loading page

How long do population level field experiments need to be? A meta-analysis across the 40-year old LTER network
  • +1
  • Sarah Cusser,
  • Jackson Helms IV,
  • Christie Bahlai,
  • Nick Haddad
Sarah Cusser
Michigan State University
Author Profile
Jackson Helms IV
Michigan State University
Author Profile
Christie Bahlai
Michigan State University W K Kellogg Biological Station
Author Profile
Nick Haddad
Michigan State University W K Kellogg Biological Station
Author Profile


Long-term experiments are important in evaluating ecosystem properties and processes that are slow to develop or require proper evaluation over an appropriately variable climate. We repurpose the wealth of data accessible through the forty-year-old Long-Term Ecological Research (LTER) network with a novel moving window algorithm and meta-analysis approach to ask if aspects of study taxa or environment alter the extent of research necessary to detect consistent results, or the proportion of spurious short-term trends. We found that experimental studies focused on plants, and those conducted in dynamic abiotic environments, were characterized by longer critical temporal thresholds and more spurious trends. Further, nearly half of the studies we investigated required 10 years or longer to reach a temporal threshold, and 4 studies (of 100) required longer than 20 years. We champion long-term data and argue that long-term experiments are more necessary than ever to understand, explain, and predict long-term trends.

Peer review status:ACCEPTED

10 Nov 2020Submitted to Ecology Letters
12 Nov 2020Submission Checks Completed
12 Nov 2020Assigned to Editor
13 Nov 2020Reviewer(s) Assigned
06 Dec 2020Review(s) Completed, Editorial Evaluation Pending
07 Dec 2020Editorial Decision: Revise Major
01 Feb 20211st Revision Received
02 Feb 2021Submission Checks Completed
02 Feb 2021Assigned to Editor
02 Feb 2021Review(s) Completed, Editorial Evaluation Pending
02 Feb 2021Editorial Decision: Accept