Site-specific Covariates Research Articles

Acquisition-invariant brain MRI segmentation with informative uncertainties

Combining multi-site data can strengthen and uncover trends, but is a task that is marred by the influence of site-specific covariates that can bias the data and, therefore, any downstream analyses. Post-hoc multi-site correction methods exist but have strong assumptions that often do not hold in real-world scenarios. Algorithms should be designed in a way that can account for site-specific effects, such as those that arise from sequence parameter choices, and in instances where generalisation fails, should be able to identify such a failure by means of explicit uncertainty modelling. This body of work showcases such an algorithm that can become robust to the physics of acquisition in the context of segmentation tasks while simultaneously modelling uncertainty. We demonstrate that our method not only generalises to complete holdout datasets, preserving segmentation quality but does so while also accounting for site-specific sequence choices, which also allows it to perform as a harmonisation tool.

Patterns of Florida Bonneted Bat Occupancy at the Northern Extent of Its Range

Abstract The Florida bonneted bat Eumops floridanus is a rare, endemic bat of South Florida that roosts in woodpecker cavities and anthropogenic structures such as roofing tiles, chimneys, and bat houses. The northernmost occurrences of the bonneted bat are from mature pine forests at the Avon Park Air Force Range, Florida. We used ultrasonic acoustic recorders to understand bonneted bat activity and habitat occupancy. We modeled occupancy using a hierarchical Bayesian analysis and included site- and time-specific covariates of detection probability and site-specific covariates of occupancy. Probability of detection was low throughout Avon Park Air Force Range but increased with Julian date. In most habitats, occupancy was poorly estimated, except for flatwood mature pinelands where occupancy was low (0.23 ± 0.06). As distance from red-cockaded woodpecker colonies increased, occupancy decreased (β = −1.19 ± 0.26 SD). At the northernmost extent of the range, and throughout much of the historic range, increasing the expanse of mature, fire-maintained forest systems will increase habitat for the bonneted bat and lead to faster population recovery.

Dynamic occupancy modelling to determine the status of a Critically Endangered lizard

AbstractMonitoring of cryptic or threatened species poses challenges for population assessment and conservation, as imperfect detection gives rise to misleading inferences about population status. We used a dynamic occupancy model that explicitly accounted for occupancy, colonization, local extinction and detectability to assess the status of the endemic Critically Endangered Bermuda skink Plestiodon longirostris. During 2015–2017, skinks were detected at 13 of 40 surveyed sites in Bermuda, two of which were new records. Ten observation-level and site-specific covariates were used to explore drivers of occupancy, colonization, extinction and detectability. Sites occupied by skinks tended to be islands with rocky coastal habitat and prickly pear cacti; the same variables were also associated with reduced risk of local extinction. The presence of seabirds appeared to encourage colonization, whereas the presence of cats had the opposite effect. The probability of detection was p = 0.45, and on average, five surveys were needed to reliably detect the presence of skinks with 95% certainty. However, skinks were unlikely to be detected on sites with cat and rat predators. Dynamic occupancy models can be used to elucidate drivers of occupancy dynamics, which in turn can inform species conservation management. The survey effort needed to determine population changes over time can be derived from estimates of detectability.

Can we model distribution of population abundance from wildlife–vehicles collision data?

Reliable estimates of the distribution of species abundance are a key element in wildlife studies, but such information is usually difficult to obtain for large spatial or long temporal scales. Wildlife–vehicle collision (WVC) data is systematically registered in many countries and could be used as a proxy of population abundance if the number of WVC in each territory increase with the population abundance. However, factors such as road density or human population should be controlled to obtain accurate abundance estimations from WVC data. Here, we propose a hierarchical modeling approach using the Royle–Nichols model for detection–non‐detection data to obtain population abundance indices from WVC. Relative abundance and individual detectability were modeled for two species, wild boar Sus scrofa and roe deer Capreolus capreolus at 10 × 10 km cells in mainland Spain from WVC data using environmental, anthropological and temporal covariates. For each cell, a detection was annotated if at least one WVC was recorded at each month (used as survey occasion). The predicted abundance indices were compared with raw hunting statistics at region level to assess the performance of the modeling approach. Site specific covariates such as road density or administrative region and the month of the year, affected individual detectability, with higher WVC probability between October and December for wild boar and between April and July for roe deer. Wild boar and roe deer abundance can be explained by both, bioclimatic and land cover covariates. Abundance indices obtained from WVC data were significantly positively correlated with regional raw hunting yields for both species. We presented empirical evidence supporting that accurate wildlife abundance indices at fine spatial resolution can be generated from WVC data when individual detectability is considered in the modeling process.

Physiological acclimation of elk during population restoration in the Missouri Ozarks, USA.

Conservation translocations—the intentional movement of animals to restore populations—have increased over the past 30 years to halt and reverse species declines and losses. However, there are many challenges translocated animals face that should be considered for restoration programs to be successful. Understanding how long it takes for translocated animals to acclimate to these challenges and their new landscape is a critical component of post-release population management. Physiological measures such as hormone responses are increasingly used to assess animal responses and acclimation to disturbances including translocation. We determined the physiological acclimation period of elk (Cervus canadensis) translocated to the Missouri Ozarks, USA, as part of a restoration effort. From 2011 to 2013, we translocated 108 GPS-radio-collared elk from Kentucky, USA, to Missouri, USA, and collected faecal samples for glucocorticoid metabolite extraction to use as an indicator of physiological acclimation. We modelled the response of population-wide faecal glucocorticoid metabolites (fGCMs) across the initial 9 years of the restoration in response to days following release and additional site-specific covariates. Presence of white-tailed deer (Odocoileus virginianus) hunts and monthly precipitation levels were positively and negatively associated with fGCM levels, respectively. Concurrent with influences from site-specific conditions on the release landscape, fGCM levels declined following release. We identified a breakpoint in fGCM decline at ~42 days following translocation releases suggesting elk acclimated physiologically relatively quickly compared to other species. The fast physiological acclimation by Missouri elk suggests effective use of temporary post-release management efforts. Determining how quickly animals acclimate following translocations allows researchers to tailor post-release management plans to each species’ needs, thus maximizing the success of future translocation efforts while minimizing costs.

Predicting the spatial distribution of soil organic carbon stock in Swedish forests using a group of covariates and site-specific data

Abstract. The status of the soil organic carbon (SOC) stock at any position in the landscape is subject to a complex interplay of soil state factors operating at different scales and regulating multiple processes resulting either in soils acting as a net sink or net source of carbon. Forest landscapes are characterized by high spatial variability, and key drivers of SOC stock might be specific for sub-areas compared to those influencing the whole landscape. Consequently, separately calibrating models for sub-areas (local models) that collectively cover a target area can result in different prediction accuracy and SOC stock drivers compared to a single model (global model) that covers the whole area. The goal of this study was therefore to (1) assess how global and local models differ in predicting the humus layer, mineral soil, and total SOC stock in Swedish forests and (2) identify the key factors for SOC stock prediction and their scale of influence. We used the Swedish National Forest Soil Inventory (NFSI) database and a digital soil mapping approach to evaluate the prediction performance using random forest models calibrated locally for the northern, central, and southern Sweden (local models) and for the whole of Sweden (global model). Models were built by considering (1) only site characteristics which are recorded on the plot during the NFSI, (2) the group of covariates (remote sensing, historical land use data, etc.) and (3) both site characteristics and group of covariates consisting mostly of remote sensing data. Local models were generally more effective for predicting SOC stock after testing on independent validation data. Using the group of covariates together with NFSI data indicated that such covariates have limited predictive strength but that site-specific covariates from the NFSI showed better explanatory strength for SOC stocks. The most important covariates that influence the humus layer, mineral soil (0–50 cm), and total SOC stock were related to the site-characteristic covariates and include the soil moisture class, vegetation type, soil type, and soil texture. This study showed that local calibration has the potential to improve prediction accuracy, which will vary depending on the type of available covariates.

There's always a better way: The application of eDNA to effectively assess biodiversity

Our ecosystem monitoring methodologies focus on data collection for reporting purposes that may not serve to identify the systematic causes of ecological change. Managers need precise and timely information at appropriate scales to build ecosystem resilience. Traditional species detection methodologies offer little information when species abundance are low, especially in large water ecosystems such as the Great Lakes. Species not found during monitoring doesn’t necessarily mean that species are absent. Moreover, even if a change in the ecosystem is detected, it is often not possible to determine its cause at a spatiotemporal scale or a trophic cascade level. As a result, we often find ourselves being reactive in our mitigation measures. Before irreversible change occurs, we must be guided by a better understanding of the actual ecological landscape which Environmental DNA (eDNA) may help provide. eDNA is a potential tool to effectively overcome traditional species survey limitations currently in use at many Parks Canada sites (Supplemental file 1). As various organisms interact with the environment, DNA is expelled and accumulates in their surroundings. Such samples can be analyzed by high-throughput DNA sequencing methods for rapid measurement and monitoring of biodiversity. Access to this genetic information makes a critical contribution to the understanding of population size, species distribution, and population dynamics for species not well documented. Despite the increasing use of eDNA in conservation practice, it requires further methodological improvement for greater influence on management decisions. The tool requires standardized protocols based on site-specific covariates and objectives. We’re working to tackle the challenge with 2 objectives: (1) to combine traditional biomonitoring knowledge and metagenomics to further develop eDNA as a reliable sampling tool for Parks Canada and (2) to support site-specific monitoring objectives for species-at-risk, invasive species, aquatic species inventories, and/or culturally significant species. The overall goal is to increase our capacity to make more informed, timely, regionally-coordinated conservation decisions through the rapid and sensitive species detection methods offered by eDNA.

Reconstructing biological invasions using public surveys: a new approach to retrospectively assess spatio-temporal changes in invasive spread

Management of biological invasions increasingly relies on the knowledge of invasive species’ dispersal pathways that operate during introduction and post-introduction dispersal. However, the early stages of biological invasions (introduction, establishment, and initial spread) are usually poorly documented, limiting our understanding of post-introduction dispersal and the role of humans in invasive spread. We aim to assess a new approach to retrospectively understand spatio-temporal patterns of introduction, establishment, dispersal, and spread in biological invasions, using the case study of an ongoing invasion of the Indian bullfrog (Hoplobatachus tigerinus) on the Andaman archipelago, Bay of Bengal. We sampled 91 villages on eight human inhabited islands of the Andaman archipelago from 2015 to 2016. We assessed the occurrence of the bullfrog using visual encounter surveys and recorded the invasion history (year of establishment, source site, and dispersal pathway) for each site by surveying 892 key informants (farmers, plantation workers, and aqua-culturists). We sought to corroborate the reconstructed invasion history with false positive occupancy modelling, using site specific covariates that corresponded to hypotheses on specific dispersal pathways. The bullfrog occurred in at least 62% of the sampled sites spread over six islands, a dramatic increase to the previously known invaded range. The bullfrog was most likely introduced in early 2000s, and its exponential expansion has occurred since 2009. ‘Contaminants’ of fish culture trade and intentional ‘release’ were reported to be the primary pathways of introduction and post-introduction dispersal, facilitating introductions from the Indian mainland and inter-island transfers. False-positive occupancy modelling confirmed that three sites on the archipelago influenced the invasion disproportionately by acting as dispersal hubs. The study elucidates the efficacy of using public surveys to identify dispersal pathways and hubs, and to understand invasive spread, when such information is typically unavailable otherwise. The proposed approach is scalable to other systems and species.

Understanding Habitat Quality for Preble's Meadow Jumping Mouse: How Survival Responds to Vegetation Structure and Composition

AbstractHabitat loss is the primary conservation concern for many rare species; yet, it is unclear what habitat components are vital for the persistence of many rare species. The Preble's meadow jumping mouse Zapus hudsonius preblei (PMJM) was listed as Threatened under the Endangered Species Act because of habitat loss in Colorado and Wyoming. The PMJM is restricted to dense shrub cover and dense herbaceous vegetation along wetland and riverine systems. Although it is well established by researchers that PMJM infrequently leave these habitat types, it is unclear what riparian vegetation structure and composition is important to PMJM survival. We collected and analyzed a 14-y PMJM vegetation-monitoring data set and a PMJM mark–recapture data set to assess the influence of vegetation structure and composition on PMJM annual survival. Using individual, group, and site-specific covariates we estimated survival and evaluated the influence of such covariates on annual survival. Annual survival for PMJM was low (< 10%), increasing with body mass, grass cover, and shrub cover, and decreasing with meadow vole Microtus pennsylvanicus captures. The PMJM use of and occupation of dense riparian habitats may increase individual survival, and likely increases population persistence. Thus, habitat modifications that reduce grass cover and shrub regeneration, such as grazing, suburban development, and development-altered hydrology, will be detrimental to PMJM populations. Given the low annual survival of PMJM, it is important for conserving PMJM populations to minimize disturbances to the vegetation structure, composition, and density, and the hydrologic processes that support them.

Site‐Specific Covariates Affecting Yield Response to Nitrogen of Late‐Sown Maize in Central Argentina

Core Ideas We used linear mixed effects models to explore maize yield response to applied N. Final models accurately described the observed data (R2 = 0.93). Best models indicated that yield response to applied N depended on soil N and soil type. Information is useful to optimize management decisions on N fertilizer rates. Resulting models are better than traditional ones based on ordinary least squares. Optimizing fertilizer rates is a common problem in modern agriculture. Frequently used response models ignore basic statistical assumptions and do not allow quantifying the effects of variables influencing yield response to fertilizer, generating uncertainty in fertilizer rate recommendations. We used linear mixed‐effects models to explore maize (Zea mays L.) yield response to applied nitrogen (N) in late sowings, and we tested different predictors for explaining yield responses across sites. Data included yield response trials to applied N at 17 different environments (combination site × year) with four to five N rates replicated twice in each trial. The best model (Model A) that included significant effect of N rate applied, sowing date, and soil N‐NO3 at sowing described grain yield variations with high accuracy (R2 = 0.93). Another best model (Model B) showed that soil type as additional variable affected significantly yield response to applied N. The final model indicated that the overall response across sites was characterized by a linear coefficient of 67 kg grain ha−1 per additional kg N ha−1 applied and a quadratic coefficient of −0.37 kg grain ha−1 per additional kg N ha−1 applied. Across all sites, soil N‐NO3 at sowing (explored range from 34 to 356 kg N ha−1) explained 46% of the variability in the linear yield response to applied N. We proposed a method and generated statistical models with site specific covariates that can help optimize farmers’ decisions on the use of optimal N fertilizer rates.

Integration of routine QA data into mega-analysis may improve quality and sensitivity of multisite diffusion tensor imaging studies.

A novel mega-analytical approach that reduced methodological variance was evaluated using a multisite diffusion tensor imaging (DTI) fractional anisotropy (FA) data by comparing white matter integrity in people with schizophrenia to controls. Methodological variance was reduced through regression of variance captured from quality assurance (QA) and by using Marchenko-Pastur Principal Component Analysis (MP-PCA) denoising. N = 192 (119 patients/73 controls) data sets were collected at three sites equipped with 3T MRI systems: GE MR750, GE HDx, and Siemens Trio. DTI protocol included five b = 0 and 60 diffusion-sensitized gradient directions (b = 1,000 s/mm2 ). In-house DTI QA protocol data was acquired weekly using a uniform phantom; factor analysis was used to distil into two orthogonal QA factors related to: SNR and FA. They were used as site-specific covariates to perform mega-analytic data aggregation. The effect size of patient-control differences was compared to these reported by the enhancing neuro imaging genetics meta-analysis (ENIGMA) consortium before and after regressing QA variance. Impact of MP-PCA filtering was evaluated likewise. QA-factors explained ∼3-4% variance in the whole-brain average FA values per site. Regression of QA factors improved the effect size of schizophrenia on whole brain average FA values-from Cohen's d = .53 to .57-and improved the agreement between the regional pattern of FA differences observed in this study versus ENIGMA from r = .54 to .70. Application of MP-PCA-denoising further improved the agreement to r = .81. Regression of methodological variances captured by routine QA and advanced denoising that led to a better agreement with a large mega-analytic study.

Understanding the Impact of Microcredit Expansions: A Bayesian Hierarchical Analysis of 7 Randomised Experiments

Bayesian hierarchical models serve as a standard methodology for aggregation and synthesis, used widely in statistics and other disciplines. I use this framework to aggregate the data from seven randomised experiments of expanding access to microcredit, assessing both the general impact of the intervention and the heterogeneity across contexts. The general impact on household profits is small, with a posterior mean of 26 USD PPP per year, and an impact of zero lies well within the central 50% posterior credible interval. Standard pooling metrics for the studies indicate 65-95% pooling on the treatment effects, suggesting that the site-specific effects are informative for each other and for the general case. Further analysis incorporating household covariates shows that the cross-study heterogeneity is almost entirely generated by heterogeneous effects for the 27% households who previously operated businesses before microcredit expansion. A cautious assessment of the correlations between site-specific covariates and treatment effects using a Bayesian Ridge procedure indicates that the interest rate on the microloans has the strongest correlation.

Patch occupancy and habitat of the hops azure (Celastrina humulus), a rare North American endemic butterfly: insights for monitoring and conservation

The hops azure (Celastrinahumulus Scott & D. Wright 1998) is a rare butterfly found along the Front Range of Colorado. Data on the prevalence of the butterfly and its preferred habitats are lacking. To describe the habitat of C.humulus at the southern part of its range, explore what factors impact C.humulus detectability, and estimate C.humulus habitat use along a riparian area known to support it, we conducted an occupancy analysis along the largest riparian system at the U.S. Air Force Academy (USAFA) in Colorado, USA. We used environmental and site-specific covariates to model the probability of detection and the probability of occupancy. Probability of detecting C.humulus was influenced by the amount of cloud cover during sampling, while the probability of occupancy was influenced by the total area of the host plant (wild hops) at the site. Probability of detection was higher during the first visit (69 %) than the second visit (64 %), and the probability of occupancy was higher (77 %) than assumed (30 %). Despite the host plant being patchily distributed throughout the butterfly’s range, the riparian areas at USAFA had a high prevalence of both wild hops and C.humulus. We use the project findings to develop future sampling efforts for the butterfly along tributaries at other locales within the butterfly’s range.

Dynamic Models for Longitudinal Butterfly Data

Abstract We present models which provide succinct descriptions of longitudinal seasonal insect count data. This approach produces, for the first time, estimates of the key parameters of brood productivities. It may be applied to univoltine and bivoltine species. For the latter, the productivities of each brood are estimated separately, which results in new indices indicating the contributions from different generations. The models are based on discrete distributions, with expectations that reflect the underlying nature of seasonal data. Productivities are included in a deterministic, auto-regressive manner, making the data from each brood a function of those in the previous brood. A concentrated likelihood results in appreciable efficiency gains. Both phenomenological and mechanistic models are used, including weather and site-specific covariates. Illustrations are provided using data from the UK Butterfly Monitoring Scheme, however the approach is perfectly general. Consistent associations are found when estimates of productivity are regressed on northing and temperature. For instance, for univoltine species productivity is usually lower following milder winters, and mean emergence times of adults for all species have become earlier over time, due to climate change. The predictions of fitted dynamic models have the potential to improve the understanding of fundamental demographic processes. This is important for insects such as UK butterflies, many species of which are in decline. Supplementary materials for this article are available online.

Efficiency of hair snares and camera traps to survey mesocarnivore populations

Mammalian carnivore communities affect entire ecosystem functioning and structure. However, their large spatial requirements, preferred habitats, low densities, and elusive behavior deem them difficult to study. In recent years, noninvasive techniques have become much more common as they can be used to monitor multiple carnivore species across large areas at a relatively modest cost. Hair snares have the potential to fulfill such requirements, but have rarely been tested in Europe. Our objective was to quantitatively assess the effectiveness of hair snares for surveying mesocarnivores in the Iberian Peninsula (Southwestern Europe), by comparison with camera-trapping. We used an occupancy modeling framework to assess method-specific detectability and occupancy estimates and hypothesized that detection probabilities would be influenced by season, sampling method, and habitat-related variables. A total of 163 hair samples were collected, of which 136 potentially belonged to mesocarnivores. Genetic identification success varied with diagnostic method: 25.2 % using mitochondrial CR, and 9.9 % using the IRBP nuclear gene. Naïve occupancy estimates were, in average, 5.3 ± 1.2 times higher with camera-trapping than with hair-snaring, and method-specific detection probabilities revealed that camera traps were, in average, 6.7 ± 1.1 times more effective in detecting target species. Overall, few site-specific covariates revealed significant effects on mesocarnivore detectability. Camera traps were a more efficient method for detecting mesocarnivores and estimating their occurrence when compared to hair snares. To improve hair snares' low detection probabilities, we suggest increasing the number of sampling occasions and the frequency at which hair snares are checked. With some refinements to increase detection rates and the success of genetic identification, hair-snaring methods may be valuable for providing deeper insights into population parameters, attained through adequate analysis of genetic information, that is not possible with camera traps.

Multi-event models reveal the absence of interaction between an invasive frog and a native endangered amphibian

Among the multiple factors involved in the decline of amphibians, competition with alien species remains understudied. In Western Europe, non-native marsh frogs (Pelophylax ridibundus and other sister species) have been widely imported for gastronomic purposes and have then often escaped from captivity. Apart from closely related species, the impact of the subsequent colonization of non-native species on native amphibians is still unknown. In this study, we analysed the response of a threatened species, the yellow-bellied toad (Bombina variegata), faced with the presence of alien frogs in mountain rivers in France. We studied the co-occurrence pattern of the native toad and the alien frog, while taking into account co-dependence of the detection rates in both species. We tested three main scenarios to explain a non-random pond occurrence for each species and to establish their respective habitat preferenda: (1) a pond-area-dependent scenario, predicting different responses from each species for a given pond size, (2) a fish-dependent scenario, predicting different responses from each species given the presence of fish in ponds, and (3) a floodplain-width-dependent scenario, predicting different responses from each species given the geomorphological characteristics of the floodplain. Taking into account site-specific covariates, we concluded that introduced frogs do not currently have an impact on the native population of the yellow-bellied toad.

Bias-corrected Pearson estimating functions for Taylor’s power law applied to benthic macrofauna data

Estimation of Taylor’s power law for species abundance data may be performed by linear regression of the log empirical variances on the log means, but this method suffers from a problem of bias for sparse data. We show that the bias may be reduced by using a bias-corrected Pearson estimating function. Furthermore, we investigate a more general regression model allowing for site-specific covariates. This method may be efficiently implemented using a Newton scoring algorithm, with standard errors calculated from the inverse Godambe information matrix. The method is applied to a set of biomass data for benthic macrofauna from two Danish estuaries.

Constraints in the identification and interpretation of ozone as a significant predictor of effects on the supersensitive indicator Nicotiana tabacum Bel-W3 in biomonitoring studies

Biomonitoring by means of the supersensitive cultivar Nicotiana tabacum Bel-W3 is frequently used to obtain information on ozone effects on plants and estimates of ozone exposure. However, most of biomonitoring surveys do not account for other environmental variables (predictors in a statistical model) and their inherent multicollinearity with ozone. We tested the relative role of different predictors (fixed: time and site; random: ozone, temperature and humidity) on height growth and on the development of visible foliar symptoms of N. tabacum Bel-W3 plants. To do this, we investigated a relatively small area (256 km 2), used a random design at every stage of the survey, controlled watering and protected plants from direct solar radiation and wind. QA/QC procedures were adopted at every stage of the investigation. Linear correlation shows that Leaf Injury Index (LII) and height increment (H.I.) positively related to ozone concentration, elevation and temperature, and negatively to relative humidity. All the predictors correlate to each other. However, relationships between response and ozone vary with the site and the monitoring week. The effect of the random factor “ozone” in combination with fixed factors “site” and “time” on the response variables was therefore formally investigated using the ANCOVA model. Besides ozone, the interactions “ozone × site” and “ozone × time” resulted always significant (0.001 < P < 0.05). While the factor “time” emphasize the inherent development of injury and growth through time, the interaction “ozone × site” pointed out the importance of local conditions. When watering, solar radiation, wind and plant characteristics were controlled, the remaining site-specific covariates of interest were temperature ( T) and humidity (RH). When T and RH were accounted for by means of partial correlation analysis, no significant relationship was found between ozone and LII. On the other side, O 3 and RH resulted significant for both absolute and relative height increment. In short, LII seemed to be not solely dependent on ozone, T and RH, but showed to integrate their combined effect. On the other side, H.I. seemed to be favoured by high RH and T, and depressed by high ozone. Based on the above results, we recommend caution when handling bioindicator data: if the purpose is to infer ozone concentrations by leaf injury data, results may be affected by a serious bias, as the frequently reported correlations may be partly an artefact due to co-variation between predictors.

The effects of survey protocol on detection probabilities and site occupancy estimates of summer breeding anurans

Recent declines in amphibian populations have created an urgent need for large-scale, long- term monitoring efforts and many anuran monitoring programs have been established that utilize calling surveys. Calling surveys can be effective monitoring tools; however, differences among survey protocols may bias survey results. Failure to take into account detection probabilities when monitoring anurans can lead to inaccurate inferences about site occupancy, since non-detections in survey data do not necessarily mean that a species is absent unless the probability of detection is 1. We used a likelihood-based method, in the form of the computer program PRESENCE, to estimate detection probabilities and site occupancy rates for summer-breeding anurans in the Western Piedmont of North Carolina. Using detection data from calling surveys, we evaluated how detectability and site occupancy for five anuran species were influenced by 1) time spent listening at each site, 2) number of surveys per site, and 3) sample- and site-specific covariates. We found considerable variation among species with regards to detection probability and site occupancy across survey duration and sampling occasion. Although 13% of all species detection occurred after 3 min, longer surveys did not significantly increase detectability of individual species. We found that detectability varied more with sampling occasion than with survey duration for each species. Covariates had differing effects on occupancy and detectability among individual species. Multiple surveys per site within a season are necessary to eliminate biased detection probabilities, but we found that 3- or 5-min surveys were adequate for detecting all species breeding at the time of the survey.

Association between serum concentrations of 25-hydroxyvitamin D and gingival inflammation

Vitamin D has been shown to have immunomodulatory effects in in vitro and in animal studies. However, data from clinical studies of inflammatory diseases are scarce. The purpose of this study was to evaluate the association between serum concentrations of 25-hydroxyvitamin D [25(OH)D] and gingival inflammation. We analyzed data from 77,503 gingival units (teeth) in 6700 never smokers aged 13 to >90 y from the third National Health and Nutrition Examination Survey. Multiple logistic regression models adjusted for subject- and site-specific covariates included age, sex, race-ethnicity, income, body mass index, diabetes, use of oral contraceptives and hormone replacement therapy among women, intake of vitamin C, missing teeth, full crown coverage, presence of calculus, frequency of dental visits, and dental examiner and survey phase. Generalized estimating equations were used to account for correlated observations within subjects. Compared with sites in subjects in the lowest 25(OH)D quintile, sites in subjects in the highest 25(OH)D quintile were 20% (95% CI: 8%, 31%) less likely to bleed on gingival probing (P for trend < 0.001). The association appeared to be linear over the entire 25(OH)D range, was consistent across racial or ethnic groups, and was similar among men and women as well as among users and nonusers of vitamin and mineral supplements. Vitamin D may reduce susceptibility to gingival inflammation through its antiinflammatory effects. Gingivitis may be a useful clinical model to evaluate the antiinflammatory effects of vitamin D.