such an integrated process can the adaptive "loop" in wildlife management be closed and management practices and policies evolve in a manner ultimately beneficial to wildlife, both in Galapagos and elsewhere.
Road mortality is suspected to have contributed to widespread population declines in turtles in the United States, a country with exceptionally high turtle diversity. We examined the issue through a modeling study that integrated road maps and traffic-volume data with simulated movements of (1) small-bodied pond turtles, (2) large-bodied pond turtles, and (3) terrestrial and semiterrestrial (“land”) turtles. Our model predicted that road networks typical of the northeastern, southeastern, and central regions have the potential to limit land-turtle populations and, to a lesser extent, populations of large-bodied pond turtles. Nowhere are populations of small-bodied pond turtles likely threatened regionally by road mortality. We conclude that the demographic traits of some turtles, in combination with their mobility, may jeopardize population persistence within road networks typical of the eastern and central United States.
W e use monitoring data routinely in our daily lives; we monitor the stock market, the weather, our blood pressure, and baseball statistics. But, does monitor-ing have a place in environmental science? Common crit-icisms of environmental monitoring are (1) that it is not really science, but merely a fishing expedition that diverts funds from "real" science, (2) that most monitoring data are never used, and (3) that we can't possibly know today what critical questions will need to be answered in the future. Some people feel that monitoring has no place in rigorous environmental science, and mindless monitoring gives the discipline a bad reputation. Who needs it? In this paper, we evaluate these common perceptions of environmental monitoring and discuss the characteristics of successful and unsuccessful monitoring programs. We define environmental monitoring as a time series of mea-surements of physical, chemical, and/or biological vari-ables, designed to answer questions about environmental change. These measurements may be taken at one or multiple locations. The meaning of "long term" depends on the time scale of the ecological process of interest, but in this paper we focus on datasets that span decades. Our discussion is particularly relevant, given the budgetary constraints on current monitoring and the ongoing debate regarding the opportunities, limitations, and costs associated with the establishment of national environ-mental observatories in the US, for example the National Ecological Observatory Network (NAS 2004), the Hydrologic Observatories, and the Cooperative Large-scale Engineering Analysis Network for Environmental Research (NRC 2006).
Efforts to design monitoring regimes capable of detecting population trends can be thwarted by observational and economic constraints inherent to most biological surveys. Ensuring that limited resources are allocated efficiently requires evaluation of statistical power for alternative survey designs. We simulated the process of data collection on a landscape, where we initiated declines over 3 sample periods in species of varying prevalence and detectability. Changing occupancy levels were estimated using a technique that accounted for effects of false-negative errors on survey data. Declines were identified within a frequentist statistical framework, but the significance level was set at an optimal level rather than adhering to an arbitrary conventional threshold. By varying the number of sites sampled and repeat visits made, we show how managers can design an optimal monitoring regime that maximizes statistical power within fixed budget constraints. Results show that 2 to 3 visits/site are generally sufficient unless occupancy is very high or detectability is low. In both cases, the number of required visits increase. In an example of woodland bird monitoring in the Mt. Lofty Ranges, South Australia, we show that, although the budget required to monitor a relatively rare species of low detectability may be higher than that for a common, easily detectable species, survey design requirements for common species may be more stringent. We discuss implications for multi-species monitoring programs and application of our methods to more complex monitoring problems.
REPTILE SPECIES ARE DECLINING ON A GLOBAL SCALE. SIX SIGNIFICANT THREATS TO REPTILE POPULATIONS ARE HABITAT
LOSS AND DEGRADATION, INTRODUCED INVASIVE SPECIES, ENVIRONMENTAL POLLUTION, DISEASE, UNSUSTAINABLE
USE, AND GLOBAL CLIMATE CHANGE.
Biodiversity monitoring programs need to be designed so that population changes can be detected reliably. This can be problematical for species that are cryptic and have imperfect detection. We used occupancy modeling and power analysis to optimize the survey design for reptile monitoring programs in the UK. Surveys were carried out six times a year in 2009-2010 at multiple sites. Four out of the six species--grass snake, adder, common lizard, slow-worm -were encountered during every survey from March-September. The exceptions were the two rarest species--sand lizard and smooth snake--which were not encountered in July 2009 and March 2010 respectively. The most frequently encountered and most easily detected species was the slow-worm. For the four widespread reptile species in the UK, three to four survey visits that used a combination of directed transect walks and artificial cover objects resulted in 95% certainty that a species would be detected if present. Using artificial cover objects was an effective detection method for most species, considerably increased the detection rate of some, and reduced misidentifications. To achieve an 85% power to detect a decline in any of the four widespread species when the true decline is 15%, three surveys at a total of 886 sampling sites, or four surveys at a total of 688 sites would be required. The sampling effort needed reduces to 212 sites surveyed three times, or 167 sites surveyed four times, if the target is to detect a true decline of 30% with the same power. The results obtained can be used to refine reptile survey protocols in the UK and elsewhere. On a wider scale, the occupancy study design approach can be used to optimize survey effort and help set targets for conservation outcomes for regional or national biodiversity assessments.
Monitoring is required of land managers and conservation practitioners to assess the success of management actions. ''Shortcuts'' are sought to reduce monitoring costs, most often consisting of the selection of a small number of species that are closely monitored to represent the status of many associated species and environmental correlates. Assumptions associated with such shortcuts have been challenged, yet alternative ap-proaches remain scant. We evaluated an approach that departs significantly from the ap-proach of selecting a few representative species. We explored two primary assertions: (1) that a coordinated multiple-species monitoring effort that collects presence–absence data on a broad range of species is a robust alternative to a few intensive single-species efforts, and (2) that the vertebrate species expected to be detected using this approach are numerous and diverse enough to represent all vertebrate species. We simulated monitoring the ver-tebrate species pool on an existing sample grid across the 7 million ha of public lands in the Sierra Nevada (USA) ecoregion. Based on the use of eight standard presence–absence protocols, we estimated the number of vertebrate species (excluding fish) with an adequate number of sample points within their range to detect 20% relative change in the proportion of points with detections between two points in time. We estimated that adequate detections would be obtained for 76% of the 465 vertebrate species, including 83% of all birds, 76% of all mammals, 65% of all reptiles, and 44% of all amphibians. Detection adequacy varied among life-history and ecological groups, but 50% of the species were adequately detected in every group with the exception of three groups: rare species, endemic species, and species of concern (33%, 24%, and 47% of associated species adequately detected, respectively). A multiple-species monitoring approach represents an effective and feasible alternative to the challenges of large-scale monitoring needs by targeting the most basic of population data for a large number and breadth of species.
Recovery efforts for threatened and endangered species often must be ini-tiated with incomplete data. The outcomes of such efforts are difficult to predict, which makes monitoring the progress of recovery efforts an integral part of the recovery process. We evaluated the role of monitoring in recovery plans for 181 species listed as threatened and endangered under the U.S. Endangered Species Act. We considered both the extent to which monitoring tasks were proposed as part of the recovery effort and the extent to which the tasks proposed were actually implemented. In general, tasks devoted to tracking the species' population trend were more likely to be proposed and implemented than were other monitoring activities (e.g., those devoted to the species' demographics, its habitat requirements, or the impact of predators, competitors, and exotics). We found that the extent and nature of the monitoring proposed and implemented appeared to reflect taxonomic biases that exist throughout the recovery process and were little influenced either by the level of understanding of the species' biology or by the recovery priority assigned to the species. In particular, monitoring efforts did not adequately address the specific threats affecting species. Proposals for, and implementation of, monitoring progress toward re-covery goals were independent of the type of criteria defined in the plans (e.g., population level and habitat extent), although population-related criteria were disproportionately com-mon. Based on these findings, we caution against an overemphasis on focal species mon-itoring, especially when such an emphasis leads to the reduction or exclusion of other types of monitoring. We also recommend that species-specific attributes factor more prominently in the development of monitoring to avoid monitoring action that is otherwise unnecessary.
Few species are likely to be so evident that they will always be detected when present. Failing to allow for the possibility that a target species was present, but undetected, at a site will lead to biased estimates of site occupancy, colonization, and local extinction probabilities. These population vital rates are often of interest in long-term monitoring programs and metapopulation studies. We present a model that enables direct estimation of these parameters when the probability of detecting the species is less than 1. The model does not require any assumptions of process stationarity, as do some previous methods, but does require detection/nondetection data to be collected in a manner similar to Pollock's robust design as used in mark–recapture studies. Via simulation, we show that the model provides good estimates of parameters for most scenarios considered. We illustrate the method with data from monitoring programs of Northern Spotted Owls (Strix occiden-talis caurina) in northern California and tiger salamanders (Ambystoma tigrinum) in Min-nesota, USA.
MARK provides parameter estimates from marked animals when they are re-encountered at a later time as dead recoveries, or live recaptures or re-sightings. The time intervals between re-encounters do not have to be equal. More than one attribute group of animals can be modelled. The basic input to MARK is the encounter history for each animal. MARK can also estimate the size of closed populations. Parameters can be constrained to be the same across re-encounter occasions, or by age, or group, using the parameter index matrix. A set of common models for initial screening of data are provided. Time effects, group effects, time x group effects and a null model of none of the above, are provided for each parameter. Besides the logit function to link the design matrix to the parameters of the model, other link functions include the log—log, complimentary log—log, sine, log, and identity. The estimates of model parameters are computed via numerical maximum likelihood techniques. The number of parameters that are estimable in the model are determined numerically and used to compute the quasi-likelihood AIC value for the model. Both the input data, and outputs for various models that the user has built, are stored in the Results database which contains a complete description of the model building process. It is viewed and manipulated in a Results Browser window. Summaries available from this window include viewing and printing model output, deviance residuals from the model, likelihood ratio and analysis of deviance between models, and adjustments for over dispersion. Models can also be retrieved and modified to create additional models. These capabilities are implemented in a Microsoft Windows 95 interface. The online help system has been developed to provide all necessary program documentation.
Int J Exerc Sci 3(1): 4-13, 2010. The purposes of this study were to a) investigate the average heart rate (HR), speed, stride length, and stride rate during moderate intensity walking and jogging in healthy young adults, b) cross validate the walking stride length calculation based on 42% of height and c) provide reliability information for measurement of walking and jogging steps, speed, stride length, and stride rate. Participants (N=130) wore two Yamax SW-200 pedometers and a Polar A-1 HR monitor while performing walking and jogging trials. The correlation between estimated (0.71 ± 0.04 m·stride-1) and actual stride length (0.78 ± 0.05 m·stride-1) was moderate (r = .46). However, a significant difference was observed between the two measurements (t(115) = -14.24, p < .001). The reliability results for speed, stride length, and stride rate showed that two or fewer trials were enough to achieve reliable estimates. In conclusion, when instructed to walk at a moderate pace, healthy young adults tend to walk at an average pace that is greater than that recommended for meeting current public health recommendations (80 m·min-1). Similarly, when instructed to jog at a comfortable pace, healthy young adults tend to jog at a speed greater than that corresponding to vigorous intensity physical activity (134 m·min-1). The results of the reliability analysis indicate that in healthy young adults, to measure typical walking and jogging patterns using a pedometer, only two trials for walking and one trial for jogging are necessary to achieve reliable estimates. Stride rate calculations requires the combination of two trials and one pedometer for both walking and jogging.
While Terrapene Carolina Carolina (Eastern Box Turtle) are found over a wide geographic range, they appear to select microhabitats based on the need for thermoregulation, minimization of water loss, and reproduction. Habitat selection and the activity patterns of Eastern Box Turtles in southwestern Virginia were studied in relation to short-term weather conditions and seasonal variation. Turtles were located using telemetry 36% of the time in the interior of mature forest habitats, 23% in edge habitats, 18% in fields, 7% in a 22-year-old Pinus strobus (Eastern White Pine) plantation, 7% in forest canopy gaps, 5% in a 4-year-old clearcut, and 4% in streams. We observed a seasonal shift in habitat use, with more turtles using mature forest habitat compared to other habitat types in all months, except for May when 47% of all turtles were located in edge habitat. Also, Eastern Box Turtles often selected canopy gaps within forests during the fall, and females moved from forests to recent clearcuts during the nesting season, which accounted for a larger home-range size. As expected, turtle activity was lower during hot, dry periods in midsummer and decreased gradually during the autumn with decreasing temperature. Rainfall increased turtle activity, especially when following prolonged dry periods.
A quantitative study of habitat selection in Terrapene carolina triunguis was conducted from August 1969 through January 1972. Thirty-three microenvironmental variables were measured in the vicinity of 201 form locations. Principal component analysis of these data indicated that temperature, cover and moisture are fundamental aspects of the environment at form locations. Canonical correlation of microhabitat and microclimatic variables showed that the microclimate surrounding forms was primarily influenced by ground cover aspects of the habitat. Moisture variables immediately surrounding the turtle were not correlated with features of the microhabitat, suggesting that turtles select humidity levels independent of prevailing conditions at the surface. Microclimatic factors were measured at the location of 55 active turtles within the study area. The similarity in temperature and humidity levels with those taken at form locations indicated that turtles exist within a relatively limited microclimatic range at all times during seasons of activity. Analysis of form location and trailing information indicated a seasonal shift in habitat utilization from grasslands in late spring and early fall to forested areas in summer, early spring and late fall. Activity in the grassland areas was found to coincide with moderate temperature and peak moisture conditions. At other seasons turtles burrowed beneath the litter in wooded areas to avoid temperature extremes and maintain high humidity.
Monitoring the abundance and stability of populations of conservation concern is often complicated by an inability to perfectly detect all members of the population. Mark-recapture offers a flexible framework in which one may identify factors contributing to imperfect detection, while at the same time estimating demographic parameters such as abundance or survival. We individually color-marked, recaptured, and re-sighted 1,635 federally listed interior least tern (Sternula antillarum; endangered) chicks and 1,318 piping plover (Charadrius melodus; threatened) chicks from 2006 to 2009 at 4 study areas along the Missouri River and investigated effects of observer-, subject-, and site-level covariates suspected of influencing detection. Increasing the time spent searching and crew size increased the probability of detecting both species regardless of study area and detection methods were not associated with decreased survival. However, associations between detection probability and the investigated covariates were highly variable by study area and species combinations, indicating that a universal mark-recapture design may not be appropriate. Published 2014. This article is a U.S. Government work and is in the public domain in the USA.
Estimates regarding population parameters are often based on data from surveys. To ensure that such estimates are as accurate as possible, it is important to know the detectability resulting from the particular survey method used. We used radiotelemetry to measure detectability of ornate box turtles (Terrapene ornata), using visual-encounter surveys in a sand prairie in northwestern Illinois, USA. We found that the overall detection probability of visual-encounter surveys was 0.03, and our high frequency of nondetection was due to a failure to detect visible turtles rather than turtles being underground or hidden in dense vegetation. Despite the substantial population density at our study site, visual-encounter surveys failed to detect box turtles on most visits, which resulted in a prohibitively high number of surveys that would be required to accurately estimate population size or to infer absence of the species from a site. Our method of using radiotelemetry to measure detection probability of a survey method could be easily applied to other small, cryptic, or rare species. However, our low detection probability and high frequency of nondetections recommend against use of visual-encounter surveys alone in estimating population parameters for ornate box turtles. © 2011 The Wildlife Society.
Movement patterns in the western box turtle, Terrapene ornata luteola, were estimated from both repeated sightings on a road and radiotelemetry. Home ranges were smaller than those expected from random movements. Turtles in the study area may have larger home ranges (mean area of 1.6 ha, minimum polygon method, mean maximum diameter of 276 m) than box turtles in more mesic habitats. No significant differences in home range size or distances moved were observed between sexes, and overlap was considerable. Turtles moved farther on the road than off the road and the road may have increased the size of the home range of some turtles. Turtles were most active when ambient temperatures were 13-24 C, solar radiation was <200 jules/cm2/h, and humidity was high (<20 mbar vapour pressure deficit). Turtles selected micro-habitats with lower soil temperatures, higher ambient temperatures, and lower humidity levels than those observed at the meteorological station on the study site.
Turtles appear to live longer than most other species of vertebrates, according to both maximal lifespans from zoo records and survivorship patterns in natural populations. Turtle longevity may reflect low metabolic activity, an absence of physiological and anatomical senility, a large investment in the adult's protective shell, and a life history with a long maturation period.
Large-scale, multispecies monitoring programs are widely used to assess changes in wildlife populations but they often assume constant detectability when documenting species occurrence. This assumption is rarely met in practice because animal populations vary across time and space. As a result, detectability of a species can be influenced by a number of physical, biological, or anthropogenic factors (e.g., weather, seasonality, topography, biological rhythms, sampling methods). To evaluate some of these influences, we estimated site occupancy rates using species-specific detection probabilities for meso- and large terrestrial mammal species on Cape Cod, Massachusetts, USA. We used model selection to assess the influence of different sampling methods and major environmental factors on our ability to detect individual species. Remote cameras detected the most species (9), followed by cubby boxes (7) and hair traps (4) over a 13-month period. Estimated site occupancy rates were similar among sampling methods for most species when detection probabilities exceeded 0.15, but we question estimates obtained from methods with detection probabilities between 0.05 and 0.15, and we consider methods with lower probabilities unacceptable for occupancy estimation and inference. Estimated detection probabilities can be used to accommodate variation in sampling methods, which allows for comparison of monitoring programs using different protocols. Vegetation and seasonality produced species-specific differences in detectability and occupancy, but differences were not consistent within or among species, which suggests that our results should be considered in the context of local habitat features and life history traits for the target species. We believe that site occupancy is a useful state variable and suggest that monitoring programs for mammals using occupancy data consider detectability prior to making inferences about species distributions or population change.
One of the most fundamental problems in monitoring animal populations is that of imperfect detection. Although imperfect detection can be modeled, studies examining patterns in occurrence often ignore detection and thus fail to properly partition variation in detection from that of occurrence. In this study, we used anuran calling survey data collected on North American Amphibian Monitoring Program routes in eastern Maryland to investigate factors that influence detection probability and site occupancy for 10 anuran species. In 2002, 17 calling survey routes in eastern Maryland were surveyed to collect environmental and species data nine or more times. To analyze these data, we developed models incorporating detection probability and site occupancy. The results suggest that, for more than half of the 10 species, detection probabilities vary most with season (i.e., day-of-year), air temperature, time, and moon illumination, whereas site occupancy may vary by the amount of palustrine forested wetland habitat. Our results suggest anuran calling surveys should document air temperature, time of night, moon illumination, observer skill, and habitat change over time, as these factors can be important to model-adjusted estimates of site occupancy. Our study represents the first formal modeling effort aimed at developing an analytic assessment framework for NAAMP calling survey data.
Designing monitoring programs to evaluate trends in low-density wildlife species at regional scales is challenging given difficulties detecting uncommon organisms distributed in potential habitats over large spatial extents. The northern goshawk (Accipiter gentilis) has been petitioned for listing under the Endangered Species Act and the review of the petition indicated a need for information on population trend. To evaluate trends in goshawk populations, the U.S. Forest Service developed the Northern Goshawk Bioregional Monitoring Design to estimate goshawk occupancy over broad spatial extents. We adapted and implemented this design to approximately 30,600 km2 of 88,128 km2 of National Forest System lands in the Forest Service Rocky Mountain Region, including portions of Colorado, Wyoming, and South Dakota. We developed a stratified random design to monitor goshawk occupancy in sampling units, defined by primary and secondary habitat quality as well as accessibility. To define habitat quality, we examined a time series for 58 previously located nesting territories. Using logistic regression, we found that the dominant conifer species and status of aspen in postfledging zones best characterized high-quality goshawk nesting habitat. We applied model results to stratify 4,445 sampling units based on habitat quality and further stratified sampling units based on accessibility into easy and difficult access categories. We conducted field sampling during the goshawk breeding season in the summer of 2006 to estimate detection probabilities and occupancy rates. Within our sampling frame, we sampled 51 sampling units and estimated goshawk occupancy of 0.329 (95% CI: 0.213–0.445). Occupancy within primary strata (high quality) sampling units was 0.811 (SE = 0.113), whereas occupancy in secondary strata (lower quality) sampling units was 0.124 (SE = 0.067). Future implementation of this monitoring program can achieve 0.8 power to detect 30–40% declines in with 140 sampling units. Our implementation of a stratified sampling design to monitor occupancy of goshawks at a region-wide scale reduced the number of sampling units in each administrative unit and focused our efforts on those areas most likely to have goshawks. © 2011 The Wildlife Society.
Effective conservation requires strategies to monitor populations efficiently, which can be especially difficult for rare or elusive species where field surveys require high effort and considerable cost. Populations of many reptiles, including Sonoran desert tortoises (Gopherus agassizii), are challenging to monitor effectively because they are cryptic, they occur at low densities, and their activity is limited both seasonally and daily. We compared efficiency and statistical power of 2 survey methods appropriate for tortoises and other rare vertebrates, line-transect distance sampling and site occupancy. In 2005 and 2006 combined, we surveyed 120 1-km transects to estimate density and 40 3-ha plots 5 times each to estimate occupancy of Sonoran desert tortoises in 2 mountain ranges in southern Arizona, USA. For both mountain ranges combined, we estimated density to be 0.30 adult tortoises/ha (95% CI = 0.17–0.43) and occupancy to be 0.72 (95% CI = 0.56–0.89). For the sampling designs we evaluated, monitoring efforts based on occupancy were 8–36% more efficient than those based on density, when contrasting only survey effort, and 17–30% more efficient when contrasting total effort (surveying, hiking to and from survey locations, and radiotracking). Occupancy had greater statistical power to detect annual declines in the proportion of area occupied than did distance sampling to detect annual declines in density. For example, we estimated that power to detect a 5% annual decline with 10 years of annual sampling was 0.92 (95% CI = 0.75–0.98) for occupancy and 0.43 (95% CI = 0.35–0.52) for distance sampling. Although all sampling methods have limitations, occupancy estimation offers a promising alternative for monitoring populations of rare vertebrates, including tortoises in the Sonoran Desert.
In the mid-Atlantic region, urban sprawl and development have resulted in habitat alterations and fragmentation; however, the effects on eastern box turtle (Terrapene carolina carolina) populations are unclear. To investigate the status of eastern box turtle populations in a fragmented landscape, we used mark—recapture and radiotelemetry to estimate population density, sex ratio, age structure, and survival on 4 study areas with differing degrees of isolation and human disturbance in northern New Castle County, Delaware, USA. We estimated adult population densities ranging from 0.81 turtles/ha to 3.62 turtles/ha among our 4 study areas. Sex ratios were male-biased at 2 study areas and balanced at 2 study areas. Proportion of juveniles ranged from 0% to 31%. Estimated annual survival rate ranged from 0.813 to 0.977. Mortality of radiotagged and marked turtles was primarily due to natural causes, but mowing was the primary cause of human-induced mortality. We found evidence of population decline at one study area due to low survival and recruitment. Human disturbances, isolation, and habitat composition appear to have the greatest influence on the box turtle populations we studied. To minimize mortality from human disturbance, we suggest planting crops adjacent to forest habitat that require no mowing or mowing at a height ≥15 cm. (JOURNAL OF WILDLIFE MANAGEMENT 72(3):745–753; 2008)
We describe an approach for estimating occupancy rate or the proportion of area occupied when heterogeneity in detection probability exists as a result of variation in abundance of the organism under study. The key feature of such problems, which we exploit, is that variation in abundance induces variation in detection probability. Thus, heterogeneity in abundance can be modeled as heterogeneity in detection probability. Moreover, this linkage between heterogeneity in abundance and heterogeneity in detection probability allows one to exploit a heterogeneous detection probability model to estimate the underlying distribution of abundances. Therefore, our method allows estimation of abundance from repeated observations of the presence or absence of animals without having to uniquely mark individuals in the population.
Over the last 30 years there has been a great deal of interest in investigating patterns of species co‐occurrence across a number of locations, which has led to the development of numerous methods to determine whether there is evidence that a particular pattern may not have occurred by random chance.
A key aspect that seems to have been largely overlooked is the possibility that species may not always be detected at a location when present, which leads to ‘false absences’ in a species presence/absence matrix that may cause incorrect inferences to be made about co‐occurrence patterns. Furthermore, many of the published methods for investigating patterns of species co‐occurrence do not account for potential differences in the site characteristics that may partially (at least) explain non‐random patterns (e.g. due to species having similar/different habitat preferences).
Here we present a statistical method for modelling co‐occurrence patterns between species while accounting for imperfect detection and site characteristics. This method requires that multiple presence/absence surveys for the species be conducted over a reasonably short period of time at most sites. The method yields unbiased estimates of probabilities of occurrence, and is practical when the number of species is small (< 4).
To illustrate the method we consider data collected on two terrestrial salamander species, Plethodon jordani and members of the Plethodon glutinosus complex, collected in the Great Smoky Mountains National Park, USA. We find no evidence that the species do not occur independently at sites once site elevation has been allowed for, although we find some evidence of a statistical interaction between species in terms of detectability that we suggest may be due to changes in relative abundances.
The fraction of sampling units in a landscape where a target species is present (occupancy) is an extensively used concept in ecology. Yet in many applications the species will not always be detected in a sampling unit even when present, resulting in biased estimates of occupancy. Given that sampling units are surveyed repeatedly within a relatively short timeframe, a number of similar methods have now been developed to provide unbiased occupancy estimates. However, practical guidance on the efficient design of occupancy studies has been lacking.
In this paper we comment on a number of general issues related to designing occupancy studies, including the need for clear objectives that are explicitly linked to science or management, selection of sampling units, timing of repeat surveys and allocation of survey effort. Advice on the number of repeat surveys per sampling unit is considered in terms of the variance of the occupancy estimator, for three possible study designs.
We recommend that sampling units should be surveyed a minimum of three times when detection probability is high (> 0·5 survey ⁻¹ ), unless a removal design is used.
We found that an optimal removal design will generally be the most efficient, but we suggest it may be less robust to assumption violations than a standard design.
Our results suggest that for a rare species it is more efficient to survey more sampling units less intensively, while for a common species fewer sampling units should be surveyed more intensively.
Synthesis and applications . Reliable inferences can only result from quality data. To make the best use of logistical resources, study objectives must be clearly defined; sampling units must be selected, and repeated surveys timed appropriately; and a sufficient number of repeated surveys must be conducted. Failure to do so may compromise the integrity of the study. The guidance given here on study design issues is particularly applicable to studies of species occurrence and distribution, habitat selection and modelling, metapopulation studies and monitoring programmes.
We present an algebraic method for estimating highway mortality in snakes, based on careful data collection during low-speed driving on paved roads. A total of 368 snakes (104 live, 264 dead) were recorded over four years on State Route 85 from Why to Lukeville, Pima Co., Arizona, during 15 525 km of road-cruising; mostly within Organ Pipe Cactus National Monument. We computed an estimate of 2383 snakes killed 13·5/km/year) during the four years on this stretch of pavement, estimate that actual numbers killed would be closer to 4000 (22·5/km/year). Two taxa of special conservation interest, the Mexican rosy boa Lichanura trivirgata and the Organ Pipe shovelnosed snake Chionactis palarostris, appear to be relatively strongly impacted by highway mortality. Overall, along our 44·1 km transect, the estimated highway mortality amounts to removal of 5 km2 of snake population over the four years of study. It is clear that roadways, especially if paved, substantially damage snake populations.
The home-range, monthly and seasonal habitat use and hibernacula selection of eastern
box turtles (Terrapene carolina carolina) were investigated in 2003 using radiotelemetric
applications at two sites (West Rock Ridge and Mountain Road) located along
the West Rock Ridge basalt trap rock ecosystem in Connecticut. Radio-tracked turtles
were re-located an average of 61.6 times each from May until they entered hibernation in
October. Turtle (n = 14) home-range size averaged 4.97 ha with a range from 0.99 - 16.2
ha. Female turtles (n = 7) had slightly smaller home-ranges (avg. 4.0 ha) than males (n =
7; avg. 6.74 ha). The general trends in monthly habitat use for West Rock Ridge (WRR)
showed variable habitat use in May with late successional upland (LSU) habitats being
preferred. In June there was a shift in the preferred habitat from LSU to open canopy
wetlands (OCW). During the months of July, August, and September similar use of
closed canopy wetland (CCW) and mature upland (MU) habitats were observed. During
October there was a large shift to MU. Similar trends were seen in the seasonal habitat
use at WRR with OCW being preferred in the spring, CCW and MU being preferred in
early to late summer, with habitat preferences towards MU in fall. General trends in
monthly habitat use for Mountain Road (MR) showed preferences towards early
successional upland (ESU) habitats during May and June and ESU and MU habitats
during the months of July and August. Habitat use shifted to predominately MU habitats
during the months of September and October. Similar trends were seen in the seasonal
habitat use at MR with ESU being preferred in the spring, ESU and MU in the early
summer, and MU in late summer and fall. All turtles from both sites hibernated in
mature upland habitats.
A bias correction to the Akaike information criterion, AIC, is derived for regression and autoregressive time series models.
The correction is of particular use when the sample size is small, or when the number of fitted parameters is a moderate to
large fraction of the sample size. The corrected method, called AICC, is asymptotically efficient if the true model is infinite dimensional. Furthermore, when the true model is of finite dimension,
AICC is found to provide better model order choices than any other asymptotically efficient method. Applications to nonstationary
autoregressive and mixed autoregressive moving average time series models are also discussed.
The information criterion AIC was introduced to extend the method of maximum likelihood to the multimodel situation. It was obtained by relating the successful experience of the order determination of an autoregressive model to the determination of the number of factors in the maximum likelihood factor analysis. The use of the AIC criterion in the factor analysis is particularly interesting when it is viewed as the choice of a Bayesian model. This observation shows that the area of application of AIC can be much wider than the conventional i.i.d. type models on which the original derivation of the criterion was based. The observation of the Bayesian structure of the factor analysis model leads us to the handling of the problem of improper solution by introducing a natural prior distribution of factor loadings.
Researchers have used occupancy, or probability of occupancy, as a response or state variable in a variety of studies (e.g., habitat modeling), and occupancy is increasingly favored by numerous state, federal, and international agencies engaged in monitoring programs. Recent advances in estimation methods have emphasized that reliable inferences can be made from these types of studies if detection and occupancy probabilities are simultaneously estimated. The need for temporal replication at sampled sites to estimate detection probability creates a trade-off between spatial replication (number of sample sites distributed within the area of interest/inference) and temporal replication (number of repeated surveys at each site). Here, we discuss a suite of questions commonly encountered during the design phase of occupancy studies, and we describe software (program GENPRES) developed to allow investigators to easily explore design trade-offs focused on particularities of their study system and sampling limitations. We illustrate the utility of program GENPRES using an amphibian example from Greater Yellowstone National Park, U.S.A.
Conserving amphibians and reptiles in the new millennium
- J W Gibbons
Gibbons, J. W., and P. W. Stangel. 1999. Conserving amphibians and
reptiles in the new millennium. Proceeding of the Partners in Amphibian
and Reptile Conservation (PARC) Conference 2. Savannah River Ecology
Laboratory Herp Outreach Publication 2, Aiken, South Carolina, USA.
A new coding system for hardshelled turtles
- C Ernst
- R Barbour
- M Hershey
Ernst, C., R. Barbour, and M. Hershey. 1974. A new coding system for
hardshelled turtles. Transactions of the Kentucky Academy of Science
35:27-28.
Box turtle road mortality estimates. Massachusetts Division of Fisheries and Wildlife, Natural Heritage and Endangered Species Program
- B Compton
Compton, B. 2010. Box turtle road mortality estimates. Massachusetts
Division of Fisheries and Wildlife, Natural Heritage and Endangered
Species Program, Westborough, USA.
Spatial ecology of eastern box turtles (Terrapene c. carolina) in central Massachusetts. Dissertation
- L L Willey
Willey, L. L. 2010. Spatial ecology of eastern box turtles (Terrapene c.
carolina) in central Massachusetts. Dissertation, University of Massachusetts, Amherst, USA.
PRESENCE 5.5-Software to estimate patch occupancy and related parameters
- J E Hines
Hines, J. E. 2012. PRESENCE 5.5-Software to estimate patch occupancy
and related parameters. U.S. Geological Survey, Patuxent Wildlife
Research Center, Laurel, Maryland, USA.
Occupancy estimation and modeling: Inferring patterns and dynamics of species occurrence
- D I Mackenzie
- J D Nichols
- J A Royle
- K H Pollock
- L L Bailey
- J E Hines
MacKenzie, D. I., J. D. Nichols, J. A. Royle, K. H. Pollock, L. L. Bailey, and
J. E. Hines. 2005. Occupancy estimation and modeling: Inferring patterns
and dynamics of species occurrence. Academic Press, New York, New
York, USA.
Estimating the number of animals in wildlife populations. Pages 106–153 in Techniques for wildlife investigations and management
- R A Lancia
- W L Kendall
- K H Pollock
- J D Nichols
Lancia, R. A., W. L. Kendall, K. H. Pollock, and J. D. Nichols. 2005.
Estimating the number of animals in wildlife populations. Pages 106–153
in C. E. Braun, editor. Techniques for wildlife investigations and
management. The Wildlife Society, Bethesda, Maryland, USA.
Terrapene carolina (Linnaeus 1758) eastern box turtle, common box turtle Account 085 in Conservation biology of freshwater turtles and tortoises: a compilation project of the IUCN/ SSC Tortoise and Freshwater Turtle Specialist Group
- A R Kiester
- L L Willey
Kiester, A. R., and L. L. Willey. 2015. Terrapene carolina (Linnaeus 1758) eastern box turtle, common box turtle. Account 085 in A. G. J. Rhodin,
P. C. H. Pritchard, P. P. van Dijk, R. A. Saumure, K. A. Buhlmann, J. B.
Iverson, and R. A. Mittermeier, editors. Conservation biology of
freshwater turtles and tortoises: a compilation project of the IUCN/
SSC Tortoise and Freshwater Turtle Specialist Group. Chelonian
Research Monographs 5:1–25.
Conservation Plan for the eastern box turtle in Massachusetts. Massachusetts Division of Fisheries and Wildlife, Natural Heritage and Endangered Species Program
- L Erb
Erb, L. 2011. Conservation Plan for the eastern box turtle in Massachusetts.
Massachusetts Division of Fisheries and Wildlife, Natural Heritage and
Endangered Species Program, Westborough, Massachusetts, USA.