Lagomorph and sheep effects on vegetation growth in dry and mesic grasslands in Otago, New Zealand

Context. Rabbits (Oryctolagus cuniculus) are major pests in the semiarid grasslands of eastern South Island. As the benefits of a viral biological control agent (rabbit haemorrhagic disease virus) introduced in 1997 begin to wane, landowners are again faced with the costs of controlling rabbits. Aim. The study aimed to measure the extent of vegetation growth as rabbit and sheep densities changed to enable land managers to better justify decisions on pest and stock management. Methods. We investigated the relationship between seasonal growth of vegetation at two replicate sites within each of three rabbit-prone areas in Otago. At these sites, the ambient densities of lagomorphs varied naturally and in response to pest-control operations, whereas the density of domestic sheep (Ovis aries) varied according to the farmers’ husbandry needs. The vegetation was highly degraded at one of the areas studied. Rabbits were present at all three sites and hares (Lepus europaeus) at two sites. At each site, we established exclosures that excluded either all herbivores or just sheep. The observed rates of vegetation growth were modelled as a function of prevailing herbivore densities, season and site. The fitted model was used to infer expected rates of change in vegetation biomass at the sites, under varying sheep and lagomorph density indices, and seasons. Key results. In the absence of herbivores, vegetation was predicted to grow in all seasons apart from winter at the degraded sites. In the absence of sheep but in the presence of up to ~10 lagomorphs seen per kilometre of spotlight transect, accumulation of vegetation biomass follows the same seasonal pattern as in the absence of herbivores. As lagomorph counts reach 50 per kilometre, vegetation biomass accumulates only in spring and summer and only at the least degraded sites. The maximum stocking rates for sheep for which positive biomass accumulation could be maintained under varying lagomorph densities was predicted from the model. Conclusions. At the least degraded sites some sheep could be grazed while maintaining positive pasture growth, even at high rabbit densities, except in winter if rabbit-density indices exceeded ~30 per kilometre. At the most degraded sites, a very low density of sheep could be maintained, but only in spring and only if lagomorph density indices were below ~10 per kilometre.