Genetic monitoring detects an overlooked cryptic species and reveals the diversity and distribution of three invasive Rattus congeners in South Africa

Abstract

Background: South Africa's long and extensive trade activity has ensured ample opportunities for exotic species introduction. Whereas the rich biodiversity of endemic southern African fauna has been the focus of many studies, invasive vertebrates are generally overlooked despite potential impacts on biodiversity, health and agriculture. Genetic monitoring of commensal rodents in South Africa which uncovered the presence of Rattus tanezumi, a South-East Asian endemic not previously known to occur in Africa, provided the impetus for expanded studies on all invasive Rattus species present.

Results: To this end, intensified sampling at 28 South African localities and at one site in Swaziland, identified 149 Rattus specimens. Cytochrome b gene sequencing revealed the presence of two R. tanezumi, seven Rattus rattus and five Rattus norvegicus haplotypes in south Africa. Phylogenetic results were consistent with a single, recent R. tanezumi introduction and indicated that R. norvegicus and R. rattus probably became established following at least two and three independent introductions, respectively. Intra- and inter-specific diversity was highest in informal human settlements, with all three species occurring at a single metropolitan township site. Rattus norvegicus and R. rattus each occurred sympatrically with Rattus tanezumi at one and five sites, respectively. Karyotyping of selected R. rattus and R. tanezumi individuals identified diploid numbers consistent with those reported previously for these cryptic species. Ordination of bioclimatic variables and MaxEnt ecological niche modelling confirmed that the bioclimatic niche occupied by R. tanezumi in south Africa was distinct from that occupied in its naturalised range in south-east Asia suggesting that factors other than climate may influence the distribution of this species.

Conclusions: This study has highlighted the value of genetic typing for detecting cryptic invasive species, providing historical insights into introductions and for directing future sampling. The apparent ease with which a cryptic species can become established signals the need for broader implementation of genetic monitoring programmes. In addition to providing baseline data and potentially identifying high-risk introduction routes, the predictive power of ecological niche modelling is enhanced when species records are genetically verified.

Figure 1

Cytochrome b gene tree inferred using full-length sequence data (1140 bp) and depicting the genetic relatedness of the three Rattus species haplotypes in South Africa. Taxon names comprise the Genbank accession number, followed by the country of origin (for South African samples this is preceded by a province code: LP = Limpopo Province, MP = Mpumalanga Province, GP = Gauteng Province, KZN = KwaZulu-Natal), haplotype and the number of individuals characterised. Nodal support values are given in percentages and are indicated ME/ML/BPP next to the relevant nodes. -- indicates nodes that were either not recovered or that had support values < 50 for ME and < 70 for ML and BPP. Haplotypes are colour-coded as follows to indicate the region and source of the data: Yellow = Southern Africa (This study), Blue = Outside Africa (This study), Green = Africa (Genbank), Grey = Outside Africa (Genbank), Purple = Laboratory strain (Genbank). Terminal nodes connecting different haplotypes and having ≥99 percent support from all three methods of inference are denoted by a black-filled circle. Country codes given in brackets behind a taxon name, indicate shared presence of a particular haplotype identified from the partial 1043 bp dataset (additional file 3, Figure S3) and are abbreviated: COM = Comores, MOZ = Mozambique, GUA = Guadeloupe, MAD = Madagascar, RSA = Republic of South Africa and SEN = Senegal.

Figure 2

Median-joining network of Rattus tanezumi cytochrome b (cyt b) haplotypes based on partial (1077 bp) gene sequences. Haplotype colour coding is consistent with that used in Figure 1, viz. Yellow = south Africa (This study), Grey = Outside Africa (Genbank). Each of the mutational steps separating haplotypes is indicated in blue and corresponds to the relevant position in the cyt b gene, whilst black nodes correspond to median vectors. Haplotype numbers are the same as those provided in Table 1. Additional haplotype numbers correspond to the laboratory numbers assigned in the Pagès et al. [8] study.

Figure 3

Median-joining network of Rattus rattus cytochrome b (cyt b) haplotypes based on partial (1043 bp) gene sequences. The circle size of south African Rattus haplotypes is proportional to the frequency of the haplotype. Haplotype colour coding is consistent with that used in Figure 1, viz. Yellow = south Africa (This study), Blue = Outside Africa (This study), Green = Africa (Genbank), Grey = Outside Africa (Genbank). Each of the mutational steps separating haplotypes is indicated in blue and corresponds to the relevant position in the cyt b gene, whilst black nodes correspond to median vectors. Haplotype numbers are the same as those provided in Table 1. Additional haplotype numbers correspond to the laboratory numbers assigned in the Tollenaere et al. [4] study.

Figure 4

Median-joining network of Rattus norvegicus cytochrome b (cyt b) haplotypes based on full-length gene sequences. The circle size of south African Rattus haplotypes is proportional to the frequency of the haplotype. Haplotype colour coding is consistent with that used in Figure 1, viz. Yellow = south Africa (This study), Blue = Outside Africa (This study), Grey = Outside Africa (Genbank), Purple = Laboratory strains (Genbank). Each of the mutational steps separating haplotypes is indicated in blue and corresponds to the relevant position in the cyt b gene, whilst black nodes correspond to median vectors. Haplotype numbers are the same as those provided in Table 1.

Figure 5

Rattus species karyotypes from South Africa. Rattus rattus (haplotype RR04, 2n = 38) and Rattus rattus (haplotype RR04, 2n = 40) above, with Rattus tanezumi (haplotype RT01, 2n = 42) depicted below.

Figure 6

Rattus distribution in south Africa and in south-east Asia. (a) Sampling sites of the three Rattus species in South Africa and Swaziland in relation to human footprint. The different species are denoted: R. tanezumi (circle), R. rattus (square) and R. norvegicus (triangle) (b) Distribution of native and naturalised occurrence records of Rattus tanezumi in south-east Asia obtained from http://www.gbif.org. Distributional range of Rattus tanezumi from Global Mammal Assessment (GMA) is overlaid and indicated by grey shading.

Figure 7

Principal component analysis (PCA) of eight bioclimatic variables, altitude and 'human footprint' in a sample of naturalised (circles), native (triangles) and invasive South African (squares) records of Rattus tanezumi occurrence.

Figure 8

MaxEnt model showing predicted geographical distribution of the invasive range of Rattus tanezumi in sub-Saharan Africa, based on the naturalised range.