The ecological and biodiversity conservation values of farm dams: A systematic review

2.2.3 What are the ecological variables used to predict the biodiversity values of farm dams (e.g., vegetation cover, and water quality)?. Studies assessing farm dams’ biodiversity success were categorized based on their objectives. Studies that included production objectives, along with biodiversity objectives, or examined production objectives in response to biodiver- sity relevant variables were included. We focused on three target groups: fauna, vegetation, and water quality. This component of our review was descriptive as there were too few samples in each class to make quantitative comparisons. Heatmaps were produced to visualise patterns of association between these variables and the objective of interest. Our approach is a standard method to summarize scientific information where inconsistent objectives and definitions are used, and data are too sparse to allow formal meta-analysis [48]. 3. Results We identified 104 relevant articles that examined the biodiversity values of farm dams. The articles came from 66 different publication outlets, with all but two being published in peer- reviewed journals, the remaining studies were theses. Studies were conducted across 27 coun- tries, with the most commonly represented countries being Spain (10%), The United States of America (9%), Japan (6%), Australia (5%), and South Korea (5%). We found there has been an increase in the number of studies examining biodiversity in farm dams since 1950 with at least one study per year from 2003. Since 2003, the number of arti-

cles published annually has fluctuated between eleven (11) in 2019 and one (1) in 2016 (Fig 2). 3.1 Risk of bias assessment Our risk of bias assessment revealed considerable variation amongst the 104 articles (see S3 Appendix). Studies with an overall ranking of “low” (i.e., low risk for all assessment criteria) made up 23% of the studies, and studies with an overall ranking of “some concern” (i.e., some concern was noted for only one assessment criteria) made up 24%. We found 32% of studies raised some concern in at least two of the assessment criteria, and 20% of studies had a high risk of bias in at least one assessment criteria. This resulted in 53% of studies with an overall ranking of “high” (Fig 3). This commonly resulted from a lack of sufficient detail to rule out the concern for the risk of bias. No studies had a critical risk of bias. 3.2 What are the characteristics of conservation-focused research on farm dams (e.g., sample size, study duration, taxonomic group coverage, publication year and country of study)? The duration of the studies in 97 of the 104 articles ranged from one day up to five years, with the remaining seven articles fitting our definition of long term, exceeding five years or more. The average farm dam study occurred over approximately 18 months. Sample size (number of water- bodies) ranged from two dams to 16,543 dams with a median sample size of 40 and a mean of 345. Ninety-eight studies utilized a sampling methodology that was observational where

researchers undertook their investigation without manipulation of the study area or subject, with the remaining six studies using interventions such as building fences around the dams. Many of the articles had research objectives, aims and goals that were exclusively ecolog- ically and/or biodiversity motivated, commonly answering questions that sought to improve the biodiversity in or around farm dams (n = 75). The remaining studies (n = 29) posed ques- tions that examined agricultural production in response to biodiversity-relevant variables hereafter referred to as production-based studies. Only one study of the 104 examined had aims and goals that addressed both ecological and agricultural outcomes. PLOS ONE The ecological and biodiversity conservation values of farm dams PLOS ONE | https://doi.org/10.1371/journal.pone.0303504 May 13, 2024 7 / 18