A test of metabarcoding for Early Detection and Rapid Response monitoring for non-native forest pest beetles (Coleoptera)

Abstract

In response to the threat of introductions of non-native forest insects, the Early Detection and Rapid Response (EDRR) program in Alaska monitors for arrivals of non-native insects, an effort that is limited by the time required to process samples using morphological methods. We compared conventional methods of processing EDRR traps with metabarcoding methods for processing the same samples. We deployed Lindgren funnel traps at three points of entry in Alaska using standard EDRR methods and trap samples were later processed using routine sorting and identification based on morphology. Samples were then processed using High Throughput Sequencing (HTS) metabarcoding methods. In three samples bycatch was included and in three samples non-native species were added. Morophological and HTS methods yielded generally similar results for scolytine and cerambycid beetle assemblages, but HTS provided more species-resolution identifications (46 species) than morphological methods (4 species plus the 3 non-native species known a priori). None of the non-native species were detected by HTS. Including bycatch did not appear to hinder identifications of scolytine and cerambycid beetles by HTS. From among the bycatch, two Palearctic species adventive to North America, Placusa incompleta Sjöberg, 1934 and Hydrophoria lancifer (Harris, 1780), are newly reported from Alaska. We do not recommend replacing our current morphological monitoring methods with HTS methods because we believe that we would be more likely to detect known non-native pest species using morphology. However, we would use HTS to increase our sample size without greatly increasing time required to process samples. We would also recommend HTS methods for surveillance monitoring where the set of target taxa is not limited to known pest species.

Keywords

• Scolytinae
• Cerambycidae
• biomonitoring
• High-T