Effects of predation and rabbit haemorrhagic disease on rabbit population dynamics in New Zealand

Abstract

This thesis investigated the effects of predation and a newly introduced biological control rabbit haemorrhagic disease (RHD) - on rabbit (Oryctolagus cuniculus) population dynamics in two bioclimatically diverse regions of New Zealand.

In New Zealand, rabbits are classified as pests because they have a negative impact on agricultural and conservation values. The chronic rabbit problem areas are primarily in the ‘semi-arid’ unimproved grasslands of the South Island, and less so in lowland, moderate rainfall areas with improved pastures. This thesis investigated the hypothesised paradigm that rabbit populations in lowland areas, are generally driven by 'top-down' processes that keep them at low stable densities. The mechanism for this process is considered to be high mortality of young rabbits. In contrast, rabbit populations in mostly semi-arid areas are thought to be driven by ‘bottom-up’ processes (i.e., food availability). The effect of RHD on these processes was also investigated. One hypothesised effect of RHD epidemics is that the virus would depress rabbit populations so that they are more often within the range of densities at which top-down processes dominate. The corollary is that the efficacy of RHD may depend on predation of juvenile rabbits that develop age-specific immune responses to RHD.

Predator-perturbation experiments were used to investigate this hypothesis and to determine the survival of juvenile rabbits. In both improved pasturelands (North Canterbury) and semiarid rangelands (Central Otago), rabbit abundance and pre- and post-emergent rabbit survival were monitored on four 30-ha sites, two of which were subject to predator removal. RHD spread through all sites during the study. The results provide support to the paradigm of top-down processes in lowland regions and bottom-up processes in semi-arid regions. In North Canterbury, rabbit abundance declined on all sites after an RHD epidemic, but the declines were more pronounced where predators were not removed. In contrast, in drier areas of Central Otago there was no apparent effect of predator reductions on rabbit abundance. There, rabbit densities declined at similar rates during the RHD epidemic regardless of the predator treatment. This suggests that predation; at least when combined with RHD, is a significantly stronger process in lowland regions than in the semi-arid regions of New Zealand.

Survival of juvenile rabbits was a key factor determining population abundance between treatments. In North Canterbury, survival of young rabbits before they left the nest was lower on the two sites where predators were present (34%) than on the two sites where they were removed (51 %). In contrast, there was no measurable effect of predator reductions on pre-emergent survival in Central Otago (predator sites, 57%; predator-removal sites, 62%). Survival of post-emergent radio-collared juveniles were low for all sites (0-22%). It is likely that predation and RHD led to these low rates of post-emergent survival. The age at which juvenile rabbits become susceptible to RHD, the timing of RHD epidemics, and the abundance of predators are likely to be important in determining survival of juvenile rabbits.

Demographic modelling of the rate of increase of rabbit populations in North Canterbury and Central Otago provided some support for the hypothesised top-down/bottom-up processes. In North Canterbury, rabbit populations on predator-removal sites had an annual rate of increase of 30%, whereas the rate of increase was effectively zero where predators were present. In contrast, there was little difference between predator and predator-removal sites in Central Otago (56-66%). Survival of juvenile rabbits and productivity were the key factors determining rate of increase.

Rabbit haemorrhagic disease appears to be most effective in semi-arid regions of New Zealand, particularly in the Mackenzie Basin, and least effective in lowland North Canterbury. From 1974 to 2001, necropsy data were collected from c. 44000 rabbits to assess the impact of RHD on demography of rabbit populations in North Canterbury, the Mackenzie Basin, and Central Otago. There were proportionately more young rabbits in all three regions post-RHD, but there was no effect on time of birth or sex ratio. There was no consistent pattern between regions for carcass weights of rabbits older than 12 months, for productivity, and for pregnancy rates. However, in general reproductive rate and carcass weight of rabbits increased where RHD worked best. Three key mechanisms may explain the observed patterns. First, reproductive rate and carcass weight may be density-dependent. Second, RHD may result in differential mortality of rabbits, resulting in RHD causing mortality in lighter, older, weaker, or less fecund individuals. Third, RHD may result in differential levels of morbidity between populations, which may have sublethal impacts on their demography. Further research is required to determine the relative importance of the proposed mechanisms.

The results of this study have significant implications for managers seeking to control ferrets to reduce the prevalence and spread of bovine Tb. Predator control in lowland regions such as North Canterbury may reduce the efficacy of RHD, and result in an increase in rabbit abundance, whereas predator control in semi-arid regions is unlikely to affect rabbit abundance. Further research is required to determine the relative importance of factors contributing to the observed top-down/bottom-up processes at these study sites, and how representative the results are for each region. Where ferret control is imperative in lowland regions, integrated control programmes that simultaneously control both ferrets and rabbits are recommended.