Spatio-temporal patterns of phenological development in Germany in relation to temperature and day length

Abstract

Phenological development of crops has been extensively studied in field experiments but less so at larger scales for which data availability is often limited. To what extent the spatio-temporal variability of crop development can be explained by relationships derived from field studies such as the temperature sum concept used in many crop models is unclear but the question could entail the large scale application of these models. The aim of this study was to analyze the spatio-temporal patterns of crop phenological development in response to temperature and day length. We used a comprehensive dataset (656,234 phenological observations at 6019 observation sites) about the phenology of oat ( Avena sativa L.) and related climate data from Germany for the period 1959–2009. Our results show that the statistically significant warming trend since 1959 resulted in an earlier onset of all phenological stages and a shortening of most phenological phases with a 17-day earlier onset of yellow ripeness and a shortening of the “sowing to yellow ripeness” phase by 14 days. There was also a distinct spatial pattern in phenological development, with differences among eco-regions in the occurrence of development stages of 15–26 days and the length of the phases between stages of 6 and 21 days. Most of this spatio-temporal variability could be explained through the effects of temperature and day length. However, temperature sums (thermal times) and day length corrected temperature sums (photo-thermal times) also varied in time and space, pointing to the use of different varieties over time and across eco-regions. A considerable part of this variability in temperature sums and photo-thermal times could be explained by the mean temperature during the development periods. This may provide a means of modelling farmers’ adaptation to climate change using varieties of different maturity types; but it requires further investigation. The good agreement of the thermal and photo-thermal requirements of oat computed in this study with relationships known from field experiments supports the use of the temperature sum concept for large scale application to simulate crop phenology in response to temperature and day length. The analysis should be extended to other crops and regions to further evaluate the observed spatio-temporal patterns in crop phenology and the relationships explaining these patterns.