- The short-term duration of most ecological studies can make it difficult to capture the long-term dynamics of ecosystems and populations. Infrequent or high-impact events can be missed, or erroneously documented as baselines. Long-term ecological research enables a deeper understanding of complex processes, provides a foundation for future insights, and can help improve research methods and management. In my research, I used long-term data collected at the H.J. Andrews Experimental Forest in Oregon’s Western Cascades, established by the U.S. Forest Service in 1948. The H.J. Andrews Experimental Forest is one of 80 U.S. Forest Service Experimental Forests and one of 28 sites in the National Science Foundation-funded Long-Term Ecological Research Network.
Air temperature is a critical variable in ecology because it regulates biological processes, influencing growth, development, reproduction, thermoregulation, and the phenology of organisms, as well as biogeochemical rates. Air temperature data has been recorded globally for hundreds of years, and is one of the most highly leveraged research investments. However, air temperature time series can include substantial random variation and systematic deviation (bias) from non-climatic artifacts, such as instrument configuration and instrument and method changes. The goal of my first study was to quantify measurement differences between the current air temperature instrument used at H.J. Andrews Experimental Forest’s primary climate station and three other instruments that have been used there over the past twenty years. The current instrument is widely considered to be the most accurate, while the other instruments are prone to warm-biased measurements. The magnitude of bias depends on siting and environmental conditions (particularly incoming solar radiation, albedo, and wind speed), time of day, and the measurement’s temporal and statistical aggregation. While other field studies have quantified differences between air temperature instruments, they were typically only a few weeks duration and did not explore how bias aggregates at different temporal and statistical resolutions. Using data collected continuously over a period of six years, we found that measurement differences for average temperatures in the daytime were as large as 7.7°C, median differences up to 0.9°C, and that as many as 68% of measurements had differences +/-0.4ºC. Daily maximum resolution was the most sensitive to differences, with median differences up to 1.2°C, and as many as 78% of measurements with differences +/-0.4ºC. H.J. Andrew’s air temperature data are publically available on the web site, and often used by researchers in trend analyses, such as for climate change. Data from warm-biased instruments, especially when combined and treated as continuous with data from more accurate instruments, can complicate analyses and may lead to erroneous conclusions. One of the practical goals of the study was to explore if incoming solar radiation, reflected solar radiation, and wind speed observations recorded concurrently with air temperature data can be used to model measurement differences. We found that differences can be relatively well-estimated using models incorporating these variables. The results of this study will be used to supplement H.J. Andrews air temperature data and metadata, and can potentially be used by researchers to adjust measurements taken by warm-biased instruments at the H.J. Andrews climate station.
The goal of my second study was to evaluate how terrestrial flying insect abundance and thermal patterns vary across landscape gradients over a period of five years, and if there is a relationship between abundance and thermal variation. Complex forest landscapes like H.J. Andrews can contain a variety of microclimates, which have the potential to act as refugia for insects and other wildlife in changing climate conditions. Even though insects are the most abundant and diverse organisms, they are underrepresented in the literature, with a bias towards species of medical and economic importance. This study contributes valuable information about the springtime abundance dynamics of a broad flying forest insect population, as well as three genera from the Diptera (true flies) and Coleoptera (beetles) orders, for which there is little information in the literature. We found that thermal patterns (represented as cumulative growing degree days) and insect abundance (broadly and for the three genera) were widely variable across sixteen sites within a year, and less variable within sites over five years. One year in particular (2011) had dramatically different abundance patterns than the other four years of the study. We also modeled the relationship between cumulative growing degree-days and insect abundance at each site, and found the relationship was widely variable in a sampling year among sites, as well as within sites among sampling years, suggesting that factors other than thermal accumulation may strongly influence abundance. These types of long-term studies are important to establish baselines and understand the environmental conditions under which terrestrial flying insect abundance within a forest can fluctuate, as well as the habitat characteristics that may be associated with abundance.