Stable isotope analysis of invasive Rattus congenerics in Gauteng Province, South Africa

Abstract

There is a prevalence of invasive synanthropic murid rodent species of Rattus in urban and peri–urban landscapes in Gauteng Province, South Africa due to a general lack of sanitary conditions and municipal services. Factors that may influence their access to, and utilisation of food resources may facilitate their successful establishment, and adverse ecological impacts in recipient ecosystems, and these can be assessed by stable isotope analysis. There are factors that may influence the evaluation of the stable isotope compositions of animals, and include tissue type, habitat, sex, ontogenetic (age) variation, and season. The current study therefore, used stable isotope analysis to investigate the effect of tissue type (liver and gastrocnemius muscle tissue, and fur), habitat, sex, age, and season on δ15N and δ13C values in the invasive synanthropic R. norvegicus and R. rattus from urban and peri–urban landscapes that included Alexandra and Tembisa Townships, and University of Pretoria (UP) Experimental Farm in Gauteng Province, South Africa. Prior to the stable isotope analysis however, an independent craniometric analysis was first undertaken to assess if there is sexual dimorphism, and to determine which age groups were represented within the sampled populations of R. norvegicus and R. rattus. The nature and extent of craniometric sexual dimorphism and age variation were evaluated using univariate (two–way analysis of variance [ANOVA] and the percent contribution of the sum of squares [%SSQ] of each source of variation to the total SSQ), and a series of multivariate analyses (principal components analysis [PCA] and Unweighted Pair–Group Method of Arithmetic Averages [UPGMA] cluster analysis). The results of both univariate and multivariate craniometric analyses were broadly congruent and showed lack of sexual dimorphism in the two species, but significant age variation between five relative age categories based on the degree of maxillary molar toothrow eruption and wear (toothwear classes I–V). When univariate and multivariate craniometric results were collectively considered, the pooling of sexes and individuals into adults (i.e., pooling of individuals of tooth–wear classes IV–V), sub–adults (tooth–wear classes II and III), and juveniles (tooth–wear class I) for subsequent analysis was justified. Few craniometric variables showed statistically significant sex–age interaction. The largest %SSQ contribution to the total variance was due to error (= residual), suggesting that apart from sex, age, and sex–age interaction, there may be other components that are responsible for the nature and extent of craniometric sexual dimorphism and age variation in R. norvergicus and R. rattus from Gauteng Province, South Africa. Stable isotope analysis was used in this study to determine the effect that tissue type, habitat, sex, age and season may have on δ15N and δ13C values in R. norvegicus and R. rattus. Tissue type was found to affect the δ15N and δ13C values of both R. norvegicus and R. rattus. There were significant differences between tissue types for both δ15N and δ13C values. Fur was therefore, used to assess the effect of habitat, sex, age, and season on δ15N and δ13C values due to its low turn–over rate compared to both liver and muscle tissues. Differences between Alexandra Township and Tembisa Township were found to be present for δ15N and δ13C. Although the differences were significant for δ15N and δ13C, they indicated that the rats from both habitats were feeding on food with a similar composition and at the same trophic level. Differences between males and females were present for R. norvegicus and R. rattus at their respective habitats. The δ15N and δ13C differences did not indicate that both sexes were subsisting on different food types, instead they indicated that males had a wider feeding niche compared to females. Ontogenetic (age) variation was similarly not present for both δ15N and δ13C. Seasonal differences were neither present for R. norvegicus from Tembisa Township nor for R. rattus from UP Experimental Farm, however the trophic niche occupied by both rat populations was wider during autumn compared to summer. Although there was no interaction among the factors, the results suggested complex effects on diet that are difficult to discern. A more holistic scenario may emerge from the use of additional techniques such as gut content analysis and telemetry data that may lead to the development of long–term rat control and management strategies.