Predicting the severity of Swiss needle cast on Douglas-fir under current and future climate in New Zealand

Abstract

Phaeocryptopus gaeumannii, the causal agent of Swiss needle cast, is widely distributed in plantations of Douglas-fir ( Pseudotsuga menziesii) throughout New Zealand, causing premature abscission of needles and significant growth losses. Data were collected from 34 sites, selected to span a broad range of environmental conditions within New Zealand, to (i) develop models of infection and foliage retention, F ret, and (ii) from these models predict F ret by region under current and future climates, using the factorial combination of 12 Global Climate Models (GCMs) and three emission scenarios (low, B1; medium, A1B; and high, A2). Pathogen abundance, as measured by a colonisation index (CI norm), was found to exhibit a significant positive linear relationship ( R 2 = 0.53; P < 0.001) with early winter (June) air temperature, and a marginally significant quadratic relationship ( R 2 = 0.19, P < 0.036) with late spring (November) rainfall that increased to a maximum at 149 mm month −1, before declining. In combination these two variables accounted for 64% of the variation in CI norm among sites. Predicted CI norm was the single variable most strongly correlated with measured F ret. The relationship between predicted CI norm and F ret was described by a significant ( P < 0.001) negative linear relationship that accounted for 51% of the variation in the data. Addition of further climatic variables did not significantly improve predictive power of this model for F ret. Under the current climate, spatial predictions of F ret were between 70% and 100% for most of the South Island, and between 40% and 70% for most of the North Island. When projected out to 2040 there was little change in F ret. However, projections out to 2090 showed substantial reductions in F ret that were positively related to the three forecast emission levels. These reductions averaged 6.5%, 10.1%, and 13.7% for the B1, A1B and A2 emission scenarios, respectively. Reductions in F ret were particularly marked within the North Island under the AIB and A2 emission scenarios. Apart from coastal and low lying regions, large areas within the South Island were projected to sustain relatively low levels of disease, and remain suitable for Douglas-fir.