Predation of cabbage pests: a combined ecological and molecular approach

Abstract

Plutella xylostella L (Lepidoptera: Plutellidae) and Crocidolomia pavonana F (Lepidoptera: Crambidae) are serious pests of Brassica crops in the highlands of West Java, Indonesia, where they co-occur, and in Southeast Queensland, Australia where C. pavonana is a serious pest early in the crop season (February–May) and P. xylostella is a pest later in the year (June-November). Introductions of parasitoids of P. xylostella, especially Diadegma semiclausum Hellen (Hymenoptera: Ichneumonidae), have had considerable success in both countries, but no effective larval or pupal parasitoids of C. pavonana are known. The use of insecticides to manage C. pavonana disrupts P. xylostella parasitoids, leading to pest outbreaks. Thus, an integrated pest management (IPM) strategy, that takes account of both pests, is needed in both regions. Although P. xylostella parasitoids have been the centre of much research over the past 50-60 years, the impact of predatory arthropods on pest populations has received much less attention. Similarly, little is known about the impact of predators on C. pavonana populations and it appears to be attacked by a particularly sparse parasitoid fauna. In recent years, considerable advances have been made in the development of DNA-based molecular techniques to detect the remains of insect prey within the guts of predatory arthropods. However, the methods have rarely been applied in conjunction with field experiments that measure the impact of natural enemies on pest/ prey populations. In this research, field studies used a combination of ecological (natural enemy exclusion experiments and the construction of life tables) and DNA-based molecular methods to quantify the impact of predatory arthropods on P. xylostella and C. pavonana populations and to identify the most important predatory groups preying on these insect pests in the different agro-ecosystems of Queensland and West Java. The specificity of previously designed primer sequences for P. xylostella, C. pavonana and Pieris rapae L. (Lepidoptera: Pieridae) mtCO1 DNA was confirmed by testing them against a wide range of herbivores and predatory arthropods collected from the field in both West Java and Queensland. Based on field studies, key groups of predators were selected to determine the detectability half-life of target P. xylostella and C. pavonana mtCO1 DNA in their guts following consumption of single prey items. For the P. xylostella mtCO1 DNA, when tested at 25 °C the detectability half-life (DHL) was lowest in Clubionidae (18 h) < Linyphiidae (42 h) ≤ Lycosidae (61 h) < Theridiidae (135 h). Similarly when tested at this temperature, the DHL for C. pavonana mtCO1 DNA was lower in Lycosidae (93 h) than in Theridiidae (193 h). Despite having a longer base pair sequence, the target C. pavonana fragment (276 bp) typically persisted for longer in the guts of predators than the P. xylostella fragment (165 bp) and in Theridiidae, the DHL of target DNA of both species decreased with increasing temperature when tested at 20, 25 and 30 °C Field studies in Australia showed that the impact of the endemic natural enemy complex could suppress P. xylostella and C. pavonana populations, but that predators had a greater impact on P. xylostella populations than on C. pavonana populations. Based on their relative abundance, the proportion sampled that tested positive for target prey DNA and the DHL for target prey DNA in gut-contents analysis, Clubionidae and Linyphiidae were identified as the major predators of P. xylostella, although Lycosidae and Theridiidae, and probably the coccinelid H. varieagata, also contributed to P. xylostella suppression. The endolarval parasitoid D. semiclausum also contributed significantly to P. xylostella mortality but overall, the combined action of predators caused greater mortality. Similar analysis identified Araneidae, Clubionidae, Lycosidae, Linyphiidae, Oxyopidae and Salticidae as predators of C. pavonana, although more limited determination of DHL for C. pavonana DNA for these groups makes it difficult to rank predators in order of importance. In Queensland, C. pavonana did not suffer significant mortality due to parasitoids. Studies in West Java investigated the impact of natural enemies on P. xylostella and C. pavonana populations simultaneously. The endemic predator complex and D. semiclausum suppressed P. xylostella populations, but predators had a greater impact than the parasitoid. Crocidolomia pavonana was not attacked by parasitoids in West Java and although predators could reduce the pest population, they had a lower and more variable effect on C. pavonana than on P. xylostella. Most foliar dwelling spiders (Araneidae, Clubionidae, Gnaphosidae, Linyphiidae, Salticidae and Tetragnathidae) consumed P. xylostella but only a small number of Araneidae and Gnaphosidae tested positive for C. pavonana DNA. When sampled at the soil surface, Gnaphosidae and Lycosidae were positive for C. pavonana DNA and only Lycosidae were positive for P. xylostella DNA. The predatory insect fauna was more diverse in West Java than in Queensland and many groups, but particularly larval Syrphidae, adult and larval Coccinellidae and Staphylinidae tested positive for both pests.