Management of an endemic New Zealand scarab pest (Costelytra zealandica (White, 1846)) in vineyards through understanding its behaviour

Background/Question/Methods Larvae of the endemic beetle C. zealandica have been causing serious damage to New Zealand pastures for more than 100 years. However, in the last 30 years, adults have started to attack the foliage of several fruit crops. Amongst those, vines are heavily damaged, resulting in winegrowers applying a range of pesticides. To reduce these inputs, we investigated the nocturnal behaviour of C. zealandica adults in terms of flying direction, flying time and landing patterns on vine plants to help develop agroecological strategies to mitigate damage. We used infra-red-sensitive video cameras and field observations in vines in the Awatere Valley, New Zealand (41°S; 173°E). The adults flew from outside the vineyard, landing mainly at the edges of the vineyard blocks. They started flying at dusk, 20 min after sunset, and this activity lasted for around 30 min. After this period, they accumulated on young vine plant tissue, mating over a 2 h period. Then they started to feed on the vines for around 3 h. From 23:00 to 01:00 h, they started to drop from the plants to the soil, in which they laid eggs in burrows under the vines. Based on these activity patterns, we tested the application of different feeding deterrents to the vine foliage, and the spreading of mussel shells on the under-vine soil as an alternative management approach. We tested diatomaceous earths, kaolin particle films and a mix of both as feeding deterrents at a rate of 400L/ha (20g/L), sprayed onto the vine foliage in a randomized block design. A generalized mixed effect model was used (error family=binomial), with the plants sampled per bay (4 vines each bay) as the random effect. The mussel shells were crushed and applied at a rate of 1.4m³/bay (0.6x7.5x0.3m) in the first bay in the outermost vine rows of the vineyard, in a linear randomized block design. A generalized linear model was used (error family=Quasi-Poisson).

Results/Conclusions The feeding deterrents significantly reduced damage on vine plants by 37% (F=16.5; df=3/56; p<0.001) compared with control, with no statistical difference between the treatments. Surprisingly, the mussel shells reduced the number of adults landing on the plants by 69% (F=78.2; df=1/28; p<0.001). We suggest that these approaches can reduce the damage produced on vine plants and affect the landing behavior of this pest, respectively. If both approaches are combined, these may operate additively or synergistically.