Article

Climatic and genetic controls of yields of switchgrass, a model bioenergy species

If you want to read the PDF, try requesting it from the authors.

Abstract

The U.S. Renewable Fuel Standard calls for 136 billion liters of renewable fuels production by 2022. Switchgrass (Panicum virgatum L.) has emerged as a leading candidate to be developed as a bioenergy feedstock. To reach biofuel production goals in a sustainable manner, more information is needed to characterize potential production rates of switchgrass. We used switchgrass yield data and general additive models (GAMs) to model lowland and upland switchgrass yield as nonlinear functions of climate and environmental variables. We used the GAMs and a 39-year climate dataset to assess the spatio-temporal variability in switchgrass yield due to climate variables alone. Variables associated with fertilizer application, genetics, precipitation, and management practices were the most important for explaining variability in switchgrass yield. The relationship of switchgrass yield with climate variables was different for upland than lowland cultivars. The spatio-temporal analysis showed that considerable variability in switchgrass yields can occur due to climate variables alone. The highest switchgrass yields with the lowest variability occurred primarily in the Corn Belt region, suggesting that prime cropland regions are the best suited for a constant and high switchgrass biomass yield. Given that much lignocellulosic feedstock production will likely occur in regions with less suitable climates for agriculture, interannual variability in yields should be expected and incorporated into operational planning.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Statistical/empirical models, particularly, have been of great use in the history of science. Their easiness of computing and usability enhanced their attractiveness among decision-makers and practitioners (Razi and Athappilly, 2005), while they allow highlighting relative importance of variables when much is uncertain (Lobell et al., 2005;Tittonell et al., 2008;Tulbure et al., 2012). Statistical models could be divided into two main subgroups: parametric and non-parametric. ...
... Wullschleger et al. (2010) used non-parametric models to establish equations of parametric ones for switchgrass yield prediction. Other non-parametric models (regression trees, Breiman et al., 1984) were utilized to analyze yield variability in maize (Tittonell et al., 2008), wheat (Lobell et al., 2005), soybean (Zheng et al., 2009), sugarcane (Ferraro et al., 2009) or switchgrass (Wullschleger et al., 2010;Tulbure et al., 2012). ...
... With our best minimum adequate model, we were able to explain up to 70% of sunflower oil concentration variability. The general performances of our models can be considered as satisfactory when compared to other existing statistical models involving a wide range of varieties/cropping conditions (R 2 : 45-61% for GAM in Tulbure et al., 2012; 51-56% for regression tree in Ferraro et al., 2009; 36-43% for multiple linear regression in Khamis et al. (2006)). The remaining unknown 30% might be linked to several causes. ...
Article
Sunflower (Helianthus annuus L.) raises as a competitive oilseed crop in the current environmentally friendly context. To help targeting adequate management strategies, we explored statistical models as tools to understand and predict sunflower oil concentration. A trials database was built upon experiments carried out on a total of 61 varieties over the 2000–2011 period, grown in different locations in France under contrasting management conditions (nitrogen fertilization, water regime, plant density). 25 literature-based predictors of seed oil concentration were used to build 3 statistical models (multiple linear regression, generalized additive model (GAM), regression tree (RT)) and compared to the reference simple one of Pereyra-Irujo and Aguirrezábal (2007) based on 3 variables. Performance of models was assessed by means of statistical indicators, including root mean squared error of prediction (RMSEP) and model efficiency (EF). GAM-based model performed best (RMSEP = 1.95%; EF = 0.71) while the simple model led to poor results in our database (RMSEP = 3.33%; EF = 0.09). We computed hierarchical contribution of predictors in each model by means of R2 and concluded to the leading determination of potential oil concentration (OC), followed by post-flowering canopy functioning indicators (LAD2 and MRUE2), plant nitrogen and water status and high temperatures effect. Diagnosis of error in the 4 statistical models and their domains of applicability are discussed. An improved statistical model (GAM-based) was proposed for sunflower oil prediction on a large panel of genotypes grown in contrasting environments.
... It is a C 4 grass that propagates through seeds and rhizomes. In Ontario, different cultivars of switchgrass are being researched in multiple locations across the province, and some cultivars are showing promising production numbers (Tubeileh et al., 2012). In addition to switchgrass, other species such as big bluestem, prairie sandreed [Calamovilfa longifolia (Hook.) ...
... Environment and management factors can alter production levels and quality for biomass use. Climate can have significant effects on the variability of switchgrass yields along with genetics, fertilizer rates, and management practices (Tulbure et al., 2012). The application of fertilizer is an important parameter in biomass production systems, as it has the potential to increase yield and can affect the CO 2 emissions, energy inputs, economics, nutrient content, and quality of the biomass. ...
... The switchgrass production levels observed are within the same range as those reported in literature from Ontario, Quebec, and upper U.S. Midwest. In northern Ontario, 3-yr old Sunburst and Cave-in-Rock switchgrass produced, respectively, 10.1 and 12.4 Mg ha -1 at 0 kg N ha -1 and 11.9 and 13.2 Mg ha -1 at 50 kg N ha -1 , with no significant differences between the two N levels (Tubeileh et al., 2012). At 65 kg N ha -1 , Rountree big bluestem stands planted in 2003 produced ~14 and 11 Mg ha -1 in 2006 and 2007, respectively (Heggenstaller et al., 2009). ...
Article
Full-text available
There is little information on the production levels of prairie C4 perennial grasses in the biomass context in eastern Canada. The objective of this study is to evaluate the effect of species/cultivar entries and N fertilizer application on biomass production of prairie perennial biomass grasses. Two switchgrass (Panicum virgatum L.) cultivars, two big bluestem (Andropogon gerardii Vit.) cultivars and one indiangrass (Sorghastrum nutans Nash.) cultivar were seeded in 2009 in eastern Ontario to evaluate their biomass production potential. The crops were seeded in a silt loam soil at 1000 pure live seeds m–2. No fertilizer was added in 2009 or 2010. In 2011 and 2012 treatment N50 received 50 kg N ha-1 while treatment N0 was kept as an unfertilized control. The crop biomass production, canopy height, leaf SPAD absorbance, leaf area index, and harvest moisture content were determined. Switchgrass production exceeded big bluestem and indiangrass, especially in the first 2 yr after establishment. Despite the drought, dry matter production for the different entries in 2012 ranged between 5.07 and 8.13 Mg ha-1. Biomass production was improved by N application only for big bluestem and indiangrass but not for switchgrass. However, N application increased leaf SPAD absorbance and leaf area index for all entries. Moisture content at harvest was species dependent. Our results suggest that switchgrass might need lower N levels than big bluestem and indiangrass. With the stands reaching their peak production, this study shows the high production potential for some of these grasses under Ontario conditions.
... Understanding how switchgrass yields will respond to projected future climate change is of fundamental importance for planning, implementation, and operational management of future biofuel production systems. A recent study by Tulbure et al (2012) showed that switchgrass yields are sensitive to climatic variability in space and time, with different cytotypes exhibiting different responses. Utilizing switchgrass yield data in bioclimatic envelope models provides improved insights into how future changes in climate will affect switchgrass productivity compared to models based on presence-absence data. ...
... Switchgrass yield data collected from 1167 observations at 45 field trials across the native range of switchgrass were used. Ten variables previously shown to influence switchgrass yields were used as explanatory variables (Tulbure et al 2012). Explanatory variables included early and late growing season precipitation, average growing season temperature, cultivar, germplasm origin, month of harvest, soil texture (per cent sand and clay), stand age, and amount of nitrogen fertilizer applied. ...
... GAMs allow for non-linear relationships between the dependent and explanatory variables thus revealing structure in the data that might otherwise be missed with linear models (Faraway 2004, Hastie andTibshirani 1990). For a description of the sources of switchgrass yield data, their location and explanatory variables please refer to (Tulbure et al 2012). The models previously developed were refitted using the WorldClim dataset. ...
Article
Full-text available
A climate envelope approach was used to model the response of switchgrass, a model bioenergy species in the United States, to future climate change. The model was built using general additive models (GAMs), and switchgrass yields collected at 45 field trial locations as the response variable. The model incorporated variables previously shown to be the main determinants of switchgrass yield, and utilized current and predicted 1 km climate data from WorldClim. The models were run with current WorldClim data and compared with results of predicted yield obtained using two climate change scenarios across three global change models for three time steps. Results did not predict an increase in maximum switchgrass yield but showed an overall shift in areas of high switchgrass productivity for both cytotypes. For upland cytotypes, the shift in high yields was concentrated in northern and north-eastern areas where there were increases in average growing season temperature, whereas for lowland cultivars the areas where yields were projected to increase were associated with increases in average early growing season precipitation. These results highlight the fact that the influences of climate change on switchgrass yield are spatially heterogeneous and vary depending on cytotype. Knowledge of spatial distribution of suitable areas for switchgrass production under climate change should be incorporated into planning of current and future biofuel production. Understanding how switchgrass yields will be affected by future changes in climate is important for achieving a sustainable biofuels economy.
... It is a C 4 grass that propagates through seeds and rhizomes. In Ontario, different cultivars of switchgrass are being researched in multiple locations across the province, and some cultivars are showing promising production numbers (Tubeileh et al., 2012). In addition to switchgrass, other species such as big bluestem, prairie sandreed [Calamovilfa longifolia (Hook.) ...
... Environment and management factors can alter production levels and quality for biomass use. Climate can have significant effects on the variability of switchgrass yields along with genetics, fertilizer rates, and management practices (Tulbure et al., 2012). The application of fertilizer is an important parameter in biomass production systems, as it has the potential to increase yield and can affect the CO 2 emissions, energy inputs, economics, nutrient content, and quality of the biomass. ...
... The switchgrass production levels observed are within the same range as those reported in literature from Ontario, Quebec, and upper U.S. Midwest. In northern Ontario, 3-yr old Sunburst and Cave-in-Rock switchgrass produced, respectively, 10.1 and 12.4 Mg ha -1 at 0 kg N ha -1 and 11.9 and 13.2 Mg ha -1 at 50 kg N ha -1 , with no significant differences between the two N levels (Tubeileh et al., 2012). At 65 kg N ha -1 , Rountree big bluestem stands planted in 2003 produced ~14 and 11 Mg ha -1 in 2006 and 2007, respectively (Heggenstaller et al., 2009). ...
... However, since second-generation biofuels have become a viable option with growing research interest, numerous field experiments (> 40) have been performed. To leverage the abundance of the data available for switchgrass at different geographic locations and over the last 40 years, meta-analyses have been performed with different purposes (Wullschleger, Davis, Borsuk, Gunderson, & Lynd, 2010, Tulbure, Wimberly, Boe & Owens, 2012a. Meta-analysis has the ability to reach broad generalizations across large number of study outcomes to provide a comprehensive conclusion and direct future primary studies (Gurevitch, Koricheva, Nakagawa, & Stewart, 2018). ...
... A metaanalysis of switchgrass biomass data collected from 39 field experiments with 1,190 observations found that switchgrass biomass was associated with growing season precipitation, annual temperature, N fertilization rates, and ecotype, but not associated with plot size, land quality, soil type and soil acidity . Tulbure, Wimberly, Boe & Owens (2012a) expanded the same data from Wullschleger et al. (2010) to 45 field experiments for 13 cultivars and their spatio-temporal analysis showed that considerable variability in switchgrass biomass can occur due to climate variables alone. Further exploration of these data at the cultivar level will help identify the environmental factors driving cultivar-specific differentiation within switchgrass. ...
... Measurements of switchgrass biomass production were compiled from the literature. Recently published studies supplemented those reported in two extensive meta-analyses performed by Wullschleger et al. (2010) and Tulbure, Wimberly, Boe & Owens (2012a). All together 900 measures of switchgrass biomass associated with 41 field locations and 26 literature references were identified for four switchgrass cultivars ( Figure 1, Table 1). ...
Article
Full-text available
Previous studies have shown that switchgrass has a wide range of genetic variation and that productivity is linked to local adaptation to the location of origin for many cultivars. In this meta‐analysis, we compiled and analyzed 900 observations associated with 41 field trials for four switchgrass cultivars (two lowlands, Alamo and Kanlow, and two uplands, Cave‐In‐Rock and Shelter). This extensive dataset and machine learning was used to identify the most influential variables impacting switchgrass productivity, to search for evidence of local adaptation to each cultivar's location of origin, and to predict change in productivity under future climate for each cultivar. In general, variables associated with climate and management are more important predictors of productivity relative to soil variables. Three climatic variables, annual mean temperature, annual precipitation, and precipitation in the wettest month, are identified as key environmental variables for productivity of all cultivars. Productivity under future climate (2041‐2060) is predicted to stay stable for all cultivars relative to the prediction under current climate (1986‐2005) across all trial locations and over a 20‐yr simulation period. However, the productivity of each cultivar varies from location to location and from year to year, although productivity varies more between locations than between years. Additionally, we observe shifts in the most productive cultivar at the local field scale depending on the combination of management practice and climates. The shape of the relationship between productivity and the annual mean temperature relative to the cultivar's location of origin is bell‐shape curve for Kanlow, Cave‐in‐Rock, and Shelter, indicative of local adaptation. Identifying influential environmental variables, and their relationships to productivity with respect to cultivar's location of origin helps predicting productivity on the local field scale, and will help with the biofuel production planning through the selection of suitable cultivars for different locations under climate changes. This article is protected by copyright. All rights reserved.
... Some data are best described by "wiggly models," and one of the widely-used models is the generalized additive model (GAM, Figure 10) while simple linear models do not fit well. In recent years, GAM [65] found its application in remote sensing for crop sustainability studies [66], plant science [67,68], plant nitrogen [69] and lignin [70], biomass modeling [71], and other areas [72], and has great potential for switchgrass and bioenergy species [69]. ...
... Some data are best described by "wiggly models," and one of the widely-used models is the generalized additive model (GAM, Figure 10) while simple linear models do not fit well. In recent years, GAM [65] found its application in remote sensing for crop sustainability studies [66], plant science [67,68], plant nitrogen [69] and lignin [70], biomass modeling [71], and other areas [72], and has great potential for switchgrass and bioenergy species [69]. ...
Article
Full-text available
Unmanned aerial vehicles (UAVs) provide an intermediate scale of spatial and spectral data collection that yields increased accuracy and consistency in data collection for morphological and physiological traits than satellites and expanded flexibility and high-throughput compared to ground-based data collection. In this study, we used UAV-based remote sensing for automated phenotyping of field-grown switchgrass (Panicum virgatum), a leading bioenergy feedstock. Using vegetation indices calculated from a UAV-based multispectral camera, statistical models were developed for rust disease caused by Puccinia novopanici, leaf chlorophyll, nitrogen, and lignin contents. For the first time, UAV remote sensing technology was used to explore the potentials for multiple traits associated with sustainable production of switchgrass, and one statistical model was developed for each individual trait based on the statistical correlation between vegetation indices and the corresponding trait. Also, for the first time, lignin content was estimated in switchgrass shoots via UAV-based multispectral image analysis and statistical analysis. The UAV-based models were verified by ground-truthing via correlation analysis between the traits measured manually on the ground-based with UAV-based data. The normalized difference red edge (NDRE) vegetation index outperformed the normalized difference vegetation index (NDVI) for rust disease and nitrogen content, while NDVI performed better than NDRE for chlorophyll and lignin content. Overall, linear models were sufficient for rust disease and chlorophyll analysis, but for nitrogen and lignin contents, nonlinear models achieved better results. As the first comprehensive study to model switchgrass sustainability traits from UAV-based remote sensing, these results suggest that this methodology can be utilized for switchgrass high-throughput phenotyping in the field.
... Random forests have been used to predict yields directly for mangoes (Fukuda et al. 2013) and have been incorporated into a complex seasonal yield forecasting model for crops in Canada (Newlands et al. 2014). Tulbure et al. (2012) used random forest regression to identify important variables for switchgrass yields across the USA. They identified nitrogen fertilizer, cultivar, rainfall, stand age, and soil silt levels as the most influential of 22 predictor variables. ...
... In this study, changes in variable importance were identified between different forecast dates while previous studies have shown that random forests can identify differences in important variables between regions (Everingham et al. 2015b). Future research should investigate if random forests can be used to provide yield estimates at a finer resolution rather than one estimate for the entire region akin to what has been accomplished for wheat (Newlands et al. 2014) and switchgrass (Tulbure et al. 2012). This would Fig. 3 Comparison of observed cane yields (black) and yields forecasted using random forest regression models. ...
Article
Foreknowledge about sugarcane crop size can help industry members make more informed decisions. There exists many different combinations of climate variables, seasonal climate prediction indices, and crop model outputs that could prove useful in explaining sugarcane crop size. A data mining method like random forests can cope with generating a prediction model when the search space of predictor variables is large. Research that has investigated the accuracy of random forests to explain annual variation in sugarcane productivity and the suitability of predictor variables generated from crop models coupled with observed climate and seasonal climate prediction indices is limited. Simulated biomass from the APSIM (Agricultural Production Systems sIMulator) sugarcane crop model, seasonal climate prediction indices and observed rainfall, maximum and minimum temperature, and radiation were supplied as inputs to a random forest classifier and a random forest regression model to explain annual variation in regional sugarcane yields at Tully, in northeastern Australia. Prediction models were generated on 1 September in the year before harvest, and then on 1 January and 1 March in the year of harvest, which typically runs from June to November. Our results indicated that in 86.36 % of years, it was possible to determine as early as September in the year before harvest if production would be above the median. This accuracy improved to 95.45 % by January in the year of harvest. The R-squared of the random forest regression model gradually improved from 66.76 to 79.21 % from September in the year before harvest through to March in the same year of harvest. All three sets of variables—(i) simulated biomass indices, (ii) observed climate, and (iii) seasonal climate prediction indices—were typically featured in the models at various stages. Better crop predictions allows farmers to improve their nitrogen management to meet the demands of the new crop, mill managers could better plan the mill’s labor requirements and maintenance scheduling activities, and marketers can more confidently manage the forward sale and storage of the crop. Hence, accurate yield forecasts can improve industry sustainability by delivering better environmental and economic outcomes.
... Two approaches to mapping biomass resources are empirical modeling and mechanistic plant growth modeling. Commonly used empirical approaches have involved statistical extrapolation of plot or field-level yield data to larger regions using climatic envelope methods (e.g., Jager et al., 2010;Wullschleger et al., 2010;Tulbure et al., 2012). The main drawback of empirical approaches has been a lack of suitable yield data (Miguez et al., 2012), and a limited ability to extrapolate beyond the range of the explanatory data (Jager et al., 2010). ...
... The issue of unusually high yields extending into the Southwest is also seen in the switchgrass map produced with a statistical model developed by Wullschleger et al. (2010); relatively high yields were estimated throughout much of Colorado, northwestern Texas and New Mexico, where precipitation is generally inadequate and moisture deficits are likely to restrict production. Tulbure et al. (2012) used a statistical model to map the yield potential for upland and lowland switchgrass in the eastern United States at 1-km resolution. The general patterns of the upland map are similar to those of PRISM-ELM, but the lowland switchgrass map is very different than all others reviewed. ...
Article
Full-text available
Several crops have recently been identified as potential dedicated bioenergy feedstocks for the production of power, fuels, and bioproducts. Despite being identified as early as the 1980s, no systematic work has been undertaken to characterize the spatial distribution of their long-term production potentials in the US. Such information is a starting point for planners and economic modelers, and there is a need for this spatial information to be developed in a consistent manner for a variety of crops, so that their production potentials can be inter-compared to support crop selection decisions. As part of the Sun Grant Regional Feedstock Partnership, an approach to mapping these potential biomass resources was developed to take advantage of the informational synergy realized when bringing together coordinated field trials, close interaction with expert agronomists, and spatial modeling into a single, collaborative effort. A modeling and mapping system called PRISM-ELM was designed to answer a basic question: how do climate and soil characteristics affect the spatial distribution and long term production patterns of a given crop? This empirical/mechanistic/biogeographical hybrid model employs a limiting factor approach, where productivity is determined by the most limiting of the factors addressed in sub-models that simulate water balance, winter low temperature response, summer high temperature response, and soil pH, salinity, and drainage. Yield maps are developed through linear regressions relating soil and climate attributes to reported yield data. The model was parameterized and validated using grain yield data for winter wheat and maize, which served as benchmarks for parameterizing the model for upland and lowland switchgrass, CRP grasses, Miscanthus, biomass sorghum, energycane, willow, and poplar. The resulting maps served as potential production inputs to analyses comparing the viability of biomass crops under various economic scenarios. The modeling and parameterization framework can be expanded to include other biomass crops.
... To analyze such a mixture of variables and the large, unbalanced amount of observation units classic statistical methods might fail to find meaningful agricultural patterns. Thus, to select the most influential variables, the classification and regression tree (CART) and random forest (RF) (Breiman et al., 1984;Tulbure et al., 2012) methods were used. As these methods are relatively new in this research field, this manuscript evaluates the potential for CART and RF to inform agronomic management decisions along with determination if environmental conditions are indeed so important in intensive winter wheat production. ...
... However, RF does not produce one tree as an output, so one CART model is much easier to interpret. The method has also been successfully used in agronomy (Tittonel et al., 2008;Tulbure et al., 2012;Fukuda et al., 2013;Everingham et al., 2016). ...
Article
The main goal of this study was to determine the major drivers of variation for the yield-scaled Global Warming Potential (GWP) and yield of winter wheat in Poland focusing on environmental, genetic and management variables. The yield-scaled GWP is the GWP calculated per grain unit expressed in kg CO2 equivalent kg−1 yield. The analysis was performed using multivariate statistical methods: CART (classification and regression tree) and RF (random forest). This is the first study in Poland to focus on variables besides those used in GWP calculations and those influencing the yield-scaled GWP variability of winter wheat production. Identification of these variables contributes to the creation of more environmentally friendly wheat cropping systems. In this study, regression-tree based analysis revealed that soil quality and water availability during crucial stages of plant growth are the most influential input variables of the yield-scaled GWP and yield variability in winter wheat production. Not surprisingly, environmentally favorable conditions for wheat growth contribute to its high yields yet require less intensified agronomic management. The strong influence of water availability in June and July, at the end of plant growth, applies to the undesirable effect of excess water leading to plant diseases which result in lower yield. N fertilizer has a strong effect on GWP of winter wheat production. However, this study also shows that nitrogen is not one of the most influential variables of wheat yield variability. Thus, increasing its average use in Polish environmental conditions from 107.3 kg ha−1 to 147.3 kg ha−1 might not increase yield sufficiently for its use to be justified. More important variables of yield variability were the use of fungicides and growth regulators, which are applied at much smaller rates (1.7 and 0.8 kg ha−1, respectively) than N fertilizer and positively influence efficient winter wheat production.
... They studied the water quality impacts of biofuel production in the Upper Mississippi River basin but did not take climate variability into consideration. Nevertheless, Brown et al. [25], Dominguez-Faus et al. [12], and Tulbure et al. [26] are amongst the very few who have analyzed the impacts of a changing climate on production of bioenergy crops. However, they do not consider the alterations in the hydrologic cycle driven by a combination of biomass feedstock production and climate change and therefore the results have limited use in assessing environmental impact and sustainability of large scale bio feedstock production systems. ...
... For this study we used the UMRB SWAT model from a previous study by a co-author [26] that was developed using the 8 digit USGS Hydrologic Unit code (HUCs), the National Hydrography Dataset (NHD), and 90 m (3-arc second) DEM data to define watershed and topographic parameters. Landuse data was generated from the Cropland Data Layer (CDL) [33] and the 2001 National Land Cover Data (NLCD 2001) [34] as illustrated in Figure 2. ...
Article
Full-text available
Impact of climate change on the water resources of the United States exposes the vulnerability of feedstock-specific mandated fuel targets to extreme weather conditions that could become more frequent and intensify in the future. Consequently, a sustainable biofuel policy should consider: (a) how climate change would alter both water supply and demand; and (b) in turn, how related changes in water availability will impact the production of biofuel crops; and (c) the environmental implications of large scale biofuel productions. Understanding the role of biofuels in the water cycle is the key to understanding many of the environmental impacts of biofuels. Therefore, the focus of this study is to model the rarely explored interactions between land use, climate change, water resources and the environment in future biofuel production systems. Results from this study will help explore the impacts of the US biofuel policy and climate change on water and agricultural resources. We used the Soil and Water Assessment Tool (SWAT) to analyze the water quantity and quality consequences of land use and land management related changes in cropping conditions (e.g., more use of marginal lands, greater residue harvest, increased yields), plus management practices due to biofuel crops to meet the Renewable Fuel Standard target on water quality and quantity.
... For that reason, robust computational tools coupled with analysis of soil physical properties and crop response have become fundamental for accurate decisions on soil management. Recently, machine learning algorithms have improved predictions of crop yield (Mishra et al. 2016;Chlingaryan et al. 2018;Maya Gopal and Bhargavi 2019) based on predictor variables related to remote sensing (Pantazi et al. 2016;Khanal et al. 2018;Richetti et al. 2018;Filippi et al. 2019), climate (Tulbure et al. 2012;Fukuda et al. 2013;Everingham et al. 2016;Mathieu and Aires 2018) and soil properties (Pantazi et al. 2016;Smidt et al. 2016;Peixoto et al. 2019a). Machine learning algorithms are capable of providing high accuracy, identifying the most important covariates for estimating yield and have the ability to model complex non-linear interactions for diverse statistical analyses (Cutler et al. 2007). ...
... Studies suggest that biomass yields generally decrease with stand age after several years as soil-water is depleted (Jager et al., 2010;Tulbure et al., 2012). However, our model results did not show such yield declines under N or water deficit after a typical life span of 10-15 years if N or water input was continuously maintained at a low level. ...
Article
The potential of switchgrass (Panicum virgatum L.) to offset large-scale greenhouse gas (GHG) emissions depends on optimizing external inputs when the crop is primarily managed as a sustainable source for renewable energy production. Due to the heterogeneity of climate and soil conditions and the complexity of agriculture, an evaluation of the effect of adopting switchgrass as a new biofuel crop into agriculture needs to be done at the regional scale. The objective of the study was to predict long-term (100-yr) GHG emissions under different N fertilization (0, 112, and 224 kg N ha⁻¹) and irrigation application (0, 25, 50, 75, and 99 cm H₂O) levels across the Central Valley of California using the DAYCENT model. Six cultivars (Alamo, Kanlow, Cave-in-Rock, Blackwell, Sunburst, and Trailblazer) were selected. The model results suggest that switchgrass productivity is primarily constrained by N inputs when no or low water stress is expected in a Mediterranean climate. In the short-term (the first decade after establishment), soil organic carbon (SOC) stocks (0–20 cm) increased by 0.42–0.92 Mg C ha⁻¹ yr⁻¹ and N₂O emissions were 1.37–2.48 kg N2O–N ha⁻¹ yr⁻¹ across the cultivars with baseline input rates of 224 kg N ha−1 yr−1 and 99 cm H2O. All cultivars were net CO₂ sinks in the near term and the potential decreased by 0.09–0.30 Mg C ha⁻¹ yr⁻¹ (15.5–52.8%) with reduced N input from baseline under varying irrigation rates. There was a reduction in N₂O emissions by 47.2–61.6% by applying less N fertilizer when irrigated at rates ≥75 cm H₂O per year over time. In general, higher-yielding cultivars (e.g., Alamo) tended to sequester more CO₂ but also led to higher N₂O emissions. In the near term, the use of N fertilizer and irrigation is needed for switchgrass systems to be a soil GHG sink, but for longer-term GHG mitigation strategies reducing both N fertilization and irrigation inputs is required.
... The characteristics which make perennial grasses attractive for biomass production are its high yield potential, the high contents of lignin and cellulose and generally anticipated positive environmental impacts [12]. Also, a number of studies have suggested recently that marginal and abandoned lands could potentially be converted to cellulosic feedstock production which would avoid a large scale conversion of current crop land to biofuel feedstock production [13,14]. ...
Article
Full-text available
Biofuel can be a potential supplement to fossil fuel and help in meeting increasing energy demand of the USA as well as rest of the world. However, it is important for the biofuel to be economically competitive and energy efficient to be established as a promising energy source. There has always been an argument about energy efficiency of biofuel production. Some studies have claimed that it requires more energy to produce ethanol then it actually produces as an output. The objective of the study is to determine energy efficiency of growing two potential cellulosic feedstock; switchgrass and Miscanthus and conversion them in to cellulosic ethanol. Energy efficiency was determined by calculating Net Energy Value (NEV), the difference between output energy obtained by ethanol and input energy used in producing ethanol. Input energy consisted of energy required to produce the cellulosic biomass, transporting it to the ethanol processing facility and processing cellulosic biomass in to ethanol. The analysis showed positive The NEV for both switchgrass and Miscanthus. NEV for Miscanthus (12.41 MJ/l of ethanol) was higher in compared to the switchgrass (7.90 MJ/l of ethanol). Monitory benefits of energy savings were also estimated to determine the impact of energy saving to the society. Results from this research showed both switchgrass and Miscanthus as potential biomass feedstock for cellulosic ethanol production.
... Reduction in anthropogenic impacts of agriculture may be accomplished through the use of perennials for bioenergy (Georgescu et al. 2011) and therefore, by extension, herbaceous perennials as grain, oilseed and potentially as other types of crops. Perennial monocultures for bioenergy are still subject to yield fluctuations due to environmental conditions, despite adequate agronomic practices (Tulbure et al. 2012). Grasslands, nature's polycultures, however are seen as important for carbon sequestration (O'Mara 2012). ...
Conference Paper
Full-text available
Population growth and demand for food continues to place pressure upon agriculture to provide for mankind. Primary staple production is currently from annual crop species. Herbaceous perennial species for grain and other food products have not been rigorously pursued. Current interest and research into the development of herbaceous perennial species for food is providing new options for food production systems. Development of perennial species will provide the cornerstones for perennial polyculture development. Production challenges including weed competition and fertility requirements may addressed by perennial polyculture. Utilization of multiple species has been demonstrated to provide greater dry matter productivity by utilizing the entire growing season. Issues concerning synchronization of production and harvest however may not be easily resolved. Historically selection of perennial grasses species for seed production has most often failed to translate into consistent enhanced productivity at the field scale. Current selection methodology and nursery design are most likely inadequate to address field level productivity issues. Competitive nurseries are suggested to provide field level gains at both the mono-and poly-culture levels. Landscape-wide utilization of narrow genetic cultivars will lead to divergent communities and reduce reliability of production. Initial development and deployment of perennial grains and oilseeds would be enhanced by the utilization of greater diversity within the crop species. Utilization of companion species will aid in many issues related to sustainability, e.g. N 2 fixation, weediness. Initial economic utilization of perennial grains and oilseeds and perennial polyculture are linked to animal production.
... Studies suggest that biomass yields generally decrease with stand age after several years as soil-water is depleted (Jager et al., 2010;Tulbure et al., 2012). However, our model results did not show such yield declines under N or water deficit after a typical life span of 10-15 years if N or water input was continuously maintained at a low level. ...
Article
The potential of switchgrass (Panicum virgatum L.) to offset large-scale greenhouse gas (GHG) emissions depends on optimizing external inputs when the crop is primarily managed as a sustainable source for renewable energy production. Due to the heterogeneity of climate and soil conditions and the complexity of agriculture, an evaluation of the effect of adopting switchgrass as a new biofuel crop into agriculture needs to be done at the regional scale. The objective of the study was to predict long-term (100-yr) GHG emissions under different N fertilization (0, 112, and 224 kg N ha⁻¹) and irrigation application (0, 25, 50, 75, and 99 cm H₂O) levels across the Central Valley of California using the DAYCENT model. Six cultivars (Alamo, Kanlow, Cave-in-Rock, Blackwell, Sunburst, and Trailblazer) were selected. The model results suggest that switchgrass productivity is primarily constrained by N inputs when no or low water stress is expected in a Mediterranean climate. In the short-term (the first decade after establishment), soil organic carbon (SOC) stocks (0–20 cm) increased by 0.42–0.92 Mg C ha⁻¹ yr⁻¹ and N₂O emissions were 1.37–2.48 kg N₂O-N ha⁻¹ yr⁻¹ across the cultivars with baseline input rates of 224 kg N ha⁻¹ yr⁻¹ and 99 cm H₂O. All cultivars were net CO₂ sinks in the near term and the potential decreased by 0.09– 0.30 Mg C ha⁻¹ yr⁻¹ (15.5–52.8%) with reduced N input from baseline under varying irrigation rates. There was a reduction in N₂O emissions by 47.2–61.6% by applying less N fertilizer when irrigated at rates ≥75 cm H₂O per year over time. In general, higher-yielding cultivars (e.g., Alamo) tended to sequester more CO₂ but also led to higher N₂O emissions. In the near term, the use of N fertilizer and irrigation is needed for switchgrass systems to be a soil GHG sink, but for longer-term GHG mitigation strategies reducing both N fertilization and irrigation inputs is required.
... Given our results, there may be a significant gap in our ability to account for how plant community composition influences the response of N 2 O fluxes to environmental drivers. Single-species monocultures have typically been used to model the broader category of herbaceous biomass crops (Surendran Nair et al., 2012), with switchgrass used as a model for exploring the properties and environmental responses of bioenergy crops (Lewandowski et al., 2003;Tulbure et al., 2012). The cropping system specificity we observed in the response of N 2 O fluxes to environmental parameters suggests it may be risky to rely on model systems to predict the behaviors of perennial and polycultural biomass cropping systems, particularly with the potential for high variability during the establishment phase. ...
Article
Full-text available
Greenhouse gas (GHG) emissions from soils are a key sustainability metric of cropping systems. During crop establishment, disruptive land-use change is known to be a critical, but under reported period, for determining GHG emissions. We measured soil N 2 O emissions and potential environmental drivers of these fluxes from a three-year establishment-phase bioenergy cropping systems experiment replicated in southcentral Wisconsin (ARL) and southwestern Michigan (KBS). Cropping systems treatments were annual monocultures (continuous corn, corn–soybean–canola rotation), perennial monocultures (switchgrass, miscanthus, and poplar), and perennial polycultures (native grass mixture, early successional community, and restored prairie) all grown using best management practices specific to the system. Cumulative three-year N 2 O emissions from annuals were 142% higher than from perennials, with fertilized perennials 190% higher than unfertilized perennials. Emissions ranged from 3.1 to 19.1 kg N 2 ON ha À1 yr À1 for the annuals with continuous corn > corn–soybean–canola rotation and 1.1 to 6.3 kg N 2 ON ha À1 yr À1 for perennials. Nitrous oxide peak fluxes typically were associated with precipitation events that closely followed fertilization. Bayesian modeling of N 2 O fluxes based on measured environmental factors explained 33% of variability across all systems. Models trained on single systems performed well in most monocultures (e.g., R 2 = 0.52 for poplar) but notably worse in polycultures (e.g., R 2 = 0.17 for early successional, R 2 = 0.06 for restored prairie), indicating that simulation models that include N 2 O emissions should be parameterized specific to particular plant communities. Our results indicate that perennial bioenergy crops in their establishment phase emit less N 2 O than annual crops, especially when not fertilized. These findings should be considered further alongside yield and other metrics contributing to important ecosystem services.
... (1) Because switchgrass is highly productive (and more productive when fertilizers and chemicals are applied to encourage its growth as a dense monoculture) and has higher biomass production than most grassland species, the total estimated switchgrass biomass productivity was assumed to be double that of the total estimated grassland biomass productivity based on previous study results (Anderson-Teixeira et al., 2012;Behrman et al., 2012;Bonin and Lal, 2014;Fike et al., 2006;Jager et al., 2010;Kiniry et al., 2008;McLaughlin et al., 2006;Schmer et al., 2010;Tulbure et al., 2012;Vogel et al., 2002;Wullschleger et al., 2010). ...
... Random forest classifiers are designed to handle correlated predictor variables (Breiman 2001), but random forest measures of predictor variable importance are biased when the variables are correlated (Hothorn, Hornik, and Zeileis 2006;Strobl et al. 2008). Consequently, in this study, a straightforward approach, similar to the procedure followed by Tulbure et al. (2012), was implemented. Recall that the predictor variables are in seven groups (fuel flammability, fuel load, fire behavior, fire seasonality, annual fire frequency, proximity to surface transportation, and local temperature) ( Table 1). ...
Article
Full-text available
The Brazilian Tropical Moist Forest Biome (BTMFB) spans almost 4 million km2 and is subject to extensive annual fires that have been categorized into deforestation, maintenance, and forest fire types. Information on fire types is important as they have different atmospheric emissions and ecological impacts. A supervised classification methodology is presented to classify the fire type of MODerate resolution Imaging Spectroradiometer (MODIS) active fire detections using training data defined by consideration of Brazilian government forest monitoring program annual land cover maps, and using predictor variables concerned with fuel flammability, fuel load, fire behavior, fire seasonality, fire annual frequency, proximity to surface transportation, and local temperature. The fire seasonality, local temperature, and fuel flammability were the most influential on the classification. Classified fire type results for all 1.6 million MODIS Terra and Aqua BTMFB active fire detections over eight years (2003–2010) are presented with an overall fire type classification accuracy of 90.9% (kappa 0.824). The fire type user’s and producer’s classification accuracies were respectively 92.4% and 94.4% (maintenance fires), 88.4% and 87.5% (forest fires), and, 88.7% and 75.0% (deforestation fires). The spatial and temporal distribution of the classified fire types are presented and are similar to patterns reported in the available recent literature.
... In their studies, the upland varieties from the U.S. Mid- west region had matured earlier and produced less bio- mass than the lowland varieties from the southern region. For lowland cytotypes, precipitation is the most impor- tant climate variable [40]. ...
Article
Full-text available
This study examined the effects of soil and switchgrass variety on sustainability and eco-friendliness of switchgrass-based ethanol production. Using the Agricultural Land Management Alternatives with Numerical Assessment Criteria (ALMANAC) model, switchgrass biomass yields were simulated for several scenarios of soils and varieties. The yields were fed to the Integrated Biomass Supply Analysis and Logistics (IBSAL) model to compute energy use and carbon emissions in the biomass supply chain, which then were used to compute Net Energy Value (NEV) and Carbon Credit Balance (CCB), the indicators of sustainability and eco-friendliness, respectively. The results showed that the values of these indicators increased in the direction of heavier to lighter soils and on the order of north-upland, south-upland, north-lowland, and south-lowland varieties. The values of NEV and CCB increased in the direction of dry to wet year. Gaps among the varieties were smaller in a dry year than in a wet year. From south to north, NEV and CCB decreased for lowland varieties but increased for upland ones. Thus, the differences among the varieties decreased in the direction of lower to higher latitudes. The study demonstrated that the sustainability and eco-friendliness of switchgrass-based ethanol production could be increased with alternative soil and variety options.
... The upland ecotype is shorter, thinner , and is more tolerable for drier and cold weather (Gunderson et al., 2008; Thomson et al., 2009). Biomass yield depend on not only the ecotypes but also the climate, soil type, etc. (Tulbure et al., 2012). Switchgrass yields by county measured in Mg/ha for both upland and lowland ecotypes were obtained from a predictive model, developed by Oak Ridge National Lab, based on the relationship between observed yields and climatic conditions (Gunderson et al., 2008 ). ...
Article
Transition of the current gasoline-based transportation system into a renewable fuel-based clean vehicle system has the potential to reduce greenhouse gas emissions and improve national energy security. However, the realized net environmental benefit or energy security improvement is tightly linked to the electricity fuel mix (for electric cars and plug-in hybrids) and fueling strategy (for cars using alternative liquid fuels). In addition, different types of transportation fuels have significantly different demands on land resources, both on land type and quantity. For example, biofuel production requires large quantities of agricultural land, while wind farms require land with sufficient wind density. Furthermore, there is substantial regional variation in the quality of necessary resources. Regions with higher wind speeds require less land to produce the same amount of electricity than those with lower wind speed, assuming the same turbine design. Similarly, regions with optimal soil conditions and climate for crop cultivation require less land to produce the same amount of biofuel. To enable comparison of land demand among different fuel choices for clean vehicles, this research provides a county-scale assessment of land demand based on a “per-vehicle-mile-traveled” basis. Potential clean vehicle fuels assessed in this study include ethanol produced from different feedstocks (corn and switchgrass), biodiesel from algae cultivated in open ponds and closed systems, and electricity produced from renewable sources (wind and solar). Our results show that, in general, engineered systems (wind electricity, solar electricity, and biodiesel from closed-system algae) are more land efficient than natural systems (corn ethanol from corn starch and stover, switchgrass ethanol, and biodiesel from open-pond algae). Solar electricity is the dominant regional optimal fuel choice from the land-use perspective for engineered systems while lowland switchgrass ethanol and biodiesel from open-pond algae are the major optimal choices for the natural systems. These results shed light on developing both federal and state level policies to minimize land-use impact for the development of a clean vehicle system.
... Even with the highest CO 2 capture scenario (1c in In the case of second-generation ligno-cellulosic perennial C4 crops, a yield of 23.6–47.2 Mg ha −1 yr −1 is required before 2055, which is about 2.1–4.2 times the currently reported average yield of switchgrass [Tulbure et al., 2012], and 1.7–3.3 times the yield of Miscanthus × giganteus [Gauder et al., 2012] depending on the technological options for pre-and postcombustion capture used in the biofuel utilization (Figure 1). ...
Article
Full-text available
Bioenergy with Carbon Capture and Storage (BECCS) is a key component of mitigation strategies in future socio-economic scenarios that aim to keep mean global temperature rise below 2 "∘ C above pre-industrial, which would require net negative carbon emissions in the end of the 21st century. Because of the additional need for land, developing sustainable low-carbon scenarios requires careful consideration of the land-use implications of deploying large-scale BECCS. We evaluated the feasibility of the large-scale BECCS in RCP2.6, which is a scenario with net negative emissions aiming to keep the 2 "∘ C temperature target, with a top-down analysis of required yields and a bottom-up evaluation of BECCS potential using a process-based global crop model. Land-use change carbon emissions related to the land expansion were examined using a global terrestrial biogeochemical cycle model. Our analysis reveals that first-generation bioenergy crops would not meet the required BECCS of the RCP2.6 scenario even with a high fertilizer and irrigation application. Using second-generation bioenergy crops can marginally fulfill the required BECCS only if a technology of full post-process combustion CO 2 capture is deployed with a high fertilizer application in the crop production. If such an assumed technological improvement does not occur in the future, more than doubling the area for bioenergy production for BECCS around 2050 assumed in RCP2.6 would be required, however, such scenarios implicitly induce large-scale land-use changes that would cancel half of the assumed CO 2 sequestration by BECCS. Otherwise a conflict of land-use with food production is inevitable.
... The delay in full establishment resulted in decreased yields of this lowland cultivar in year 1 in particular. Tulbure et al. [30] utilized numerous variables to model switchgrass yields across the USA and concluded that genetics or ecotype, among other parameters, was one of the most important factors for explaining switchgrass yield variability. ...
Article
Full-text available
The Regional Feedstock Partnership is a collaborative effort between the Sun Grant Initiative (through Land Grant Universities), the US Department of Energy, and the US Department of Agriculture. One segment of this partnership is the field-scale evaluation of switchgrass (Panicum virgatum L.) in diverse sites across the USA. Switchgrass was planted (11.2 kg PLS ha−1) in replicated plots in New York, Oklahoma, South Dakota, and Virginia in 2008 and in Iowa in 2009. Adapted switchgrass cultivars were selected for each location and baseline soil samples collected before planting. Nitrogen fertilizer (0, 56, and 112 kg N ha−1) was applied each spring beginning the year after planting, and switchgrass was harvested once annually after senescence. Establishment, management, and harvest operations were completed using field-scale equipment. Switchgrass production ranged from 2 to 11.5 Mg ha−1 across locations and years. Yields were lowest the first year after establishment. Switchgrass responded positively to N in 6 of 19 location/year combinations and there was one location/year combination (NY in Year 2) where a significant negative response was noted. Initial soil N levels were lowest in SD and VA (significant N response) and highest at the other three locations (no N response). Although N rate affected some measures of biomass quality (N and hemicellulose), location and year had greater overall effects on all quality parameters evaluated. These results demonstrate the importance of local field-scale research and of proper N management in order to reduce unnecessary expense and potential environmental impacts of switchgrass grown for bioenergy.
... This result also confirms that factors in addition to cultivar affected differences in biomass yield among locations. Tulbure et al. [30] utilized numerous variables to model switchgrass yields across the USA and concluded that fertilizer application (N primarily), genetics, precipitation, and other management practices were the most important for explaining switchgrass yield variability despite the fact that precipitation was not significant in this study. ...
... Grassland biomass productivity (kg ha -1 year -1 ) = 9936.5 x GSN -1554 (1) Based on the previous study results, switchgrass has higher biomass production than most grassland species; therefore, the total estimated switchgrass biomass productivity gain from the identified biofuel potential areas in the GP was assumed to be double that of the total estimated grassland biomass productivity (Anderson-Teixeira et al., 2012, Behrman et al., 2012, Bonin & Lal, 2014, Fike et al., 2006, Gu & Wylie, 2016, Jager et al., 2010, Kiniry et al., 2008, McLaughlin et al., 2006, Schmer et al., 2010, Tulbure et al., 2012, Vogel et al., 2002. Fig. 5). ...
Article
Growing cellulosic feedstock crops (e.g., switchgrass) for biofuel is more environmentally sustainable than corn-based ethanol. Specifically, this practice can reduce soil erosion and water quality impairment from pesticides and fertilizer, improve ecosystem services and sustainability (e.g., serve as carbon sinks), and minimize impacts on global food supplies. The main goal of this study is to identify high risk marginal croplands that are potentially suitable for growing cellulosic feedstock crops (e.g., switchgrass) in the U.S. Great Plains (GP). Satellite-derived growing season Normalized Difference Vegetation Index, a switchgrass biomass productivity map obtained from a previous study, U.S. Geological Survey (USGS) irrigation and crop masks, and U.S. Department of Agriculture (USDA) crop indemnity maps for the GP were used in this study. Our hypothesis was that croplands with relatively low crop yield but high productivity potential for switchgrass may be suitable for converting to switchgrass. Areas with relatively low crop indemnity (crop indemnity < $2,157,068) were excluded from the suitable areas based on low probability of crop failures. Results show that approximately 650,000 ha of marginal croplands in the GP are potentially suitable for switchgrass development. The total estimated switchgrass biomass productivity gain from these suitable areas is about 5.9 million metric tons. Switchgrass can be cultivated in either lowland or upland regions in the GP depending on the local soil and environmental conditions. This study improves our understanding of ecosystem services and the sustainability of cropland systems in the GP. Results from this study provide useful information to land managers for making informed decisions regarding switchgrass development in the GP. This article is protected by copyright. All rights reserved.
... Random forest measures of variable importance may become biased if the predictor variables are correlated [Hothorn et al., 2006;Strobl et al., 2008]. In this study a simple approach, similar to the procedure followed by Tulbure et al., [2012], to reduce this bias was implemented. The predictor variables were grouped as variables concerned with fuel flammability, fuel load, fire behavior, fire seasonality, fire diurnal cycle, proximity to people, and local temperature ( Table 1). ...
Thesis
Full-text available
The Brazilian Tropical Moist Forest Biome (BTMFB) is “Earth’s greatest biological treasure and a major component of the earth system” and forest degradation and deforestation by fire is a serious issue in this region. Fires in the BTMFB can be broadly classified as maintenance, deforestation and forest fire types. Spatially and temporally explicit information on the incidences of fire types are important as they have widely varying atmospheric emissions and ecological impacts. Satellite based remote sensing is a practical means of monitoring the BTMFB that spans almost 4 million km2. However, there has been no way to reliably classify satellite active fire type to date. In this work, methods to characterize MODIS active fire detections are developed using physically based and geographic context/proximity approaches. The research methodology is developed by addressing four hypotheses concerning differences among active fire type characteristics including factors that drive and mediate fire in the BTMFB. Differences in the active fire characteristics among different fire types are presented and discussed. The spatio-temporal distribution of fire types over 8 year (2003-2010) period is documented, analyzed and presented. This dissertation has, to date, resulted in one published, one in press, and one submitted paper. http://openprairie.sdstate.edu/etd/1078/
... Considering natural disasters like floods that occur almost every year, these scenarios help discover vulnerability of farmers to undesirable combinations of crop prices and production costs, which are likely to increase in the face of The authors coded the model using the General Algebraic Modelling System GAMS® language with the CONOPT nonlinear solver. This solver has seen widespread use in the agriculture and water resources literature (Ahrends et al., 2008;Carvallo et al., 1998;Ernoult et al., 2006;Nainggolan et al., 2012;Tulbure et al., 2012;Wang et al., 2013;Zhang et al., 2015). A mathematical description of the most important model elements is presented in Appendix A. ...
Article
CONTEXT A classical challenge in the search for principles and guidelines to protect food security affordably in the developing world comes from the well-known dilemma faced by farmers who incur economic losses from low prices brought on by excess production combined with poorly developed policies to anticipate, adapt to, and manage production and price fluctuations. This challenge is elevated in the face of growing evidence of climate change. OBJECTIVE This work performs a series of policy experiments to discover an affordable set of measures to protect food grain security as well as assure profitability of high valued crops in Bangladesh. METHODS This work uses primary and secondary data from Bangladesh agriculture to develop an empirical mathematical programming economic optimization model to achieve each of four alternative food policy objectives. RESULTS AND CONCLUSIONS Results show total economic welfare achievable under each policy objectives: protecting baseline observed farm sector outcomes, minimum protection of food grain security, protection of farm income from producing high value crops, and unconstrained food welfare optimization. The application is to an important farming region of northwest Bangladesh. We find that policymakers can achieve both food grain security and farm income from high value crops at a minimum cost of economic welfare displaced. SIGNIFICANCE The use of empirical models that describe and predict farm economic optimization behavior can provide guidance in the search for affordable policies for protecting food grain security while protecting farm income in the developing world.
... Compared with other grass species, switchgrass is characterized by higher aboveground biomass production, lower herbicide and fertilizer input requirements, and more widespread adaptability to climatic conditions, and hence has stronger ability to sequester atmospheric carbon and to mitigate climate change (Gelfand et al., 2013;Eichelmann et al., 2016). As a result, the U.S. Department of Energy (DOE), partnering with the U.S. Department of Agriculture (USDA), has selected switchgrass as the model feedstock to be used for bioenergy production (McLaughlin and Kszos, 2005;Tulbure et al., 2012). Accordingly, the scope of switchgrass lands has rapidly increased in recent decades (Parrish and Fike, 2005;Schmer et al., 2008), and the U.S. switchgrass yield was expected to double or even triple for the goals of 36 billion gallons of biofuels production annually by 2022 (McLaughlin et al., 2006). ...
... Switchgrass (Panicum virgatum L.) is one of the leading second generation perennial crops capable of maintaining or enhancing soil quality through impacts on features such as soil carbon [1,56,60]. Previous studies of switchgrass have largely focused on modeling biomass production by parameterizing aboveground growth, because the most commonly measured parameters are those governing leaf area development and aboveground biomass accumulation [6,11,23,32]. For example, Behrman et al. [6] modeled the differential growth for two lowland cultivars (Alamo and Kanlow) and two upland cultivars (Cave-in-Rock and Blackwell) using ALMANAC (Agricultural Land Management and Numerical Assessment Criteria) by parameterizing potential heat units during the growing season, maximum leaf area index, and light extinction coefficient while all belowground parameters remained constant for all the cultivars. ...
Article
Full-text available
Crop simulation models are increasingly being used to understand the feasibility of large-scale cellulosic biofuel production along with the multi-dimensional impacts on environmental sustainability. However, how the uncertainty in model parameters impacts model performance for sustainability is unclear. In this case study, sensitivity analyses were conducted for three switchgrass sustainability metrics: total biomass production, nitrogen loss, and soil carbon change using the APEX (Agricultural Policy/Environmental eXtender) model. Fifteen out of the 45 parameters (25 crop growth (CROP) parameters and 20 additional model parameters (PARM)) were identified as influential for the three sustainability metrics for three lowland genotypes (WBC, AP13, and KAN) across two locations (Temple, TX, and Austin, TX). Our sensitivity results showed that parameter importance was not dependent on the genotypes but depended on the variables of interest, and differed only slightly between locations. Influential belowground-related CROP and PARM parameters were identified for each sustainability metric, indicating that belowground-related parameters are just as important as commonly measured aboveground CROP parameters. Further investigation of the linear or non-linear relationships and the two-way interactions between each of the individual influential parameters with the three sustainability metrics reflected the functions and characteristics within the APEX model and the interrelations among different processes. Strong interactions between the most influential parameters for total biomass, nitrogen loss, and soil carbon change also highlighted the importance of accurately setting these parameters. Identification of influential model parameters for switchgrass sustainability may help guide field measurements and provide further understanding of the interrelated processes in the APEX model. Furthermore, future field experiments can be designed to measure these influential parameters and understand the non-linear relationships identified between influential parameters and response variables. More accurate model parameterization will help improve APEX model performance and our understanding of the possible underlying physiological mechanisms.
... Random forest classifiers are designed to handle correlated predictor variables (Breiman 2001), but random forest measures of predictor variable importance are biased when the variables are correlated (Hothorn, Hornik, and Zeileis 2006;Strobl et al. 2008). Consequently, in this study, a straightforward approach, similar to the procedure followed by Tulbure et al. (2012), was implemented. Recall that the predictor variables are in seven groups (fuel flammability, fuel load, fire behavior, fire seasonality, annual fire frequency, proximity to surface transportation, and local temperature) ( Table 1). ...
Conference Paper
Abstract Vegetation fires in the Brazilian Tropical Moist Forest Biome can be broadly classified into three types: i) Deforestation fires, lit to aid deforestation by burning of slashed, piled and dried forest biomass, ii) Maintenance fires, lit on agricultural fields or pasture areas to maintain and clear woody material and to rehabilitate degraded pasture areas, iii) Forest fires, associated with escaped anthropogenic fires or, less frequently, caused by lightning. Information on the incidence and spatial distribution of fire types is important as they have widely varying atmospheric emissions and ecological impacts. Satellite remote sensing offers a practical means of monitoring fires over areas as extensive as the Brazilian Tropical Moist Forest Biome which spans almost 4 million square kilometers. To date, fire type has been inferred based on the geographic context and proximity of satellite active fire detections relative to thematic land cover classes, roads, and forest edges, or by empirical consideration of the active fire detection frequency. In this paper a classification methodology is presented that demonstrates a way to classify the fire type of MODerate Resolution Imaging Spectroradiometer (MODIS) active fire detections. Training and validation fire type data are defined conservatively for MODIS active fire detections using a land cover transition matrix that labels MODIS active fires by consideration of the PRODES 120m land cover for the previous year and the year of fire detection. The training data are used with a random forest classifier and remotely sensed predictor variables including the number of MODIS Aqua and Terra satellite detections, the maximum and median Fire Radiative Power (FRP) [MW km-2], the scaling parameter of the FRP power law distribution, the number of day and night detections, and the fire surrounding "background" surface brightness temperature [K]. In addition, the total rainfall over periods from 1 to 24 months prior to fire detection and the fire detection proximity to official and unofficial roads and navigable rivers are included as predictor variables. Results are illustrated for eight years (2003-2010) of MODIS active fire detections with a cross validation showing greater than 70% fire type classification accuracy. The spatio-temporal distribution of fire types across the Brazilian Tropical Moist Forest Biome are presented with higher incidences of deforestation fires in the "arc of deforestation" and similar proportions of forest and maintenance fires for all years except for 2007 and 2010 that exhibited a relatively higher proportion of forest fires.
... RF methods have been widely adopted for certain agricultural problems, such as remote sensing analysis (Lebourgeois et al., 2017;Parente et al., 2017), leaf nitrogen levels (Abdel-Rahman and Ahmed, 2008) and classifying sugarcane varieties (Everingham et al., 2007). RF were used in many problems of yield estimation (Park et al., 2005;Tulbure et al., 2012;Fukuda et al., 2013;Newlands et al., 2014;Jeong et al., 2016), particularly in sugarcane fields (Everingham et al., 2009;Everingham et al., 2015a;Everingham et al., 2015b;Everingham et al., 2016). RF algorithms can handle large volumes of data, use categorical variables as predictors, measure the degree of importance of the predictive variables, and output the class probability and is robust against overfitting, even for slightly imbalanced datasets (Khoshgoftaar et al., 2007). ...
Article
The adoption of information technology (IT) and precision agriculture (PA) has converted agricultural fields into data sources. However, the transformation of data into knowledge for decision making remains a major challenge. In the Brazilian sugarcane industry, the current use of PA technology is very far from its full potential for site-specific management, mainly because yields are not temporally or spatially monitored. The objective of the present study was to investigate the relationship between the physical and chemical properties of soils and sugarcane yield, thereby identifying the soil parameters that determine the final productivity of the field. Two sugarcane fields were monitored from 2011 to 2014. During the crop season, soil samples and yield data were collected annually. A random forest algorithm was applied to investigate the influence of different soil attributes on yield using data that were collected spatially over the study period. The results showed that the amount of available soil organic matter (OM), clay content and cation exchange capacity (CEC) are important factors impacting sugarcane yield variation. Furthermore, it was found that the temporal variability in yield is caused mainly by the variability in soil pH over the study period. The results indicated that when OM increased over time, there was greater phosphorus availability. Large volumes of spatial and temporal data, together with data mining techniques, allowed the extraction of knowledge and the creation of specific management zones in the field to support the decision-making process for producers
... Random forests have been involved in predicting mango yields (Fukuda et al., 2013), and in the predicting model of crops for complex seasonal yield. Using the random forest to recognize more significant variables including rainfall, soil silt, fertilizer, and so on, in the USA for forecasting the yields of switchgrass are presented in Tulbure et al. (2012). Data analysis using random forests is also utilized to examine agriculture-related problems like emission of nitrous oxide (Philibert et al., 2013), the amount of nitrogen present in the leaves, prediction of droughts are another examples (Rhee and Im, 2017). ...
Article
The advancement in science and technology has led to a substantial amount of data from various fields of agriculture to be incremented in the public domain. Hence a desideratum arises from the investigation of the available data and integrating them with a process like a crop improvement, yield prediction, crop disease analysis, identifying water stress etc. Computing techniques like Machine learning is a novel advent for the analysis and resoluteness of these intricate issues. Various analytical models like Decision Trees, Random Forests, Support Vector Machines, Bayesian Networks, and Artificial Neural Networks etc. have been utilized for engendering the models and analyze the results. These methods enable to analyze soil, climate, and water regime which are significantly involved in crop growth and precision farming. This survey incorporates an overview of some of the existing supervised and unsupervised machine learning models associated with the crop yield in literature. Moreover, this survey compares one approach with other by means of various error measures like Root Mean Square Error (RMSE), Relative Root Mean Square Error (RRMSE), Mean Absolute Error (MAE), and Coefficient of Determination (R2).
... RF is one of the most used machine learning techniques in agriculture thanks to its non-parametric nature, high predictive ability, internal evaluation of attributes, robustness to noise, and lack of proneness to overfitting (Rupnik et al., 2019). To date, applications of RF for yield prediction targeted, among others, mango (Fukuda et al., 2013) and switchgrass (Tulbure et al., 2012), whereby Jeong et al. (2016) proved that RF outperformed multiple linear regression for yield prediction of staple food crops at the regional and global level. In our study, other than being essential in improving system performances, RF provided a robust ranking of the predictors, confirming the added value of the process-based model application, whose simulated yields were top-ranked, as well as the relevant contribution of ground information on the targeted hazelnut systems. ...
Article
Full-text available
Crop yield forecasting activities are essential to support decision making of farmers, private companies and public entities. While standard systems use georeferenced agro-climatic data as input to process-based simulation models, new trends entail the application of machine learning for yield prediction. In this paper we present HADES (HAzelnut yielD forEcaSt), a hazelnut yield prediction system, in which process-based modeling and machine learning techniques are hybridized and applied in Turkey. Official yields in the top hazelnut producing municipalities in 2004–2019 are used as reference data, whereas ground observations of phenology and weather data represent the main HADES inputs. A statistical analysis allows inferring the occurrence and magnitude of biennial bearing in official yields and is used to aid the calibration of a process-based hazelnut simulation model. Then, a Random Forest algorithm is deployed in regression mode using the outputs of the process-based model as predictors, together with information on hazelnut varieties, the presence of alternate bearing in the yield series, and agro-meteorological indicators. HADES predictive ability in calibration and validation was balanced, with relative root mean square error below 20%, and R2 and Nash-Sutcliffe modeling efficiency above 0.7 considering all municipalities together. HADES paves the way for a next-generation yield prediction system, to deliver timely and robust information and enhance the sustainability of the hazelnut sector across the globe.
... The RF algorithm, proposed by Breiman [58], is a bagging-based method that employs a regression tree method. It has been widely applied for prediction via the "RandomForest" package within the R software environment [59], although it lacks efficiency, especially when dealing with our large training set [60]. The final predicted value is the mean fitted response from all individual trees [61]. ...
Article
Full-text available
Citation: Chen, X.; Feng, L.; Yao, R.; Wu, X.; Sun, J.; Gong, W. Prediction of Maize Yield at the City Level in China Using Multi-Source Data. Remote Sens. Publisher's Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons At-tribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Abstract: Maize is a widely grown crop in China, and the relationships between agroclimatic parameters and maize yield are complicated, hence, accurate and timely yield prediction is challenging. Here, climate, satellite data, and meteorological indices were integrated to predict maize yield at the city-level in China from 2000 to 2015 using four machine learning approaches, e.g., cubist, random forest (RF), extreme gradient boosting (Xgboost), and support vector machine (SVM). The climate variables included the diffuse flux of photosynthetic active radiation (PDf), the diffuse flux of shortwave radiation (SDf), the direct flux of shortwave radiation (SDr), minimum temperature (Tmn), potential evapotranspiration (Pet), vapor pressure deficit (Vpd), vapor pressure (Vap), and wet day frequency (Wet). Satellite data, including the enhanced vegetation index (EVI), normalized difference vegetation index (NDVI), and adjusted vegetation index (SAVI) from the Moderate Resolution Imaging Spectroradiometer (MODIS), were used. Meteorological indices, including growing degree day (GDD), extreme degree day (EDD), and the Standardized Precipitation Evapotranspiration Index (SPEI), were used. The results showed that integrating all climate, satellite data, and meteorological indices could achieve the highest accuracy. The highest estimated correlation coefficient (R) values for the cubist, RF, SVM, and Xgboost methods were 0.828, 0.806, 0.742, and 0.758, respectively. The climate, satellite data, or meteorological indices inputs from all growth stages were essential for maize yield prediction, especially in late growth stages. R improved by about 0.126, 0.117, and 0.143 by adding climate data from the early, peak, and late-period to satellite data and meteorological indices from all stages via the four machine learning algorithms, respectively. R increased by 0.016, 0.016, and 0.017 when adding satellite data from the early, peak, and late stages to climate data and meteorological indices from all stages, respectively. R increased by 0.003, 0.032, and 0.042 when adding meteorological indices from the early, peak, and late stages to climate and satellite data from all stages, respectively. The analysis found that the spatial divergences were large and the R value in Northwest region reached 0.942, 0.904, 0.934, and 0.850 for the Cubist, RF, SVM, and Xgboost, respectively. This study highlights the advantages of using climate, satellite data, and meteorological indices for large-scale maize yield estimation with machine learning algorithms.
... Random forest is one of the most significant machine learning algorithms that has contributed solutions to numerous agricultural applications. The random forest has been engaged with forecasting yield for mango and forecasting the crop model for complex seasonal yield (Fukuda et al. 2013) Utilizing the random forest progressively significant factors including precipitation, soil sediment and manure for determining the switchgrass yields are perceived (Tulbure et al. 2012) Mapping of crop type in affiliation with random forest and crop statistics is a convincing application for monitoring the crop yield for food security . Enhanced prediction of poppy crop capsule volume is necessary for optimal crop yield management. Estimating poppy crop yield based on random forest machine learning algorithm using an unmanned aircraft system provided significant results with better predictive accuracy (Iqbal et al. 2018). ...
Article
Full-text available
The development in technology and science has contributed to a vast volume of data from various agrarian fields to be aggregated in the public domain. Predicting the crop yield based on climate, soil and water parameters has been a potential re- search subject. Therefore an objective arises in integrating the data with agrarian processes for crop enhancement. In this paper, a new hybrid regression-based algorithm, Reinforcement Random Forest is proposed which displays significantly enhanced performance over traditional machine learning techniques like the random forest, decision tree, gradient boosting, artificial neural network and deep Q-learning. The new strategy executes reinforcement learning at every selection of a splitting attribute amid the process of tree construction for the efficient utilization of the available samples. They analyze the variable significance measure to select the most substantial variable for node splitting process in the model development and promotes efficient utilization of training data. This integrated hybrid procedure provides significant enhancement over prevailing strategies specifically for sparse model structures. Besides performing internal cross-validation, the proposed approach requires less parameter tuning, reduces over-fitting, faster calculation and more transparent. The experimented models are evaluated with various assessment metrics like Root Mean Squared Error, Mean Squared Error, Determination Coefficient and Mean Absolute Error. The results obtained delineated that the proposed approach performs better with reduced error measures and improved accuracy of 92.2%.
... c o m / l o c a t e / e n e e c o environmental impacts (Lewandowski et al., 2003). Also, a number of studies have suggested recently that marginal and abandoned lands could potentially be converted to cellulosic feedstock production which would avoid a large scale conversion of current crop land to biofuel feedstock production (Tulbure et al., 2012). The contribution of renewable energy resources is vital for economic growth as well as an effective tool for the confrontation of climate change (Zafeiriou et al., 2016). ...
Article
There has been an increasing interest in the use of perennial grasses as potential feedstock for ethanol production. The characteristics which make perennial grasses attractive for bioenergy feedstock development initiative are their high yield potential and the high contents of lignin and cellulose. The objective of the study is to model energy input and output and simulate Net Energy Value (NEV) of producing ethanol from native warm season grasses. According to simulated results, the mean NEV of ethanol production from native warm season grasses considered in the analysis was positive. Mean NEV for switchgrass and eastern gammagrass was higher compared to Indiangrass and big bluestem. Although there is a higher probability of having positive NEV, there is a risk of having negative energy balance under low output scenarios.
... In the context of modeling crop yields, random forest approaches have been applied to predict yields in corn, wheat, potato, sugarcane, sorghum, and groundnut based on climate variables (Everingham, Sexton, Skocaj, & Inman-Bamber, 2016;Hoffman, Kemanian, & Forest, 2017;Jeong et al., 2016). Tulbure, Wimberly, Boe, and Owens (2012) leveraged random forest regression to identify variables related to weather, genetics, and management practices that were underlying spatiotemporal variability in switchgrass yield. Smidt et al. (2016) used a similar approach to identify the key predictors of soybean yield among topographical and soil field attributes under a variable planting rate experimental design. ...
Article
Full-text available
In recent years, planting machinery that enables precise control of the planting rates has become available for corn (Zea mays L.) and soybean (Glycine max L.). With increasingly available topographical and soil information, there is a growing interest in developing variable rate planting strategies to exploit variation in the agri‐landscape in order to maximize production. A random forest regression‐based approach was developed to model the interactions between planting rate, hybrid/variety, topography, soil characteristics, weather variables, and their effects on yield by leveraging on‐farm variable rate planting trials for corn and soybean conducted at 27 sites in New York between 2014 and 2018 (57 site‐years) in collaboration with the New York Corn and Soybean Growers Association. Planting rate ranked highly in terms of random forest regression variable importance while explaining relatively minimal yield variation in the linear context, indicating that yield response to planting rate likely depends on complex interactions with agri‐landscape features. Random forest models explained moderate levels of yield within site‐years, while the ability to predict yield in untested site‐years was low. Relatedly, variable importance measures for the predictors varied considerably across sites. Together, these results suggest that local testing may provide the most accurate optimized planting rate designs due to the unique set of conditions at each site. The proposed method was extended to identify the optimal variable rate planting design for maximizing yield at each site given the underlying conditions, and empirical validation of the resulting designs is currently underway.
... Several outstanding features that render perennial grasses attractive in bioenergy production include higher yield potential, high cellulose content, and positive environmental impacts (Lewandowski et al., 2003). In addition, marginal and abandoned land may be exploited for the cultivation of lignocellulosic crops, which can avoid the competition with farm land in biofuel feedstock production (Tulbure et al., 2012). ...
Article
Full-text available
As an herbaceous perennial, Miscanthus has attracted extensive attention in bioenergy refinery and ecological remediation due to its high yield and superior environmental adaptability. This review summarizes current research advances of Miscanthus in several aspects including biological properties, biofuels production, and phytoremediation of contaminated soil. Miscanthus has relatively high biomass yield, calorific value, and cellulose content compared with other lignocellulosic bioenergy crops, which make it one of the most promising feedstocks for the production of second‐generation biofuels. Moreover, Miscanthus can endure soil pollutions caused by various heavy metals and survive in a variety of adverse environmental conditions. Therefore, it also has potential applications in ecological remediation of contaminated soil, and reclamation of polluted soil and water resources. Nevertheless, more endeavors are still needed in the genetic improvement and elite cultivar breeding, large‐scale cultivation on marginal land, and efficient conversion to biofuels, when utilizing Miscanthus as a bioenergy crop. Furthermore, more efforts should also be undertaken to translate Miscanthus into a bioenergy crop with the phytoremediation potential.
Conference Paper
High biomass production potential, wide adaptability, low input requirement, and low environmental risk make switchgrass an economically and ecologically viable energy crop. The inherent variability in switchgrass productivity due to variations in soil and variety could affect the sustainability and eco-friendliness of switchgrass-based ethanol production. This study examined the soil and variety effects on these variables. Three locations in Mississippi were selected based on latitude and potential acreage. Using ALMANAC, switchgrass biomass yields were simulated for several scenarios of soils and varieties. The simulated yields were fed to IBSAL to compute energy use and CO2 emissions in various operations in the biomass supply chain. From the energy and emissions values, the sustainability and eco-friendliness of ethanol production were determined using net energy value (NEV) and carbon credit balance (CCB) as indicators, respectively. Soil and variety effects on NEV and CCB were analyzed using the Kruskal-Wallis test. Results showed significant differences in NEV and CCB across soils and varieties. Both NEV and CCB increased in the direction of heavier to lighter soils and on the order of north-upland, south-upland, north-lowland, and south-lowland varieties. Only north-upland and south-lowland varieties were significantly different because they were different in both cytotype and ecotype. Gaps between lowland and upland varieties were smaller in a dry year than in a wet year. The NEV and CCB increased in the direction of dry to wet year. From south to north, they decreased for lowland cytotypes but increased for upland cytotypes. Thus, the differences among varieties decreased northwards.
Article
Full-text available
In northern China, precipitation that is primarily concentrated during the fallow period is insufficient for the growth stage, creates a moisture shortage, and leads to low, unstable yields. Yield prediction in the early growth stages significantly informs field management decisions for winter wheat (Triticum aestivum L.). A 10-year field experiment carried out in the Loess Plateau area tested how three tillage practices (deep ploughing (DP), subsoiling (SS), and no tillage (NT)) influenced cultivation and yield across different fallow periods. The experiment used the random forest (RF) algorithm to construct a prediction model of yields and yield components. Our results revealed that tillage during the fallow period was more effective than NT in improving yield in dryland wheat. Under drought condition, DP during the fallow period achieved a higher yield than SS, especially in drought years; DP was 16% higher than SS. RF was deemed fit for yield prediction across different precipitation years. An RF model was developed using meteorological factors for fixed variables and soil water storage after tillage during a fallow period for a control variable. Small error values existed in the prediction yield, spike number, and grains number per spike. Additionally, the relative error of crop yield under fallow tillage (5.24%) was smaller than that of NT (6.49%). The prediction error of relative meteorological yield was minimum and optimal, indicating that the model is suitable to explain the influence of meteorological factors on yield.
Article
Full-text available
Nitrogen is the most limiting nutrient for turfgrass growth. Instead of pursuing the maximum yield, most turfgrass managers use nitrogen (N) to maintain a sub-maximal growth rate. Few tools or soil tests exist to help managers guide N fertilizer decisions. Turf growth prediction models have the potential to be useful, but the currently existing turf growth prediction model only takes temperature into account, limiting its accuracy. This study developed machine-learning-based turf growth models using the random forest (RF) algorithm to estimate short-term turfgrass clipping yield. To build the RF model, a large set of variables were extracted as predictors including the 7-day weather, traffic intensity, soil moisture content, N fertilization rate, and the normalized difference red edge (NDRE) vegetation index. In this study, the data were collected from two putting greens where the turfgrass received 0 to 1,800 round/week traffic rates, various irrigation rates to maintain the soil moisture content between 9 and 29%, and N fertilization rates of 0 to 17.5 kg ha –1 applied biweekly. The RF model agreed with the actual clipping yield collected from the experimental results. The temperature and relative humidity were the most important weather factors. Including NDRE improved the prediction accuracy of the model. The highest coefficient of determination (R ² ) of the RF model was 0.64 for the training dataset and was 0.47 for the testing data set upon the evaluation of the model. This represented a large improvement over the existing growth prediction model ( R ² = 0.01). However, the machine-learning models created were not able to accurately predict the clipping production at other locations. Individual golf courses can create customized growth prediction models using clipping volume to eliminate the deviation caused by temporal and spatial variability. Overall, this study demonstrated the feasibility of creating machine-learning-based yield prediction models that may be able to guide N fertilization decisions on golf course putting greens and presumably other turfgrass areas.
Article
Full-text available
47 个气象要素和小麦种植区经纬度、高程 3 个空间要素,共计 50 个参数作为特征变量集,以实际 单产、气象产量和相对气象产量分别作为目标变量,构建多种变量组合模型对冬小麦产量进行回归预测,并结合袋外数 据重要性结果对产量影响因子进行分析。研究结果表明:1)使用气象产量和相对气象产量作为目标变量建模的预测效果 优于单产模型,决定系数 R 2 均达到 0.8 以上,气象产量的平均绝对误差(mean absolute error,MAE)和均方根误差(root mean square error,RMSE)分别为 415 和 558 kg/hm 2 ,相对气象产量的 MAE 和 RMSE 分别为 0.07 和 0.09;2)相较于气 象特征,空间特征在产量预测中起决定性作用,且小花分化期以及抽穗开花期的气象特征产量预测精度高于其他穗分化 期;3)在气象特征中,利用袋外数据变量重要性得出平均温度、最低温度、负积温、最高温度在不同生育阶段对产量的 影响程度。该研究结果为冬小麦生育早期产量预测提供了新的思路和方法。 关键词:模型;温度;随机森林;产量预测;冬小麦 doi:10.
Conference Paper
Full-text available
Background/Question/Methods The recent U.S. Renewable Fuel Standard calls for 36 billion gallons of ethanol production by 2022 with over half produced from plant biomass. The goals of our research are to evaluate different empirical modeling approaches of feedstock crop yields and assess the sensitivity of biomass feedstock production to climate change. Switchgrass (Panicum virgatum L.), a warm season perennial grass native to North America was chosen as model bioenergy species. Yield data from 1,345 observation points associated with 37 field trial locations across the United States were gathered from 21 reference papers. We compared different modeling approaches, including generalized linear models, generalized additive models, and recursive partitioning to explore the relationships between switchgrass yield and environmental variables. Explanatory variables include climate variables, soil type, and origin of the switchgrass cultivar. We used PRISM climate data on minimum and maximum temperature (tmin and tmax) and precipitation (ppt) to model variation in switchgrass biomass yields as a function of climate. Climate data were summarized by month, growing season (April-September), and year before harvesting. To determine the most relevant explanatory variables we performed independent second order polynomial regressions between switchgrass biomass yield explanatory variables and determined their predictive power based on the coefficient of determination (R2). Results/Conclusions Preliminary results show that average March tmin and tmax, followed by average February tmin and tmax had the strongest association with switchgrass biomass yields. These results suggest that winter/early spring temperature may influence winter survival and the length of the growing season, which might be especially important for the northern distribution of the species. Annual precipitation followed by April precipitation and average growing season precipitation had the strongest associations with switchgrass yield. Future modeling efforts will use a climate-envelope approach to model switchgrass yields as a function of climate and soils and will incorporate downscaled GCM data for future climate change scenarios from the Community Climate System Model (CCSM) to predict potential changes in switchgrass productivity.
Article
Spatially explicit information about stand-level Time Since the last stand-replacing Disturbance (TSD) is fundamental for modelling many forest ecosystem processes, but most of the current satellite remote sensing mapping approaches are based on change detection and time series analysis, and can detect only disturbances that have occurred since the start of the optical satellite data record. The spatial legacy of stand-replacing disturbances can however persist on the landscape for several decades to centuries, in the form of distinct horizontal and vertical stand structure features. We propose a new approach to reconstruct the long-term disturbance history of a forest, estimating TSD through stand-level analysis of LiDAR data, which are highly sensitive to the three-dimensional forest canopy structure. The study area is in the Nez Perce-Clearwater National Forest in north-central Idaho, where airborne LiDAR covering about 52,000 ha and ancillary TSD reference data for a period of more than 140 years were available. The root mean square difference (RSMD) between predicted and reference TSD was 17.5 years with a BIAS of 0.8 years; and on 72.8% of the stands the predicted TSD was less than 10 years apart from the reference TSD (78.2% of the stands when considering only disturbances occurred in the last 100 years). The results demonstrate that airborne LiDAR-derived data have enough explanatory power to reconstruct the long-term, stand-replacing disturbance history of temperate forested areas at regional scales.
Article
The biomass-based energy sector will play an important role in our drive towards renewable energy. Switchgrass has been identified as one of the candidate-dedicated energy crops in the USA because of its various desirable traits. However, many challenges still exist in ensuring a commercially viable switchgrass crop that is sustainable and techno-economically feasible. This article reviews the existing literature to highlight those challenges and potential solutions and recommendations to address those in the future. Various stages in the switchgrass-based biofuel system, namely crop production, provision and logistics, and conversion, are analysed. For each of these stages, the important challenges as reported in the literature are mentioned, and research gaps are identified. The systems-level issues as well as the challenges associated with the practical implementation of switchgrass production are also highlighted. The review illustrates that although switchgrass is a viable biofuel feedstock, considerable progress is necessary to achieve large-scale practical success.
Article
Switchgrass (Panicum virgatum) is one of the leading candidates for sustainable lignocellulosic biofuel production in North America. Most current management recommendations for switchgrass include applications of synthetic nitrogen (N) fertilizers to increase production, particularly when grown on marginal lands. However, this management can be costly to growers and have negative impacts on ecosystem functioning. Also, N-fertilization does not always result in higher yield in switchgrass and may have unintended effects on plant biomass quality, including cell wall composition, that can affect the efficiency of fermentation processes for biofuel production. Drought stress may reduce biomass responses to N-fertilization, further reducing the value of fertilizer application. To examine whether N-fertilization and reduced precipitation affected switchgrass productivity and cell wall composition, we conducted a two-year field experiment in mature stands of two switchgrass cultivars grown for bioenergy at the W.K. Kellogg Biological Station Long Term Ecological Research Site in Michigan, USA. Nitrogen was added at a rate of 56 kg N ha⁻¹ (urea and ammonium nitrate), and precipitation was reduced using rainout shelters. Overall, we did not observe any effect of N-fertilization on biomass production. However, under ambient rainfall conditions, N-fertilization altered switchgrass biomass quality by reducing hemicellulose. Reduced precipitation minimized the effects of N-fertilization on switchgrass cell wall composition. Switchgrass is a relatively drought-tolerant species, and our results indicate that this crop will be a viable bioenergy feedstock even in a changing climate. However, in this study, N-fertilization had no effect on biomass quality or quantity under drought conditions.
Article
Soil water supply plays a key role in driving switchgrass (Panicum virgatum L.) yield, and therefore is an important parameter for crop‐model accuracy. ALMANAC (Agricultural Land Management Alternative with Numerical Assessment Criteria) model has been applied to simulating switchgrass growth with mixed results. The objective was to develop and test a calibration for ALMANAC simulating soil water dynamics in a switchgrass (‘Alamo’) stand under Arkansas conditions. Soil volumetric water content (SW) profiles were measured daily in switchgrass from May 2009 to February 2013. Soil, crop, and weather input data were developed based on in situ measurements. After identifying the most sensitive parameters in SW simulation, a calibration method was proposed, and the parameters initial soil water (FFC), permanent wilting point (U), field capacity (FC), sand content (SAN), silt content (SIL), pH, and maximum stomatal conductance (GSI) were modified. Daily SW simulation outputs from default and calibrated runs were compared to SW observations. Default Willmott agreement d‐index values were lower than the calibrated d‐index values in all years. Therefore, calibration improved simulation accuracy. Calibration accuracy was greater in 2009 and 2010 than in 2011 and 2012. Lower root mean square error of calibrated versus observed SW data confirmed the elevated d‐index values. Lower accuracy in the latter years was related to drought periods when ALMANAC was unable to mimic switchgrass drought adaption by lowering GSI. Calibration of SW in ALMANAC was aided by using site‐specific soil data. Improvement of SW calibration for drought‐resistant plants may be achieved by quantifying GSI dynamics.
Chapter
Different forms of carbon (C) cycle continuously through several pools in natural and managed ecosystems and spheres. Carbon's recent “commodification,” as a negative environmental externality, has rendered it a “scarce” and “tradable” element. Although C supply in nature is not limited, energy is required to make it available as a plant nutrient, assimilate it in plant tissues, and sequester it as temporary or recalcitrant C in soils. Human demand for C-based energy and plant-fixed C has accelerated, and altered its global cycle, raised its atmospheric content, contributed to climate change through global warming, and impacted several provisioning, regulating, and supporting ecosystem services. Agroecosystems are both sources and sinks of C. In a carbon dioxide-constrained world, plant-fixed and sequestered C in natural and managed ecosystems has a potential role in mitigating climate change, providing C-neutral and renewable bioenergy, and positively affecting ecosystem services. Due to the intricacies of the complex, interconnected biogeochemical cycles involving C, nitrogen, and water in which soils play an important role, bioenergy crops do not provide an easy solution to climate change mitigation, they may contribute to it. This chapter presents a critical review assessing the state of knowledge, and exploring opportunities and challenges of the role of C in bioenergy crops in mitigating global climate change, while sustainably providing other ecosystem services.
Article
Full-text available
The rapid depletion of fossil fuel reserves and environmental concerns with their combustion necessitate looking for alternative sources for long term sustainability of the world. These concerns also appear serious in developing countries who are striving for rapid economic growth. The net biomass growing potential on the global land surface is 10 times more than the global food, feed, fiber, and energy demands. This study investigates whether the developing countries have sufficient land resource to meet the projected energy demand towards 2035 by planting energy crops on surplus agricultural land after food and feed production. The annual yields of four commonly grown energy crops specifically jatropha, switchgrass, miscanthus, and willow have been used to make scenarios and estimate land requirements against each scenario. This paper first performs literature reviews on the availability of land resource, past and future trends in land use changes, demand of lands for food production, and potential expansion of croplands. The energy demands towards 2035 are compiled from energy scenarios derived by the International Energy Agency (IEA) and the British Petroleum (BP). This paper also reviewed bio-physiological characteristics of these energy crops to determine whether they are cultivable under tropical climatic conditions in developing regions. This paper found that projected energy demand through 2035 in developing regions could be provided by energy crops grown on a portion of surplus croplands or upgraded grasslands (27% and 22% respectively for miscanthus scenario). Sustainable land management practices, improved agricultural productivity, and adopting suitable energy crops cultivation can potentially supply increasing energy demands.
Article
Full-text available
Most prior studies have found that substituting biofuels for gasoline will reduce greenhouse gases because biofuels sequester carbon through the growth of the feedstock. These analyses have failed to count the carbon emissions that occur as farmers worldwide respond to higher prices and convert forest and grassland to new cropland to replace the grain (or cropland) diverted to biofuels. By using a worldwide agricultural model to estimate emissions from land-use change, we found that corn-based ethanol, instead of producing a 20% savings, nearly doubles greenhouse emissions over 30 years and increases greenhouse gases for 167 years. Biofuels from switchgrass, if grown on U.S. corn lands, increase emissions by 50%. This result raises concerns about large biofuel mandates and highlights the value of using waste products.
Article
Full-text available
A model for forage yield with adequate details for leaf area, biomass, nutrients, and hydrology would be valuable for making management decisions. The objectives of this study were to develop Alamo switchgrass (Panicum virgatum L.) parameters for the Agricultural Land Management Alternatives with Numerical Assessment Criteria (ALMANAC) model and demonstrate its accuracy across a wide range of environments. Derived plant parameters included potential leaf area index (LAI), potential biomass growth per unit intercepted light, optimum nutrient concentrations, and growth responses to temperature. The model's simulated yields accounted for 79% of the variability in measured yields for one-cut and two-cut harvest systems from six diverse sites in Texas in 1993 and 1994. Simulated yields for three locations differed in sensitivity to potential LAI, heat units to maturity, radiation use efficiency (RUE), and soil depth. The ALMANAC model shows promise as a management tool for this important forage and bioenergy crop.
Article
Full-text available
Switchgrass (Panicum virgatum L.) has been planted on land en- rolled in the Conservation Reserve Program (CRP). Management strategies for conversion of this land from CRP to biomass energy require evaluation. Objectives of this study were to: (i) determine the effect of harvest timing and N rate on biomass production and char- acteristics of switchgrass land enrolled in or managed similarly to CRP and (ii) evaluate the impact of harvest management on species com- position and persistence. Five N rates (spring applications of 0, 56, 112, and 224 kg ha 21 and 224 kg ha 21 split between spring and post- harvest) and two harvest timings (anthesis and post-killing frost) were applied to plots from 2001 to 2003 at three South Dakota locations. Harvestingafterakillingfrostproducedhighertotalyieldsandimproved switchgrass persistence compared with anthesis harvests. The concen- trationofneutraldetergentfiber(NDF),aciddetergentfiber(ADF),and acid detergent lignin (ADL) increased between anthesis and killing- frost harvests, while total nitrogen (TN) and ash decreased. Nitrogen applied at 56 kg ha 21 increased total biomass without affecting switch- grass persistence, but therewas no additional benefit with N above 56 kg ha 21 .Harvestinglong-establishedswitchgrassstandsonceperyearaftera killing frost and applying N at 56 kg ha 21 was an effective system for switchgrass biomass production and persistence on land enrolled in or managed similarly to CRP in South Dakota.
Article
Full-text available
Fundamental to deriving a sustainable supply of cellulosic feedstock for an emerging biofuels industry is understanding how biomass yield varies as a function of crop management, climate, and soils. Here we focus on the perennial switchgrass (Panicum virgatum L.) and compile a database that contains 1190 observations of yield from 39 field trials conducted across the United States. Data include site location, stand age, plot size, cultivar, crop management, biomass yield, temperature, precipitation, and information on land quality. Statistical analysis revealed the major sources of variation in yield. Frequency distributions of yield for upland and lowland ecotypes were unimodal, with mean ({+-}SD) biomass yields of 8.7 {+-} 4.2 and 12.9 {+-} 5.9 Mg ha-1 for the two ecotypes, respectively. We looked for, but did not find, bias toward higher yields associated with small plots or preferential establishment of stands on high quality lands. A parametric yield model was fit to the data and accounted for one-third of the total observed variation in biomass yields, with an equal contribution of growing season precipitation, annual temperature, N fertilization, and ecotype. The model was used to predict yield across the continental United States. Mapped output was consistent with the natural range of switchgrass and, as expected, yields were shown to be limited by precipitation west of the Great Plains. Future studies should extend the geographic distribution of field trials and thus improve our understanding of biomass production as a function of soil, climate, and crop management for promising biofuels such as switchgrass.
Article
Full-text available
Switchgrass (Panicum virgatum L.) is currently being evaluated as a raw material for producing fuel, chemicals, and electricity. Switchgrass biomass is bound by the growing environment that includes fertility. More information is needed on sustainable switchgrass production as influenced by nitrogen fertility and harvest management. Field experiments were initiated at Chickasha and Perkins, OK in 1996 and 1998, respectively to evaluate switchgrass response to applied nitrogen (N) at rates of 0, 112, 224, 448, and 896 kg ha. In addition, harvest frequency and time of N application were evaluated. Yield maximums and the greatest N, potassium (K), phosphorus (P), and sulfur (S) uptake values were achieved with 448 kg N ha applied all in April and harvested three times. In fact, harvest frequency was the most important factor affecting yields over the course of these studies with average dry matter yields of 16.3, 14.7, and 12.9 Mg ha yr for three, two, and one harvest yr, respectively. No significant change in soil organic carbon was detected over time. Although dry matter yields were found to decline with time, total N uptake did not. Forage N concentration was found to be greater in later years, thus increasing production costs. While yields were highest (18.0 Mg ha) with 448 kg N ha applied all in April and three harvests, applying 0 N and harvesting three times produced almost as much total biomass (16.9 Mg ha). This limited response to N is possibly explained by the evolution of switchgrass under low N conditions. Increasing forage concentrations of K, magnesium (Mg), P, and S were noted with increasing yields, indicating a potential for response to these nutrients. Contribution from the Oklahoma Agricultural Experiment Station.
Article
Full-text available
Management practices for biomass production of bioenergy grasses may differ from management for forage. Our objective was to determine the yield and stand responses of 'Alamo' switchgrass (Panicum virgatum L.) to N and P fertilization as affected by row spacing. A combination of five rates each of N and P were applied to plots during 1992 to 1998 at Stephenville, TX and 1993 to 1995 at Beeville, TX. Three row-spacing treatments were applied as subplots. Biomass production was determined each year with a single harvest in late summer. Tiller density and tiller mass were measured during 1993 to 1996 at Stephenville. Biomass production was not influenced by the addition of P. Biomass production response to N at Beeville was greater in narrow rows than wide rows during the establishment year only. Biomass production responses to N were quadratic in 5 of 7 yr at Stephenville and linear at Beeville. A maximum yield of 22.5 Mg ha-1 occurred during 1995 at Stephenville at 168 kg N ha-1 Lodging occurred at both locations but only at the 224 kg N ha-1 rate. Tiller density and mass increased as row width increased. Tiller mass also increased with increasing N fertility at Stephenville. This response was more important in determining biomass production than was tiller density. Average biomass production at 168 kg N ha-1 yr-1 was 14.5 and 10.7 Mg ha-1 yr-1 at Stephenville and Beeville, respectively. Biomass production without applied N tended to decline over the years. Our data indicated that switchgrass biomass production is sustainable at Stephenville only with the application of at least 168 kg N ha-1 yr-1 but P application and row spacing are not crucial.
Article
Full-text available
Warm-season grasses are increasingly being cultivated in North America for summer forage and biomass production. The cooler temperatures and shorter growing seasons typical of Canadian production areas, are major limiting factors to warm-season grass production in these areas. This research assessed the morphological development and relationship of growing degree-days (GDD) to plant morphology and tiller characteristics in nine cultivars of switchgrass (Panicum virgatum L.; Blackwell, Cave-in-Rock, Dakota, Forestburg, Pathfinder, Shelter, Sunburst, ND3743, and New Jersey 50) and in 'Niagara' big bluestem (Andropogon gerardii Vitman). The study was conducted for three years on a St. Bernard sandy clay loam (Typic Hapludalf) in southwestern Quebec. Stand cover, plant morphology, tiller number, height, and diameter, and leaf number per tiller were all assessed during the season. All entries persisted through the three years of the study and showed increases in tiller number (from an average of 565 to 683 m-2) from one year to the next. Dakota, Cave-in-Rock, and Shelter switchgrass had the highest ground cover ratings after three years (85, 85, and 84%, respectively). Dakota, ND3743, and Forestburg were early maturing; New Jersey 50 was the latest in maturity. Niagara big bluestem had the tallest tillers (183 cm) and largest rates of increase in height (2.8 cm d-1), followed by Cave-in-Rock (2.0 cm d-1) and Blackwell (1.9 cm d-1). The shortest tillers were recorded for Dakota (111 cm) and ND3743 (118 cm). Changes in leaf number per tiller with GDD were best described by quadratic (r2 = 0.80-0.97). These models were stable over two regression models years. Cultivars varied in the number of GDD required for maximum number of leaves per tiller, with later-maturing cultivars generally requiring greater GDD accumulation. These data indicate that warm-season grasses can be grown successfully in eastern Canada.
Article
Full-text available
Switchgrass (Panicum virgatum L.)—a perennial, warm-season (C4) species—evolved across North America into multiple, divergent populations. The resulting natural variation within the species presents considerable morphological diversity and a wide range of adaptation. The species was adopted as a crop—initially as a forage—only in the last 50 yr. Its potential uses have recently been expanded to include biofuels. Management of switchgrass for biofuels is informed by an understanding of the plant's biology. Successful establishment requires attention to seed dormancy and weed control as well as proper depth and date of planting. The plant's growth rate is closely tied to temperature, but timing of reproductive development is linked to photoperiod. Accordingly, the period of vegetative growth can be extended by planting lower-latitude cultivars at higher latitudes. This strategy may provide a yield advantage, but cold tolerance can become limiting. Switchgrass is thrifty in its use of applied N; it appears able to obtain N from sources that other crops cannot tap. The N removed in harvested biomass is often greater than the amount of N applied. In areas with sufficient rainfall, sustainable yields of ∼15 Mg ha yr may be achievable by applying ∼50 kg N ha yr. Harvesting biomass once per season—after plants have senesced and translocated N into perennial tissues—appears to allow plants to maintain an internal N reserve. Two harvests yr may increase yields in some cultivars, but a single annual harvest maximizes yields in many cases. If two harvests are taken, more N must be applied to compensate for the N removed in the midseason harvest. Taking more than two harvests yr often adversely affects long-term productivity and persistence. Switchgrass has potential as a renewable fuel source, but such use will likely require large infrastructural changes; and, even at maximum output, such systems could not provide the energy currently being derived from fossil fuels.
Article
Full-text available
The task of modeling the distribution of a large number of tree species under future climate scenarios presents unique challenges. First, the model must be robust enough to handle climate data outside the current range without producing unacceptable instability in the output. In addition, the technique should have automatic search mechanisms built in to select the most appropriate values for input model parameters for each species so that minimal effort is required when these parameters are fine-tuned for individual tree species. We evaluated four statistical models—Regression Tree Analysis (RTA), Bagging Trees (BT), Random Forests (RF), and Multivariate Adaptive Regression Splines (MARS)—for predictive vegetation mapping under current and future climate scenarios according to the Canadian Climate Centre global circulation model. To test, we applied these techniques to four tree species common in the eastern United States: loblolly pine (Pinus taeda), sugar maple (Acer saccharum), American beech (Fagus grandifolia), and white oak (Quercus alba). When the four techniques were assessed with Kappa and fuzzy Kappa statistics, RF and BT were superior in reproducing current importance value (a measure of basal area in addition to abundance) distributions for the four tree species, as derived from approximately 100,000 USDA Forest Service’s Forest Inventory and Analysis plots. Future estimates of suitable habitat after climate change were visually more reasonable with BT and RF, with slightly better performance by RF as assessed by Kappa statistics, correlation estimates, and spatial distribution of importance values. Although RTA did not perform as well as BT and RF, it provided interpretive models for species whose distributions were captured well by our current set of predictors. MARS was adequate for predicting current distributions but unacceptable for future climate. We consider RTA, BT, and RF modeling approaches, especially when used together to take advantage of their individual strengths, to be robust for predictive mapping and recommend their inclusion in the ecological toolbox.
Article
Full-text available
Greenhouse gas abatement policies will increase the demand for renewable sources of energy, including bioenergy. In combination with a global growing demand for food, this could lead to a food-fuel competition for bio-productive land. Proponents of bioenergy have suggested that energy crop plantations may be established on less productive land as a way of avoiding this potential food-fuel competition. However, many of these suggestions have been made without any underlying economic analysis. In this paper, we develop a long-term economic optimization model (LUCEA) of the U.S. agricultural and energy system to analyze this possible competition for land and to examine the link between carbon prices, the energy system dynamics and the effect of the land competition on food prices. Our results indicate that bioenergy plantations will be competitive on cropland already at carbon taxes about US $20/ton C. As the carbon tax increases, food prices more than double compared to the reference scenario in which there is no climate policy. Further, bioenergy plantations appropriate significant areas of both cropland and grazing land. In model runs where we have limited the amount of grazing land that can be used for bioenergy to what many analysts consider the upper limit, most of the bioenergy plantations are established on cropland. Under the assumption that more grazing land can be used, large areas of bioenergy plantations are established on grazing land, despite the fact that yields are assumed to be much lower (less than half) than on crop land. It should be noted that this allocation on grazing land takes place as a result of a competition between food and bioenergy production and not because of lack of it. The estimated increase in food prices is largely unaffected by how much grazing land can be used for bioenergy production.
Article
Full-text available
Switchgrass (Panicum virgatum L.) is being developed into a perennial, herbaceous, cellulosic feedstock crop for use in temperate regions of the USA. Information on spatial and temporal variation for stands and biomass yield among and within fields in large agroecoregions is not available. Spatial and temporal variation information is needed to model feedstock availability for biorefineries. In this 5-yr study, the spatial and temporal variation for biomass yield and stands was determined among and within 10 fields located in North Dakota, South Dakota, and Nebraska. Switchgrass fields were managed for bioenergy from 2000 to 2004 for the Nebraska locations and 2001 to 2005 for the South Dakota and North Dakota locations. A global positioning system (GPS) receiver was used to repeatedly measure within field quadrat sites for switchgrass stands using frequency grid (2.25 m2) measurements in June for five growing seasons. Sixteen quadrat (≥1 m2) yield samples were taken post-killing frost in the establishment year and in August in subsequent years at each location. Topographic within field effects on switchgrass stand frequency and biomass yields were largely insignificant. Stands tended to increase from establishment year to year 3 and then begin to plateau. Weather factors, which were the principal source of temporal variation, were more important in switchgrass yield variation than on switchgrass stand frequencies. Temporal standard deviations for yield were higher on quadrat sites with higher than average field means while temporal standard deviations were smaller in quadrat sites that had lower than average field means at six locations. In the Northern Great Plains agroecoregion, there is greater temporal and spatial variation for switchgrass biomass yields among fields than within fields. Results indicate that modeling feedstock availability for a biorefinery can be based on field scale yields.
Article
Full-text available
Switchgrass (Panicum virgatum L.) shows potential as a sustainable herbaceous energy crop from which a renewable source of transportation fuel and/or biomass-generated electricity could be derived. In 1992, a new research program focused on developing switchgrass as a biomass energy feedstock was initiated by the U.S. Department of Energy in five of the southern United States. The multifaceted, multi-institution research addresses breeding for improved biomass yields, regional field tests, cultural practices, physiology and tissue culture. Recent progress is highlighted in this paper. Preliminary results from the breeding program indicate that recurrent restricted phenotypic selection could lead to development of new cultivars. A technique for regenerating switchgrass plants via tissue culture has been proven and new populations of regenerated plants have been established in the field. Performance trials at three regional cultivar testing centers in Virginia, Alabama and Texas have shown that ‘Alamo’ switchgrass has higher biomass yield and broader adaptability than other cultivars tested. Research on management practices designed to maximize biomass yield has shown that multiple harvests of switchgrass may reduce total seasonal yields in some instances and that responses to fertilizer inputs vary with the environment. Seed dormancy often retards rapid establishment of competitive stands of switchgrass. Our research has indicated that seed dormancy can be modified, resulting in increased seed germination and a greater number of switchgrass plants. Research on the physiology of switchgrass has shown that lowland and upland ecotypes differ in photosynthetic rate but not in respiration rate. Findings in each of these areas can contribute to development of switchgrass as a sustainable bioenergy crop. Future research will address molecular biology techniques for exploiting genetic variation, explore canopy architecture and carbon allocation patterns affecting biomass yield, elucidate key factors in successful establishment of switchgrass and provide technology transfer that facilitates scale-up of switchgrass production for commercial energy production.
Article
Full-text available
The utilization of energy crops produced on American farms as a source of renewable fuels is a con-cept with great relevance to current ecological and economic issues at both national and global scales. De-velopment of a significant national capacity to utilize perennial forage crops, such as switchgrass (Panicum virgatum, L., Poaceae) as biofuels could benefit our agricultural economy by providing an important new source of income for farmers. In addition energy production from perennial cropping systems, which are compatible with conventional farming practices, would help reduce degradation of agricultural soils, lower national dependence on foreign oil supplies, and reduce emissions of greenhouse gases and toxic pollutants to the atmosphere (McLaughlin 1998). Interestingly, on-farm energy production is a very old concept, extending back to 19th century America when both transportation and work on the farm were powered by approximately 27 million draft animals and fueled by 34 million hectares of grasslands (Vogel 1996). Today a new form of energy production is envisioned for some of this same area. The method of energy production is exactly the same, solar energy captured in photosynthesis, but the subsequent modes of energy conversion are vastly different, leading to the production of electricity, transportation fuels, and chemicals from the renewable feedstocks. While energy prices in the United States are among the cheapest in the world, the issues of high depen-dency on imported oil, the uncertainties of maintaining stable supplies of imported oil from finite reserves, and the environmental costs associated with mining, processing, and combusting fossil fuels have been im-portant drivers in the search for cleaner burning fuels that can be produced and renewed from the landscape. At present biomass and bioenergy combine provide only about 4% of the total primary energy used in the US (Overend 1997). By contrast, imported oil accounts for approximately 44% of the foreign trade deficit in the US and about 45% of the total annual US oil consumption of 34 quads (1 quad = 10, Btu, Lynd et al. 1991). The 22 quads of oil consumed by transportation represents approximately 25% of all energy use in the US and exceeds total oil imports to the US by about 50%. This oil has environmental and social costs, which go well beyond the purchase price of around $15 per barrel. Renewable energy from biomass has the potential to reduce dependency on fossil fuels, though not to totally replace them. Realizing this potential will require the simultaneous development of high yielding biomass production systems and bioconversion technologies that efficiently convert biomass energy into the forms of energy and chemicals usable by industry. The endpoint criterion for success is economic gain for both agricultural and industrial sectors at reduced environmental cost and reduced political risk. This paper reviews progress made in a program of research aimed at evaluating and developing a perennial forage crop, switchgrass as a regional bioenergy crop. We will highlight here aspects of research progress that most closely relate to the issues that will determine when and how extensively switchgrass is used in commercial bioenergy production.
Article
Full-text available
Optimizing feedstock production from switchgrass (Panicum virgatum L.) requires careful matching of genotype to environment, especially for southern U.S. regions. Nine genotypes from four combinations of ecotype and morphological type were harvested once yearly in autumn for 3 or 4 yr at five locations across Texas, Arkansas, and Louisiana that varied in latitude and precipitation. Genotypes were evaluated for dry matter yield (DMY), plant density, tiller density, lodging, and rust (caused by Puccinia spp.) infection. Genotype x environment (GxE) interactions were identified for most traits. Biomass yield of all genotypes tended to increase with latitude, but lowland morphological types may have been more sensitive than upland morphological types to differences in moisture availability. Yield (5.82 vs. 14.97 Mg ha(-1), respectively) and persistence (final stand density, 3.99 vs. 5.96 plants m(-2)) were lower for upland than for lowland genotypes, particularly at higher rainfall and more southern sites. Lowland genotypes were often able to compensate for stand thinning by increasing individual plant size, but upland genotypes were not. Lodging and rust scores were higher for upland than for lowland genotypes. Yield (13.65 vs. 9.75 Mg ha(-1)) and final plant density (5.58 vs. 4.95 plants m(-2)) were higher for southern than northern ecotypes. The southern-lowland combination exhibited the best yield and persistence over the study region, and genotypes within this group exhibited variability in yield among sites. Therefore, development of switchgrass cultivars for biomass production in the southern USA should focus on the southern-lowland genotypes.
Article
Full-text available
Increased demand for corn grain as an ethanol feedstock is altering U.S. agricultural landscapes and the ecosystem services they provide. From 2006 to 2007, corn acreage increased 19% nationally, resulting in reduced crop diversity in many areas. Biological control of insects is an ecosystem service that is strongly influenced by local landscape structure. Here, we estimate the value of natural biological control of the soybean aphid, a major pest in agricultural landscapes, and the economic impacts of reduced biocontrol caused by increased corn production in 4 U.S. states (Iowa, Michigan, Minnesota, and Wisconsin). For producers who use an integrated pest management strategy including insecticides as needed, natural suppression of soybean aphid in soybean is worth an average of $33 ha(-1). At 2007-2008 prices these services are worth at least $239 million y(-1) in these 4 states. Recent biofuel-driven growth in corn planting results in lower landscape diversity, altering the supply of aphid natural enemies to soybean fields and reducing biocontrol services by 24%. This loss of biocontrol services cost soybean producers in these states an estimated $58 million y(-1) in reduced yield and increased pesticide use. For producers who rely solely on biological control, the value of lost services is much greater. These findings from a single pest in 1 crop suggest that the value of biocontrol services to the U.S. economy may be underestimated. Furthermore, we suggest that development of cellulosic ethanol production processes that use a variety of feedstocks could foster increased diversity in agricultural landscapes and enhance arthropod-mediated ecosystem services.
Book
The first edition of this book has established itself as one of the leading references on generalized additive models (GAMs), and the only book on the topic to be introductory in nature with a wealth of practical examples and software implementation. It is self-contained, providing the necessary background in linear models, linear mixed models, and generalized linear models (GLMs), before presenting a balanced treatment of the theory and applications of GAMs and related models. The author bases his approach on a framework of penalized regression splines, and while firmly focused on the practical aspects of GAMs, discussions include fairly full explanations of the theory underlying the methods. Use of R software helps explain the theory and illustrates the practical application of the methodology. Each chapter contains an extensive set of exercises, with solutions in an appendix or in the book’s R data package gamair, to enable use as a course text or for self-study.
Article
Mass field collections and plants grown under various environmental conditions were analyzed to determine the extent of the morphological differences occurring between upland and lowland types of switchgrass, Panicum virgatum L., in McClain County, Oklahoma. The bases for these differences were also investigated. A comparison of field populations revealed differences in clonal habit, in the size of clones, and in the gross morphology of their vegetative organs. Lowland plants were larger in most aspects than those of upland areas although the differences between them were somewhat nullified by variations within each of the two types. Genetically controlled morphological differences became evident when the effects of environmental differences were eliminated. Although upland and lowland switchgrasses were somewhat modified under common greenhouse conditions, they exhibited essentially the same morphological differences as did field populations. A physiological difference in water requirement also occurred between the two types. Lowland switchgrass grew best under flooded conditions, whereas upland plants reached their greatest development under more moderate conditions of soil water. The morphological differences between field populations occurred largely, therefore, as a result of the morphological expression of this physiological difference in water requirement which was superimposed on lesser genetic differences. Upland and lowland switchgrasses also differed in their requirement for nitrogen, the latter having a somewhat lower requirement than the former. This factor probably did not contribute significantly to the existing morphological differences. The results of reciprocal transplants substantiated these conclusions. Plants of the lowland type were tetraploids. Hexaploids and octoploids occurred in the upland with hexaploids being the most common.
Article
Switchgrass (Panicum virgatum L.) has been identified as a potential biofuel crop for the northern Great Plains region of the USA. Biomass yield and survival percentage in western North Dakota were measured for 3 yr at three field sites, and plant development was monitored for 2 yr at one site to determine adaptation and stability of performance for eight diverse switchgrass cultivars and experimental strains. Harvest treatments were single annual cuttings in mid-August or mid-September. Except for ‘Dacotah’, ND3743, and ‘Sunburst’, all other entries originated greater than 500 km south of the evaluation sites and were subject to winter injury. Sunburst, from southern South Dakota, ranked first or second in biomass yield in all environments and was the top-yielding entry in all environments in the third production year, a drought year at all sites. ‘Trailblazer’ ranked first, second, or third in biomass yield in all environments while yield ranking of the other entries was not consistent. Genotype × environment interactions occurred for biomass yield and would be expected based on the wide range in origin among the eight populations. Survival percentage was equal for the two harvest dates, but all eight populations averaged greater biomass yields at the mid-September (5.98 Mg/ha) than the mid-August (5.51 Mg/ha) harvest. Biomass yield of Sunburst at the site with the greatest yield potential ranged from 3.20 Mg/ha in a drought year to 12.48 Mg/ha in a year with above-average precipitation. Biomass yield of adapted switchgrass cultivars fluctuated widely in western North Dakota, depending in large part on available soil water.
Article
Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The generalization error for forests converges a.s. to a limit as the number of trees in the forest becomes large. The generalization error of a forest of tree classifiers depends on the strength of the individual trees in the forest and the correlation between them. Using a random selection of features to split each node yields error rates that compare favorably to Adaboost (Y. Freund & R. Schapire, Machine Learning: Proceedings of the Thirteenth International conference, ***, 148–156), but are more robust with respect to noise. Internal estimates monitor error, strength, and correlation and these are used to show the response to increasing the number of features used in the splitting. Internal estimates are also used to measure variable importance. These ideas are also applicable to regression.
Article
Growing demand for alternative energy sources has contributed to increased biofuel production, but the effects on biodiversity of land-use change to biofuel crops remain unclear. Using a meta-analysis for crops being used or considered in the US, we find that vertebrate diversity and abundance are generally lower in biofuel crop habitats relative to the non-crop habitats that these crops may replace. Diversity effects are greater for corn than for pine and poplar, and birds of conservation concern experience greater negative effects from corn than species of less concern. Yet conversion of row-crop fields to grasslands dedicated to biofuels could increase local diversity and abundance of birds. To minimize impacts of biofuel crops on biodiversity, we recommend management practices that reduce chemical inputs, increase heterogeneity within fields, and delay harvests until bird breeding has ceased. We encourage research that will move us toward a sustainable biofuels economy, including the use of native plants, development of robust environmental criteria for evaluating biofuel crops, and integrated cost-benefit analysis of potential land-use change.
Article
The objective of this study was to determine the yield, chemical composition and production costs of eight switchgrass (Panicum virgatum L.) varieties grown in the southeastern USA. Plots were seeded during the summer of 1988 and were harvested twice in each of the two subsequent years. Samples taken in 1990 were analyzed for nitrogen and several fiber components. Total 1989 dry biomass yields of “Alamo” and “Kanlow” were 17.5 and 13.8 Mg ha−1, respectively. Yields provided by the other six varieties averaged 8.6 Mg ha−1. In 1990, total yield almost doubled for “Alamo” (34.6 Mg ha−1) and increased 68% for “Kanlow” (23.2 Mg ha−1), over that achieved the previous year. Average yields of the other varieties increased slightly to 9.4 Mg ha−1. Variation in chemical composition among varieties was low. In 1989, production costs per Mg DM ranged from $19.70 and $24.91 for “Alamo” and “Kanlow”, respectively, to an average of $40.28 for the other six varieties. In 1990, costs fell to $9.94 and $14.82 Mg−1 for “Alamo” and “Kanlow”, and to $37.65 Mg−1 for the other varieties.