Abstract: From 1989 to 1996, surveys were made in most of the historic range of African wild asses in Somalia, Ethiopia, and Eritrea. From the 1970s to the mid 1990s populations of African wild asses (Equus africanus, Fitzinger, 1857) in Somalia and Ethiopia have declined from approximately 6 to 30 per 100 km2 to 1 or 2 per 100 km2. Given the current IUCN criteria, they are Critically Endangered (CR) and face extremely high risk of extinction in the wild in the immediate future, as their populations have been reduced by at least 80% over the last 10+ years (IUCN, 1994). Basic research is needed on this species as scientific information on its reproductive biology, behavior, ecology, and genetics is very limited. Improved support needs to be provided to existing parks and reserves and new multiple use reserves need to be established.

Abstract: Feral asses have been studied in the arid habitats of the southwestern United States [Moehlman, P.D., 1974. Behavior and ecology of feral asses (Equus asinus). PhD dissertation, University of Wisconsin, Madison, 251 pp.; Moehlman, P.D., 1979. Behavior and ecology of feral asses (Equus asinus). Nat. Geogr. Soc. Res. Reports 1970, 405-411.; Woodward, S.L., 1979. The social system of feral asses (Equus asinus). Z. Tierpsychol. 49, 304-316] and in the mesic habitat of Ossabaw Island, Georgia [Moehlman, P.D., 1979, ibid; McCort, W.D., 1980. The feral asses (Equus asinus) of Ossabaw Island, Georgia. PhD dissertation, Pennsylvania State University, University Park, 219 pp]. The feral ass populations in these two locales exhibited intraspecific variation in polygynous mating systems and social organization which were consistent with the ecological classification of mating systems of Emlen and Oring (1977) [Emlen, S.T., Oring, S.W., 1977. Ecology, sexual selection, and the evolution of mating systems. Science 197 (4300), 215-223]. Feral asses in the arid environment have a `resource defense' polygynous mating system, and those in the mesic habitat exhibit `female (harem) defense' polygyny. The intraspecific variation observed in feral asses encompasses the interspecific variation observed in the family Equidae.

Abstract: Models of costly signalling are commonly employed in evolutionary biology in order to explain how honest communication between individuals with conflicting interests can be stable. These models have focused primarily on a single type of honest signalling equilibrium, the separating equilibrium in which any two different signallers send distinct signals, thereby providing signal receivers with complete information. In this paper, we demonstrate that in signalling among relatives (modelled using the Sir Philip Sidney game), there is not one but a large number of possible signalling equilibria, most of which are pooling equilibria in which different types of signallers may share a common signal. We prove that in a general Sir Philip Sidney game, any partition of signallers into equi-signalling classes can have a stable signalling equilibrium if and only if it is a contiguous partition, and provide examples of such partitions. A similar (but slightly stricter) condition is shown to hold when signals are transmitted through a medium with signalling error. These results suggest a solution to a problem faced by previous signalling theory models: when we consider the separating equilibrium, signal cost is independent of the frequency of individuals sending that signal and, consequently, even very rare signaller types can drastically affect signal cost. Here, we show that by allowing these rare signallers to pool with more common signallers, signal cost can be greatly reduced. Copyright 1998 Academic Press.

Abstract: Leptospirosis is a zoonotic bacterial disease caused by Leptospira interrogans. There is a serologic evidence that horses are exposed to L. interrogans and, as a shedder of these organisms, can be a threat to humans. We examined risk factors associated with the risk of testing seropositive to three L. interrogans serovars (L. icterohaemorrhagiae, L. grippotyphosa, and L. canicola) in the horses of New York State, in order to understand the epidemiology of the disease and suggest strategies to control and prevent equine leptospirosis. To carry out this study, blood samples were collected from a random sample of 2551 horses and tested for the presence of antibodies to the above serovars using the microscopic agglutination test. Samples with a titer $100 were considered positive. Clinical and demographic data were collected on each horse, the farms' management practices and ecology. Logistic regression analysis was used to develop a multivariate indexing system and to identify factors significantly associated with the risk of leptospirosis. Four indices were developed based on the possible sources of exposure: rodent exposure index; wildlife exposure index; soil and water index; and management index. The soil and water index was significantly associated with the risk of exposure to all three serovars. Management was positively associated with L. icterohaemorrhagiae and L. canicola. Density of horses turned out together was positively associated with the risk of exposure to L. grippotyphosa. We concluded that indirect exposure of horses to L. interrogans through contaminated soil and water appears to be significantly associated with the risk of exposure to all three serovars. Management appears to play an important role in the exposure to L. interrogans. Modification of management practices might reduce the horses' risk of exposure and hopefully minimize the human hazards.