No abstract

The causative agent of cutaneous leishmaniasis, Leishmania major, was studied in a Tunisian population of the fat sand rat, Psammomys obesus. Seasonal changes in the abundances of parasite and host were monitored in a longitudinal field survey lasting 21 months. Overall, 566 P. obesus, collected during 10 trapping sessions between May 1995 and January 1997, were examined. Analysis of prevalence, using logistic regression, revealed that extrinsic factors, such as season and climatic conditions, and intrinsic factors, such as host age, have a combined effect. Leishmania major showed a seasonal pattern, with most transmission occurring in late summer and in autumn, when prevalences were 7.5- and 6.6-fold higher, respectively, than in spring. Prevalence peaked, at 70%, in September 1995 and then decreased to 0% in June 1996. The same temporal fluctuation was observed during the second study year, albeit among prevalences that were relatively low because of unusually dry conditions. Prevalence was highly dependent on the age of the P. obesus, and season and age acted in synergy so that the rodents were highly infected in late summer and in autumn. Prevalence was not correlated with the relative density of the P. obesus and also appeared independent of gender. Although the ear lesions observed on 378 sand rats during a 1-year survey were closely associated with Leishmania infection, such lesions were not good predictors of infection, as 35% of the rodents found to be infected had no visible lesions on their ears. The prevalences of Leishmania infection observed in this study, among P. obesus living in monospecific colonies, were generally lower than those observed in other studies of P. obesus. It seems possible that P. obesus living in monospecific colonies could have a lesser role in propagating the parasite than those living in plurispecific colonies of rodents, and act as an 'epidemiological sink'.

Summary1. The rodent Psammomys obesus is the main reservoir host for Leishmania major, the causative agent of cutaneous leishmaniasis in Tunisia, much of North Africa and mid-western Asia. An understanding of the population dynamics of this rodent is essential to establish a preventive control strategy based on the early prediction of rodent outbreaks. 2. The study of P. obesus dynamics at a regional scale requires index-based sampling. Rodent numbers were monitored twice per year at the beginning and end of the breeding season, using transects around the Sidi Bouzid region in central Tunisia. 3. Two di erent types of dynamics occurred in two drainage basins. Rodents living in the northern basin were at low density and those living in the southern basin were at high density. 4. At the scale of a plot, occupied burrows were generally associated with the presence of three Chenopodiaceae: Arthrocnemum glaucum, Salsola tetrandra and Suñda fruticosa, rarely in monospeci®c formations. However, in dry periods, occupied burrows drifted to A. glaucum formation linked with high moisture and salinity of the soil (w 2 MacNemar 6Á26). 5. The risk of a P. obesus outbreak can be assessed by a simultaneous knowledge of¯ooding regimes in the drainage basins, the distribution of halophytic plant formations, and the progressive movements of the rodents as drying out proceeds. The transect indices of rodents are easily repeatable, economical and would be applicable in other developing countries where cutaneous leishmaniasis occurs. 6. This work illustrates that simple ecological methods can assist the assessment of spatial and temporal components of epidemiological risk such as the proximity between rodent colonies and human habitats at the time of outbreaks.

The tapeworm Raillietina trapezoides was studied in a Tunisian population of the fat sand rat Psammomys obesus. Seasonal changes in the abundance of parasite and host were monitored in a longitudinal field survey lasting 20 months. In total 582 intestinal samples were collected during 10 trapping sessions between May 1995 and January 1997 and examined. The impact of abiotic and biotic factors on the prevalence and parasite burden were explored, using generalized linear models. R. trapezoides showed a seasonal pattern with most transmission occurring in summer and autumn. In August-September when the rodents were at low density and most adult (90-100 %) were infected. In winter, infections with R. trapezoides were highly prevalent in the cohort of adults only. The worm burden by number was also high in late summer and autumn. Prevalence and parasite burden by number or by biomass were highly age dependent. Season and age act in synergy so that the rodents were highly infected in late summer and autumn. Unusually dry conditions in winter seemed to increase prevalence and to reduce the number of worms. There was no difference between males and females in prevalence or parasite burden (by number or by biomass). Prevalence was not correlated with the relative density of the hosts whereas parasite burden depended upon their densities. The more numerous the worms were, the smaller they were individually, reflecting a crowding effect. The possibility of an immune response regulating P. obesus is discussed, in regard to the very high natural prevalence and the variation in worm burden, and compared with other parasite-host systems.

The fat sand rat Psammomys obesus is the main reservoir host of cutaneous leishmaniasis in North Africa and western Asia. In order to understand better the transmission of this disease to humans, a longitudinal study of a population of the rodent was carried out in Tunisia. Rodents were collected in 10 trapping sessions between May 1995 and January 1997, using mole traps and cage traps. The structure of the population was established from eye lens weights and the reproductive state was determined on dissection. Reproduction took place continuously from autumn 1995 to spring 1996 with sexual activity rates from 60±65% in females and 23±100% in males. The mean litter size rose from 3.6 in September to 6.1 in March. Reproduction ceased completely in the summer of 1996, and did not start again till January 1997, following an apparent failure in the autumn. Animals born early in the breeding season bred at an early age, but those born later delayed breeding till the following season. At most times the population clearly consisted of two cohorts, one born in the most recent breeding season, and one born in the preceding one. In response to the reproductive cycle, population numbers showed an increase in winter and a decline in summer. The peak observed in March 1996 was due to the sexual activity of multiparous females and young breeding females. At this time there was a female bias in the sex ratio. In a winter dry period, favourable habitat for breeding was con®ned to an area in which several species of Chenopodiaceae were associated. Sexual activity in females was clearly correlated with rainfall and rain days, with a 1 month delay (r = 0.78; P = 0.007). Male sexual activity was independent of weather conditions. This result is comparable with other rodents of Sahelian or tropical conditions. The increase in fertility was unrelated to age and is discussed in the light of food availability as a proximate factor. The persistence of the parasite causing cutaneous leishmaniasis in host populations is discussed.