School of BioSciences - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemContrasting population responses of ecologically-similar sympatric species to multiple threatening processes( 2015)Understanding drivers of population change is crucial when species are declining. However, uncertainty about species declines and causal factors are pervasive problems hampering conservation efforts. This is because populations can naturally fluctuate and pre-decline population data are often insufficient to diagnose causes of decline which may be driven by multiple threats, environmental processes and possible interactions. Globally, amphibians are declining at a rate exceeding other assessed vertebrate taxa. Chytrid fungus (Batrachochytrium dendrobatidis) is a key threat that has caused rapid amphibian declines. However, species and population responses to chytrid can vary and may be influenced by interactions with sympatric species, climatic conditions or other threats. Understanding why species respond differently to chytrid is crucial as few options currently exist to manage the threat. I examined a case for the critically endangered Spotted Tree Frog (Litoria spenceri) and the non-threatened Lesueur’s Frog (Litoria lesueurii). Litoria spenceri has been recorded at 50 sites at elevations of 300 - 1110m asl in south-eastern Australia. Litoria lesueurii co-occurs at some sites but has a much broader distribution. Both are stream-breeding species that respond differently to introduced trout (Salmo trutta and Oncorhynchus mykiss) and were suspected to respond differently to chytrid. I constructed a two-species dynamic occupancy model to examine the historic and future changes in site occupancy of both frog species, and evaluated factors that influence their probabilities of local extinction and colonization. Models combined fragmented historic data and more intensive post-decline data collected over a 55-year period (1958-2012) at 49 historic L. spenceri sites. Trout influence could not be examined as they occurred at all but one site. My analysis revealed that L. spenceri has declined from approximately 50% of known historic sites. This decline could be influenced by either (or both) the presence of chytrid or the presence of L. lesueurii, but was most severe when both chytrid and L. lesueurii were present at sites. In contrast, L. lesueurii tended to become more prevalent over time and changes were uncorrelated with the occurrence of chytrid or L. spenceri. Without intervention, the model extrapolated that L. spenceri will continue to decline, and may be extinct by as early as 2035. I hypothesize that L. spenceri decline is most severe at sites when the presence of pathogen-host-reservoir species (or total frog density) facilitates chytrid persistence and maintains chytrid transmission. Several PCR tests exist to detect chytrid but their results may be imperfect. Prior to evaluating the impact of chytrid on wild frog populations, I assessed the diagnostic sensitivity and specificity of a Qiagen quantitative PCR method to detect chytrid using spiked samples. The Qiagen quantitative PCR was selected as preliminary results indicated equal or superior sensitivity to other published methods. The Qiagen diagnostic sensitivity was high (mean generally > 0.937) but varied in response to the concentration of zoospores in the sample. Diagnostic sensitivity could be improved by classifying equivocal results as positive, with minimal impact on diagnostic specificity. Mean diagnostic specificity was 0.961 (95% CI: 0.897-0.995). The results highlight a risk of misclassifying samples due to imperfect diagnostic sensitivity and specificity which should be considered by practitioners when evaluating the impact of disease on wildlife. To clarify the response of L. spenceri and L. Lesueurii to chytrid I assessed the infection state of all frogs each time they were encountered during a four – six year mark-recapture study at 4 sites. I quantified individual’s probabilities of infection, recovery from infection and survival in each infection state using multi-infection-state models. Nine candidate models were compared to account for four factors that could bias parameter estimates: 1) potential misclassification of disease state, 2) disease state affecting detection, 3) impacts of marking on survival or capture probabilities, and 4) the assumed ability of individuals to recover from an infection. The model with the most support (lowest DIC value) for both species assumed some misclassification of an individual’s disease state, equal detection of infected and uninfected frogs, no effect of marking on return rates, and that frogs could recover from an infection. Marking did not reduce the probability of capture or survival of L. lesueurii and the impact upon L. spenceri was uncertain. Chytrid infection reduced the survival of both species. Litoria lesueurii had the higher probability of becoming infected and the lower probability of survival when infected. Both species had similar probabilities of recovery from infection. Discrete-time deterministic multi-state population models were constructed using multi-state mark-recapture and published parameter estimates to assess the combined influence of chytrid and introduced trout upon each frog species. High, mid and low elevation site models were constructed for each species to reflect demographic differences that are known to be correlated with climate and elevation. Age to maturation, clutch size and egg-year1 survival influenced interspecific and intraspecific population-level responses of the two frogs to threats. Crucially, L. spenceri populations were non-viable at high and mid elevations but had some capacity to persist at low elevations when impacted by both threats (in the absence of stochastic processes and pathogen reservoir species). In contrast, L. lesueurii populations had a greater capacity to persist at both low and mid elevation sites. This thesis highlights that population dynamics and the causes of decline must be carefully evaluated to understand species’ contrasting responses to multiple threats and to identify effective conservation management solutions. Chytrid can severely impact both threatened and non-threatened frogs at an individual-level. However, species population-level responses to chytrid are influenced by their ability to compensate for pathogen-induced mortality through recruitment. Interspecific and intraspecific differences in amphibian populations’ risk of extinction can occur when species recruitment is influenced by site-specific environmental processes and other threats. Furthermore site-specific extinction risk may also be exacerbated by the presence of native sympatric species that act as pathogen reservoir hosts or competitors. In this case, mitigation of threats to L. spenceri might be achieved at some sites by enhancing recruitment through trout management. However, trout mitigation may increase chytrid infection risk, which should be monitored and potentially concurrently managed. Prudent management action is required for species like L. spenceri that are facing multiple threats to avoid undesirable conservation outcomes and to prevent extinction.