STATISTICAL ANALYSIS OF MAMMALIAN MOVEMENTS, WITH EMPHASIS ON THE TREATMENT OF AUTOCORRELATED OBSERVATIONS (HOME RANGE, MICROTUS OCHROGASTER, RADIOTELEMETRY, SIGMODON HISPIDUS).

摘要

Most studies of mammalian movements rely on indirect methods of observation, primarily livetrapping and radiotelemetry. I develop procedures for analyzing movement data collected by each of these methods, and illustrate their applications using data on cotton rats (Sigmodon hispidus) and prairie voles (Microtus ochrogaster).; Home range indices of cotton rats were calculated for trap-revealed movements cross-classified by age, sex, reproductive status, residency status, season, and habitat. The influence of man-made structures (brush piles and roads) on dispersion and movement also were examined. I found that reproductively mature adult and subadult males occupied larger home ranges than their nonreproductive counterparts and that adult males inhabiting a more sparsely populated abandoned pasture displayed larger home ranges than did those in an old field. Cotton rats aggregated around brush piles during winter and spring, presumably using them for shelter in inclement weather. Cotton rats and prairie voles were extremely reluctant to cross a 3-m wide dirt road, suggesting that such structures might hinder dispersal in small mammals.; Radiotelemetry permits a nearly continuous record of an animal's position but the short sampling intervals that result also yield autocorrelated (dependent) observations. I show that autocorrelation leads to underestimation of home range size. Although this bias occurs for both nonstatistical and statistical home range measures and indices, the problem is most severe for statistical estimates. I develop tests for independence between successive observations for use with both telemetric and trapping data. When the null hypothesis of independence is rejected, I present a procedure for determining the time interval at which autocorrelation becomes negligible. The test can be used to design an efficient sampling schedule for movement studies, and it is also useful in revealing behavioral phenomena such as home range shifting and any tendency of animals to follow prescribed routes in their daily activities. In addition, I demonstrate the utility of the test in studies of physiological time scaling of movements.