Linking pasture, livestock productivity and vertebrate pest management

The study of pastures, pests (rabbits and caterpillars) and sheep productivity by Barlow (1987; New Zeal J Ecol 10: 43–55) is reviewed, updated and extended. Pasture growth was modelled as a dynamic process, though sheep and rabbit abundance were not dynamic. The model predicted that there was a parabolic relationship between sheep productivity and sheep stocking rate with the effects of rabbits being to shift the relationship lower and to the left. The relationship is analogous to a model of sustained harvest of a wild population, and the shape parameter of each has similar numerical values (2 to 3). The Barlow model also predicted a negative relationship between sheep productivity and rabbit abundance, with the relationship being curved with fixed stocking rates. Analysis of data from an independent experimental study shows more support for a negative quadratic (concave down, R2 = 0.509) than a negative linear (R2 = 0.416, ΔAICc = 2.770) relationship. The Barlow (1987) study assumed a positive linear relationship between the area of denuded pasture and pest abundance. A model selection analysis of a priori models of disturbance by feral pigs provides support for a positive curved relationship (R2 = 0.854) and a positive linear relationship (R2 = 0.357; ΔAICc = 0.03) between the area of denuded pasture (as disturbed ground) and pig abundance. The general results and their implications are discussed.