Potential consequences of climate change on crop production can be studied using mechanistic crop simulation models. While a broad variety of maize simulation models exist, it is not known whether different models diverge on grain yield responses to changes in climatic factors, or whether they agree in their general trends related to phenology, growth, and yield. With the goal of analyzing the sensitivity of simulated yields to changes in temperature and atmospheric carbon dioxide concentrations [CO2 ], we present the largest maize crop model intercomparison to date, including 23 different models. These models were evaluated for four locations representing a wide range of maize production conditions in the world: Lusignan (France), Ames (USA), Rio Verde (Brazil) and Morogoro (Tanzania). While individual models differed considerably in absolute yield simulation at the four sites, an ensemble of a minimum number of models was able to simulate absolute yields accurately at the four sites even with low data for calibration, thus suggesting that using an ensemble of models has merit. Temperature increase had strong negative influence on modeled yield response of roughly -0.5 Mg ha(-1) per °C. Doubling [CO2 ] from 360 to 720 μmol mol(-1) increased grain yield by 7.5% on average across models and the sites. That would therefore make temperature the main factor altering maize yields at the end of this century. Furthermore, there was a large uncertainty in the yield response to [CO2 ] among models. Model responses to temperature and [CO2 ] did not differ whether models were simulated with low calibration information or, simulated with high level of calibration information.

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.

No abstract

No abstract

long been routinely used in soil mapping (Northcote, 1954). Geomorphometry was proposed as a data source Digital elevation models were proposed and used as a data source to predict soil properties (Moore et al., 1993; McKenzie to estimate soil properties. This study evaluated variability of texture and water retention of soils for a gently sloping 3.7-ha field located and Austin, 1993; McSweeney et al., 1994). in the long-term precision farming research site at the Beltsville Ag-Two basic approaches to relate soil properties to landricultural Research Center, MD. The specific objectives of this rescape position have been suggested to date. The first is search were (i) to characterize variability of water retention across based on separating hillslopes into distinct sections, i.e., the hillslope, and (ii) determine and describe any correlations of soil summit, upper and lower interfluve, shoulder, backwater retention with soil texture and surface topography. Soil was slope, upper and lower linear, footslope, toeslope, etc. sampled along four 30-m transects and in 39 points within the study It has been shown that soil properties within a section area. Textural fraction contents, bulk density, and water retention at vary much less than between sections, so that distinct 0, 2.5, 5.0, 10, 33, 100, 500, and 1500 kPa were measured in samples values of soil properties can be assigned to each section taken from 4-to 10-cm depth. A 30-m digital elevation model (DEM) (Ovalles and Collins, 1986). Section-specific regression was constructed from aerial photography data. Slopes, profile curvatures, and tangential curvatures were computed in grid nodes and equations can also be developed to correlate soil properinterpolated to the sampling locations. Regressions with spatially cor-ties (Brubaker et al., 1994). related errors were used to relate water retention and texture to The second approach to relate soil properties to landcomputed topographic variables. Sand, silt, and clay contents descape positions is to use topographic variables, or terrain pended on slope and curvatures. Soil water retention at 10 and 33 attributes, i.e., mathematical characteristics of the land kPa correlated with sand and silt contents. The regression model surface shape, such as slope, profile, plan and tangential relating water retention to the topographic variables explained more curvatures, and aspect (Evans, 1980; Mitá sova and Hothan 60% of variation in soil water content at 10 and 33 kPa, and only fierka, 1993; Shary, 1995). These variables can be com-20% of variation at 100 kPa. Increases in slope values and decreases in puted directly at the nodes of a grid and used for statistitangential curvature values, i.e., less concavity or more convexity cal correlation with soil properties at these nodes across the slope, led to the decreases in water retention at 10 and 33

The effects of CO 2 enrichment on the growth and physiology of maize were investigated at the molecular, biochemical, leaf, and canopy levels. Maize plants were grown in sunlit soil-plant-atmosphere research (SPAR) chambers at ambient (370 lmol mol À1 ) or elevated (750 lmol mol À1 ) atmospheric carbon dioxide concentration (C a ) under wellwatered and fertilized conditions. Canopy gas exchange rates and leaf temperatures were monitored continuously during the growing season. CO 2 enrichment did not enhance the growth or canopy photosynthesis of maize plants. However, canopy evapotranspiration rates decreased by 22% and daytime leaf temperatures were increased about 1 1C in response to CO 2 enrichment. Leaf carboxylation efficiency and leaf nitrogen concentration also decreased at elevated C a . Transcription profiling using maize cDNA microarrays revealed that approximately 5% of tested genes responded to CO 2 enrichment. Of the altered transcripts, several were known to encode proteins involved in stomatal development or photosynthesis. For the majority of the altered transcripts, however, it was difficult to link their functions with specific physiological factors partly because many of these genes encoded unknown proteins. We conclude that maize did not exhibit enhanced growth or photosynthesis in response to CO 2 enrichment but a number of molecular and physiological processes including those involved in stomatal relations were affected by growth in elevated C a .

Knowledge of temperature effects on whole canopy photosynthesis, growth, and development of potato (Solanum tuberosum L.) is important for crop model development and evaluation. The objective of this study was to quantify the effects of temperature on canopy photosynthesis, development, growth, and partitioning of potato cv. Atlantic under elevated atmospheric CO 2 concentration (700 mL L 21 CO 2 ). Potato plants were grown in day-lit plant growth chambers at six constant day/night temperatures, (12, 16, 20, 24, 28, and 32°C) during a 52-d experimental period in 1999 in Beltsville, MD. Main stem length and main stem expanded leaf number were measured nondestructively at 4 d intervals while leaf, stem, root, and tuber weights were obtained by destructive harvesting at biweekly time intervals. Canopy level net photosynthesis (P N ) was obtained from gas exchange measurements. The optimum temperature for canopy photosynthesis was 24°C early in the growth period and shifted to lower temperatures as the plants aged. Total end-of-season biomass was highest in the 20°C treatment. End-of-season tuber mass and the ratio of tuber to total biomass decreased with increasing temperature above 24°C. Accumulated biomass was a linear function of total C gain with a common slope for all treatments. However, the proportion of C allocated to tubers decreased with increasing temperatures. High respiration losses decreased total C gain at higher temperatures. When simulating photosynthesis and C assimilation in crop models, source-sink relationships with temperature and photosynthesis need to be accounted for.

Accurate prediction of phenological development in maize (Zea mays L.) is fundamental to determining crop adaptation and yield potential. A number of thermal functions are used in crop models, but their relative precision in predicting maize development has not been quantified. The objectives of this study were (i) to evaluate the precision of eight thermal functions, (ii) to assess the effects of source data on the ability to differentiate among thermal functions, and (iii) to attribute the precision of thermal functions to their response across various temperature ranges. Data sets used in this study represent >1000 distinct maize hybrids, >50 geographic locations, and multiple planting dates and years. Thermal functions and calendar days were evaluated and grouped based on their temperature response and derivation as empirical linear, empirical nonlinear, and process-based functions. Precision in predicting phase durations from planting to anthesis or silking and from silking to physiological maturity was evaluated. Large data sets enabled increased differentiation of thermal functions, even when smaller data sets contained orthogonal, multi-location and -year data. At the highest level of differentiation, precision of thermal functions was in the order calendar days < empirical linear < process based < empirical nonlinear. Precision was associated with relatively low temperature sensitivity across the 10 to 26°C range. In contrast to other thermal functions, process-based functions were derived using supra-optimal temperatures, and consequently, they may better represent the developmental response of maize to supra-optimal temperatures. Supra-optimal temperatures could be more prevalent under future climate-change scenarios, but data sets in this study contained few data in that range.