CLIMALERT − Methods and Results: Interviews with Farmers
21 face-to-face interviews were conducted with farmers between June 2018 and February 2019. These interviews aimed to gain an understanding of the context in which farmers make decisions and the influence that different factors – i.e. weather, biodiversity, politics, history and the economy – have on the types of agricultural management decisions that are made.
Results
In a first step, we focused on three particular crops that are common in Central Germany: canola, wheat and corn. The motivation for doing so was to investigate whether different weather-related information is required for different crops. We focused on four decision-making processes (i.e. sowing, fertilising, pesticide application and harvesting). For each of these decisions, information was collected about the timing of the decisions, the information used and required as well as the sources of information used and the quality of those sources. Finally, we were interested in the level of flexibility that farmers have in regards to each decision and how they deal with the risk that the weather does not turn out as forecasted / expected.
Only 11 of the 21 interviews are included in the following results (we are currently in the process of analysing the remaining interviews).
Decisions taken by farmers at each phase of a crop’s life are the result of a mixture of factors of which weather-related information is just one (I6, I8, I10). Decisions are based on crop rotation plans, the quality of the seeds on offer, previous experience with certain types of seeds, weather history, whether temporary workers need to be organised, and what sort of weather conditions occurred before the current crop phase, etc. – “our business is complicated” (I8). In addition, farmers were quick to point out the myriad of weather services that already exist (listed in the tables above). What the results also show is a general lack of trust in weather forecasts – “… what sometimes amazes me [is] the difference between weather forecasts. The difference is sometimes enormous!” (I9). As a result, there was a level of scepticism about the added value of a new app as well as weather-related apps more generally – “There is so much flexibility that is expected of us, that can’t be delivered by an app” (I6).
However, farmers still use weather-related information, including forecasts, to inform their decisions. How often farmers check weather-related information depends on the types of decisions that need to be taken. In regards to forecasts, I9 stated that when decisions have to be made, forecasts become an important point of reference – “Okay, I have already, the harvest is complete, everything is finished, everything is okay and then … I can relax a little, at that moment the weather doesn’t matter. But, as soon as something needs to be done, like a hay harvest … sowing, then it could be that I check the weather four, five, six times. Just to see if, has something changed?” (I9).
All farmers use weather forecasts but there is a general agreement about their limited accuracy. Generally, 24 hour forecasts are seen to be reliable (I7), some interviewees extended this to 3-day forecasts (I2, I5). But anything over 3 – 5 days is seen to be like playing lotto (I5, I6, I7). In addition, storms are seen as problematic and rarely behave as forecasted (I6, I7). Therefore, the future is perceived as difficult to predict. However, this belief also appears to be key motivation for why farmers farm. The “game” as it is sometimes referred to by the interviewees (I5, I7), requires farmers to play by some rules but also the need to be flexible and take well-informed choices in order to remain profitable. Generally, weather information is seen to be a guide or an indication of how the day might turn out, weather-wise. This, combined with the number of service providers and the differences between weather forecasts leads to a certain type of optimism – “there are so many services … you look at each of them and you choose the best forecast” (I8). Optimism in the face of the unknown is a reoccurring theme throughout the interviews. For example, I9 explained that “it might turn out to be a very rainy summer. That wouldn’t be good. But you can’t do anything anyway and I am optimistic”.
Although decisions are influenced by the weather, each phase (i.e. sowing, fertilising, pesticide application and harvesting) still takes place despite of the weather. Often farmers will take a chance and hope that their decision pays-off. This includes taking decisions based on weather forecasts. For example, when rain is predicted and farmers postpone their work and then it does not rain, they have lost precious work-time (I3, I5). In other words, agricultural decision-making is a mixture of strategy and chance. The crop rotation plan provides a long-term structure and the crop phases need to take place. The weather poses risks that the fields are too wet to plough, too cold or dry for the seedlings to thrive, too wet to fertilise or apply pesticides, or too wet to harvest. Regardless of the weather, all these phases take place. However, the profitability of the harvest depends on the weather at each phase of the crops life as well as the decisions that the farmer takes in order to overcome the challenges the weather poses.
The aim of the CLIMALERT project is to understand how farmers and water managers use weather-related information to inform their decisions and how the UFZ might be able to assist in the improvement of such information. The UFZ has developed a Drought Monitor and would like to use CLIMALERT as an opportunity to further develop this tool and make it more user-friendly. The first question, however, is if information about drought and, therefore, soil moisture is relevant for farmers. The results of the interviews show that soil moisture is of great importance. Moreover, there is a potential gap in access to this information. None of the farmers that we interviewed reported measuring their own soil moisture. This is because it is seen to be too complicated (all interviewees). Generally, this information is received through literature and the press (I1) or UFZ (I1, I2, I5).
When referring to soil moisture farmers tend to focus on the soil being too wet or too dry. In regards to whether it is too wet, the general consensus is that in regards to machine access to the field, one does not need an official measurement, instead it is clear when one goes out to the field if it is possible to achieve the planned work or not. However, for decisions related to irrigation, pesticide use and fertilisation, information about soil moisture could be very helpful. For farmers with irrigation such information could help to save money − “if we had a forecast model that told us that 20 litres of rain is going to fall tomorrow, then we could turn the irrigation system off because we can save money” (I6). It could also help to predict potential infections – “I could have the situation where the soil is overgrown, that it is light grey and is not able to absorb the required nutrients … this information is based on the experience that one has. Long-term information [about soil moisture] is very important for the fertilisation process” (I3).
In regards to information about potential droughts, interviewees saw the potential of receiving information about trends which could inform them of historical crop development in previous times of water stress (I5). On the other hand, in regards to drought forecasts, information about soil moisture could either improve or dampen one’s optimism. If the models show that there is still water in the system, farmers can remain optimistic but if they do not have access to an irrigation system there is little that they can do to change the situation (I5, I9) – “It is naturally still only a glimmer of hope: aha, we still have 80, 60, 40% field capacity. When nothing comes, then we can see that the soil reserves are dissolving. But there is nothing we can do about it … it is of course, interesting to know how the field capacity has changed. At least during years such as this one; that would be very interesting. When one could say: Okay, we have one more week, when it doesn’t rain then it will become critical. But it is only an estimate of the situation. We can’t change it” (I5). In other words, having access to historical trends is seen as desirable but forecasts are viewed with scepticism (I3, I5, I6).
This perceived inability to control the weather and the future is a strong theme throughout the interviews. I7 stated that it is interesting to gain an impression of past trends but it does not reduce uncertainty about the future − “it remains a game of poker, yeah? It is a reference, a good reference, if one could say: ‘look, a few years ago on 15th June Leipzig-Holzhausen experienced this figure and the plant developed in a similar fashion. Or, the plant is very advanced or experienced a slow development … that isn’t bad … it is a statistic. But, it doesn’t take away the uncertainty … if we knew that the weather from now on would develop in the same way that it did one or two years ago, then we could work with that. But, you don’t know that” (I10).
Finally, in order to improve weather-related information, the CLIMALERT team set out to understand the weather-related thresholds that force farmers to take certain decisions. However, the results of the interviews show that such thresholds are difficult to provide – “During the autumn sowing period, everything within 5 litres per day, that doesn’t bother us so much … except when the soil is already full. Then every additional 5 litres is really annoying” (7). In addition when asked if it was possible to provide thresholds, I10 stated – “No, it is not possible [to name thresholds] because it is dependent on different factors. For example, on the crop rotation plan, on the quality of the soil” (I10).