Using PhenoCams to track crop phenology and explain the effects of different cropping systems on yield

CONTEXT: Crop phenology integrates information of how environmental drivers and management practices affect plant performance and crop yield. However, little is known about the impact of cropping systems (CS) on crop phenology and how this relates to differences in yield. OBJECTIVES: We assessed the applicability of PhenoCams to track crop phenology, how four CS, i.e., organic vs. conventional farming with either intensive or conservation (no/reduced) tillage affect the phenology of a pea-barley mixture and winter wheat, how crop phenology is related to harvest characteristics, e.g., grain yield and total N uptake, and explains CS effects on these characteristics. METHODS: We used time-lapse cameras (PhenoCams) to track vegetation changes in the two crops and extracted the green chromatic coordinate (GCC) to estimate different phenological metrics, i.e., dates with major changes in GCC (PhenoTimePoints), the duration between those (PhenoPhases), and the rate of increasing or decreasing GCC (PhenoSlopes). We assessed how phenological metrics were affected by different CS, and related pheno-logical metrics to harvest characteristics. RESULTS AND CONCLUSIONS: CS significantly affected phenological metrics of both crops, with less pronounced effects in the unfertilized pea-barley mixture compared to the fertilized winter wheat, and stronger effects for early-season than for late-season PhenoTimePoints. For winter wheat, organic compared to conventional farming caused an initial growth lag (up to 7 days) and a shorter duration (approximately 10 days) of the period of stable GCC. Winter wheat in reduced/no-tillage systems showed a tendency of delayed phenology (up to 5 days) compared to intensive tillage. While phenological metrics explained harvest characteristics of winter wheat well, they were almost unrelated to those of pea-barley, most likely because pea-barley yields were similar among CS. For winter wheat, effects of CS on harvest characteristics could be well explained by phenological metrics (max. R2 = 0.9). Thus, we demonstrated that delayed phenology acted as an important factor causing lower yield in organic compared to conventional farming. SIGNIFICANCE: PhenoCams are valuable tool for high-resolution temporal monitoring of crop phenology. As different CS have been proposed as a tool for climate change adaptation, we suggest that the effects of CS on crop phenology need to be considered as they may impact yield via changes in crop phenology, particularly in organic agriculture.