From climate perceptions to actions: A case study on coffee farms in Ethiopia

Increasing temperatures and shifting precipitation patterns have major consequences for smallholder farmers, especially in the Global South. Our study examined spatial patterns and climatic drivers of farmers' perceptions of climate change, and how these perceptions translated into adaptation actions. We interviewed 56 farmers in southwestern Ethiopia and analyzed ERA5-Land reanalysis climate data from 1971 to 2020. The majority of farmers perceived the recorded temperature increase as well as a decrease and shift in the timing of rainfall. Perceived climate change varied with local climate factors and not with the rate of climate change itself. Farmers' adaptation practices showed associations with local temperature, but not with farmers’ perceptions of climate change. Our findings highlight that even if farmers perceive climate change, perceptions are most common in areas where climate action is already urgent, and perceptions may not translate into adaptation. Thus, targeted and timely information and extension programs are crucial. Supplementary Information The online version contains supplementary material available at 10.1007/s13280-024-01990-0.

Relocation of coffee to more suitable area * Farmers were asked about "planting shade trees," but the responses and interpretation of this question referred to the same practices listed under the category of "shade management".
For simplicity and to reduce reader confusion by using two different terms for the same practice, I have referred to this question as "shade management" in this study.(Walter et al., 1975).
The diagrams show the average precipitation (blue line) and temperature (red line) as well as the drier conditions (red dots), humid conditions (blue shaded) and wet conditions (blue filled) for each grid cell of the study area.The circles indicate the temperature distribution below (blue) and above (red) the mean temperature across all study sites of 16.8°C.The stress value is 0.102 (Clarke, 1993).

Table S1. Statistical overview of the relationships between perceived changes in climatic
variables during the last 30 years and spatial variation in local climatic variables in Jimma zone in southwestern Ethiopia.Shown are the odds ratio (OR, with 1°C, 100 mm and 0.01 in SPEI as one unit, respectively), p-value, R 2 -value and number of observations.Values are obtained from generalized linear models with a binary distribution and logit link.

Farmers' perception of changes in:
Local so, how do you think the weather/climate has changed during the last 30 years?(In terms of temperature, rainfall etc.) 3. Generally, how do you think the following weather variables have changed during the last 30 years?Please place an "X" at the best answer.observed changes in the climate, how has this impacted your farm?1.How did it affect yield? 2. Is there any other effect of climate on your coffee farm? 3.In your opinion, how informed are you about climate change?a. Very well informed ("I know everything about climate change") b.Informed ("I know the basics of future climate change") c.Not very well informed ("I have heard about climate change before") d.Not informed at all ("no idea what climate change entails") Section B, part 1 1.How did you change your management practice since the fall of the Derg?Why? 2. Have you changed your input use (fertilizer, herbicide)?Why? 3. Have you changed the varieties you plant?Why? 4. Have you managed the shade trees (such as number and choice of trees, shade regulation i.e., thinning or planting trees)? 5. Do you intercrop any other crops or trees in your coffer these days? A. Yes B. No a.If yes: what crops / trees?

Figure S1 :
Figure S1: Seasonal changes in temperature in the Jimma zone in southwestern Ethiopia.

Figure S2 :
Figure S2: Distribution of temperature in °C across the study area based on the ERA5-Land

Figure S3 :
Figure S3: Distribution of precipitation in millimeters along the study area based on the

Figure S4 :
Figure S4: Walter diagrams along the study area in the Jimma zone in southwestern

Figure S5 :
Figure S5: Distribution of temperature gradient in °C per decade along the study area based

Figure S6 :
Figure S6: Walter diagram for grid cell C04 of the ERA5-Land dataset showing the average

Figure S7 :
Figure S7: Walter diagram for grid cell C12 of the ERA5-Land dataset showing the average

Figure S8 .
Figure S8.Relationship between the composition of management practices and

Table S2 .
Statistical overview of the relationships between perceived changes in climatic variables during the last 30 years and spatial variation in the rate of climate change in Jimma zone in southwestern Ethiopia.(OR, with 1°C, 100 mm and 0.01 in SPEI as one unit, respectively), p-value, R 2 -value and number of observations.Values are obtained from generalized linear models with a binary distribution and logit link.The rate of climate change is expressed as the Sen's slope.

Table S3 .
Current SPEI index, and rate of change in the SPEI index, for each of the grid cells in the study area.SPEI values shown are the average monthly SPEI values for the dry period from November to April from 1991-2020, separately for each grid cell in the study area in

Table S4 :
Permutational multivariate analysis of variance (PERMANOVA) on the composition of management practices as a function of spatial variation in temperature and precipitation as well as farmers' perceptions of changes in temperature and precipitation since 1991.The model was run separately for each explanatory variable.PERMANOVAs were based on the binomial deviance dissimilarity measure.P-values were obtained using 999 permutations.