Introduction
Crop production through organic farming (OF) has been promoted as sustainable and ecologically safe (Reganold and Wachter, Reference Reganold and Wachter2016; Muller et al., Reference Muller, Schader, El Hage, Bruggemann, Isensee, Erb, Smith, Klocke, Leiber, Stolze and Niggli2017). The potential of OF to feed the current and future population has received mixed opinions while some studies support feeding sustainably through OF (Liu et al., Reference Liu, Meng, Bo, Cheng, Li, Guo, Li, Zheng, Liu, Ning, Wu, Yu, Feng, Wuyun, Li, Li, Zeng, Liu Shi and Jiang2016; Muneret et al., Reference Muneret, Mitchell, Seufert, Aviron, Djoudi El, Petillon, Plantegenest, Thiery and Rusch2018; Eyhorn et al., Reference Eyhorn, Muller, Reganold, Frison, Herren Hans, Luttikholt, Mueller, Sanders and El-Hage2019; Krauss et al., Reference Krauss, Berner, Perrochet, Frei, Niggli and Mader2020) and others negate it due to relatively lower yields (Kirchmann et al., Reference Kirchmann, Bergstrom, Katterer, Andrean, Andersson, Kirchmann and Bergstrom2008, Reference Kirchmann2019; de Ponti et al., Reference de Ponti, Rijk and Van Ittersum2012; Seufert et al., Reference Seufert, Ramankutty and Foley2012; Gabriel et al., Reference Gabriel, Sait, Kunin and Benton Tim2013; Ponisio et al., Reference Ponisio, Gonigle Leithen, Mace, Palomino, Valpine and Kremen2015). OF has special relevance to small and marginal household farmers practicing farming through traditional systems. Maize is one of the major crops and cardinal to the agricultural systems of North Eastern hill (NEH) region of India with diverse uses both as food and feed. The region is blessed with unparalleled maize diversity harbouring unique landraces (LR; Dhawan, Reference Dhawan1964; Singh, Reference Singh1977; Sachan and Sarkar, Reference Sachan and Sarkar1986; Prasanna and Sharma, Reference Prasanna and Sharma2005; Prasanna et al., Reference Prasanna, Sharma, Wasala, Singode, Kumar, Guleria, Sekhar, Karuppaiyan, Srinivasan and Gupta2009; Prasanna, Reference Prasanna2010; Rahman and Karuppaiyan, Reference Rahman and Karuppaiyan2011; Sharma et al., Reference Sharma, Partap, Sharma, Rasul and Awasthe2016; Prakash et al., Reference Prakash, Zungare, Muthusamy, Chand, Kamboj, Bhat and Hossain2019; Kapoor et al., Reference Kapoor, Singh, Avasthe, Raj, Singh and Lepcha2022). Despite of rich genetic wealth, the NEH region harvests an average productivity of 1785.71 kg/ha as compared to 3065 kg/ha at national level (Anonymous, 2020). Among NE states, Sikkim besides achieving the feat of first organic state globally is also a haven for maize enthusiasts owing to remarkable maize diversity (Prasanna and Sharma, Reference Prasanna and Sharma2005; Karuppaiyan and Avasthe, Reference Karuppaiyan and Avasthe2006; Prassana, Reference Prasanna2010; Kapoor et al., Reference Kapoor, Avasthe, Gopi, Kalita, Singh, Babu and Das2014).
Remarkable yield gains have been achieved through improved maize varieties and hybrids under conventional systems (CS) with chemical-based fertilizers and plant protectants. Still large yield disparities are observed in areas dominated by OF or low-input marginal areas.
Genetic improvement for organic agriculture has been advocated to be an essential strategy for development of cultivars for organic conditions (Van Bueren et al., Reference Van Bueren, Goldringer, Scholten and Ostegard2007; Van Bueren et al., Reference Van Bueren, Jones, Tamm, Murphy, Myers, Leifert and Messmer2011; Herrera and Ortiz, Reference Herrera Leonardo and Ortiz2015; Nuijten et al., Reference Nuijten, Messmer and Lammerts van Bueren2017). Cultivars derived from high-input conditions may fail to show same response under organic conditions or may cause variation in unique trait expression (Murphy et al., Reference Murphy, Campbell, Lyon and Jones2007; Loschenberger et al., Reference Loschenberger, Fleck, Grausgruber, Hetzendorfer, Hof, Lafferty, Marn, Neumayer, Pfaffinger and Birschitzky2008; Reid et al., Reference Reid, Yang, Salmon, Navabi and Spaner2011; Van Bueren et al., Reference Van Bueren, Jones, Tamm, Murphy, Myers, Leifert and Messmer2011). Moreover, majority of the crop varieties (95%) currently under organic production that are conventionally bred under high-input conditions generally lack traits for organic and low-input agriculture (Van Bueren et al., Reference Van Bueren, Jones, Tamm, Murphy, Myers, Leifert and Messmer2011). While some studies supported a separate maize breeding programme for organic systems (Burger et al., Reference Burger, Schloen, Schmidt and Geiger2008; Revilla et al., Reference Revilla, de Galarreta, Malvar, Landa and Ordas2015; Hoffman et al., Reference Hoffman, Abel, Pollak, Goldstein, Pratt, Smith, Montgomery, Grant, Edwards and Scott2018), others suggested it to be a part of on-going conventional maize improvement programme (Lorenzana and Bernerdo, Reference Lorenzana and Bernerdo2008) with essential traits (Schmidt and Burger, Reference Schmidt and Burger2008, Reference Schmidt and Burger2010; Goldstein et al., Reference Goldstein, Schmidt, Burger, Messmer, Pollak, Smith, Goodman, Kutka, Pratt, ETL and Myers2011). Though Coordinated Maize Improvement Program in India has been entirely focused on the development and promotion of single-cross hybrids (SCH) for different ecologies, LR and improved open-pollinated varieties (OPVs) cover a major share of maize cultivars in hills of Eastern and North Eastern India. Maize LR still support subsistence farming in the region mainly due to yield stability, low disease incidence, superior sensory attributes, minimal seed maintenance cost and various ritualistic and cultural values attached to the local cultivars.
Although enhancing maize productivity via system approach under maize-based cropping system has shown encouraging results in hill ecology (Babu et al., Reference Babu, Singh, Avasthe, Yadav and Rajkhowa2016, Reference Babu, Singh, Avasthe, Yadav, Das, Singh, Mohapatra, Rathore, Chandra and Kumar2020; Singh et al., Reference Singh, Babu, Avasthe, Yadav, Das, Mohapatra, Kumar, Singh and Chandra2021), studies with respect to genetic improvement for low-input organic systems, especially for hill ecologies, are still lacking. Low soil fertility, heavy rainfall, low sunshine hours, heavy weed infestation and dominance of local maize cultivars are major deterrents in achieving high yields in NEH region. Maize breeding programme specifically for organic systems requires separate resources, specific testing conditions hence reasons for selection specifically under organic systems be determined. As of now the current strategy relies on testing conventionally bred varieties/hybrids under organic conditions and their promotion based on yield superiority and quality traits. Knowledge on phenological and agronomic traits is much required beforehand for utilizing them in maize improvement programme. Realizing yield gains in maize that hybrids have achieved under CS, our study explored the data to examine if hybrids have significant yield advantage over OPVs and LR under low-input organic systems too. The study explored yield and yield-related traits and their interrelationships under organic system utilizing diverse genetic resources comprised of LR, OPVs and SCH.
Materials and methods
Experimental site
Field experiments were conducted at a research farm of ICAR Research Complex for NEH Region, Sikkim Centre, Tadong, Gangtok, India located at an altitude of 1320 m above mean sea level at 27o32’N latitude and 88o60’E longitude (Supplementary Fig. S1). Climatic conditions during experimental crop season during each year are shown in Supplementary Table S1. The field experimental site is a certified organic farm and has been under organic management since 2003. The soil of the experimental site is Haplumbrept and sandy loam in texture, moderately deep with pH of 5.97 and soil organic carbon content of 1.38.
Experimental material and layout
Maize trials for the proposed study were a part of our ongoing maize improvement programme for organic conditions for hills where a new set of germplasm which includes LR, improved OPVs and hybrids is tested every year at our centre. The experimental material consisted of locally cultivated LR, improved composites/OPVs and single-cross F1 hybrids. Data from a total of 373 maize genotypes that comprised of 171 LR, 51 OPV and 151 SCH that were tested over a period of 8 years during spring season of 2009–2015 and 2019 were taken for the present study. Each year, the trials were laid out in a randomized block design with plant spacing of 60 cm × 20 cm. Basal dose of Farm Yard Manure (FYM) at 10 t/ha was applied in experimental fields a fortnight before sowing. A mixture of Bio-NPK at 5 kg/100 kg FYM was applied during earthing up at knee high stage. Prophylactic spray of neem oil was done for pest management. Data on 12 quantitative traits, i.e. plant height (PH), days to 50% tasselling (DF), days to 50% silking (DS), days to 75% dry husk (DM), grain yield per ha (GYPHa, plot yield converted to per ha basis), anthesis–silking interval (ASI), cob length (CL), cob diameter (CD), kernel rows per cob (KRPC), kernels per row (KPR), number of cobs per plot (NCPT) and test weight (TW, 100 seed weight) were recorded each year on the experimental material.
Data analysis
The significance of trait values among LR, OPVs and hybrids was examined using one-way analysis of variance. Pearson correlations and standardized major axis (SMA) slopes were carried out to explore bivariate trait associations. The significance of regression coefficients between grain yield and other related agronomic traits was tested using SMA regression (Michaletz et al., Reference Michaletz, Cheng, Kerkhoff and Enquist2014). Variables were standardized to have mean 0 and standard deviation of 1 before SMA analysis. Structural equation model (SEM), a combination of factor analysis and multiple regression analysis, was constructed to quantify the multivariate trait interrelationships (Grace and Bollen, Reference Grace and Bollen2005). SEM analysis was performed using R software platform (R Core Team, 2015). Pairwise deletion was done to handle missing data. Adjusted goodness-of-fit statistic and the standardized root mean square residual were used to evaluate overall model fit (Hooper et al., Reference Hooper, Coughlan and Mullen2008).
Results
Differences in mean trait values among maize cultivar groups
Maize genotypes evaluated under different cultivar groups displayed variation for quantitative traits. Summary statistics of trait values for OPVs, LR and hybrids has been shown in Table 1. SCH had significantly higher grain yield as compared to both OPVs and LR. In phenological traits, LR showed greater variability for days to tasselling, days to silking, anthesis–silking interval and days to maturity. LR and SCH were at par for test weight, while OPV displayed significant variation for test weight. OPV exhibited higher variability for number of cobs per plot as compared to LR and SCH. LR were the tallest in height followed by OPV and SCH. High anthesis–silking interval (7.18 days) was recorded in LR while OPV and SCH displayed similar ASI. SCH were the earliest to tassel, silk and mature. OPV in comparison displayed a higher mean value for traits related to cob, i.e. cob length, number of cobs per plot and kernels per row. Mean values for different traits over years in LR, OPV and SCH are summarized in Supplementary Table S1.
Table 1. Summary statistics for trait values in OPVs, landraces and hybrids
Relationship between grain yield and other quantitative traits
Correlation coefficient of different traits with grain yield in LR, OPVs and hybrids is shown in Table 2 and graphically in Supplementary Fig. S2(a)–(d). SCH showed a negative relationship of grain yield with days to tasselling (r = −0.22, P < 0.001), days to silking (r = −0.25, P < 0.001), anthesis–silking interval (r = −0.19, P < 0.018) and days to maturity (r = −0.26, P < 0.001) while LR showed a significantly positive association of grain yield with days to tasselling (r = 0.26, P < 0.001), silking (r = 0.22, P < 0.001) and maturity (r = 0.29, P < 0.001). Plant height displayed a positive relationship with grain yield in all three cultivar groups with a significantly positive association in LR (r = 0.33, P < 0.001) and SCH (r = 0.34, P < 0.001). Number of cobs per plot showed a very strong association with grain yield in SCH (r = 0.96, P < 0.001) while a weak association was recorded in LLR (r = 0.26, P < 0.001). OPV exhibited a non-significant association in majority of the traits except ASI, cob diameter and test weight. Grain yield in OPV showed a significantly positive association with test weight (r = 0.37, P < 0.001) while non-significant in LR and SCH. In all, traits showing positive correlation with grain yield in LR are days to tasselling, days to silking, days to maturity, plant height, cob length, cob diameter, number of cobs per plot, kernel rows per cob and kernels per row while only cob diameter and test weight were associated positively with grain yield in OPV. In SCH, plant height, cob length, cob diameter, number of cobs per plot, kernel rows per cob and kernels per row showed a significantly positive association with grain yield. Details of trait relationship using SMA regression are shown in Supplementary Table S3.
Table 2. Correlation coefficient of traits in OPVs, landraces and single-cross hybrids
***Significant at 1% level, **significant at 5% level of significance. 1OPV, 2landraces, 3single-cross hybrids.
Direct and indirect effects of quantitative traits on grain yield
Based on the adjusted goodness-of-fit statistic, the models fitted the data with AGFI values >0.99 and standardized root mean square residual values <0.08 (Fig. 1(a)–(c)). For LR, traits ASI, DM, NCPT, TW and KPR accounted for 97% of the variation in grain yield. KPR had the highest positive effect on grain yield (β = 0.50, P < 0.05) followed by DM (β = 0.49, P < 0.05). For OPV, kernel rows per cob (β = 0.35, P < 0.05) and test weight (β = 0.52, P < 0.05) contributed highest to the grain yield contributing to 24% of the variation. In case of SCH, days to tasselling, silking, anthesis–silking interval, plant height and number of cobs per plot contributed 97% to the grain yield; among which number of cobs per plot contributed maximum (β = 1.0, P < 0.05) followed by days to tasselling (β = 0.50, P < 0.05).
Figure 1. (a) Structural equation model showing association among grain yield and its associated traits open-pollinated varieties. (b) Structural equation model showing association among grain yield and its associated traits in landraces. (c) Structural equation model showing association among grain yield and its associated traits in single-cross hybrids.
Traits having indirect effects have a crucial role in contributing to grain yield. In LR, kernel rows per cob, cob diameter, days to flowering, plant height and cob length contributed indirectly towards grain yield. Cob length contributed positively (β = 0.76) towards kernel rows per cob while days to flowering contributed negatively towards days to maturity. In OPV, cob diameter has a positive effect on test weight (β = 0.66) while kernel per row had a negative effect on kernel rows per cob. In SCH, test weight, kernel per row, cob length, days to maturity and cob diameter had an indirect influence on yield contributing to traits. Days to maturity had a positive effect on days to flowering, days to silking and ASI. Cob diameter had a positive influence on number of cobs per plot and plant height. Test weight had a negative influence on kernel rows per cob.
Discussion
Results of the study highlighted the trait variability in LR, OPV and SCH and the contribution of different agronomic traits in the formation of grain yield under organic conditions. The findings of the study have special relevance to low-input organic systems in hills where apart from low soil fertility other environmental and economic factors are a deterrent in achieving high grain yields in maize. LR showed maximum variability for most of the traits recorded. These represent collections from farmers’ field located at different locations of Eastern and NEH region which are highly heterogeneous due to their breeding behaviour and population mixtures commonly observed in cross-pollinated crops. LR from NEH region have shown high phenotypic variability for leaf length, width, ear length, ear width, kernel rows per cob, kernels per row and test weight (Goyanka et al., Reference Goyanka, Yadav, Kumari, Tiwari and Kumar2021) and divergence in inbred lines developed from local populations for exploiting heterosis (Naveenkumar et al., Reference Naveenkumar, Sen, Vashum and Sanjenbam2020). Plant height, ear height, grain yield, ear width, number of kernel rows and kernels per row were major contributors to phenotypic diversity of maize LR (Kumar et al., Reference Kumar, Kumari, Rana, Chaudhary, Kumar, Singh, Singh and Dutta2015; Kumari et al., Reference Kumari, Kumar, Singh, Bhatt, Mishra, Semwal, Sharma and Rana2016).
From our local survey, we could infer reasons for dominance of local maize cultivars in local maize varietal system mostly due to their luxuriant vegetative growth exhibited in the form of tall plant height, sweet texture of grains, multiuse of maize grains for local culinary, yield ensurity under stress conditions, unavailability/scarcity of hybrid or OPV seed during crop season, high cost of hybrid seed, less remunerative and predominantly subsistence maize farming. As predicted, hybrids out yielded OPV and LR, where hybrids on an average yielded 95.36% higher than the LR and 58.60% higher than the improved OPVs. This significant yield gap signifies the heterotic effects displayed by SCH under low-input organic environment too. Hybrids bear high number of ears per unit area with better uniform plant stand and plant type which ultimately resulted in high yield gains. Yield superiority of hybrids as compared to OPVs has been reported in earlier studies (Pixley and Banziger, Reference Pixley and Banziger2001; Van Asselt et al., Reference Van Asselt, Battista, Kolavalli and Udry2018; Ndoli et al., Reference Ndoli, Baudron, Sida, Schut, Van heerwaarden and Giller2019; Landoni et al., Reference Landoni, Scapin, Cassani, Borlini, Follador, Giupponi, Ghidoli, Hejna, Rossi and Pilu2020). SCH had the lowest anthesis–silking interval (2.98) as compared to OPV (3.09) and LR (7.18). ASI in maize is an important trait deciding much of the performance of cultivar under stress conditions. High ASI has been related to low cob formation, less seed set and low number of filled cobs. The experimental location is a high rainfall area coinciding with flowering/silking stage. Long anthesis–silking interval in high rainfall areas causes poor pollen availability for higher seed set. Long ASI has more exposure to external factors and hence affected seed set in maize LR.
ASI at flowering accounts for larger variation (70–80%) in grain yield (Bolanos and Edmeades, Reference Bolanos and Edmeades1993; Chapman and Edmeades, Reference Chapman and Edmeades1999). Drought, plant density and shading, low fertility, adaptation and photoperiod response have been analysed as some of the major factors for high ASI (Edmeades et al., Reference Edmeades, Bolanos, Elings, Ribaut and Banziger2000). Selection for shorter ASI significantly enhanced grain yield under low nitrogen (Banziger and Lafitte, Reference Banziger and Lafitte1997).
Grain yield is a complex trait regulated by both genetic and non-genetic factors. Variation in degree and significance of trait association in LR, OPV and SCH indicates relative importance of yield-contributing traits in different maize cultivar groups. Phenology had a significant positive effect on grain yield in LR while it was associated negatively in SCH. LR are typically tall and show luxuriant vegetative growth and took full season to mature. Hybrids are strictly homogenous in nature with predicted duration of phenological stages. Hybrids accumulate high biomass at shorter time even under low-input organic conditions and partitioned to growing cob unlike in LR where partitioning is poor and slow and much of the assimilates are utilized simultaneously in plant growth. Hence, medium duration hybrids with maturity duration from 115 to 120 days are suitable for achieving high yields. Plant height also emerged out to be an important trait showing positive correlation in all groups which pertains to need for high biomass required to synthesize assimilates under areas with high rainfall and low sunshine hours. Magar et al. (Reference Magar, Acharya, Gyawali, Timilsena, Upadhayaya and Shrestha2021) reported positive and significant correlation of grain yield with test weight (r = 0.706), cob length (r = 0.671), cob diameter (r = 0.573) and number of rows per cob (r = 0.539). In traditional farming systems, improved OPVs are quite popular due to their low seed maintenance cost with stable yields for 2–3 years in farmers’ field. Grain yield in OPV is contributed by test weight and cob diameter only. Cultivars with high test weight and cob diameter shall be looked for during selection for improved OPVs for the region. Few traits exhibit positive influence on grain yield in OPV due to less number of OPV cultivars taken in the study as compared to LR and SCH.
Implications for maize improvement for organic farming in hills
Hybrids are generally not promoted under OF with the pre-conceived concept of poor performance under low-input organic system. Dependence on external sources for supply of hybrid seed in each season makes farmers reluctant to adopt hybrids. Yield gains achieved through maize hybrids under OF too have indicated a large scope of yield enhancement in low-input systems. Emphasis shall be made on selection of hybrids showing high plant stand with high number of cobs per unit area. Local maize populations have their own merit as these are evolving populations having broad genetic base with superior quality and stress-resilient traits. Their real worth shall be assessed by screening for tolerance to disease and pests, quality traits, abiotic stresses such as cold, heat and drought which are much needed to identify potential donor/source population for tolerance to biotic and abiotic stress. High ASI in LR influenced grain yield negatively. Factors for high ASI in LR under organic conditions are a topic of detailed separate study and factors that affected the ASI have to be looked in to. OPV showed that moderate yield levels are already in varietal system but require seed replacement every 2–3 years for maintaining varietal purity. Hence, in short term, conventionally bred maize hybrids showing stable and high yields under organic management shall be utilized for organic systems until a maize improvement programme specifically for organic systems is in place or an organically bred hybrid/OPV replaces conventionally bred variety in terms of yield or quality traits. Majority of the farmers in hills are low to marginal landholders; therefore, demand for hybrid seed can be met through ‘Seed Village’ scheme which has already been promoted for meeting seed demand locally.
Conclusion
The study indicates significant morphological variation in maize LR of NEH region; however, their low yield needs to be improved via breeding interventions. Yield superiority of single-cross maize hybrids over OPVs and LR points to superiority of heterotic effects under low-input organic environment. This suggests for popularization and adoption of suitable maize hybrids in NEH region for raising per unit productivity along with safeguarding local maize genetic resources in-situ. Breeding maize programme for organic environment needs to be carried out utilizing local diverse LR.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1479262123000333.
Acknowledgements
The authors thank The Director, ICAR Research Complex for NEH Region, Umiam, Meghalaya for providing necessary facilities for carrying out the study.
Author contributions
C. K. and S. S. conceived and designed the study. M. S. performed the statistical analysis. R. K. A. reviewed the article. A. P., C. R. and M. S. performed and recorded field data. C. K. wrote the article.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
None.
Ethical standards
Not applicable.
