Accurate predictions of crop yield are critical for developing effective agricultural and food policies at the regional and global scales. We evaluated a machine-learning method, Random Forests (RF), for its ability to predict crop yield responses to climate and biophysical variables at global and regional scales in wheat, maize, and potato in comparison with multiple linear regressions (MLR) serving as a benchmark. We used crop yield data from various sources and regions for model training and testing: 1) gridded global wheat grain yield, 2) maize grain yield from US counties over thirty years, and 3) potato tuber and maize silage yield from the northeastern seaboard region. RF was found highly capable of predicting crop yields and outperformed MLR benchmarks in all performance statistics that were compared. For example, the root mean square errors (RMSE) ranged between 6 and 14% of the average observed yield with RF models in all test cases whereas these values ranged from 14% to 49% for MLR models. Our results show that RF is an effective and versatile machine-learning method for crop yield predictions at regional and global scales for its high accuracy and precision, ease of use, and utility in data analysis. RF may result in a loss of accuracy when predicting the extreme ends or responses beyond the boundaries of the training data.

Characterization of leaf physiology is an important step for understanding the ecophysiology of a crop as well as for developing a process-based crop simulation model. We determined photosynthetic and transpiration responses to photosynthetic photon flux (PPF), carbon dioxide concentrations, and temperature, and parameterized a coupled leaf gas-exchange model for hardneck garlic (Allium sativum). The parameterized model performed with high accuracy and precision in predicting photosynthetic responses [r2 = 0.95, bias = 1.7 μmol·m−2·s−1, root mean square error (RMSE) = 2.4 μmol·m−2·s−1] when tested against independent data that were not used for model calibration. The model performance for transpiration rates was less satisfactory (r2 = 0.49, bias = –0.14 mmol·m−2·s−1, RMSE = 0.94 mmol·m−2·s−1). In addition, we characterized the relationships among chlorophyll meter readings, leaf photosynthetic capacity (Amax), and leaf nitrogen content in garlic leaves. The chlorophyll meter readings were a reasonable indicator of both Amax (r2 = 0.61) and leaf nitrogen (N) status (r2 = 0.51) for garlic leaves we studied. The garlic leaf gas-exchange model developed in this study can serve as a key component in ecophysiological crop models for garlic. Similarly, the quantitative relationship identified between chlorophyll meter readings and Amax in this study can provide useful information for non-destructively assessing leaf photosynthetic capacity in garlic.

Garlic (Allium sativum) is a commercially and culturally important crop worldwide. Despite the importance of garlic, there have been few studies investigating how garlic growth and development will be affected by the atmospheric enrichment of carbon dioxide (CO2). A split-plot experiment with CO2 concentrations as main plot and nitrogen (N) fertilization as subplot was carried out to examine the effects of elevated CO2 at (mean ± sd) 745 ± 63 µmol·mol−1 across three levels of N: high-N (16.0 mm), mid-N (4.0 mm), and low-N (1.0 mm). Three hypotheses were tested: 1) garlic plants will allocate proportionally more biomass to bulb when grown in elevated CO2 compared with the plants grown in ambient CO2; 2) plants will sustain improved photosynthesis without downregulation in elevated CO2, irrespective of N; and 3) elevated CO2 will improve plant water use efficiency (WUE) across N fertilization levels. We found that proportional biomass allocation to bulb was not significantly enhanced by CO2 enrichment in garlic. Overall biomass accumulation represented by leaf, stem, and bulb did not respond significantly to CO2 enrichment but responded strongly to N treatments (P < 0.001). Contrary to our hypothesis, photosynthetic downregulation was apparent for garlic plants grown in elevated CO2 with a decrease in Rubisco capacity (P < 0.01). Instantaneous leaf WUE improved in response to elevated CO2 (P < 0.001) and also with increasing N fertilization (P < 0.001). Finally, our results indicate that bulbing ratio is likely to remain unchanged with CO2 or N levels and may continue to serve as a useful nondesctructive metric to estimate harvest timing and bulb size.