Aspects of the ecology and behaviour of Batrachochytrium dendrobatidis on amphibians in South Africa

Abstract

Amphibians have been experiencing threats to their survival in terms of habitat destruction and disease. Some species are more sensitive than others, especially towards disease and the lineages of the disease that may impact on them. Batrachochytrium dendrobatidis (Bd) is a fungal organism causing chytridiomycosis in amphibians and has been responsible for declines and even extinctions of amphibian species around the globe. This fungus has been studied since 1998 and great progress has been made in terms of understanding the conservation implications and resolving its genome. Recent discovery of recombinant lineages, although rare, brought the possibility that new lineages may disrupt the equilibrium seen between amphibian populations and prevalence of this fungus therein. In the past lineage identification could only be done with whole genome sequencing, and this limited our investigation into the influence infection with Bd may have on the host. The development of lineage-specific primers for qPCR now enables us to study lineage level effects of Bd on the host as well as allowing for the study of co-infection with multiple lineages. Our experiments showed that although South Africa is host to both the hypervirulent BdGPL and the hypovirulent BdCAPE, the latter cannot be considered of less danger to amphibians, as we found that BdCAPE had more transmission success to naïve individuals during our experiment, as compared to BdGPL. Our experiments also showed that co-infection with multiple lineages on a single individual is possible, when individuals infected with different lineages of Bd inhabit the same habitat. This is a unidirectional situation, as only individuals already infected with BdCAPE became infected with BdGPL as well. When studying the behaviour of Bd infected anuran larvae, it was seen that differences in behaviour could be seen based on the lineage of infection, hence not only is the virulence and transmission ability lineage-specific for the different lineages found in South Africa, but also the behavioural influences an infection has on the host. BdCAPE was seen to increase certain movement parameters compared to the control animals, while BdGPL decreased the same parameters when compared to the control animals. The influence of these altered behavioural patterns can potentially influence successful predation upon infected tadpoles, as well as their exploration and foraging efficiency in order to cope with their changing environments or physiological needs. Since South Africa hosts more than one lineage of Bd, unidirectional co-infection was proven through experimentation and a recombinant lineage has already been isolated originating from genetic exchange between BdGPL and BdCAPE, it became necessary to model the lineage-specific predicted distribution for the different lineages within the South African landscape. Using a database consisting of Bd records from 26 African countries as well as xii field data collected within South Africa, we were able to determine that large areas of South Africa are suitable for BdGPL, including virtually all areas where BdCAPE is currently found and predicted to occur. This is of conservation concern as BdCAPE areas can potentially become infected with BdGPL as well, increasing the risk for the amphibian host populations as well as increasing the risk of uncharacterised recombinant lineages forming. The modelling of the South African Bd landscape also showed that low to moderate likelihood areas were still Bd positive, with these sites often infected with both lineages. We also constructed a Bd-distribution prediction map for the African continent, that showed where previous models predicted that Bd can have a disjunct distribution between different countries, these areas are connected by low to moderate likelihood Bd prevalence areas, similar to areas often testing positive in South Africa. In the past the identity of lineages infecting amphibian populations could only be done by whole genome sequencing and required significant financial input and skill. The development of lineage-specific qPCR primers reduced the need for whole genome sequencing and also the time needed to determine the lineage of infection during field surveys. This was tested by using these newly developed primers together with isolation attempts on a large scale transect survey of South Africa spanning from the Drakensberg, along the Orange River to the Namibian border. It showed that these genetic tools were highly successful in identifying the correct lineage of infection at the sample sites and also allowed for the faster recognition of sites that showed multiple lineage co-infection. It is also possible to use these new genotype identification tools in most laboratories previously capable of testing for Bd presence using ITS-qPCR. This, thus, increased the capacity to determine the lineage of infection during surveys, and thus increase our knowledge about Bd lineage distributions globally.Based on recent genomic analysis of all Bd lineages available, it became apparent that Bd did not originate in Africa as previously thought. This necessitated the need to change the hypotheses that were accepted for approximately 15 years that Bd is an endemic pathogen to African amphibian species, to it being an introduced pathogen. The review that formed part of this showed that both BdCAPE and BdGPL can be found in disjunct reaches in Africa, and under the novel pathogen approach species vulnerability need to be tested and conservation programs need to increase their focus on the less virulent BdCAPE as well.