A comparison of sampling methods and temporal patterns of arthropod abundance and diversity in a mature, temperate, Oak woodland

Abstract

Arthropods underpin fundamental ecological processes such as herbivory, pollination and nutrient cycling, and are often responsive to subtle changes in environmental conditions. Thus, changes in their abundance and phenology may be crucial indicators of system-wide responses to climate change.

The new Birmingham Institute for Forest Research (BIFoR) Free Air Carbon Dioxide Enrichment (FACE) facility provides a unique opportunity assess arthropod diversity and abundance in mature deciduous forest and the effect of sampling method and seasonality. This is an essential first step before attempting to measure the potential impacts of climate change, such as elevated CO₂, on arthropod populations. Two fundamental criteria are: i) diverse sampling methods in order to effectively assess diversity and in particular, differences between structural layers of the woodland system, e.g., ground, sub-canopy and canopy layers, ii) a temporal resolution that can identify seasonal patterns of change (phenology). This paper sets out the methodological approaches employed to achieve these objectives.

A total of 22,568 invertebrates from 108 families were sampled across 12 months of continuous sampling using a range of techniques from forest floor to canopy. Diptera were the most abundant order sampled and had the greatest number of families represented (45). Phenology patterns generally followed the anticipated seasonal cycle, with increasing abundance and diversity from spring to summer. Temperature was the best environmental predictor of abundance within Malaise and pitfall traps. Precipitation was not correlated with any monthly patterns of trap data. Yellow pan traps collected more arthropods than white or blue traps. Canopy beating yielded a greater diversity than that in the understory samples.

These data provide an important baseline from which to assess any future impacts of eCO₂ over the 10-year BIFoR FACE experiment, and highlight the importance of employing diverse sampling methods, temporal replication and measuring environmental factors over appropriate timescales.