Long Term Can Trapping for Population Analyses of Ground-Surface, Arid- Land Arachnids

Abstract

This investigation was designed to estimate statistically the number of can traps required to study populations of ground-surface, arid-land arachnids. Data were obtained from 16 traps operated for 1 year in the pinyon-juniper life zone of the Pinos Altos Mountains in southwestern New Mexico. The formula N = S^2/SE^2 (where N = number of traps, S^2 = variance, and SE = standard error of the mean) was used to estimate the number of traps required for assigned precision limits of 10, 20, 30, and 50% of the mean (x) number per trap. Only precision limits within 20% or less of the x are considered meaningful. Thus, population studies on the local scorpion, Vejovis coahuilae Williams, would require 67 traps, whereas 25 traps would be sufficient for similar studies on local gross solpugid populations. Mastigoproctus giganteus Lucas populations might be estimated using only 21 traps per area. Finally, the number of traps per study area required for estimates of local, arid-land spider populations varied from 11 for the abundant cursorial gnaphosid, Zelotes tuobus Chamberlin, through 12 for the abundant web-building pholcid, Psilochorus imitatus Gertsch and Mulaik, to 22 for the abundant wolf spider, Schizocosa n. sp. (nr. avida Walckenaer). Data on less common species indicated that 20-70 traps per area might be required for study, even of male cursorial activity. This evaluation of can trap population estimates suggests that a knowledge of the life cycle, food habits, ecological requirements, and behavior of target species, and the use of this and other methods of population estimation should be combined for valid population studies.