Search

Nutrient management, as described in NRCS Code 590, has been intensively investigated, with research largely focused on crop yields and water quality. Yet, due to complex processes and mechanisms in nutrient cycling (especially the nitrogen (N) cycle), there are many challenges in evaluating the effectiveness of nutrient management practices across site conditions. We therefore synthesized data from peer-reviewed publications on subsurface-drained agricultural fields in the Midwest U.S. with corn yield and drainage nitrate-N (NO3-N) export data published from 1980 to 2019. Through literature screening and data extraction from 43 publications, we obtained 577 site-years of data with detailed information on fertilization, corn yields, precipitation, drainage volume, and drainage NO3-N load/concentration or both. In addition, we estimated flow-weighted NO3-N concentrations ([NO3-N]) in drainage for those site-years where only load and volume were reported. Furthermore, we conducted a cost analysis using synthesized and surveyed corn yield data to evaluate the cost-effectiveness of different nutrient management plans. Results from the synthesis showed that N fertilizer rate was strongly positively correlated with corn yields, NO3-N loads, and flow-weighted [NO3-N]. Reducing N fertilizer rates can effectively mitigate NO3-N losses from agricultural fields; however, our cost analysis showed negative economic returns for continuous corn production at lower N rates. In addition, organic fertilizers significantly boosted corn yields and NO3-N losses compared to inorganic fertilizers at comparable rates; however, accurate quantification of plant-available N in organic fertilizers is necessary to guide appropriate nutrient management plans because the nutrient content may be highly variable. In terms of fertilizer application methods, we did not find significant differences in NO3-N export in drainage discharge. Lastly, impact of fertilization timing on NO3-N export varied depending on other factors such as fertilizer rate, source, and weather. According to these results, we suggest that further efforts are still required to produce effective local nutrient management plans. Furthermore, government agencies such as USDA-NRCS need to work with other agencies such as USEPA to address the potential economic losses due to implementation of lower fertilizer rates for water quality improvement.

The emergence of new technologies to synthesize and analyze big data with high-performance computing has increased our capacity to more accurately predict crop yields. Recent research has shown that machine learning (ML) can provide reasonable predictions faster and with higher flexibility compared to simulation crop modeling. However, a single machine learning model can be outperformed by a "committee" of models (machine learning ensembles) that can reduce prediction bias, variance, or both and is able to better capture the underlying distribution of the data. Yet, there are many aspects to be investigated with regard to prediction accuracy, time of the prediction, and scale. The earlier the prediction during the growing season the better, but this has not been thoroughly investigated as previous studies considered all data available to predict yields. This paper provides a machine leaning based framework to forecast corn yields in three US Corn Belt states (Illinois, Indiana, and Iowa) considering complete and partial in-season weather knowledge. Several ensemble models are designed using blocked sequential procedure to generate out-of-bag predictions. The forecasts are made in county-level scale and aggregated for agricultural district and state level scales. Results show that the proposed optimized weighted ensemble and the average ensemble are the most precise models with RRMSE of 9.5%. Stacked LASSO makes the least biased predictions (MBE of 53 kg/ha), while other ensemble models also outperformed the base learners in terms of bias. On the contrary, although random k-fold cross-validation is replaced by blocked sequential procedure, it is shown that stacked ensembles perform not as good as weighted ensemble models for time series data sets as they require the data to be non-IID to perform favorably. Comparing our proposed model forecasts with the literature demonstrates the acceptable performance of forecasts made by our proposed ensemble model. Results from the scenario of having partial in-season weather knowledge reveals that decent yield forecasts with RRMSE of 9.2% can be made as early as June 1 st . Moreover, it was shown that the proposed model performed better than individual models and benchmark ensembles at agricultural district and state-level scales as well as county-level scale. To find the marginal effect of each input feature on the forecasts made by the proposed ensemble model, a methodology is suggested that is the basis for finding feature importance for the ensemble model. The findings suggest that weather features corresponding to weather in weeks 18-24 (May 1 st to June 1 st ) are the most important input features., (Copyright © 2020 Shahhosseini, Hu and Archontoulis.)

Temperature is known to be correlated with crop yields, causing reduction of crop yield with climate warming without adaptations or CO 2 fertilization effects. The historical temperature-crop yield relation has often been used for informing future changes. This relationship, however, may change over time following alternations in other environmental factors. Results show that the strength of the relationship between the interannual variability of growing season temperature and corn yield (R GST_CY ) has declined in the United States between 1980 and 2010 with a loss in the statistical significance. The regression slope which represents the anomalies in corn yield that occur in association with 1 degree temperature anomaly has decreased significantly from -6.9%/K of the first half period to -2.4%/K--3.5%/K of the second half period. This implies that projected corn yield reduction will be overestimated by a fact of 2 in a given warming scenario, if the corn-temperature relation is derived from the earlier historical period. Changes in R GST_CY are mainly observed in Midwest Corn Belt and central High Plains, but are partly reproduced by 11 process-based crop models. In Midwest rain-fed systems, the decrease of negative temperature effects coincides with an increase in water availability by precipitation. In irrigated areas where water stress is minimized, the decline of beneficial temperature effects is significantly related to the increase in extreme hot days. The results indicate that an extrapolation of historical yield response to temperature may bias the assessment of agriculture vulnerability to climate change. Efforts to reduce climate impacts on agriculture should pay attention not only to climate change, but also to changes in climate-crop yield relations. There are some caveats that should be acknowledged as the analysis is restricted to the changes in the linear relation between growing season mean temperature and corn yield for the specific study period., (Copyright © 2017 Elsevier B.V. All rights reserved.)