Are local plague endemicity and ecological characteristics of vectors and reservoirs related? A case study in north-east Tanzania

The pattern of plague endemicity in Tanzania is characterized by continuous re-appearance of the disease in some locations,while in other neighbouring villages the disease has never or seldom been observed.To understand the reasons for this pattern,we studied small mammal and flea species composition,diversity and relative abundances in two plague-endemic and two plague-free locations.We asked(a)whether fleas more abundant in plague-endemic locations differ in their characteristic abundance and the degree of their host specificity from fleas more abundant in plague-free locations and(b)whether hosts most abundant in plague-endemic locations differ in the diversity of their flea assemblages from hosts most abundant in plague-free locations.We characterized(a)each host species by species richness and degree of taxonomic relatedness of its flea assemblages and(b)each flea species by its mean abundance and size,and degree of taxonomic relatedness of its host spectrum and compared their relative abundances between locations.No significant difference between plague-endemic and plague-free locations in either host density or any variable related to flea abundance or diversity was found.However,there was marginally significant effect of taxonomic distinctness of a flea assemblage harboured by a host on its probability to be more abundant in either plague-endemic or plague-free locations.Furthermore,hosts more abundant in plague-endemic locations tended to harbour closely-related fleas.Finally,while opportunistic and specialist fleas were equally distributed in both sets of locations,fleas exploiting distantly-related hosts were found mainly in plague-free locations during the rainy season.We suggest that the observed patterns might arise due to seasonal and spatial differences in local microclimatic conditions and landscape connectivity.

Demography, reproductive biology and diet of the bushveld gerbil Tatera leucogaster (Rodentia: Gerbillinae) in the Lake Rukwa valley, south-western Tanzania

Seasonal abundance, reproductive biology and feeding ecology of the bushveld gerbil Tatera leucogaster (Peters,1852) were investigated in small-scale maize field–fallow land mosaics in south-western Tanzania. The gerbilswere collected over a 2-year period using Sherman live and Victor hold-fast snap traps in permanent 4.5-ha grids. Atotal of 664 individuals were captured over 13 650 trap nights, giving an overall trap success rate of 4.9%. Trapsuccess varied between seasons with and without crops in the field but not between habitat types. At this site, thebreeding activity of this species is seasonal. All individuals whose stomachs were analyzed ate a wide range ofitems, indicating omnivory in this species at this site;however, seeds were the most preferred diet category, with amean contribution of 50.4%, followed by arthropods, with a mean contribution of 25.7%. Other plant materialsbecame important during the very dry periods.

Relationship between sampling intensity and precision for estimating damage to maize caused by rodents

In this study we aimed to determine the relationship between sampling intensity and precision for estimating rodentdamage. We used the systematic row sampling technique to provide data to achieve precision and accuracy inestimations of rodent damage in maize fields at the planting and seedling stages. The actual rodent damage to maizein 15 fields, each 0.5 ha in size, in Morogoro, Tanzania, was established at the seedling stage. These data were usedto simulate the sampling intensities that would provide precision and accuracy. The variations between estimateswere plotted against the sampling intervals. The results of this study show that the relationship between averagestandardized variances and sampling intervals is linear. The heterogeneous distribution of damage in some plotscaused variations in the accuracy of the estimates between plots, but a sampling interval of five rows consistentlyproduced estimates with a variance of less than 10%. We provide a standard curve that will allow a decision to bemade on the sampling intensity as a function of required precision using the systematic row sampling technique inmaize fields.

Survival and maturation rates of the African rodent, Mastomys natalensis: density-dependence and rainfall

Survival and maturation rates of female Mastomys natalensis were analysed based on a ten-year monthlycapture-recapture data set. We investigated whether direct and delayed density dependent and independent (rainfall)variables accounted for the considerable variation in demographic traits. It was estimated that seasonal and annualcovariates accounted for respectively 29 and 26% of the total variation in maturation rates and respectively 17 and11% of the variation in survival rates. Explaining the between-year differences in maturation rates with annual pastrainfall or density did not improve the model fit. On the other hand we showed that maturation rates were correlatednegatively with density the previous month and positively to cumulative rainfall over the past three months.Survival estimates of both adults and subadults varied seasonally, with higher estimates during the increase phase(dry season). The subadults were characterised by a very high survival rate (> 0.95) during this phase. In thedecrease phase only minor differences were found between survival rates of subadults and adults. We found that39% of the between-year variation in survival can be explained by accumulated rainfall over the past year.

Soil type limits population abundance of rodents in crop fields:case study of the multimammate rat Mastomys natalensis Smith,1834 in Tanzania

Studies of populations of the multimammate rat Mastomys natalensis in Morogoro, Tanzania, show that soil textureappears to influence the population abundance and distribution of these rats in agricultural fields. The lowest rodentpopulation abundance was found on sandy clay soils (F(2, 5) = 8.42;P = 0.025). The population abundances of M.natalensis on sandy clay loam and sandy loam soils did not differ significantly (P ≤ 0.05), possibly because thesesoils have a very similar texture. The results of this study suggest that M. natalensis prefers loam-textured soilswith a high percentage of sand, which are probably better than clay soils for burrowing and nesting, particularly inthe rainy season. The lower preference for clay soils is probably related to the poor aeration in these soils and thewaterlogging that occurs during the wet season.

Seasonal and habitat dependence of fleas parasitic on small mammals in Tanzania

We investigated host and flea species composition across different habitats during dry and rainy seasons in the Western Usambara Mountains in Tanzania.During both seasons,similarity in flea species composition increased with an increase in the similarity in host species composition.Nevertheless,between-season within-habitat as well as within-season between-habitat similarity in host species composition was higher than similarity in flea species composition.Ordination of habitats according to their host and flea species composition demonstrated that the pattern of between-habitat similarity in both host and flea species composition varied seasonally.Despite the relatively rich mammal and flea fauna of the study region,the major contribution to variation in species composition between seasons and among habitats was due to a few species only.Flea assemblages on Lophuromys kilonzoi Verheyen et al.,2007 and Praomys delectorum Thomas,1910 in different habitats were equally similar in either season.In contrast,flea assemblages on Mastomys natalensis(Smith,1834)occurring in different habitats were more similar in the dry than in the rainy season,whereas the opposite was the case for fleas on Grammomys sp.In different hosts,the main differences in species composition of flea assemblages between seasons as well as among habitats were due to different flea species.Although our results support the earlier idea that parasite species composition is determined by both host species composition and habitat properties,the former appears to explain variance in flea species composition between localities in the tropics better than between localities in temperate and arid zones.