Responses of gross primary production of grasslands and croplands under drought, pluvial, and irrigation conditions during 2010-2016, Oklahoma, USA

To accurately estimate the terrestrial carbon cycle and food production, it is essential to understand how gross primary production (GPP) of irrigated and non-irrigated grasslands and croplands respond to drought and pluvial events. This study analyzed annual GPP of irrigation-permitted and non-permitted grasslands, winter wheat(Triticwn aesttwan L), other C3 croplands, and C4 croplands in Caddo County of western Oklahoma from 2010 through 2016, a period which consisted of extreme drought (2011) and pluvial events (2015). First, we compared GPP from the Vegetation Photosynthesis Model (GPP(VPM)) and GPP data from the Moderate Resolution Imaging Spectroradiometer (GPP(MOD17)) with GPP estimates from three eddy covariance towers (GPP(EC)) in Oklahoma. GPP(VPM) more accurately estimated mean daily GPP at each of the three sites than GPP(MOD17). Second, we analyzed the seasonal and interannual dynamics of GPP(VPM) for eight pixels, one each for the four irrigation-permitted and non-permitted land types. The interannual variation of GPP(VPM) was due to the complexity of decision making and practice for irrigation, cropping intensity, and crop types. Finally, at the county scale, annual GPP(VPM) from the 2011 drought and pluvial 2015 were compared with mean annual GPP(VPM) from the other 5 years of the study period. The results show that for the 2011 drought: 1) non-permitted C4 croplands had the largest percentage decrease in GPP, but permitted C4 croplands had the smallest decrease; 2) regardless of water rights, GPP was significantly lower than the 5-year reference mean for grasslands, winter wheat, and other C3 crops; and 3) non-permitted lands were more affected by drought than irrigation-permitted lands, except for grasslands, which had similar percentage reductions in GPP. Results for the pluvial year 2015 show that: 1) GPP was significantly higher for grasslands, winter wheat, and non permitted C3 croplands than the 5-year reference mean, but there was no significant difference in GPP for irrigation permitted C3 croplands or non-permitted C4 croplands; and 2) GPP for C4 irrigation-permitted croplands was lower than the 5-year reference mean. Crop-specific responses to drought and pluvial events largely depend on a landowner's ability to irrigate, and caution should be used when assessing or generalizing how crops respond to climate variability, drought, and pluvial conditions in the absence of irrigation-related data.