Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: March 12, 2009
Publication Date: March 12, 2009
Citation: Jaradat, A.A. 2009. Metrics of Climate Change Impact and Adaptation of Physiologically-diverse Crops [abstract]. IOP Conference Series: Earth and Environmental Science. Paper No. 372021. Technical Abstract: Climate change is predicted to be most pronounced at high latitudes and is expected to cause a net loss of agro-ecosystem carbon (C) and a positive feedback to global warming. Nevertheless, demand for food, feed, fiber and bioenergy continues to be a major driver of global climate change, especially in the cropping systems of the upper Midwest (UMW) of the USA. Precipitation, temperature and length of the growing season (measured in growing degree days (GDD)) are expected to impact agriculture in these high-production cropping systems in which average farm yields ~70% of yield potential presently. Significant scientific advances, critical to understand and quantify response indicators of crop plants to climate and management stresses, are crucial in formulating regulatory, technological, and policy changes needed to successfully mitigate the contribution of agriculture to climate change and to adapt agriculture to climate change. The objectives of a series of long-term studies were to quantify the impact of climate change on (1) plant architecture, seed characteristics and yield interrelationships; and on (2) dry matter (DM) partitioning and the C:Nitrogen:Phosphorus (C:N:P) ratio in the mature seed and its impact on yield. Additionally, measured and simulated data were used to (3) predict the climate-adjusted genetic yield potential ceiling for carbohydrates-, protein-, and oil-producing crops; and to (4) develop a model-based risk assessment of crop production under the most likely climate change scenarios by 2058. Multi-year fractional factorial experiments were conducted to study the impact of multiple climatic variables and management factors on corn (Zea mays L.); sweet sorghum [Sorghum bicolor (L.) Moench]; soybean [Glycine max (L.) Merr.]; wheat (Triticum aestivum L. and Triticum durum L.); chickpea (Cicer arietinum L.); safflower (Cartahmus tinctorius L.); and Cuphea spp. (Cuphea viscosissima x C. lanceolata), a semi-domesticated oil-producing bioenergy crop in the UMW (45º 41' N, 95º 48' W, elevation 370 m). Multi-year data were collected from greenhouse experiments and from permanent geo-referenced sampling sites established within long-term field experiments on soil (physical and chemical properties), and plant variables (phenotypic plasticity, allocation of DM, C, N, and P to roots, stems, leaves, and seed of two contrasting genotypes of each crop species). Current and simulated weather variables [temperature, solar irradiance, rainfall, and actual (ET) and potential (PET) evapo-transpiration] were used in conjunction with current (380 ppm; 2008) and projected (~800 ppm; 2058) levels of carbon dioxide in simulating crop response to likely climate change scenarios. Yield potential under the relatively short growing season in the UMW was determined for each crop in the field by the amount of incident solar radiation, temperature, and population density. Dry matter, C, and N partitioned to plant organs and to the mature seed under no-stress were estimated under controlled conditions in growth chambers and used as a baseline for comparison with their respective estimates under stress conditions and simulation results. Results of empirical analyses [fractal dimension (Do) of plant architecture, partial least squares (PLS) regression, and restricted maximum likelihood (REML) for variance components estimation] and modeling [feed-forward, back-propagation Artificial Neural Networks (ANN)] studies indicated large inter- and intra-species contrasts in response to single and multiple, interacting stress factors and their impact on seed characteristics (e.g., weight and C:N:P ratio) and grain yield. The variance in DM, C, N, and P partitioned among and within species, and among organs within species (VAR) indicated that crops differed as to their adaptive capacity to likely climate change scenarios, and that VAR plays a critical role in determining the magnitude of response to climate change and management options. The Do estimates of monocots (corn<sorghum <wheat) were significantly less impacted by stress factors than those of dicots (chickpea< soybean<safflower<Cuphea spp.). Variation in Do accounted for a larger portion of variance in grain yield of dicots (R2=0.87) than monocots (R2=0.73) and among dicots of the more- (soybean and Cuphea sp.; R2=0.92) than the less-indeterminate (chickpea and safflower; R2=0.85) species. The largest portion of variation in DM, C, N, and P partitioned to the seed were generally accounted for by species, followed, in decreasing order, by GDD, temperature, population density, and their 2- and 3-way interactions with years. Climate factors and management options that would maximize seed weight and seed-sequestered C, N, and P in the carbohydrates-, protein-, and oil-producing species were species-specific with temperature and population density being common to all species. Seed weight and seed C:N:P ratio were strongly correlated, and were the least impacted by stress factors. However, they exhibited significant and variable, moderate to strong correlation with grain yield, especially under stress. Management options that create the most appropriate micro-climate and promote optimum expression of plant architecture (i.e., large Do) were more important for DM accumulation in the early-planted (e.g., wheat) than the late-planted (e.g., soybean and Cuphea spp.) crops. The gap between actual and climate-adjusted genetic yield potential ceiling was the smallest for the carbohydrates-producing crops (corn and sorghum), and steadily increased for the carbohydrates and protein- (wheat and chickpea), protein and oil- (soybean), and oil-producing crops (safflower and Cuphea spp.). Feed-forward, back-propagation ANNs approximated the nonlinear yield function relating crop yield to stress factors. The optimum combination of input factors was crop-specific and allowed ANNs to predict yields 50-65% larger than the maximum measured yield used in the training phases of the ANNs. Trained ANNs predicted crop yields with large R2 (0.70-0.86) values; when subjected to sensitivity, analyses indicated that mid-July rainfall followed in decreasing order by ET/PET, N, GDD, mid-August rainfall, and planting density were the most important factors impacting the accuracy of yield predictions. Risk analysis of possible impacts of climate change on crop production was based on a conditional probability of not exceeding the critical yield threshold for each crop in at least 50% of years using 100 climate change scenarios for all crops. Very low (<25), low (26-50), medium (51-75) and large (>75%) conditional probabilities were encountered 52, 23, 15, and 10% of the time. Wheat, corn, and soybean had the smallest; safflower and sorghum had intermediate; whereas, chickpea and Cuphea spp. had the largest conditional probabilities. A composite index, based on a normalized matrix of impact and adaptation metrics under the most likely climate change scenario in the study area was developed and validated for single crops and crop sequences and is being used to predict effectiveness of adaptation in a dynamic spatio-temporal context.