Michael J. Conroy

Parameter estimation, reliability, and model improvement for spatially explicit models of animal populations

We address model specification, parameter estimation, and model reliability for spatially explicit population models (SEPMs). We assume that these models have the complementary goals of understanding the processes that influence the number and dis- tribution of animals in space and time, and forecasting the effect of management or other human activities on population abundance and distribution. Incorrect model structure, pa- rameter estimates, or both will result in unreliable model output. Spatially explicit models require knowledge of population spatial structure, dispersal, and movement rates, in addition to the usual demographic parameters and structural assumptions such as density-depen- dence, and are thus potentially very vulnerable to propagation of model uncertainty. Sen- sitivity analysis and validation can both be used to evaluate the reliability of SEPMs, but the level of spatiotemporal resolution at which the model should be evaluated is often not clear. Many SEPMs are very complex, and validation may only be possible or meaningful on a sub-model basis. Forecasting, that is, prediction under a different set of conditions than that under which the model was built, will provide a stronger test of model reliability. Forecasts from SEPMs can be used to generate hypotheses that can then be tested as parts of large-scale adaptive management experiments. In this way resource management goals can be achieved, while providing enhanced understanding of systems and improved pre- dictability of future scenarios.

Mapping of Species Richness for Conservation of Biological Diversity: Conceptual and Methodological Issues

Biodiversity mapping (e.g., the Gap Analysis Program [GAP]), in which vegetative features and categories of land use are mapped at coarse spatial scales, has been proposed as a reliable tool for land use decisions (e.g., reserve identification, selection, and design). This implicitly assumes that species richness data collected at coarse spatio-temporal scales provide a first-order approximation to community and ecosystem representation and persistence. This assumption may be false because (1) species abundance distributions and species richness are poor surrogates for community/ecosystem processes, and are scale dependent; (2) species abundance and richness data are unreliable because of unequal and unknown sampling probabilities and species-habitat models of doubtful reliability; (3) mapped species richness data may be inherently resistant to scaling up or scaling down: and (4) decision-making based on mapped species richness patterns may be sensitive to errors from unreliable data and models, resulting in suboptimal conservation decisions. We suggest an approach in which mapped data are linked to management via demographic models, multiscale sampling, and decision theory. We use a numerical representation of a system in which vegetation data are assumed to be known and mapped without error, a simple model relating habitat to predicted species persistence, and statistical decision theory to illustrate use of mapped data in conservation decision-making and the impacts of uncertainty in data or models on the decision outcome.

Simultaneous use of mark-recapture and radiotelemetry to estimate survival, movement, and capture rates

Biologists often estimate separate survival and movement rates from radiotelemetry and mark -recapture data from the same study population. We describe a method for combining these data types in a single model to obtain joint, potentially less biased estimates of survival and movement that use all available data. We furnish an example using wood thrushes (Hylocichla mustelina) captured at the Piedmont National Wildlife Refuge in central Georgia in 1996. The model structure allows estimation of survival and capture probabilities, as well as estimation of movements away from and into the study area. In addition, the mode structure provides many possibilities for hypothesis testing. Using the combined model structure, we estimated that weekly survival of wood thrushes was 0.989 ± 0.007 (±SE). Survival rates of banded and radiomarked individuals were not different (α[S radioed , S banded ] = log[S radioed /S banded ] = 0.0239, 95% CI = -0.0196 to 0.0486) Fidelity rates (weekly probability of remaining in a stratum) did not differ between geographic strata (ψ = 0.911 ± 0.020; α[ψ 11 ψ 22 ] = 0.0161, 95% CI = -0.0309 to 0.0631). and recapture rates (p = 0.097 ± 0.016) of banded and radiomarked individuals were not different (α[P radioed , P banded ] = 0.145, 95% CI = -0.510 to 0.800). Combining these data types in a common model resulted in more precise estimates of movement and recapture rates than separate estimation, but ability to detect stratum or mark-specific differences in parameters was weak. We conducted simulation trials to investigate the effects of varying study designs on parameter accuracy and statistical power to detect important differences. Parameter accuracy was high (relative bias [RBIAS] <2%) and confidence interval coverage close to nominal, except for survival estimates of banded birds for the off study area stratum, which were negatively biased (RBIAS -7 to -15%) when sample sizes were small (5-10 banded or radioed animals released per time interval). To provide adequate data for useful inference from this model, study designs should seek a minimum of 25 animals of each marking type observed (marked or observed via telemetry) in each time period and geographic stratum.

Effects of forest management on density, survival, and population growth of wood thrushes

Loss and alteration of breeding habitat have been proposed as causes of declines in several Neotropical migrant bird populations. We conducted a 4-year study to determine the effects of winter prescribed burning and forest thinning on breeding wood thrush (Hylocichla mustelina) populations at the Piedmont National Wildlife Refuge (PNWR) in Georgia. We estimated density, adult and juvenile survival rates, and apparent aunual survival using transect surveys, radiotelemetry, and mist netting. Burning and thinning did not cause lower densities (P = 0.25); wood thrush density ranged from 0.15 to 1.30 pairs/10 ha. No radiomarked male wood thrushes (n = 68) died during the 4 years, but female weekly survival was 0.981 ± 0.014 (SE) fo lemales (n = 63) and 0.976 ± 0.010 for juveniles (n = 38). Apparent annual adult survival was 0.579 (SE = 0.173) Thinning and prescribed burning did not reduce adult or juvenile survival during the breeding season or apparent annual adult survival. Annual population growth (λ) at PNWR was 1.00 (95% confidence interval [CI] = 0.32-1.63), and the considerable uncertainty in this prediction underscores the need for long-term monitoring to effectively manage Neotropical migrants, Population growth increased on experimental compartments after the burn and thin (95% CI before = 0.91-0.97, after = 0.98-1.05), while control compartment λ declined (before = 0.98-1.05, after = 0.87-0.92). We found no evidence that the current management regime at PNWB, designed to improve red-cockaded woodpecker (Picoides borealis) habitat, negatively affected wood thrushes.

WILD BIRD MORTALITY AND WEST NILE VIRUS SURVEILLANCE: BIASES ASSOCIATED WITH DETECTION, REPORTING, AND CARCASS PERSISTENCE

Surveillance targeting dead wild birds, in particular American crows (Corvus brachyrhynchos), plays a critical role in West Nile virus (WNV) surveillance in the United States. Using crow decoy surrogates, detection and reporting of crow carcasses within urban and rural environments of DeKalb County, Georgia were assessed for potential biases that might occur in the countys WNV surveillance program. In each of two replicated trials, during July and September 2003, 400 decoys were labeled with reporting instructions and distributed along randomly chosen routes throughout designated urban and rural areas within DeKalb County. Information-theoretic methods were used to compare alternative models incorporating the effects of area and trial on probabilities of detection and reporting. The model with the best empirical support included the effects of area on both detection and reporting of decoys. The proportion of decoys detected in the urban area (0.605, SE=0.024) was approximately twice that of the rural area (0.293, SE=0.023), and the proportion of decoys reported in the urban area (0.273, SE=0.023) was approximately three times that of the rural area (0.103, SE=0.028). These results suggest that human density and associated factors can substantially influence dead crow detection and reporting and, thus, the perceived distribution of WNV. In a second and separate study, the persistence and fate of American crow and house sparrow (Passer domesticus) carcasses were assessed in urban and rural environments in Athens-Clarke, Madison, and Oconee counties, Georgia. Two replicated trials using 96 carcasses of each species were conducted during July and September 2004. For a portion of the carcasses, motion sensitive cameras were used to monitor scavenging species visits. Most carcasses (82%) disappeared or were decayed by the end of the 6-day study. Carcass persistence averaged 1.6 days in rural areas and 2.1 days in urban areas. We analyzed carcass persistence rates using a known-fate model framework in program MARK. Model selection based on Akaikes Information Criteria (AIC) indicated that the best model explaining carcass persistence rates included species and number of days of exposure; however, the model including area and number of days of exposure received approximately equal support. Model-averaged carcass persistence rates were higher for urban areas and for crow carcasses. Six mammalian and one avian species were documented scavenging upon carcasses. Dead wild birds could represent potential sources of oral WNV exposure to these scavenging species. Species composition of the scavenger assemblage was similar in urban and rural areas but “scavenging pressure” was greater in rural areas.

STATE-SPECIFIC DETECTION PROBABILITIES AND DISEASE PREVALENCE

Investigations of disease dynamics in wild animal populations often use estimated prevalence or incidence as a measure of true disease frequency. Such indices, almost always based solely on raw counts of infected and uninfected individuals, are often used as the basis for analysis of temporal and spatial dynamics of diseases. Generally, such studies do not account for potential differences in observer detection probabilities of host individuals stratified by biotic and/or abiotic factors. We demonstrate the potential effects of heterogeneity in state-specific detection probabilities on estimated disease prevalence using mark-recapture data from previous work in a House Finch (Carpodacus mexicanus) and Mycoplasma gallisepticum system. In this system, detection probabilities of uninfected finches were generally higher than infected individuals. We show that the magnitude and seasonal pattern of variation in estimated prevalence, corrected for differences in detection probabilities, differed markedly from uncorrected (apparent) prevalence. When the detection probability of uninfected individuals is higher than infected individuals (as in our study), apparent prevalence is negatively biased, and vice versa. In situations where state-specific detection probabilities strongly interact over time, we show that the magnitude and pattern of apparent prevalence can change dramatically; in such cases, observed variations in prevalence may be completely spurious artifacts of variation in detection probability, rather than changes in underlying disease dynamics. Accounting for differential detection probabilities in estimates of disease frequency removes a potentially confounding factor in studies seeking to identify biotic and/or abiotic drivers of disease dynamics. Given that detection probabilities of different groups of individuals are likely to change temporally and spatially in most field studies, our results underscore the importance of estimating and incorporating detection probabilities in estimated disease prevalence (specifically), and more generally, any ecological index used to estimate some parameter of interest. While a mark-recapture approach makes it possible to estimate detection probabilities, it is not always practical, especially at large scales. We discuss several alternative approaches and categorize the assumptions under which analysis of uncorrected prevalence may be acceptable.

A model to predict breeding-season productivity for multibrooded songbirds

Breeding-season productivity (the per capita number of offspring surviving to the end of the breeding season) is seldom estimated for multibrooded songbirds because of cost and logistical constraints. However, this parameter is critical for predictions of population growth rates and comparisons of seasonal productivity across geographic or temporal scales. We constructed a dynamic, stochastic, individual-based model of breeding-season productivity using demographic data from Wood Thrushes (Hylocichla mustelina) in central Georgia from 1993 to 1996. The model predicts breeding-season productivity as a function of adult survival, juvenile survival, nesting success, season length, renesting interval, and juvenile-care intervals. The model predicted that seasonal fecundity (number of fledglings produced) was 3.04, but only 2.04 juveniles per female survived to the end of the breeding season. Sensitivity analyses showed that differences in renesting interval, nesting success, fledglings per successful nest, and adult and juvenile survival caused variation in breeding-season productivity. Contrary to commonly held notions, season length and fledgling-care interval length did not cause variation in breeding-season productivity. This modeling exercise emphasizes the need for demographic data for songbird species, and we encourage biologists to use similar models to evaluate productivity in songbird populations.

Home range, movements, and habitat use of coyotes in southcentral Georgia

Most previous studies of coyotes (Canis latrans) have been conducted in western North America, and inferences about their ecology probably are not relevant to the southeastern U.S. Consequently, we radiotracked 12 coyotes from September 1987 through December 1988 to study movement and habitat use in southcentral Georgia. Home ranges averaged 12.4, and 10.1 km 2 for the 95% harmonic contour and 95% minimum convex polygon methods, respectively. Home range size was smallest during the gestation season

WOOD THRUSH MOVEMENTS AND HABITAT USE: EFFECTS OF FOREST MANAGEMENT FOR RED-COCKADED WOODPECKERS

Abstract We monitored adult and juvenile breeding-season movements and habitat use of radio-tagged Wood Thrushes (Hylocichla mustelina) at the Piedmont National Wildlife Refuge, central Georgia, USA. We investigated the effects that management for Red-cockaded Woodpeckers (Picoides borealis), thinning and burning >30 year old loblolly pine (Pinus taeda) habitat, had on Wood Thrushes, a ground-foraging and midstory-nesting species. Adult Wood Thrush pairs regularly moved long distances between nesting attempts (range 1 to 17,388 m). The only experimental effect we found on adult movements was a decrease in weekly emigration rates (Ψ) from thinned and burned compartments after silvicultural management. Adult males preferred riparian hardwoods with sparse to moderate cover and those preferences increased following management. Juveniles remained near their nest site (x̄ = 177 m, SE = 113) for an average 24 days (SE = 6.3), and then dispersed a mean 2,189 m (SE = 342). Before dispersal, juveniles preferred upland hardwood–pine mixed habitat (P < 0.05) with moderate overstory cover (P < 0.05). We found no management effects on dispersal distances or predispersal habitat use. However, juveniles from thinned and burned compartments dispersed to hardwood habitats with dense cover, whereas birds from control compartments dispersed to pine-dominated habitats with sparse cover. All juveniles dispersed to areas with habitat similar to what they used before dispersal. Small-scale thinning and burning for Red-cockaded Woodpeckers may have had little effect on Wood Thrush habitat use and movements because typical movements were often larger than the scale (stand or compartment) targeted for management.

Geographic and temporal differences in band reporting rates for American black ducks

Importance du denombrement des reprises de bagues de gibier a plume migrateur pour etablir la reglementation de la chasse (les baguages sont effectues aux Etats Unis et au Canada)