How long do population level field experiments need to be? Utilising data from the 40-year old LTER network

Authors: CUSSER, SARAH - University Of Vermont; Helms Iv, Jackson; BAHLAI, CHRISTIE - Kent State University; HADDAD, NICK - Michigan State University

Submitted to: Ecology Letters
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/2/2021
Publication Date: 2/22/2021
Citation: Cusser, S., Helms Iv, J.A., Bahlai, C., Haddad, N.M. 2021. How long do population level field experiments need to be? Utilising data from the 40-year old LTER network. Ecology Letters. 24(5):1103-1111. https://doi.org/10.1111/ele.13710.
DOI: https://doi.org/10.1111/ele.13710

Interpretive Summary: We analyzed data from experiments in ecosystems across the US to see how long field experiments in ecology need to run to be confident in the results. We find that nearly half of the trends needed at least 10 years of study to generate consistent results, and sometimes needed more than 20 years.

Technical Abstract: We utilize the wealth of data accessible through the forty-year-old Long-Term Ecological Research (LTER) network to ask if aspects of study environment or taxa alter the duration of research necessary to detect consistent results. To do this, we use a moving-window algorithm. We limit our analysis to long-term (>10 year) press experiments recording organismal abundance. We find that studies conducted in dynamic abiotic environments need longer periods of study to reach consistent results, as compared to those conducted in more moderated environments. Studies of plants were more often characterized by spurious results than those on animals. Nearly half of the studies we investigated required 10 years or longer to become consistent, where all significant trends agreed in direction, and 4 studies (of 100) required longer than 20 years. Here, we champion the importance of long-term data and bolster the value of multi-decadal experiments in understanding, explaining, and predicting long-term trends.