Understanding scientists’ communication challenges at the intersection of climate and agriculture

Statistical analyses were conducted in R (version 3.5.1). Since the responses were not nor- mally distributed (determined by a Shapiro-Wilk normality test), a non-parametric Kruskal- Wallis test by ranks was used to determine if the distribution of responses were significantly different across stakeholders overall by question. To determine if there was a significant differ- ence between two stakeholders by question, a Wilcoxon-Mann-Whitney test with Bonferroni correction method adjusting for family-wise error rate was employed to conduct pairwise comparison. Results Response demographics The overall scientist survey response rate was 42.8% (N = 265). The intent of the initial survey was to collect perspective from all roles within organizations; however, respondents were pri- marily scientists. Only eight respondents identified as nonscientists and these were removed from analysis and response rate calculation. Surveys were considered complete if they answered at least one question (S1 Table). Not all respondents answered all questions; there- fore, response rates vary by question (“N” henceforth indicates the total number of respon- dents). Each of the contacted organizations (government/AACS) had at least one respondent, with the majority of responses from the USDA, with the following distribution: AASC = 9.5% (n = 10), EPA = 1.0% (n = 1), LCC = 27.6% (n = 29), NOAA = 12.4% (n = 13), USDA = 33.3% (n = 35), USGS = 16.2% (n = 17). The following results are presented as a combined audience since no statistical differences were found when aggregated into separate groups. Mean

respondent age was 52.5 ± 10.6 sd (N = 216) and were majority male (74.4%, N = 215). Respondents most frequently identified as a life, climate, and/or physical scientist (S1 Fig); since participants were not all climate scientists by education/degree but conducted climate- related work, respondents are not referred to as “climate scientists” but rather “scientists.” Experience The respondents indicated being in their current role for 11.6 ± 10.1 sd (N = 220) years and spent on average 16.9 ± 11.6 sd years (N = 218) in a climate-related field as their occupation. The majority (86.4%; N = 213) worked with two or more stakeholder groups, most commonly the “general public” (S2 Fig). The respondents that indicated experience with multiple stake- holder groups were queried on the group they selected as their main priority. Of those, most (88.9%; N = 181) indicated their focus was “producers,” “general public,” or “policymakers.” Communication training Nearly all (90.7%) of the respondents (N = 225) agreed that scientists should receive training on how to communicate scientific findings to nonscientists. However, the majority (64.9%; N = 225) were not formally trained on how to communicate with stakeholders. Open ended responses were coded for formal and informal training in which they had participated; these included formal training (N = 74): “workshops” (n = 28), “classes” (n = 11), and “in-service” (n = 11) and informal (N = 162): “on-the-job” (n = 101) and “publications” (n = 31). Perception of stakeholders

Scientists ranked their level of agreement with each climate change statement. They were also asked to rate their perspective on how the five stakeholders would respond to the same climate change statements. All but one statement (“Earth’s climate conditions occur in a cyclical pat- tern”) was found to have responses that were statistically different across the stakeholder PLOS ONE Understanding scientists’ communication challenges at the intersection of climate and agriculture PLOS ONE | https://doi.org/10.1371/journal.pone.0269927 August 2, 2022 7 / 22