- Email: [email protected]
Potential for dual-purpose maize varieties to meet changing maize demands: Synthesis
Field Crops Research 153 (2013) 107–112
Contents lists available at ScienceDirect
Field Crops Research journal homepage: www.elsevier.com/locate/fcr
Potential for dual-purpose maize varieties to meet changing maize demands: Synthesis Michael Blümmel a,∗ , Elaine Grings b , Olaf Erenstein c a b c
International Livestock Research Institute (ILRI), c/o ICRISAT, Patancheru 502 324, Andhra Pradesh, India South Dakota State University, Department of Animal Science, Box 2170, Brookings, SD 57007, USA International Maize and Wheat Improvement Centre (CIMMYT), c/o ILRI, PO Box 5689, Addis Ababa, Ethiopia
1. Introduction Maize—or corn (Zea mays L.)—now is the most important global cereal in terms of production reflecting its versatility in use, including human food, animal feed and fodder, industrial products and biofuel. Most uses revolve around maize grain as the primary product, although whole plant utilization for silage is also common in industrialized agriculture (e.g. Klopfenstein et al., 2013). Despite being a versatile crop, maize production and maize breeding efforts over time have typically had a single-purpose orientation. For instance, maize breeding has focused on overcoming biotic and abiotic stresses so as to generate high yielding, stress-tolerant and widely-adapted maize varieties through judicious combination of conventional and molecular breeding approaches (Muttoni et al., 2013; Shiferaw et al., 2011). Even smallholders within mixed maize–livestock systems typically focus on maize grain yield (De Groote et al., 2013), with maize stover as additional byproduct and benefit. Although farmers may still try to increase fodder off-take, they still try to minimize maize grain yield loss (Byerlee et al., 1989; Lukuyu et al., 2013). Critical to the development and acceptance of dual-purpose maize are the potential trade-offs between grain and fodder both in terms of quantity and quality. Single-purpose maize can focus on either grain or fodder, with grain reflecting partial utilization of maize biomass against potentially full (above ground) biomass utilization for fodder (as in case of silage). Dual-purpose maize focuses on both grain and fodder (particularly stover), and typically the full (above ground) biomass. Optimization for one trait (grain) is inherently easier than for two traits (grain and fodder), and may be pursued in instances where the fodder/stover value is negligible. However, the widespread use of maize stover suggests it indeed has value, although typically less than the maize grain. Given that both traits have value and improvement for each is likely subject to diminishing returns, dual-purpose breeding could potentially make economic sense. Furthermore, similar to many cereals, maize varieties appear to have a wide variety of (grain
∗ Corresponding author. Tel.: +91 40 30713653; cell: +91 9866315505. E-mail address: [email protected] (M. Blümmel). 0378-4290/$ – see front matter © 2013 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.fcr.2013.10.006
and stover) yields and stover quality, and most interestingly, there are prospects within the range of stover quality to increase fodder quality without compromising grain yield. It is this potential of dual-purpose varieties that has reignited research interest and some of the research underlying this special issue. Indeed, despite earlier skepticism only a decade ago, substantial progress has been made in developing dual-purpose maize options for both grain and fodder purposes as reflected in the papers contained in this special issue (Erenstein et al., 2013). This paper synthesizes the key findings presented in 12 papers in this Special Issue around the potential for dual-purpose maize varieties to meet changing maize demands. We summarize the key findings around three thematic areas: (1) demand for dual-purpose maize cultivars and associated targeting domains; (2) quality traits, whole plant utilization and phenotyping; and (3) exploiting trait variation for maize improvement. In a final section we provide some of the lessons learned and guidance for further dual-purpose maize research and development (R&D). 2. Demand for dual-purpose maize cultivars and associated targeting domains While plant breeders have the technical capability to alter maize plant characteristics, adoption of new technologies or varieties requires an understanding of farmers’ preferences for crop characteristics. Issues of trade-offs for other values, such as soil improvement and fuel, market forces, disease resistance, and availability of alternative inexpensive feed resources affect farmers’ willingness to accept new varieties with improved stover quantity and fodder quality. Four papers discussed the potential demand for dual-purpose maize cultivars in North and Meso-America and Eastern and Southern Africa and the potential trade-offs in the use of maize plant parts for food, feed, fodder and soil improvements. Ways are discussed to better match specifically identified and desired traits to farming and production systems. System contexts are provided by agro-ecologies, cropping pattern, human and livestock densities, importance of livestock and alternative feed and fodder resources in the form of rangelands and common property resources to help delineate targeting domains for dual-purpose maize cultivars.
108
M. Blümmel et al. / Field Crops Research 153 (2013) 107–112
For the United States of America (USA), Klopfenstein et al. (2013) review the history of maize (corn) use and the current conditions impacting both maize production and its use as livestock feed. Starting in the middle of the 20th century, plentiful maize grain was used increasingly as livestock feed. Changes in maize use for ethanol production have altered demands for grain and resulted in new by-products that provide valuable feed sources and that have been used to replace maize in livestock feeding in the U.S.A. Maize has been used as a roughage source in two primary ways in the U.S.A., either as silage or through use of the stover remaining after the grain harvest (used ex situ or in situ). Crop breeding strategies have primarily focused on grain but breeding programs specifically for silage also exist. Klopfenstein et al. do report on one study in which varietal differences in morphological traits are compared in stovers, which showed some variation. They also discuss the value of grazed maize stover as fodder for beef cattle and suggest that maize stover can be well utilized as a roughage source for both gestating cows and growing calves as long as adequate protein is provided to meet the needs of rumen microbes. Research is cited that shows gains of up to 1 kg/d for calves supplemented with up to 3.5 kg of distillers grains while grazing maize stalks. These authors also discuss research that indicates that allowing cattle to graze maize stover during winter does not affect subsequent maize grain yields. Hellin et al. (2013) take us to Mexico, the centre of origin of maize and U.S.A.’s southern neighbour. They use farmer group/community surveys in three contrasting agro-ecologies (semi-arid, temperate highland and tropical sub-humid) to explore maize stover use. Mixed maize–livestock smallholder farming systems prevail, with maize grain produced for home consumption and the market and maize stover as an important by-product, primarily as feed. Although in situ grazing is found in all three study sites, it represented the bulk of stover use in only one site (with an estimated 70% of stover in the sub-humid tropics), with ex situ feed dominating in the other two sites (estimated >80%). Maize stover commercialization was limited and mainly restricted to households with no livestock and often within the immediate local context. They also explore potential trade-offs in stover use, particularly its use as feed against its potential use as mulch (soil cover) to manage soil health within the context of conservation agriculture. They report that farmers are generally hesitant to adopt conservation agricultural practices that require the retention of stover as mulch, as this competes with their livestock feed needs and purchased feed is expensive. The authors go on to explore a portfolio of options to reduce these trade-offs, including partial residue retention, cover and feed crops and sustainable intensification. Promising (Muttoni et al., 2013) but yet to be explored further in the context of Mexico, are investments in generating dual-purpose (grain-fodder) maize varieties. De Groote et al. (2013) investigated the potential demand for dual purpose maize in Ethiopia and Tanzania using participatory rural appraisals, farm household surveys and formal farmer evaluations of maize varieties, the latter focusing on commercially available material that were planted side-by-side in trial sites in the study areas as purposively developed dual-purpose maize varieties are not yet on the market. They show that maize stover is an important element of livestock feed in all the study areas. Desirable traits farmers consistently mentioned were related to grain yield and pest resistance, ease of cooking and taste characteristics and stover yield and fodder quality. Analysis of adoption patterns of existing maize varieties by farm households shows that varieties that score well on feed characteristics have a higher probability of being adopted. The authors concluded that there is a demand for varieties with superior stover quantity and fodder quality as long as grain yield and consumer preferences were not compromised. They further hypothesise that such varieties have the potential to increase the
productivity of maize-livestock systems and the income of these farmers, provided these varieties are taken up in the seed portfolio of the emerging seed sector in the sub-region. Homann-Kee Tui et al. (2013) explored the use of a multilevel approach and associated data sources to assess the potential of dual-purpose maize cultivars in Southern Africa. The authors construct recommendation domains for dual-purpose maize using maize mega-environments and demand estimates derived from livestock and human population densities and potential feed supply, including biomass contributions from range and croplands. These estimates were subjected to ground truthing through survey data collected from 480 households in contrasting sites in Malawi, Mozambique and Zimbabwe. In addition the authors’ reported on maize varieties, both landraces and advanced hybrids, for variations in grain and stover yield and fodder quality traits. Their study showed that where livestock densities were high and feed resources from rangeland limited, maize cultivars with superior stover yield and fodder quality can have substantial impact on livestock productivity. Stover with higher fodder quality could provide sufficient energy for providing livestock maintenance requirements and support about 200 gram of live weight gain daily. The authors further proposed an approach to target grain-fodder maize cultivars to demand domains for either stover quantity or stover fodder quality based on maize stover intake estimations derived from Ravi et al. (2013). Homann-Kee Tui et al. complement and build on Notenbaert et al. (2013) who focused on eastern Africa to develop similar demand domains for dual-purpose maize as part of an integrated targeting approach based on maize mega environments, livestock numbers, population densities and alternative feed resources. Research such as that reported in these four papers reinforces the need to evaluate the specific contexts for technology adoption. For example maize improvement/cultivar choice can increase stover quantity, stover quality or perhaps both. It is important to realize that from a livestock nutrition viewpoint, an increase in stover quantity is only useful (unless making stover cheaper) if livestock can respond with increased intake, which is stover quality dependent.
3. Quality traits, whole plant utilization and phenotyping Evaluating the nutritional fodder value for livestock in vivo can be a complex and time consuming process. Additionally, livestock may be fed complex diets in which feedstuffs impact the utilization of one another. Nutritionists have long worked on developing laboratory indicators that well reflect animal response. Validation of rapid laboratory measures is key to evaluation of large collections of germplasm for screening. A set of four papers looked into the variations in grain and stover traits in existing maize cultivars, their meaning for whole maize plant optimization, and laboratory phenotyping capability. These papers thus investigated exploitable variations in a range of food-feed-fodder traits in existing and widely used cultivars, the implication of these traits for livestock productivity and/or tradability of maize stover, and techniques for selecting and predicting specific traits. Anandan et al. (2013) investigated six maize hybrids commonly grown in South Asia for grain and stover yields and for stover morphological (residual green leaf area, plant height, stem diameter and proportions of leaf blade: leaf sheath: stem), chemical (nitrogen (N), neutral (NDF) and acid (ADF) detergent fiber, acid detergent lignin (ADL)) and biological (in vitro organic matter digestibility (IVOMD) and metabolizable energy (ME) content) fodder quality traits. The stovers were also fed to sheep as major part (90%) of their feed. The authors observed significant cultivar differences
M. Blümmel et al. / Field Crops Research 153 (2013) 107–112
109
Table 1 Comparisons between measured digestibility (DMDmeas , %), intake (DMImeas , g/d) and digestible intake (DDMImeas , g/d) and DMDpred , DMIpred and DDMIpred predicted by cross validation procedures based on stover nitrogen (N), neutral (NDF) and acid detergent (ADF) fiber, acid detergent lignin (ADL), in vitro organic matter digestibility (IVOMD) and metabolizable energy (ME) including simple linear relationships between traits and in vivo measurements. Derived from a combination of the data set of Ravi et al. (2013) and Anandan et al. (2013) using observations from 16 maize stovers.
N NDF ADF ADL ME IVOMD
DMDpred versus DMDmeas
DMIpred versus DMImeas
DDMpred I versus DDMImeas
y = 22.8 + 0.60x; [r = 0.75; P = 0.0008; Syx = 1.9] y = 37.4 + 0.34x; [r = 0.53; P = 0.04; Syx = 2.0] y = 19.2 + 0.66x; [r = 0.79; P = 0.0003; Syx = 1.9] y = 28.3 + 0.50x; [r = 0.66; P = 0.005; Syx = 2.1] y = 40.6 + 0.28x; [r = 0.45; P = 0.08; Syx = 2.1] y = 39.4 + 0.3x; [r = 0.47; P = 0.07; Syx = 2.1]
y = 20.5 + 0.13x; [r = 0.24; P = 0.36; Syx = 1.8] y = 18.4 + 0.22x; [r = 0.37; P = 0.16; Syx = 2.0] y = 14.8 + 0.37x; [r = 0.55; P = 0.03; Syx = 2.0] y = 8.7 + 0.63x; [r = 0.78; P = 0.0004; Syx = 1.8] y = 4.2 + 0.82x; [r = 0.90; P < 0.0001; Syx = 1.4] y = 5.3 + 0.77x; [r = 0.87; P < 0.0001; Syx = 1.6]
y = 9.0 + 0.33x; [r = 0.49; P = 0.05; Syx = 1.59] y = 8.8 + 0.34x; [r = 0.51; P = 0.04; Syx = 1.58] y = 5.6 + 0.57x; [r = 0.73; P = 0.001; Syx = 1.47] y = 3.5 + 0.73x; [r = 0.85; P < 0.0001; Syx = 1.27] y = 2.5 + 0.81x; [r = 0.89; P < 0.0001; Syx = 1.1] y = 2.9 + 0.78x; [r = 0.88; P < 0.0001; Syx = 1.2]
Simple linear regression between stover laboratory traits and in vivo measurement in sheep
N NDF ADF ADL ME IVOMD
DMD (%)
DMI g/kg LW
DDMI g/kg LW
y = 50.4 + 7.0x; r = 0.81; P = 0.0001 y = 85.4 − 0.4x; r = 0.66; P = 0.005 y = 79.4 − 0.56x; r = 0.85; P < 0.0001 y = 66.8 − 2.4x; r = 0.76; P = 0.0007 y = 39.3 + 2.3x; r = 0.62; P = 0.009 y = 37 + 0.39x; r = 0.64; P = 0.008
y = 19.4 + 4.4x; r = 0.51; P = 0.04 y = 48.9 − 0.35x; r = 0.59; P = 0.017 y = 41.8 − 0.46x; r = 0.69; P = 0.003 y = 34.4 − 2.6x; r = 0.83; P < 0.0001 y = −2.7 + 3.4x; r = 0.93; P < 0.0001 y = −4.8 + 0.55x; r = 0.90; P < 0.0001
y = 9.4 + 4.3x; r = 0.67; P = 0.005 y = 35.4 − 0.3x; r = 0.67; P = 0.005 y = 29.8 − 0.41x; r = 0.81; P = 0.0001 y = 22.3 − 2.1x; r = 0.89; P < 0.0001 y = −6.4 + 2.6x; r = 0.92; P < 0.0001 y = −8.2 + 0.42x; r = 0.91; P < 0.0001
in important animal response variables such as digestible feed intake and nitrogen balance. For example, nitrogen balances ranged from 0.47 to 1.2 g per day per sheep. No significant relationships (P = 0.32–0.77) were observed between animal response variables and grain yields and the authors identified a cultivar that had high grain and stover yield and for which animal response variables not differ from that of the (numerically) best stover. The authors related the morphological, chemical and biological stover traits to the animal response variables and reported that these were more highly correlated with intake and digestible feed intake than with in vivo digestibility and nitrogen balance. Certain chemical (particularly NDF and ADF) and biological (IVOMD and ME) traits were more closely correlated with the in vivo measurements than were the morphological traits. NDF and ADF were highly negatively related with digestible feed intake (the product of stover digestibility and stover intake), accounting for 90% of the variation therein, while IVOMD and ME were positively associated with digestible feed intake, accounting for 90 and 94% of the variation therein, respectively. The authors designed complete diets from stovers of the lowest and highest IVOMD maize cultivars grown by farmers and observed about 30% greater weight gain in sheep fed the diet with the higher IVOMD stover. The authors suggest that the impact of the difference in IVOMD of about 4% units agreed well with ex ante impact assessment of sorghum and pearl millet stover improvement (Kristjanson and Zerbini, 1999) and fodder market studies with sorghum stover (Blümmel and Rao, 2006) in which a 1% unit improvement in stover digestibility could reap benefits of 5–8% in terms of higher livestock productivity. Ravi et al. (2013) investigated ten maize stovers (including two cultivars also used by Anandan et al., 2013) and confirmed the usefulness of ADF, IVOMD and ME measurements in ranking stover for in vivo digestibility, feed intake and digestible feed intake. For the current synthesis we asked the authors to combine their data sets since the animal experimental facilities used were identical and the different in vivo experimentations were standardized and very similar. Statistical cross-validation procedures were employed on the combined data set where the predicted values were not used for the development of the regression equations (“blind-predictions”). These findings are summarized in Table 1. As in the individual data sets, DMD was less well predicted than DMI and DDMI possibly because the concentrate used as supplement (about 10–15% of the ration) had affected DMD while only maize stover was considered for DMI (Ravi et al., 2013) Prediction of DMD of maize stover-based diets from stover N and ADF may still be meaningful
since the intercepts of their regression equations were relatively low and the correlation between measured and predicted DMD were 0.75–0.78. It seems quite clear that IVOMD and ME are well suited for use in DMI and DDMI predictions of maize stover-based diets (Table 1). Digestibility measurements (in vitro and in vivo) provide an approximation of the proportion of the feed the animal can use. These measures are an easier nutritional concept to grasp for non-animal nutritionists than is ME, which requires advanced understanding of ruminant physiology. Similarly understanding the nutritional implications of variation in NDF, ADF and ADL requires some familiarity with livestock nutritional concepts. As shown by Ramana Reddy et al. (2013) essentially all of the laboratory traits used in Table 1 can be predicted accurately by near infrared spectroscopy (NIRS). The R2 for comparison between measured N, NDF and ADF were very good (0.94–0.96) and quite good for the more difficult to analyze ADL, IVOMD and ME (0.81–0.82). The NIRS software allows predictions of 36 traits at one time, and in practice all or most of the traits listed in Table 1 would be predicted if the pertinent NIRS equations were available. The IVOMD and ME analysis requires ruminally-fistulated animals which are more likely to be available in livestock than crop institutions, while analyses for N, NDF, ADF and ADL are well within the grasp of crop institute laboratories. In the latter case traits such as IVOMD and ME can be predicted by the chemical traits using, for example, the summative equations of Van Soest (1994). Lukuyu et al. (2013) illustrate whole maize plant optimization less by choice of cultivar but rather by management options. Smallholder farmers in mixed maize-dairy systems in Kenya used maize management strategies such as high planting densities with subsequent timely thinning to supply critical livestock feed during the growing season without compromising grain yield. They explore the effects of variations in maize management to increase quantity and quality of fodder off-take from maize (thinnings and stover). Their work also focused on the effects of maize streak virus disease (MSVD), a major disease in these systems that lowers grain yields and also impacts fodder yields. Increases in fodder quality are not adequate to overcome decreased yields associated with MSVD infection. MSVD tolerant maize varieties offer promise but such maize improvement efforts should consider fodder quantity and quality impacts. 4. Exploiting trait variation for maize improvement Great advances in technologies available in plant breeding, such as rapid genotyping have been made within the past decade
110
M. Blümmel et al. / Field Crops Research 153 (2013) 107–112
allowing more rapid assessment of trait variation and associated selection. This adds to the value in furthering our understanding of the feasibility and how best to pursue improvement in fodder quality traits in maize. The remaining 4 papers focused on variations in grain and stover traits in maize breeding work and how to exploit this through conventional and molecular breeding approaches for targeted concomitant improvement of grain and stover traits. Ertiro et al. (2013a) investigated a total of 335 experimental highland hybrids targeted for eastern Africa in multiyear (2004–2006) and multilocation (Ethiopia and Tanzania) trials for grain and stover yield and stover fodder quality traits. Across years and locations they observed the existence of exploitable genetic variation not only for grain yield but also for stover fodder quality and quantity. Their study pinpointed the possibility for simultaneous improvement of grain yield and stover traits to address the high demand existing for dual purpose grain-fodder type of maize genotypes in maize-livestock mixed farming system of Eastern Africa. The authors observed a minimum of 5.6 percentage units difference in stover IVOMD that could be exploited across environments. Across all 335 experimental hybrids, grain yield and stover quality were either statistically unrelated or the associations were weak (r = −0.24–0.33). In their follow up work, Ertiro et al. (2013b) investigated the trend in variability and association between grain and stover traits in inbred parents and the hybrids derived from them. The authors used sixteen inbred lines to generate sixty single cross hybrids which were evaluated for grain and stover yield and stover N, NDF, ADF, ADL, IVOMD and ME across three environments in Ethiopia. Genotypes in both hybrids and inbred trials showed highly significant variations for all the traits studied. The authors reported substantial variations in key stover traits, for example IVOMD, which varied by 9.2 percentage units among hybrids and by 11 percentage units among inbred parents while IVOMD and grain yield in stover were quite independent (P = 0.56). At high grain yields of about 10–12 t/ha, IVOMD could range from about 53 to 61%. The authors noted that grain–stover relationships were less rigid then often assumed. Thus while stover and grain yield were highly correlated in hybrids and inbred lines the variation in the one accounted for about only 58 (hybrids) and 53% (inbred parents) of the variation in the other. In other words substantial parts of the variation in stover yields were not explained by variations in grain yields. Using harvest indices to predict one from the other is probably not satisfactory. The authors reported correlation between mid-parent and hybrid values for stover N, NDF, ADF, ADL, ME and IVOMD to range from r = 0.42 for stover ADF to r = 0.79 for stover ME and argued that the significant positive relationships observed between inbred lines per se and hybrid performances for these fodder quality traits suggest the feasibility of predicting hybrid performance from the performance of the inbred lines. The authors also observed that the general combining abilities (GCA) of both lines and testers and specific combining ability (SCA) of line by tester interactions were significant for most traits studied. The highly significant GCA effects observed for most traits and the greater relative importance of GCA (lines and testers) as compared to SCA for grain yield and most stover fodder quality traits suggest the importance of additive gene effects in controlling grain and stover yield as well as stover fodder quality. Zaidi et al. (2013) used cluster analysis to identify the top ranking lines among the existing elite lines from CIMMYT-Asia with respect to superior stover IVOMD (in addition to good grain and stover yield performance) for initiating high stover fodder quality breeding activities. The authors observed that the IVOMD of the crosses did not significantly deviate from the mid-parental values which would suggest a simple genetic basis with predominant additive effects for IVOMD. The authors further compared the performance of their bi-parental crosses targeting stover qual-
ity improvement with one of the leading commercial hybrids in India (900 M Gold from Monsanto) and all but two crosses showed favourable deviation from the commercial hybrid for stover quality parameters. The authors observed a wide range in productive performance of grain and stover traits. For example grain and stover yields ranged from 0.94 to 7.3 t/ha and from 2.0 to 8.5 t/ha (P < 0.0001), respectively. Stover IVOMD ranged from 46.9 to 55.5% (P = 0.08). Grain and stover traits showed a considerable degree of independency. Grain and stover yield were significantly positively correlated (r = 0.53; P = 0.003) but high grain yields at around 7 t/ha could be associated with stover yields varying approximately from 3.5 to 6 t/ha. The authors observed no significant correlation was observed between IVOMD and grain yield (r = −0.12, P = 0.52) or IVOMD and stover yield (r = 0.03; P = 0.98) and noted that the cross with the third highest grain yield (7.1 t/ha) had the second highest stover digestibility (close to 55%). Maize stover quality traits thereby are at least at par with sorghum stover which is widely marketed and still preferred for fodder in southern India. Vinayan et al. (2013) used Genome-wide Association Studies (GWAS) and Genotyping by Sequencing (GBS) on a panel of 276 inbred lines from CIMMYT’s tropical and sub-tropical program test crossed to maize line CML312. Their single crosses were evaluated for grain and stover yields and morphological/physiological crop traits and a range of chemical (N, NDF, ADF, ADL) and biological (IVOMD, ME) stover traits. Using a 55K SNPs genotypic dataset showed several regions of significant association for key maize stover fodder traits such as N, ADF and IVOMD, each explaining from 3 to 9% of phenotypic variance for these traits. SYN7725 from the 55K chip on chromosome 4 explained the largest proportion of phenotypic variance (∼9%) for ADF and had a robust minor allele frequency of 0.35. A specific genomic region on chromosome 3 (132.7–149.2 Mb) was found to be significantly associated with all the three fodder quality traits, with the largest effect on IVOMD. The authors suggest that this region merits attention for further validation and marker-assisted introgressions. A cellulose-related candidate gene, Xyloglucan endotransglucosylase/hydrolase (xth1, GRMZM2G119783) was also identified closer to the peak on chr.10 (∼76.9 Mb) for ADF, which has been previously demonstrated to have a significant role in fiber elongation in cotton. The authors found strong negative associations between desirable (N, IVOMD, ME) and undesirable (NDF, ADF, ADL) components and hypothesize that this suggests proximity of genomic regions governing these traits and that this association could be extremely beneficial in breeding perspective, as close proximity between these regions would indicate high linkage disequilibrium resulting in co-segregation of positive and negative alleles of desirable and undesirable traits respectively. Above compiled papers show that exploitable variations in grain-fodder traits in a wide range of maize breeding lines, experimental hybrids and mapping populations are generally larger than the variations observed in released and commonly used hybrids. The findings from the hybridisation experiments suggest that the exploitable variations in grain-fodder traits in released hybrids can be increased by considering stover traits in parental lines in maize breeding work.
5. Lessons and way forward The various papers in this special issue allow us to draw out some lessons and provide guidance for further dual-purpose maize R&D. A consistent finding was the confirmation of the relatively favorable feed value of maize stover vis-à-vis other coarse cereal residues – having at least par if not better feed quality traits compared to sorghum and millet, which have been the focus of prior dual-purpose crop improvement research and have been reported
M. Blümmel et al. / Field Crops Research 153 (2013) 107–112
to contribute substantially to gross crop production values. Promising too is the confirmation of being able to rely on a few key laboratory indicators (IVOMD, ME) as good proxies for feed quality (Ravi et al., 2013) as this enhances the ease of screening for feed quality traits. Encouraging too were the findings of (1) generally positive association between grain and stover yield; and (2) the independence between fodder quality on the one hand and stover and grain yield on the other (Anandan et al., 2013; Ertiro et al., 2013a); Zaidi et al., 2013). This is similar to reports for sorghum and millet (Sharma et al., 2010). In other words, maize germplasm differences in fodder quality can be exploited without compromising on grain yield. The four studies in this special issue that looked into the potential demand for dual-purpose maize cultivars all reiterated the widespread maize stover use as fodder across the American- and African-based studies, including in USA’s commercial agriculture. This reiterates the paradox of limited maize stover use in India reported earlier (Erenstein et al., 2011). The latter study also raised the issue of tradition and food preferences as helping explain the stickiness of crop residue use patterns, which possibly could undermine the potential of dual-purpose maize varieties in nontraditional maize producing areas. This special issue substantiates that dual-purpose maize varieties are technically feasible and have a large potential market, particularly in many emerging markets. The reported findings argue the case for continued investments in maize stover R&D and thus reigniting earlier dual-purpose crop research in general. Despite the highlighted prospects of selecting for stover quantity and quality traits, challenges remain in terms of actually releasing some of such dual-purpose maize varieties and having them widely adopted. Within the cereal seed sector across the globe, the maize seed sector tends to be the most developed given the availability of maize hybrids. Nonetheless, commercializing dual-purpose maize varieties (including hybrids) may pose new challenges – for instance in terms of premium-pricing of dual-purpose seed or using such additional trait as embodied valueaddition to enhance or secure market share. This may challenge existing business models and bring in market segmentation implications in view of spatial and economic heterogeneity – be it by continent, country, ecology or farmer type. The development and utilization of dual-purpose maize varieties thus calls for further multi-disciplinary and multi-stakeholder R&D. Further fine-tuning R&D is also needed – including monitoring of the eventual release and uptake of purposively developed dual-purpose maize varieties. Enhanced processing may also open new opportunities. For instance, Anandan et al. (2013) reported that feed mash proved superior to feed blocks in terms of animal productivity. The core focus of the special issue has been on grain-fodder trade-offs for dual-purpose maize varieties. However, there are other potential trade-offs, including alternative stover uses such as the potential use of biomass for soil health within conservation agriculture systems, as explored in the case of Mexico (Hellin et al., 2013) and elsewhere (Valbuena et al., 2012). The findings presented by Homann-Kee Tui et al. (2013) however question some of the assumptions of Valbuena et al. (2012) – particularly that simple increases in quantity of biomass, for example through fertilizer application, will automatically reduce competition between biomass usage for livestock feeding and soil improvement. Livestock have limited intake capacities for bulky feed such as stover and increasing quantitative availability of a given stover above the intake capacity of livestock will not improve fodder resources beyond the possibility of making fodder cheaper. Therefore, quantitative and quality aspects of stover work, or more generally crop residue work, in crop improvement cannot strictly be separated
111
either in the context of livestock feeding or biomass allocation and partitioning for different purposes. Biomass use trade-offs are also not limited to the direct competition between stover for feed or soil health and may actually be further entangled by dual-purpose maize varieties. For instance, the higher fodder value of dual-purpose maize may increase stover off-take (in situ or ex situ) and thereby exacerbate competition between uses. Furthermore, dual-purpose maize varieties typically have higher digestibility as feed, but this may also have implications for their digestibility by soil biota and susceptibility to weathering vis-à-vis their retention as mulch within conservation agriculture systems (Erenstein, 2002). Some of these trade-offs are inherently complex and also may not be limited to maize – and merit further research. The focus of the special issue and R&D thus far has been on dual-purpose maize for grain-fodder. As illustrated by the USA case, biofuel considerations may add uses and challenges, both in terms of alternative grain use (bioethanol, Klopfenstein et al., 2013) or alternative stover use for second generation (cellulosic) biofuels (Lorenz et al., 2010). The latter may imply new trade-offs on stover use, but also may give rise to: 1) new byproducts with higher feed value, as was the case with the first generation grainbased biofuels; 2) spin-off technologies for upgrading the fodder value of ligno-cellulose biomass for livestock. These developments also have potential implications for moving beyond the current focus on dual-purpose to the future development and use of multipurpose maize varieties (food-feed-fuel) in industrial agriculture and beyond.
Acknowledgements The special issue including this synthesis paper benefited from support from various sources – as variously acknowledged in the preceding papers. Here we would like to particularly refer to the more detailed acknowledgements in the overview paper (Erenstein et al., 2013). The views expressed here are those of the authors and do not necessarily reflect the views of the donor or the authors’ institution. The usual disclaimer applies.
References Anandan, S., Khan, A.A., Ravi, D., Sai Bucha Rao, M., Ramana Reddy, Y., Blümmel, M., 2013. Identification of a superior dual purpose maize hybrid among widely grown hybrids in South Asia and value addition to its stover through feed supplementation and feed processing. Field Crop Res. 153, 52–57. Blümmel, M., Rao, P.P., 2006. Economic value of sorghum stover traded as fodder for urban and peri-urban dairy production in Hyderabad, India. Int. Sorgh. Mill. Newslett. 47, 97–100. Byerlee, D., Iqbal, M., Fischer, K.S., 1989. Quantifying and valuing the joint production of grain and fodder from maize fields: evidence from northern Pakistan. Exp. Agr. 25, 435–445. De Groote, H., Dema, G., Sonda, G.B., Gitonga, Z.M., 2013. Maize for food and feed in East Africa–The farmers’ perspective. Field Crop Res. 153, 22–36. Erenstein, O., 2002. Crop residue mulching in tropical and semi-tropical countries: an evaluation of residue availability and other technological implications. Soil Till. Res. 67, 115–133. Erenstein, O., Samaddar, A., Teufel, N., Blümmel, M., 2011. The paradox of limited maize stover use in India’s smallholder crop-livestock systems. Exp. Agr. 47, 677–704. Erenstein, O., Blümmel, M., Grings, E., 2013. Potential for dual-purpose maize varieties to meet changing maize demands: Overview. Field Crops Res. 153, 1–4. Ertiro, B.T., Twumasi-Afriyie, S., Blümmel, M., Friesen, D., Negera, D., Worku, M., Abakemal, D., Kitenge, K., 2013a. Genetic variability of maize stover quality and the potential for genetic improvement of fodder value. Field Crop Res. 153, 79–85. Ertiro, B.T., Zeleke, H., Friesen, D., Blümmel, M., Twumasi-Afriyie, S., 2013b. Relationship between the performance of parental inbred lines and hybrids for food-feed traits in maize (Zea mays L.) in Ethiopia. Field Crop Res. 153, 86–93. Hellin, J., Erenstein, O., Beuchelt, T., Camacho, C., Flores, D., 2013. Maize stover use and sustainable crop production in mixed crop–livestock systems in Mexico. Field Crop Res. 153, 12–21.
112
M. Blümmel et al. / Field Crops Research 153 (2013) 107–112
Homann-Kee Tui, S., Blümmel, M., Valbuena, D., Chirima, A., Masikati, P., van Rooyen, A.F., Kassie, G.T., 2013. Assessing the potential of dual-purpose maize in southern Africa: a multi-level approach. Field Crop Res. 153, 37–51. Klopfenstein, T.J., Erickson, G.E., Berger, L.L., 2013. Maize is a critically important source of food, feed, energy and forage in the USA. Field Crop Res. 153, 5–11. Kristjanson, P.M., Zerbini, E., 1999. Genetic enhancement of sorghum and millet residues fed to ruminants: An ex ante assessment of returns to research. Impact Assessment Series 3. ILRI, Nairobi, Kenya. Lorenz, A.J., Gustafson, T.J., Coors, J.G., Leon, N.D., 2010. Breeding maize for a bioeconomy: A literature survey examining harvest index and stover yield and their relationship to grain yield. Crop Sci. 50, 1–12. Lukuyu, B.A., Murdoch, A.J., Romney, D., Mwangi, D.M., Njuguna, J.G.M., McLeod, A., Jama, A.N., 2013. Integrated maize management options to improve forage yield and quality on smallholder farms in Kenya. Field Crop Res. 153, 70–78. Muttoni, G., Palacios-Rojas, N., Galicia, L., Rosales, A., Pixley, K.V., de Leon, N., 2013. Cell wall composition and biomass digestibility diversity in Mexican maize (Zea mays L) landraces and CIMMYT inbred lines. Maydica 58, 21–33. Notenbaert, A., Herrero, M., De Groote, H., You, L., Gonzalez-Estrada, E., Blümmel, M., 2013. Identifying recommendation domains for targeting dual-purpose maizebased interventions in crop-livestock systems in East Africa. Land use policy 30, 834–846. Ramana Reddy, Y., Ravi, D., Ramakrishna Reddy, C., Prasad, K.V.S.V., Zaidi, P.H., Vinayan, M.T., Blümmel, M., 2013. A note on the correlations between
maize grain and maize stover quantitative and qualitative traits and the implications for whole maize plant optimization. Field Crop Res. 153, 63–69. Ravi, D., Khan, A.A., Saibutcharao, M., Blümmel, M., 2013. A note on suitable laboratory stover quality traits for multidimensional maize improvement. Field Crop Res. 153, 58–62. Sharma, K., Pattanaik, A.K., Anandan, S., Blümmel, M., 2010. Food-Feed Crops Research: A Synthesis. Animal Nutrition and Feed Technology 10s, 1–10. Shiferaw, B., Prasanna, B., Hellin, J., Banziger, M., 2011. Crops that feed the world 6. Past successes and future challenges to the role played by maize in global food security. Food Sec. 3, 307–327. Valbuena, D., Erenstein, O., Homann-Kee TUI, S., Abdoulaye, T., Claessens, L., Duncan, A.J., Gerard, B., Rufino, M.C., Teufel, N., van Rooyen, A., van Wijk, M.T., 2012. Conservation Agriculture in mixed crop-livestock systems: Scoping crop residue trade-offs in Sub-Saharan Africa and South Asia. Field Crop Res. 132, 175– 184. Van Soest, P.J., 1994. Nutritional Ecology of the Ruminant. University Press, Cornell. Vinayan, M.T., Babu, R., Jyothsna, T., Zaidi, P.H., Blümmel, M., 2013. A note on potential candidate genomic regions with implications for maize stover fodder quality. Field Crop Res. 153, 102–106. Zaidi, P.H., Vinayan, M.T., Blümmel, M., 2013. Genetic variability of tropical maize stover quality and the potential for genetic improvement of food-feed value in India. Field Crop Res. 153, 94–101.
Contents lists available at ScienceDirect
Field Crops Research journal homepage: www.elsevier.com/locate/fcr
Potential for dual-purpose maize varieties to meet changing maize demands: Synthesis Michael Blümmel a,∗ , Elaine Grings b , Olaf Erenstein c a b c
International Livestock Research Institute (ILRI), c/o ICRISAT, Patancheru 502 324, Andhra Pradesh, India South Dakota State University, Department of Animal Science, Box 2170, Brookings, SD 57007, USA International Maize and Wheat Improvement Centre (CIMMYT), c/o ILRI, PO Box 5689, Addis Ababa, Ethiopia
1. Introduction Maize—or corn (Zea mays L.)—now is the most important global cereal in terms of production reflecting its versatility in use, including human food, animal feed and fodder, industrial products and biofuel. Most uses revolve around maize grain as the primary product, although whole plant utilization for silage is also common in industrialized agriculture (e.g. Klopfenstein et al., 2013). Despite being a versatile crop, maize production and maize breeding efforts over time have typically had a single-purpose orientation. For instance, maize breeding has focused on overcoming biotic and abiotic stresses so as to generate high yielding, stress-tolerant and widely-adapted maize varieties through judicious combination of conventional and molecular breeding approaches (Muttoni et al., 2013; Shiferaw et al., 2011). Even smallholders within mixed maize–livestock systems typically focus on maize grain yield (De Groote et al., 2013), with maize stover as additional byproduct and benefit. Although farmers may still try to increase fodder off-take, they still try to minimize maize grain yield loss (Byerlee et al., 1989; Lukuyu et al., 2013). Critical to the development and acceptance of dual-purpose maize are the potential trade-offs between grain and fodder both in terms of quantity and quality. Single-purpose maize can focus on either grain or fodder, with grain reflecting partial utilization of maize biomass against potentially full (above ground) biomass utilization for fodder (as in case of silage). Dual-purpose maize focuses on both grain and fodder (particularly stover), and typically the full (above ground) biomass. Optimization for one trait (grain) is inherently easier than for two traits (grain and fodder), and may be pursued in instances where the fodder/stover value is negligible. However, the widespread use of maize stover suggests it indeed has value, although typically less than the maize grain. Given that both traits have value and improvement for each is likely subject to diminishing returns, dual-purpose breeding could potentially make economic sense. Furthermore, similar to many cereals, maize varieties appear to have a wide variety of (grain
∗ Corresponding author. Tel.: +91 40 30713653; cell: +91 9866315505. E-mail address: [email protected] (M. Blümmel). 0378-4290/$ – see front matter © 2013 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.fcr.2013.10.006
and stover) yields and stover quality, and most interestingly, there are prospects within the range of stover quality to increase fodder quality without compromising grain yield. It is this potential of dual-purpose varieties that has reignited research interest and some of the research underlying this special issue. Indeed, despite earlier skepticism only a decade ago, substantial progress has been made in developing dual-purpose maize options for both grain and fodder purposes as reflected in the papers contained in this special issue (Erenstein et al., 2013). This paper synthesizes the key findings presented in 12 papers in this Special Issue around the potential for dual-purpose maize varieties to meet changing maize demands. We summarize the key findings around three thematic areas: (1) demand for dual-purpose maize cultivars and associated targeting domains; (2) quality traits, whole plant utilization and phenotyping; and (3) exploiting trait variation for maize improvement. In a final section we provide some of the lessons learned and guidance for further dual-purpose maize research and development (R&D). 2. Demand for dual-purpose maize cultivars and associated targeting domains While plant breeders have the technical capability to alter maize plant characteristics, adoption of new technologies or varieties requires an understanding of farmers’ preferences for crop characteristics. Issues of trade-offs for other values, such as soil improvement and fuel, market forces, disease resistance, and availability of alternative inexpensive feed resources affect farmers’ willingness to accept new varieties with improved stover quantity and fodder quality. Four papers discussed the potential demand for dual-purpose maize cultivars in North and Meso-America and Eastern and Southern Africa and the potential trade-offs in the use of maize plant parts for food, feed, fodder and soil improvements. Ways are discussed to better match specifically identified and desired traits to farming and production systems. System contexts are provided by agro-ecologies, cropping pattern, human and livestock densities, importance of livestock and alternative feed and fodder resources in the form of rangelands and common property resources to help delineate targeting domains for dual-purpose maize cultivars.
108
M. Blümmel et al. / Field Crops Research 153 (2013) 107–112
For the United States of America (USA), Klopfenstein et al. (2013) review the history of maize (corn) use and the current conditions impacting both maize production and its use as livestock feed. Starting in the middle of the 20th century, plentiful maize grain was used increasingly as livestock feed. Changes in maize use for ethanol production have altered demands for grain and resulted in new by-products that provide valuable feed sources and that have been used to replace maize in livestock feeding in the U.S.A. Maize has been used as a roughage source in two primary ways in the U.S.A., either as silage or through use of the stover remaining after the grain harvest (used ex situ or in situ). Crop breeding strategies have primarily focused on grain but breeding programs specifically for silage also exist. Klopfenstein et al. do report on one study in which varietal differences in morphological traits are compared in stovers, which showed some variation. They also discuss the value of grazed maize stover as fodder for beef cattle and suggest that maize stover can be well utilized as a roughage source for both gestating cows and growing calves as long as adequate protein is provided to meet the needs of rumen microbes. Research is cited that shows gains of up to 1 kg/d for calves supplemented with up to 3.5 kg of distillers grains while grazing maize stalks. These authors also discuss research that indicates that allowing cattle to graze maize stover during winter does not affect subsequent maize grain yields. Hellin et al. (2013) take us to Mexico, the centre of origin of maize and U.S.A.’s southern neighbour. They use farmer group/community surveys in three contrasting agro-ecologies (semi-arid, temperate highland and tropical sub-humid) to explore maize stover use. Mixed maize–livestock smallholder farming systems prevail, with maize grain produced for home consumption and the market and maize stover as an important by-product, primarily as feed. Although in situ grazing is found in all three study sites, it represented the bulk of stover use in only one site (with an estimated 70% of stover in the sub-humid tropics), with ex situ feed dominating in the other two sites (estimated >80%). Maize stover commercialization was limited and mainly restricted to households with no livestock and often within the immediate local context. They also explore potential trade-offs in stover use, particularly its use as feed against its potential use as mulch (soil cover) to manage soil health within the context of conservation agriculture. They report that farmers are generally hesitant to adopt conservation agricultural practices that require the retention of stover as mulch, as this competes with their livestock feed needs and purchased feed is expensive. The authors go on to explore a portfolio of options to reduce these trade-offs, including partial residue retention, cover and feed crops and sustainable intensification. Promising (Muttoni et al., 2013) but yet to be explored further in the context of Mexico, are investments in generating dual-purpose (grain-fodder) maize varieties. De Groote et al. (2013) investigated the potential demand for dual purpose maize in Ethiopia and Tanzania using participatory rural appraisals, farm household surveys and formal farmer evaluations of maize varieties, the latter focusing on commercially available material that were planted side-by-side in trial sites in the study areas as purposively developed dual-purpose maize varieties are not yet on the market. They show that maize stover is an important element of livestock feed in all the study areas. Desirable traits farmers consistently mentioned were related to grain yield and pest resistance, ease of cooking and taste characteristics and stover yield and fodder quality. Analysis of adoption patterns of existing maize varieties by farm households shows that varieties that score well on feed characteristics have a higher probability of being adopted. The authors concluded that there is a demand for varieties with superior stover quantity and fodder quality as long as grain yield and consumer preferences were not compromised. They further hypothesise that such varieties have the potential to increase the
productivity of maize-livestock systems and the income of these farmers, provided these varieties are taken up in the seed portfolio of the emerging seed sector in the sub-region. Homann-Kee Tui et al. (2013) explored the use of a multilevel approach and associated data sources to assess the potential of dual-purpose maize cultivars in Southern Africa. The authors construct recommendation domains for dual-purpose maize using maize mega-environments and demand estimates derived from livestock and human population densities and potential feed supply, including biomass contributions from range and croplands. These estimates were subjected to ground truthing through survey data collected from 480 households in contrasting sites in Malawi, Mozambique and Zimbabwe. In addition the authors’ reported on maize varieties, both landraces and advanced hybrids, for variations in grain and stover yield and fodder quality traits. Their study showed that where livestock densities were high and feed resources from rangeland limited, maize cultivars with superior stover yield and fodder quality can have substantial impact on livestock productivity. Stover with higher fodder quality could provide sufficient energy for providing livestock maintenance requirements and support about 200 gram of live weight gain daily. The authors further proposed an approach to target grain-fodder maize cultivars to demand domains for either stover quantity or stover fodder quality based on maize stover intake estimations derived from Ravi et al. (2013). Homann-Kee Tui et al. complement and build on Notenbaert et al. (2013) who focused on eastern Africa to develop similar demand domains for dual-purpose maize as part of an integrated targeting approach based on maize mega environments, livestock numbers, population densities and alternative feed resources. Research such as that reported in these four papers reinforces the need to evaluate the specific contexts for technology adoption. For example maize improvement/cultivar choice can increase stover quantity, stover quality or perhaps both. It is important to realize that from a livestock nutrition viewpoint, an increase in stover quantity is only useful (unless making stover cheaper) if livestock can respond with increased intake, which is stover quality dependent.
3. Quality traits, whole plant utilization and phenotyping Evaluating the nutritional fodder value for livestock in vivo can be a complex and time consuming process. Additionally, livestock may be fed complex diets in which feedstuffs impact the utilization of one another. Nutritionists have long worked on developing laboratory indicators that well reflect animal response. Validation of rapid laboratory measures is key to evaluation of large collections of germplasm for screening. A set of four papers looked into the variations in grain and stover traits in existing maize cultivars, their meaning for whole maize plant optimization, and laboratory phenotyping capability. These papers thus investigated exploitable variations in a range of food-feed-fodder traits in existing and widely used cultivars, the implication of these traits for livestock productivity and/or tradability of maize stover, and techniques for selecting and predicting specific traits. Anandan et al. (2013) investigated six maize hybrids commonly grown in South Asia for grain and stover yields and for stover morphological (residual green leaf area, plant height, stem diameter and proportions of leaf blade: leaf sheath: stem), chemical (nitrogen (N), neutral (NDF) and acid (ADF) detergent fiber, acid detergent lignin (ADL)) and biological (in vitro organic matter digestibility (IVOMD) and metabolizable energy (ME) content) fodder quality traits. The stovers were also fed to sheep as major part (90%) of their feed. The authors observed significant cultivar differences
M. Blümmel et al. / Field Crops Research 153 (2013) 107–112
109
Table 1 Comparisons between measured digestibility (DMDmeas , %), intake (DMImeas , g/d) and digestible intake (DDMImeas , g/d) and DMDpred , DMIpred and DDMIpred predicted by cross validation procedures based on stover nitrogen (N), neutral (NDF) and acid detergent (ADF) fiber, acid detergent lignin (ADL), in vitro organic matter digestibility (IVOMD) and metabolizable energy (ME) including simple linear relationships between traits and in vivo measurements. Derived from a combination of the data set of Ravi et al. (2013) and Anandan et al. (2013) using observations from 16 maize stovers.
N NDF ADF ADL ME IVOMD
DMDpred versus DMDmeas
DMIpred versus DMImeas
DDMpred I versus DDMImeas
y = 22.8 + 0.60x; [r = 0.75; P = 0.0008; Syx = 1.9] y = 37.4 + 0.34x; [r = 0.53; P = 0.04; Syx = 2.0] y = 19.2 + 0.66x; [r = 0.79; P = 0.0003; Syx = 1.9] y = 28.3 + 0.50x; [r = 0.66; P = 0.005; Syx = 2.1] y = 40.6 + 0.28x; [r = 0.45; P = 0.08; Syx = 2.1] y = 39.4 + 0.3x; [r = 0.47; P = 0.07; Syx = 2.1]
y = 20.5 + 0.13x; [r = 0.24; P = 0.36; Syx = 1.8] y = 18.4 + 0.22x; [r = 0.37; P = 0.16; Syx = 2.0] y = 14.8 + 0.37x; [r = 0.55; P = 0.03; Syx = 2.0] y = 8.7 + 0.63x; [r = 0.78; P = 0.0004; Syx = 1.8] y = 4.2 + 0.82x; [r = 0.90; P < 0.0001; Syx = 1.4] y = 5.3 + 0.77x; [r = 0.87; P < 0.0001; Syx = 1.6]
y = 9.0 + 0.33x; [r = 0.49; P = 0.05; Syx = 1.59] y = 8.8 + 0.34x; [r = 0.51; P = 0.04; Syx = 1.58] y = 5.6 + 0.57x; [r = 0.73; P = 0.001; Syx = 1.47] y = 3.5 + 0.73x; [r = 0.85; P < 0.0001; Syx = 1.27] y = 2.5 + 0.81x; [r = 0.89; P < 0.0001; Syx = 1.1] y = 2.9 + 0.78x; [r = 0.88; P < 0.0001; Syx = 1.2]
Simple linear regression between stover laboratory traits and in vivo measurement in sheep
N NDF ADF ADL ME IVOMD
DMD (%)
DMI g/kg LW
DDMI g/kg LW
y = 50.4 + 7.0x; r = 0.81; P = 0.0001 y = 85.4 − 0.4x; r = 0.66; P = 0.005 y = 79.4 − 0.56x; r = 0.85; P < 0.0001 y = 66.8 − 2.4x; r = 0.76; P = 0.0007 y = 39.3 + 2.3x; r = 0.62; P = 0.009 y = 37 + 0.39x; r = 0.64; P = 0.008
y = 19.4 + 4.4x; r = 0.51; P = 0.04 y = 48.9 − 0.35x; r = 0.59; P = 0.017 y = 41.8 − 0.46x; r = 0.69; P = 0.003 y = 34.4 − 2.6x; r = 0.83; P < 0.0001 y = −2.7 + 3.4x; r = 0.93; P < 0.0001 y = −4.8 + 0.55x; r = 0.90; P < 0.0001
y = 9.4 + 4.3x; r = 0.67; P = 0.005 y = 35.4 − 0.3x; r = 0.67; P = 0.005 y = 29.8 − 0.41x; r = 0.81; P = 0.0001 y = 22.3 − 2.1x; r = 0.89; P < 0.0001 y = −6.4 + 2.6x; r = 0.92; P < 0.0001 y = −8.2 + 0.42x; r = 0.91; P < 0.0001
in important animal response variables such as digestible feed intake and nitrogen balance. For example, nitrogen balances ranged from 0.47 to 1.2 g per day per sheep. No significant relationships (P = 0.32–0.77) were observed between animal response variables and grain yields and the authors identified a cultivar that had high grain and stover yield and for which animal response variables not differ from that of the (numerically) best stover. The authors related the morphological, chemical and biological stover traits to the animal response variables and reported that these were more highly correlated with intake and digestible feed intake than with in vivo digestibility and nitrogen balance. Certain chemical (particularly NDF and ADF) and biological (IVOMD and ME) traits were more closely correlated with the in vivo measurements than were the morphological traits. NDF and ADF were highly negatively related with digestible feed intake (the product of stover digestibility and stover intake), accounting for 90% of the variation therein, while IVOMD and ME were positively associated with digestible feed intake, accounting for 90 and 94% of the variation therein, respectively. The authors designed complete diets from stovers of the lowest and highest IVOMD maize cultivars grown by farmers and observed about 30% greater weight gain in sheep fed the diet with the higher IVOMD stover. The authors suggest that the impact of the difference in IVOMD of about 4% units agreed well with ex ante impact assessment of sorghum and pearl millet stover improvement (Kristjanson and Zerbini, 1999) and fodder market studies with sorghum stover (Blümmel and Rao, 2006) in which a 1% unit improvement in stover digestibility could reap benefits of 5–8% in terms of higher livestock productivity. Ravi et al. (2013) investigated ten maize stovers (including two cultivars also used by Anandan et al., 2013) and confirmed the usefulness of ADF, IVOMD and ME measurements in ranking stover for in vivo digestibility, feed intake and digestible feed intake. For the current synthesis we asked the authors to combine their data sets since the animal experimental facilities used were identical and the different in vivo experimentations were standardized and very similar. Statistical cross-validation procedures were employed on the combined data set where the predicted values were not used for the development of the regression equations (“blind-predictions”). These findings are summarized in Table 1. As in the individual data sets, DMD was less well predicted than DMI and DDMI possibly because the concentrate used as supplement (about 10–15% of the ration) had affected DMD while only maize stover was considered for DMI (Ravi et al., 2013) Prediction of DMD of maize stover-based diets from stover N and ADF may still be meaningful
since the intercepts of their regression equations were relatively low and the correlation between measured and predicted DMD were 0.75–0.78. It seems quite clear that IVOMD and ME are well suited for use in DMI and DDMI predictions of maize stover-based diets (Table 1). Digestibility measurements (in vitro and in vivo) provide an approximation of the proportion of the feed the animal can use. These measures are an easier nutritional concept to grasp for non-animal nutritionists than is ME, which requires advanced understanding of ruminant physiology. Similarly understanding the nutritional implications of variation in NDF, ADF and ADL requires some familiarity with livestock nutritional concepts. As shown by Ramana Reddy et al. (2013) essentially all of the laboratory traits used in Table 1 can be predicted accurately by near infrared spectroscopy (NIRS). The R2 for comparison between measured N, NDF and ADF were very good (0.94–0.96) and quite good for the more difficult to analyze ADL, IVOMD and ME (0.81–0.82). The NIRS software allows predictions of 36 traits at one time, and in practice all or most of the traits listed in Table 1 would be predicted if the pertinent NIRS equations were available. The IVOMD and ME analysis requires ruminally-fistulated animals which are more likely to be available in livestock than crop institutions, while analyses for N, NDF, ADF and ADL are well within the grasp of crop institute laboratories. In the latter case traits such as IVOMD and ME can be predicted by the chemical traits using, for example, the summative equations of Van Soest (1994). Lukuyu et al. (2013) illustrate whole maize plant optimization less by choice of cultivar but rather by management options. Smallholder farmers in mixed maize-dairy systems in Kenya used maize management strategies such as high planting densities with subsequent timely thinning to supply critical livestock feed during the growing season without compromising grain yield. They explore the effects of variations in maize management to increase quantity and quality of fodder off-take from maize (thinnings and stover). Their work also focused on the effects of maize streak virus disease (MSVD), a major disease in these systems that lowers grain yields and also impacts fodder yields. Increases in fodder quality are not adequate to overcome decreased yields associated with MSVD infection. MSVD tolerant maize varieties offer promise but such maize improvement efforts should consider fodder quantity and quality impacts. 4. Exploiting trait variation for maize improvement Great advances in technologies available in plant breeding, such as rapid genotyping have been made within the past decade
110
M. Blümmel et al. / Field Crops Research 153 (2013) 107–112
allowing more rapid assessment of trait variation and associated selection. This adds to the value in furthering our understanding of the feasibility and how best to pursue improvement in fodder quality traits in maize. The remaining 4 papers focused on variations in grain and stover traits in maize breeding work and how to exploit this through conventional and molecular breeding approaches for targeted concomitant improvement of grain and stover traits. Ertiro et al. (2013a) investigated a total of 335 experimental highland hybrids targeted for eastern Africa in multiyear (2004–2006) and multilocation (Ethiopia and Tanzania) trials for grain and stover yield and stover fodder quality traits. Across years and locations they observed the existence of exploitable genetic variation not only for grain yield but also for stover fodder quality and quantity. Their study pinpointed the possibility for simultaneous improvement of grain yield and stover traits to address the high demand existing for dual purpose grain-fodder type of maize genotypes in maize-livestock mixed farming system of Eastern Africa. The authors observed a minimum of 5.6 percentage units difference in stover IVOMD that could be exploited across environments. Across all 335 experimental hybrids, grain yield and stover quality were either statistically unrelated or the associations were weak (r = −0.24–0.33). In their follow up work, Ertiro et al. (2013b) investigated the trend in variability and association between grain and stover traits in inbred parents and the hybrids derived from them. The authors used sixteen inbred lines to generate sixty single cross hybrids which were evaluated for grain and stover yield and stover N, NDF, ADF, ADL, IVOMD and ME across three environments in Ethiopia. Genotypes in both hybrids and inbred trials showed highly significant variations for all the traits studied. The authors reported substantial variations in key stover traits, for example IVOMD, which varied by 9.2 percentage units among hybrids and by 11 percentage units among inbred parents while IVOMD and grain yield in stover were quite independent (P = 0.56). At high grain yields of about 10–12 t/ha, IVOMD could range from about 53 to 61%. The authors noted that grain–stover relationships were less rigid then often assumed. Thus while stover and grain yield were highly correlated in hybrids and inbred lines the variation in the one accounted for about only 58 (hybrids) and 53% (inbred parents) of the variation in the other. In other words substantial parts of the variation in stover yields were not explained by variations in grain yields. Using harvest indices to predict one from the other is probably not satisfactory. The authors reported correlation between mid-parent and hybrid values for stover N, NDF, ADF, ADL, ME and IVOMD to range from r = 0.42 for stover ADF to r = 0.79 for stover ME and argued that the significant positive relationships observed between inbred lines per se and hybrid performances for these fodder quality traits suggest the feasibility of predicting hybrid performance from the performance of the inbred lines. The authors also observed that the general combining abilities (GCA) of both lines and testers and specific combining ability (SCA) of line by tester interactions were significant for most traits studied. The highly significant GCA effects observed for most traits and the greater relative importance of GCA (lines and testers) as compared to SCA for grain yield and most stover fodder quality traits suggest the importance of additive gene effects in controlling grain and stover yield as well as stover fodder quality. Zaidi et al. (2013) used cluster analysis to identify the top ranking lines among the existing elite lines from CIMMYT-Asia with respect to superior stover IVOMD (in addition to good grain and stover yield performance) for initiating high stover fodder quality breeding activities. The authors observed that the IVOMD of the crosses did not significantly deviate from the mid-parental values which would suggest a simple genetic basis with predominant additive effects for IVOMD. The authors further compared the performance of their bi-parental crosses targeting stover qual-
ity improvement with one of the leading commercial hybrids in India (900 M Gold from Monsanto) and all but two crosses showed favourable deviation from the commercial hybrid for stover quality parameters. The authors observed a wide range in productive performance of grain and stover traits. For example grain and stover yields ranged from 0.94 to 7.3 t/ha and from 2.0 to 8.5 t/ha (P < 0.0001), respectively. Stover IVOMD ranged from 46.9 to 55.5% (P = 0.08). Grain and stover traits showed a considerable degree of independency. Grain and stover yield were significantly positively correlated (r = 0.53; P = 0.003) but high grain yields at around 7 t/ha could be associated with stover yields varying approximately from 3.5 to 6 t/ha. The authors observed no significant correlation was observed between IVOMD and grain yield (r = −0.12, P = 0.52) or IVOMD and stover yield (r = 0.03; P = 0.98) and noted that the cross with the third highest grain yield (7.1 t/ha) had the second highest stover digestibility (close to 55%). Maize stover quality traits thereby are at least at par with sorghum stover which is widely marketed and still preferred for fodder in southern India. Vinayan et al. (2013) used Genome-wide Association Studies (GWAS) and Genotyping by Sequencing (GBS) on a panel of 276 inbred lines from CIMMYT’s tropical and sub-tropical program test crossed to maize line CML312. Their single crosses were evaluated for grain and stover yields and morphological/physiological crop traits and a range of chemical (N, NDF, ADF, ADL) and biological (IVOMD, ME) stover traits. Using a 55K SNPs genotypic dataset showed several regions of significant association for key maize stover fodder traits such as N, ADF and IVOMD, each explaining from 3 to 9% of phenotypic variance for these traits. SYN7725 from the 55K chip on chromosome 4 explained the largest proportion of phenotypic variance (∼9%) for ADF and had a robust minor allele frequency of 0.35. A specific genomic region on chromosome 3 (132.7–149.2 Mb) was found to be significantly associated with all the three fodder quality traits, with the largest effect on IVOMD. The authors suggest that this region merits attention for further validation and marker-assisted introgressions. A cellulose-related candidate gene, Xyloglucan endotransglucosylase/hydrolase (xth1, GRMZM2G119783) was also identified closer to the peak on chr.10 (∼76.9 Mb) for ADF, which has been previously demonstrated to have a significant role in fiber elongation in cotton. The authors found strong negative associations between desirable (N, IVOMD, ME) and undesirable (NDF, ADF, ADL) components and hypothesize that this suggests proximity of genomic regions governing these traits and that this association could be extremely beneficial in breeding perspective, as close proximity between these regions would indicate high linkage disequilibrium resulting in co-segregation of positive and negative alleles of desirable and undesirable traits respectively. Above compiled papers show that exploitable variations in grain-fodder traits in a wide range of maize breeding lines, experimental hybrids and mapping populations are generally larger than the variations observed in released and commonly used hybrids. The findings from the hybridisation experiments suggest that the exploitable variations in grain-fodder traits in released hybrids can be increased by considering stover traits in parental lines in maize breeding work.
5. Lessons and way forward The various papers in this special issue allow us to draw out some lessons and provide guidance for further dual-purpose maize R&D. A consistent finding was the confirmation of the relatively favorable feed value of maize stover vis-à-vis other coarse cereal residues – having at least par if not better feed quality traits compared to sorghum and millet, which have been the focus of prior dual-purpose crop improvement research and have been reported
M. Blümmel et al. / Field Crops Research 153 (2013) 107–112
to contribute substantially to gross crop production values. Promising too is the confirmation of being able to rely on a few key laboratory indicators (IVOMD, ME) as good proxies for feed quality (Ravi et al., 2013) as this enhances the ease of screening for feed quality traits. Encouraging too were the findings of (1) generally positive association between grain and stover yield; and (2) the independence between fodder quality on the one hand and stover and grain yield on the other (Anandan et al., 2013; Ertiro et al., 2013a); Zaidi et al., 2013). This is similar to reports for sorghum and millet (Sharma et al., 2010). In other words, maize germplasm differences in fodder quality can be exploited without compromising on grain yield. The four studies in this special issue that looked into the potential demand for dual-purpose maize cultivars all reiterated the widespread maize stover use as fodder across the American- and African-based studies, including in USA’s commercial agriculture. This reiterates the paradox of limited maize stover use in India reported earlier (Erenstein et al., 2011). The latter study also raised the issue of tradition and food preferences as helping explain the stickiness of crop residue use patterns, which possibly could undermine the potential of dual-purpose maize varieties in nontraditional maize producing areas. This special issue substantiates that dual-purpose maize varieties are technically feasible and have a large potential market, particularly in many emerging markets. The reported findings argue the case for continued investments in maize stover R&D and thus reigniting earlier dual-purpose crop research in general. Despite the highlighted prospects of selecting for stover quantity and quality traits, challenges remain in terms of actually releasing some of such dual-purpose maize varieties and having them widely adopted. Within the cereal seed sector across the globe, the maize seed sector tends to be the most developed given the availability of maize hybrids. Nonetheless, commercializing dual-purpose maize varieties (including hybrids) may pose new challenges – for instance in terms of premium-pricing of dual-purpose seed or using such additional trait as embodied valueaddition to enhance or secure market share. This may challenge existing business models and bring in market segmentation implications in view of spatial and economic heterogeneity – be it by continent, country, ecology or farmer type. The development and utilization of dual-purpose maize varieties thus calls for further multi-disciplinary and multi-stakeholder R&D. Further fine-tuning R&D is also needed – including monitoring of the eventual release and uptake of purposively developed dual-purpose maize varieties. Enhanced processing may also open new opportunities. For instance, Anandan et al. (2013) reported that feed mash proved superior to feed blocks in terms of animal productivity. The core focus of the special issue has been on grain-fodder trade-offs for dual-purpose maize varieties. However, there are other potential trade-offs, including alternative stover uses such as the potential use of biomass for soil health within conservation agriculture systems, as explored in the case of Mexico (Hellin et al., 2013) and elsewhere (Valbuena et al., 2012). The findings presented by Homann-Kee Tui et al. (2013) however question some of the assumptions of Valbuena et al. (2012) – particularly that simple increases in quantity of biomass, for example through fertilizer application, will automatically reduce competition between biomass usage for livestock feeding and soil improvement. Livestock have limited intake capacities for bulky feed such as stover and increasing quantitative availability of a given stover above the intake capacity of livestock will not improve fodder resources beyond the possibility of making fodder cheaper. Therefore, quantitative and quality aspects of stover work, or more generally crop residue work, in crop improvement cannot strictly be separated
111
either in the context of livestock feeding or biomass allocation and partitioning for different purposes. Biomass use trade-offs are also not limited to the direct competition between stover for feed or soil health and may actually be further entangled by dual-purpose maize varieties. For instance, the higher fodder value of dual-purpose maize may increase stover off-take (in situ or ex situ) and thereby exacerbate competition between uses. Furthermore, dual-purpose maize varieties typically have higher digestibility as feed, but this may also have implications for their digestibility by soil biota and susceptibility to weathering vis-à-vis their retention as mulch within conservation agriculture systems (Erenstein, 2002). Some of these trade-offs are inherently complex and also may not be limited to maize – and merit further research. The focus of the special issue and R&D thus far has been on dual-purpose maize for grain-fodder. As illustrated by the USA case, biofuel considerations may add uses and challenges, both in terms of alternative grain use (bioethanol, Klopfenstein et al., 2013) or alternative stover use for second generation (cellulosic) biofuels (Lorenz et al., 2010). The latter may imply new trade-offs on stover use, but also may give rise to: 1) new byproducts with higher feed value, as was the case with the first generation grainbased biofuels; 2) spin-off technologies for upgrading the fodder value of ligno-cellulose biomass for livestock. These developments also have potential implications for moving beyond the current focus on dual-purpose to the future development and use of multipurpose maize varieties (food-feed-fuel) in industrial agriculture and beyond.
Acknowledgements The special issue including this synthesis paper benefited from support from various sources – as variously acknowledged in the preceding papers. Here we would like to particularly refer to the more detailed acknowledgements in the overview paper (Erenstein et al., 2013). The views expressed here are those of the authors and do not necessarily reflect the views of the donor or the authors’ institution. The usual disclaimer applies.
References Anandan, S., Khan, A.A., Ravi, D., Sai Bucha Rao, M., Ramana Reddy, Y., Blümmel, M., 2013. Identification of a superior dual purpose maize hybrid among widely grown hybrids in South Asia and value addition to its stover through feed supplementation and feed processing. Field Crop Res. 153, 52–57. Blümmel, M., Rao, P.P., 2006. Economic value of sorghum stover traded as fodder for urban and peri-urban dairy production in Hyderabad, India. Int. Sorgh. Mill. Newslett. 47, 97–100. Byerlee, D., Iqbal, M., Fischer, K.S., 1989. Quantifying and valuing the joint production of grain and fodder from maize fields: evidence from northern Pakistan. Exp. Agr. 25, 435–445. De Groote, H., Dema, G., Sonda, G.B., Gitonga, Z.M., 2013. Maize for food and feed in East Africa–The farmers’ perspective. Field Crop Res. 153, 22–36. Erenstein, O., 2002. Crop residue mulching in tropical and semi-tropical countries: an evaluation of residue availability and other technological implications. Soil Till. Res. 67, 115–133. Erenstein, O., Samaddar, A., Teufel, N., Blümmel, M., 2011. The paradox of limited maize stover use in India’s smallholder crop-livestock systems. Exp. Agr. 47, 677–704. Erenstein, O., Blümmel, M., Grings, E., 2013. Potential for dual-purpose maize varieties to meet changing maize demands: Overview. Field Crops Res. 153, 1–4. Ertiro, B.T., Twumasi-Afriyie, S., Blümmel, M., Friesen, D., Negera, D., Worku, M., Abakemal, D., Kitenge, K., 2013a. Genetic variability of maize stover quality and the potential for genetic improvement of fodder value. Field Crop Res. 153, 79–85. Ertiro, B.T., Zeleke, H., Friesen, D., Blümmel, M., Twumasi-Afriyie, S., 2013b. Relationship between the performance of parental inbred lines and hybrids for food-feed traits in maize (Zea mays L.) in Ethiopia. Field Crop Res. 153, 86–93. Hellin, J., Erenstein, O., Beuchelt, T., Camacho, C., Flores, D., 2013. Maize stover use and sustainable crop production in mixed crop–livestock systems in Mexico. Field Crop Res. 153, 12–21.
112
M. Blümmel et al. / Field Crops Research 153 (2013) 107–112
Homann-Kee Tui, S., Blümmel, M., Valbuena, D., Chirima, A., Masikati, P., van Rooyen, A.F., Kassie, G.T., 2013. Assessing the potential of dual-purpose maize in southern Africa: a multi-level approach. Field Crop Res. 153, 37–51. Klopfenstein, T.J., Erickson, G.E., Berger, L.L., 2013. Maize is a critically important source of food, feed, energy and forage in the USA. Field Crop Res. 153, 5–11. Kristjanson, P.M., Zerbini, E., 1999. Genetic enhancement of sorghum and millet residues fed to ruminants: An ex ante assessment of returns to research. Impact Assessment Series 3. ILRI, Nairobi, Kenya. Lorenz, A.J., Gustafson, T.J., Coors, J.G., Leon, N.D., 2010. Breeding maize for a bioeconomy: A literature survey examining harvest index and stover yield and their relationship to grain yield. Crop Sci. 50, 1–12. Lukuyu, B.A., Murdoch, A.J., Romney, D., Mwangi, D.M., Njuguna, J.G.M., McLeod, A., Jama, A.N., 2013. Integrated maize management options to improve forage yield and quality on smallholder farms in Kenya. Field Crop Res. 153, 70–78. Muttoni, G., Palacios-Rojas, N., Galicia, L., Rosales, A., Pixley, K.V., de Leon, N., 2013. Cell wall composition and biomass digestibility diversity in Mexican maize (Zea mays L) landraces and CIMMYT inbred lines. Maydica 58, 21–33. Notenbaert, A., Herrero, M., De Groote, H., You, L., Gonzalez-Estrada, E., Blümmel, M., 2013. Identifying recommendation domains for targeting dual-purpose maizebased interventions in crop-livestock systems in East Africa. Land use policy 30, 834–846. Ramana Reddy, Y., Ravi, D., Ramakrishna Reddy, C., Prasad, K.V.S.V., Zaidi, P.H., Vinayan, M.T., Blümmel, M., 2013. A note on the correlations between
maize grain and maize stover quantitative and qualitative traits and the implications for whole maize plant optimization. Field Crop Res. 153, 63–69. Ravi, D., Khan, A.A., Saibutcharao, M., Blümmel, M., 2013. A note on suitable laboratory stover quality traits for multidimensional maize improvement. Field Crop Res. 153, 58–62. Sharma, K., Pattanaik, A.K., Anandan, S., Blümmel, M., 2010. Food-Feed Crops Research: A Synthesis. Animal Nutrition and Feed Technology 10s, 1–10. Shiferaw, B., Prasanna, B., Hellin, J., Banziger, M., 2011. Crops that feed the world 6. Past successes and future challenges to the role played by maize in global food security. Food Sec. 3, 307–327. Valbuena, D., Erenstein, O., Homann-Kee TUI, S., Abdoulaye, T., Claessens, L., Duncan, A.J., Gerard, B., Rufino, M.C., Teufel, N., van Rooyen, A., van Wijk, M.T., 2012. Conservation Agriculture in mixed crop-livestock systems: Scoping crop residue trade-offs in Sub-Saharan Africa and South Asia. Field Crop Res. 132, 175– 184. Van Soest, P.J., 1994. Nutritional Ecology of the Ruminant. University Press, Cornell. Vinayan, M.T., Babu, R., Jyothsna, T., Zaidi, P.H., Blümmel, M., 2013. A note on potential candidate genomic regions with implications for maize stover fodder quality. Field Crop Res. 153, 102–106. Zaidi, P.H., Vinayan, M.T., Blümmel, M., 2013. Genetic variability of tropical maize stover quality and the potential for genetic improvement of food-feed value in India. Field Crop Res. 153, 94–101.