Search

The probability of surviving and moving between 'states' is of greatinterest to biologists. Robust estimation of these transitions usingmultiple observations of individually identifiable marked individuals has received considerable attention in recent years. However, in some situations, individuals in a sample can be assigned to a particular state (e.g. breeding or non-breeding) without error. In such cases,only aggregate data (number of individuals in a given state at each occasion) are available. If the underlying matrix of transition probabilities does not vary through time and aggregate data are available for several time periods, then it is possible to estimate these parameters using least-squares methods. Even when such data are available,this assumption of stationarity will usually be deemed overly restrictive and, frequently, data will only be available for two time periods. In these cases, the problem reduces to estimating the most likelymatrix (or matrices) leading to the observed frequency distribution of individuals in each state. An entropy maximization approach has been previously suggested. In this paper, we show that the entropy approach rests on a particular limiting assumption, and does not provide estimates of latent population parameters (the transition probabilities), but rather predictions of realized rates. [ABSTRACT FROM AUTHOR]

Heterogeneous capture probabilities are a common problem in many capture-recapture studies. Several methods of detecting the presence of such heterogeneity are currently available, and stratification of data has been suggested as the standard method to avoid its effects. However, few studies have tried to identify sources of heterogeneity, orwhether there are interactions among sources. The aim of this paper is to suggest an analytical procedure to identify sources of capture heterogeneity. We use data on the sex and age of Great Tits captured fn baited funnel traps, at two localities differing in average temperature. We additionally use 'recapture' data obtained by videotaping at a feeder (with no associated trap), where the tits ringed with different colors were recorded. This allowed us to test whether individuals in different classes (age, sex and condition) are not trapped because of trap shyness or because of a reduced use of the bait. We used logistic regression analysis of the capture probabilities to test forthe effects of age, sex, condition, location and 'recapture' method.The results showed a higher recapture probability in the colder locality. Yearling birds (either males or females) had the highest recapture probabilities, followed by adult males, while adult females had the lowest recapture probabilities. There was no effect of the method of 'recapture' (trap or video-tape), which suggests that adult females are less often captured in traps not because of trap-shyness but because of less dependence on supplementary food. The potential use o this methodological a roach in other studies is discussed. [ABSTRACT FROM AUTHOR]

The past several years have seen the development of a number of statistical models and methods for drawing inferences about bird movementusing data from marked individuals. It can be difficult to keep up with this rapid development of new methods, so our purpose here is to categorize and review methods for drawing inferences about avian movement. We also outline recommendations about future work, dealing bothwith methodological developments and with studies directed at hypotheses about bird movement of interest from conservation, management, or ecological perspectives. [ABSTRACT FROM AUTHOR]

The size of the breeding population of Brandt's Cormorants Phalacrocorax penicillatus on Southeast Farallon Island, off the coast of central California, USA, shows large annual fluctuations and has declinedby 50% since the 1970s. We investigated patterns of variation in adult survival, breeding probability, resighting probability, juvenile survival and reproductive success, to determine the extent to which variation in demographic processes reflected variation in environmentalconditions (e.g. food availability) versus density- dependence, using observations of breeders and non-breeders from 2976 to 1995. Resighting and breeding probabilities varied significantly among years, andboth were positively correlated with an annual index of prey availability -the abundance of juvenile rockfish. Adult survival between years was significantly correlated with changes in juvenile rockfish abundance and differed between sexes (0.77, males; 0.72, females). Female, but not male, survival appeared age specific: older females showedreduced survival, especially during El Nino years. Intermittent breeding appeared common in this population: 54% of males and 49% of females estimated to be alive in a given year were not observed breeding in that year. Reproductive success varied in relation to the juvenilerockfish index but not in relation to population size. The return rate of juveniles, an index of survival during the first three years oflife, was negatively related to sea surface temperature, both in theyear of hatching and in the third year of life, when individuals first return to the colony to breed. We conclude that this population isstrongly susceptible to environmental fluctuation. All parameters ofsurvival and reproduction deteriorate under poor environmental conditions, with no evidence of buffering, and there was no evidence of density-dependence. Thus extrinsic, not intrinsic, forces appear most important in explaining recent population fluctuations. [ABSTRACT FROM AUTHOR]

This study categorizes and reviews methods used to estimate stopoverlength in migratory birds. Stopover length is an important variable in analyzing migration strategies. The analysis of the stopover pattern of migratory birds typically attempts to investigate variability in the number of stopovers and in the length of time spent at stopoversites in relation to species, age, sex and body condition. Although the number of capture-recapture studies analyzing stopover length hasincreased substantially, most are analysed using the simple last first capture method which gives biased estimates. ,The use of mathematical models to calculate stopover length is uncommon. Most researchersassume homogeneity in stopover length and do not determine the capture probability. Intensive studies are available for many migratory passerines, waders and geese. Some of the data sets are highly standardized and could be reanalyzed with new applications and innovative uses of existing techniques, such as the CJS models. [ABSTRACT FROM AUTHOR]

Radiotelemetry has become a very popular technique for avian survival studies in recent years. Usually the survival function is estimated. However, the hazard function (the instantaneous rate of change in mortality) is also very important in wildlife ecology. We explain important reasons for estimating the hazard function in wildlife ecology,and show how to estimate the hazard function by applying smoothing techniques (also card non-parametric regression) to telemetry data. Weillustrate the smoothing methodology with an example using survival data from a study on wild turkeys in West Virginia. [ABSTRACT FROM AUTHOR]

The great flexibility of capture-recapture models to estimate survival and recruitment parameters is described. Recruitment models appearas analogous to survival models applied to capture histories reversed in time. The available flexibility is used at its best by basing model selection on the minimization of Akaike's information criterion within a set of biologically plausible models. However; to proceed to model selection, one needs to determine the number of identifiable parameters. We explain how SURGE version 5.0 (available on the Internetby anonymous ftp from ftp.cefe.cnrs-mop.fr) makes it possible to fitsurvival and recruitment models and determines numerically how many and which parameters are concerned with identifiability problems. Model selection for survival and recruitment models is illustrated usingEuropean Dipper Cinclus cinclus data. The advantages and shortcomings are discussed. [ABSTRACT FROM AUTHOR]

More than one attribute group of animals can be modeled. 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 ofmodel 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 Ale 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 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 capabilitiesare implemented in a Microsoft Windows 95 interface. The online helpsystem has been developed to provide all necessary program documentation. [ABSTRACT FROM AUTHOR]

Methods for diagnosing causes of bird population declines are described, paying special attention to methods that rely on comparisons among geographical regions and/or time periods with different populationtrends. Comparisons of breeding productivity and survival among regions or periods provide a valuable check on the diagnosis of external causes of a decline indicated by correlation with population trend. Ringing or tagging programs can give information on population levels and are the main source of information on survival. Age ratios from counts or catches of juveniles and adults in standardized trapping programs can provide data on breeding productivity. Nest recording schemes yield estimates of productivity if females make one breeding attempt per year or if the number of attempts per female can be estimated from supplementary data. Population models should be used to explore hypotheses about the external causes and demographic mechanisms of population declines. Models are described which have combinations of parameter values set to give rates of population decline similar to those seen in various passerines in Britain during the 1980s and 1990s. These models indicate that demographic changes associated with declines are subtle and difficult to detect if the demographic rate which changes is density-dependent. Analytical procedures should be devised to compare the fit of models with different underlying demographic mechanisms for the decline to data on survival and breeding success recorded before, during and after declines. [ABSTRACT FROM AUTHOR]