Keywords

Finger millet is an annual, small grain self-pollinated allotetraploid (2n = 4x = 36) plant mainly grown in two major continents, Africa and Asia (Sood et al. 2017, 2019). It has wide adaptability and its cultivation extends from sea level to higher elevations in the Himalayas (Gupta et al. 2012). Finger millet has the ability to grow under harsh conditions in diverse environments and has great food value in terms of nutritional profile. It is grown in dry and semi-dry regions for both grains and forage. The crop has exceptional adaptation under low moisture conditions and provides assured harvest under dry spells in marginal areas, suitable for contingency crop planning (Sood et al. 2016). The grains can be stored for years and have many health-promoting benefits besides a very good nutritional profile. The finger millet forage is also highly palatable and nutritious. At the global level among millets, finger millet occupies the fourth place, after major coarse grains, i.e., sorghum, pearl millet, and a minor millet, foxtail millet (Gupta et al. 2012).

Finger millet is a crop under the Poaceae family and Chloridiodeae subfamily. It is the only millet that belongs to the tribe Chlorideae, whereas Piniceae is the tribe for all other millets. The finger millet panicles resemble the shape of the human thumb and fingers, therefore its English name has been given as “finger millet”.

Global estimates for precise area and production data on finger millet are not available. However, the literature estimates reveal that 5 million tons of finger millet grains were produced from 4 to 4.5 million ha area globally. The total production of finger millet in Africa was about 2 million tons, which was slightly lower than in India (2.2 million tons) (Sood et al. 2019). In Africa, finger millet is cultivated in eastern and southern African countries mainly Ethiopia, Kenya, Malawi, Tanzania, Uganda, Zaire, Zambia, and Zimbabwe. India and Nepal are the major finger millet producers in Asia, but the crop is also grown to some extent in China, Bhutan, Japan, and Sri Lanka. The latest estimates on area, production, and productivity of the crop in India are 67.2 thousand ha, 61.6 thousand tons, and 1332 kg/ha, respectively, which indicate a considerable decline in area, production, and productivity in comparison to previous years (Directorate of Economics & Statistics, Government of India, 2020, https://eands.dacnet.nic.in/PDF/At%20a%20Glance%202019%20Eng.pdf). Among various finger millet-producing states of India, Karnataka tops the list with  >50% area, followed by Maharashtra and Uttarakhand (Chandra et al. 2020).

1.1 Origin and Phylogeny

Earlier botanists argued and suggested the origin of finger millet as India based on historical records and mention of finger millet by Sanskrit writers as ragi or rajika (De Candolle 1886; Dixit et al. 1987). Burkill (1935) proposed that E. coracana is the cultivated form ascended through selection in India from wild species E. indica (L) Gaertn. It was further stated that finger millet originated in India and Africa independently (Vavilov 1951). More precise studies later explained that E. coracana is of African origin, domesticated in Western Uganda and Ethiopian highlands around 5000 years BC. The crop reached the Western Ghats in southern parts of India  ~3000 BC (Hilu and De Wet 1976a; Hilu et al. 1979a, b; Fuller 2014).

The archaeologic studies in Ethiopia dating back to the third millennium BC confirm the African origin of finger millet (Hilu et al. 1979a, b). Finger millet has two discrete races, African highland and Afro-asiatic lowland race. The former seems to be a derivative from Eleusine africana, which also resulted in the African lowland race. As per Hilu and De Wet (1976b), this African lowland race reached India as an Afro-Asiatic lowland race around 3000 BC. Wide phenotypic variability has been reported in African germplasm collections in comparison to Indian collections in many studies, strengthening the claim that Africa is the primary center of origin of finger millet. During its cultivation for years in the Indian subcontinent, the gene flow has resulted in great diversity in local and primitive crop cultivars, making India the secondary center of origin of the crop (Padulosi et al. 2009; Sood et al. 2016).

The cultivated finger millet species (E. coracana) is tetraploid with a basic chromosome number of 9 (2n = 4x = 36) (Sood et al. 2019). The species E. Africana (2n = 36) exhibits great similarity with E. coracana in morphological features and gene flow occurs between them (Hilu and de Wet 1976b). First, the cytological studies showed that E. indica is one of the genome (AA) contributors to the cultivated species E. coracana and later chloroplast genome studies revealed E. indica to be the maternal genome donor of E. africana. The cytological studies also confirmed that E. intermedia and E. tristachya also belong to the genomic group “A” along with E. indica and these three species have a close genetic grouping (Mallikharjun et al., 2005). The results of genomic in situ hybridization and ribosomal DNA sites comparison on the chromosomes of diploid and polyploid species inferred E. indica and E. floccifolia as two progenitors of E. coracana and E. africana (Bisht and Mukai 2000, 2001). Later, Neves et al. (2005) refuted E. floccifolia as a B genome donor based on genome analysis using nuclear internal transcribed spacers and plasmid trnT-trnF sequences. The results of molecular markers of Pepc4 gene inferred that the species E. coracana, E. africana and E. kigeziensis are of allopolyploid origin (Liu et al. 2011), and strengthened the claim of two separate allopolyploidization origins for E. africana-E. coracana group and E. kigeziensis. However in both the cases, the diploid species group, E. indica-E. tristachya was recognized as the maternal parent (Liu et al. 2014), the paternal parents could not be traced for both the events as they might not exist now (Zhang et al. 2019). The placement of Eleusine in the subfamily Chloridoideae is undisputed.

1.2 Taxonomy and Classification

Eleusine is a small genus with 9–10 species distributed across continents mainly Africa, Asia, and South America in tropical and subtropical habitats (Hilu 1981; Phillips 1972). Out of the nine species, eight species, E. coracana, E. africana, E. indica, E. kigeziensis, E. intermedia, E. multiflora, E. floccifolia, and E. jaegeri (Phillips 1972) belong to East Africa, which is the center of diversity for the genus, Eleusine. The only species which has emerged outside Africa is E. tristachya (Neves 2011). This species is native to South America. The three species under the genus Eleusine, E. coracana, E. tristachya, and E. indica has wide adaptation ranging from sea level to high hills, while E. jaegeri, E. floccifolia, E. kigeziensis, E. intermedia, and E. multiflora are adapted to upland habitats and grow well in areas above 1,000 m amsl. Both diploid and polyploidy species are found in the genus Eleusine with three basic chromosome numbers (x = 8, 9, 10). The species has been classified into two separate groups, annual and perennial based on their growth habit.

The species under the genus Eleusine lack clear separation based on the taxonomic relationships, therefore, the gene pool does not have defined boundaries with respect to the primary, secondary, and tertiary gene pool species. However, phylogenetic studies in the Eleusine genus have categorized the species into three classes. Domesticated and wild forms of finger millet have been placed in the primary gene pool while diploid wild species progenitors constitute the secondary gene pool and all other species belong to the tertiary gene pool (Sood et al. 2019). The primary gene pool includes E. coracana subsp. africana and Eleusine coracana subsp. coracana, secondary gene pool comprises E. indica, E. floccifolia, and E. tristachya and the species E. intermedia, E. jaegeri, E. kigeziensis, E. multiflora, and E. semisterlis (syn. E. compressa) form the tertiary gene pool (Table 1.1).

Table 1.1 Eleusine species, habitat, and salient features
Full size table

The cultivated Eleusine form can be easily distinguished from wild forms based on its firm spikes and large and ball-shaped grains (Neves 2010). As discussed above, eight species of the Eleusine genus are native to Africa, which also includes the wild species E. coracana subspecies africana which has moved to America and Asia, particularly in warmer parts. Due to natural interbreeding between cultivated and wild finger millet species, many new hybrid combinations have appeared, most of which are lookalike companion weeds of the crop. This has been studied and demonstrated through scientific evidence using molecular markers that gene flow between subsp. africana and subsp. coracana has happened in nature (Dida et al. 2008).

1.3 Eleusine Germplasm Collections

India holds the largest germplasm collections of 10,507 accessions in the National Bureau of Plant Genetic Resources, New Delhi, under long-term conservation. Although most of these accessions belong to cultivated species and are indigenous, the collection also contains 6 wild species. ICRISAT in India holds about 5,957 global accessions, of which 105 are of wild species. The major collection of wild species of finger millet is conserved and maintained at Agricultural Research Station, Griffin, Georgia, USDA, which has 17 wild species (E. floccifolia, E. indica, E. jaegeri, E. multiflora, and E. tristachya) out of the total collection of 766 accessions. Eastern Africa, which is the primary center of origin of the crop, Kenya, Zimbabwe, Uganda, and Zambia hold about 1902, 1158, 1155, and 497 accessions. Besides, many other South Asian and African countries hold small germplasm collections (Sood et al. 2019). The global finger millet collection at ICRISAT has been characterized and core, as well as the mini-core set, have been developed for use in breeding and genomics studies (Upadhyaya et al. 2006). Although global diversity of finger millet has been conserved and important accessions have been identified, the wild species particularly, E. coracana subsp. africana and progenitors also need due attention (Neves 2010).

Based on compactness and shape of inflorescence, finger millet germplasm has been classified into races and subraces. The salient characters of races and subraces under each species are given in Table 1.2 (Prasada Rao et al. 1993; Bharathi 2011).

Table 1.2 Races and subraces in finger millet germplasm and their features
Full size table

1.4 Crop Adaptation and Floral Biology

Finger millet has wide adaptation and can be grown in a wide habitat because of its hardy nature and short growing season. Being a C4 crop, it is highly efficient in adapting to environmental fluctuations and climate change. It can be grown from coastal plains to high hills, between 500 and 2,400 m above sea level (Fig. 1a). The genotype response although varies with agro-ecologies. Short-duration varieties are generally adapted to highlands and medium to long-duration cultivars do well in plains and tropical areas. The crop completes its life cycle in 75–160 days. It is generally grown in drylands as a rainfed crop but irrigated crop does well in terms of grain yield and the potential yield under irrigated conditions is around 5–6 t/ha. The crop can tolerate some waterlogging, but water stagnation severely affects crop productivity. Finger millet volunteers, shattering types are common in crop fields and difficult to identify early in the season. They look like normal plants but their seed starts shattering even in the immature stage itself.

Fig. 1.1
figure 1

Finger millet crop in Uttarakhand hills, India. a Crop stand of improved variety VL 376. b Immature panicles of finger millet variety VL 376. c Mature panicles of finger millet variety VL 376

Full size image

The height of finger millet plants varies from 30 to 150 cm and mostly medium height cultivars are grown in India (100–130 cm). Finger millet inflorescence is in the whorl of 2–11 digitate, straight or slightly curved spikes (Fig. 1b and c). The spike is 8–15 cm long and 1.3 cm broad. In each spike, about 50–70 spikelets are arranged alternatively on one side of the rachis (Gupta et al. 2012). Each spikelet contains 3–13 florets. The florets have three stamens and the gynoecium is bi-carpellary, uni-locular with a superior ovary having two styles with feathery branched stigma (Seetharam et al. 2003). The anthers surround the stigma, which ensures self-pollination. Finger millet grains vary in shape from round-oblong/oval, and white -reddish-brown in grain color (Fig. 1.2). The surface of finger millet grains is finely grooved and its pericarp is fused to the surface of the grain. Finger millet wild relatives have seed shattering trait, and at maturity seeds disperse naturally from the panicle, the cultivated species lost the seed shattering trait during domestication but it varies from cultivar to cultivar (de wet et al. 1984). Some cultivars are hard threshers while others still disperse some seed naturally at maturity.

Fig. 1.2
figure 2

Cleaned finger millet grains after threshing. a White grains of finger millet variety VL 382. b Reddish-brown grains of finger millet variety VL 376

Full size image

The pollination system studies in finger millet revealed that pollen dust covers the stigma before it comes out of the lemma, leaving no or little chance for cross-pollination (Gupta et al. 2012). The spikelets opening follows the top to the bottom pattern in each spike, and florets in the spikelets open from bottom to top. The studies suggest that one floret in the spikelet opens per day. The flowering completes in around 5–7 days. Anthesis happens early morning between 1.00 and 5.00 a.m., when anthers dehisce to pollinate their stigmas (Gowda 1997). Dodake and Dhonukshe (1998) reported that pollen grains remain viable in finger millet for about 20 min, while stigma receptivity stays for up to 5 h. The estimation of natural crossing does not exceed 1% in finger millet (Seetharam 1998). Inter-varietal hybridization using the contact method (Ayyangar 1934) is the simplest and easiest way for recombination breeding. For successful hybridization, genotypes having dominant character such as pigmentation on nodes are used as the male parent. This helps in the identification of true hybrids in the F1 generation. However, inducing male sterility through hot water treatment for 5 min at a temperature of 48–52 °C of immature inflorescence on the 3rd to 4th day of emergence was effective in getting few true hybrid seeds (Sood et al. 2019). Genetic male sterility (GMS) and partial GMS source have been identified in the crop but are of little use due to maintenance problem and varying level of sterility/fertility in different locations (Gupta et al. 1997; Gowda et al. 2014; Sood et al. 2019).

1.5 Genome Size

Using Feulgen microspectrophotometry, nuclear DNA content of Eleusine spp. was first reported by Hiremath and Salimath (1991), which was later substantiated by Mysore and Baird (1997) with more accurate analysis using laser flow cytometry. The 2C DNA content of different Eleusine species varied from 1.51 to 3.87 pg. The cultivated species, E. coracana was found to have 3.36–3.87 pg 2C DNA followed by E. coracana subsp. africana (3.34 pg), E. indica (1.61–1.76 pg), E. tristachya (1.51 pg), E. floccifolia (2.0 pg), E. multiflora (2.65 pg), and E. jaegeri (1.90 pg). A recent study reported 1.20, 1.84, 1.14, 1.21, 2.52 pg 2C DNA content in E. jaegeri, E. multiflora, E. tristachya, E. indica, and E. coracana subsp. africana, respectively (Hittalmani et al. 2017). In their study, the cultivated species E. coracana was found to have 3.01 pg 2C DNA content and 1453 Mb genome size. The analysis of genome size of wild species showed a range of 580 Mb in E. jaegeri to 1217 Mb in E. coracana subsp. africana. E. coracana subsp. Coracana, and E. coracana subsp. africana were found to have almost similar genome size which was attributed to the domestication of E. coracana subsp. coracana from E. coracana subsp. africana (Hittalmani et al. 2017). Although in comparison to many types of grass and other plants, the genome size of the Eleusine species is small, still it is too large for genomics studies (Neves 2010).

1.6 Genetic Improvement

This spatial isolation of the crop in India and Africa has led to the appearance of two genetically and morphologically diverse gene pools. However, studies conducted on genetic diversity in African and Indian collections presented much larger variation for inflorescence color in African accessions in comparison to Indian collections. Many studies conducted on the phenotypic evaluation of Indian and African germplasm showed wide variation for inflorescence types in both gene pools. The studies reported that most Indian accessions inflorescence belong to race vulgaris, i.e., they have semi-compact to compact ears while varied ear types extending from open to fist-shaped, mostly belonging to two major races plana and compacta were found in African accessions. Accessions in both the gene pools also vary for many quantitative traits (Naik et al. 1993). More diversity in the African gene pool has been attributed to gene flow from wild species E. africana into cultivated finger millet.

Various DNA-based molecular markers have been used to study the genetic diversity of the finger millet gene pool. Both genomic and genic simple sequence repeat (SSR) markers have been used markers for profiling finger millet accessions to study the genetic diversity. Due to the nonavailability of SSRs earlier studies used random amplified polymorphic DNA (RAPD) markers. Most of these studies clustered the finger millet accessions into two major groups, belonging to two distinct gene pools, i.e., African and Indian gene pools. In a study of Indian accessions, DNA markers could clearly classify accessions of North India and southern India. The South Indian accessions were found to be genetically close to African accessions. The results of genetic diversity studies substantiate that accessions of southern India are closer to African genotypes due to their origin from wild species E. africana, however, the north and northeast accessions are different and the uniqueness of such gene pool needs to be explored (Panwar et al. 2010a, b).

The introduction and use of African germplasm in India resulted in a higher genetic gain in finger millet breeding. Indo-African crosses in finger millet generated more variability and diverse parents resulted in higher productivity of finger millet, which increased >50% in Karnataka State and around 60 percent in Tamil Nadu State in India (Seetharam 1982; Nagarajan and Raveendran 1983). Blast is the major biotic stress affecting finger millet productivity and the identification of stable sources of resistance is the key to developing resistant genotypes. Screening of diverse germplasm particularly African germplasm has resulted in the identification of stable sources of resistance for the blast in finger millet, which has been used to develop resistant varieties through recombination breeding (Seetharam 1998). To date, more than 30 finger millet varieties have been released in India, where the major breeding objectives were maturity duration, grain yield, fodder yield, and disease resistance. The least emphasis was laid on nutritional quality traits earlier but now the nutritional quality is the integral component of finger millet breeding programs in India and Africa.