The fact is, most models or approaches measuring the effects of climate change on agriculture are basically assumptions. It is almost impossible to hold other non-climate change related factors to constant in order to state categorically what variables and extent of such climate change factors that affect agriculture.
However, these models or approaches are (or could at least be) nearly correct. Using a controlled closed environment versus an open space experiment will almost convince you of the effects of climate change on such experimented crop. Then you also have to consider for example in crops, edaphic factors, seedling itself and other natural and unexplanable factors.
Finally, to support the reported case of climate change on agriculture, change in temperature is often cited as a leading effect alongside precipitation rate/rainfall. Example; In Nigeria, rainfall usually begins in early March and peaks at June/July with a break in August then falls steadily till late October then very less to non from November till early March. Today, it's almost hard to predict. Local farmers who relied on this knowledge and planted in recent years faced or are still facing great economic loss due to mismatched rainfall-planting. Now they have to adjust.
Article Impacts of Climate Change on Animal and Crop Production in t...
The simple relationship between agroecosystems production and drought is likely to be veiled by practices used to increase productivity in affected areas. To eliminate climate component, the agroecosystems productivity (Yi) has been examined through a sum of two components: Yi = Yi (T) + Δ Yi (T), where Yi (T) is a contribution of technological factors. which are expressed using a time depended function (trend) that describes the contribution of the socio-economic variables to the agroecosystems productivity in the average climate of a given location. The second (Δ Yi (T)) component is controlled by climate. Herewith, it is assumed that in case of positive values of Δ Yi (T) there is no significant negative climate impact on the agroecosystems productivity, and, on the contrary, decreased (negative values) productivity, exceeding the certain level, is conditioned by excessive or insufficient quantity of the main climate factors (see "New Drought Products: transforming drought information to facilitate decision making" by Daradur at al.)