The present study arrived at investigating the influence of supplementary food on a developing rodent community on the coastal dune forests north of Richards Bay, South Africa. Four sites were selected for the field experiment. Each of these study sites comprised a treatment and control grid. Sunflower seeds were provided on the treatment grids and the response thereof on rodent population and community variables were measured. Each of the grids were trapped on a monthly bases as to follow the cumulative effects that supplementary food may have had on the rodent populations and the community. The amount of seed taken was measured. The minimum number of animals alive, as well as the survival and reproductive rates were calculated for the rodent community and each species on all the grids. Various community variables (richness, diversity and Bray-Curtis dissimilarities) were calculated. The diet of each species was investigated as to confirm that the species present take seeds as a food item. Five rodent species were recorded on the newly rehabilitating areas. They include Otomys angoniensis, Mastomys natalensis, Mus minutoides, Saccostomus campestris and Dendromus melanotis. Most of these species are granivores and consume seeds as part of their diet with the exception of 0. angoniensis that is a herbivore. Seeds form the main part of the diet of M natalensis, M minutoides and S. campestris and it could be assumed that sunflower seeds can form part of their diet. Bray-Curtis dissimilarity values indicated that treatment and control grids became more dissimilar as time progressed. Supplementary food had no significant effect on the species richness and rodent community densities. However, the supplementary food on treatment grids resulted in the significant decrease in species diversity in contrast with control grids. This decrease in species diversity was not as a result in the change in species richness, but was rather the result of the proportional increase in the number of individuals of M natalelensis on the treatment grids. Mastomys natalelensis were the most dominant species. Their numbers increased on the treatment grids where they continued to dominate the rodent community. Mus minutoides was the second most abundant on the study grids. This species, however, did not responded to the food supplied and decreased in numbers on both the treatment and control grids. No relationship could be demonstrated between any pair of species for treatment or control grids. Interspecific interactions between species could therefore not be illustrated. The possible existence of intraspecific interactions for M natalensis is a more plausible explanation as to why their numbers increase on treatment grids were food was no more limiting than on control grids. Species specific survival rate of M minutoides decreased with time for the treatment and control grids combined. This coincide with the decrease in the minimum number of M minutoides alive and may be explained by the successional changes in habitat characteristics. Factors other than the availability of sunflower seeds must therefore influence M. minutoides numbers on the newly rehabilitating areas. The survival probability of M natalensis individuals were not influenced by the supplementary food. Factors other than supposedly increased survival are responsible for their high numbers that were maintained on treatment grids. Supplementary food had no influence on the reproductive rates of adults and the proportion of juveniles present in each of the species present. This imply that the effects that supplementary food had on the rodent community and the species it contain, were not established through the reproductive effort and output. This suggests that immigration may be the reason for increased numbers of M natalensis individuals rather than their survival probability or reproductive performance. Sunflower seeds were provided as supplementary food to increase only a small part of the resource spectrum. Food supplementation for this developing rodent community resulted in a decrease in diversity as MacArthur's (1972) model predicted. However, the decrease in diversity 76 was not as a result of a decrease in species richness, but rather the result of the increase and maintained dominance of the pioneer species M natalensis. No interspecific interactions could be demonstrated, and it is speculated that intraspecific interactions play an important role in the M. natalensis population. The rodent community is therefore shaped by factors other than interspecific interactions, with supplementary food having no effect on species other than M natalensis.
Dissertation (MSc (Zoology))--University of Pretoria, 2006.