Although sorghum is the third important crop next to teff and maize in Ethiopia, the yield is below the world average. The sorghum worldwide average yield is 1314 kg/ha; developed countries is 3056 kg/ha and that of developing countries is 1127 kg/ha. The Ethiopian national average yield account for up to 2000 kg/ha (EIAR, 2014). One of the reasons for low yield productivity was related to seed delivery system.

In Ethiopia, sorghum formal seed industry is not developed yet (Eltayeb and Sana, 2010). Due to the fact that seed is an essential input in farming and most of the farmers in marginal agro-ecologies solely rely on local seed system, this study explored the indigenous on-farm seed selection, maintenance and seed physiological quality attributes.

The assessment was conducted in Jijiga and Gursum districs of Fafen Zone, Somali Regional State in 2013/2014 cropping seasons. The districts received mean annual rainfall ranges from 380.1 to 756 mm and the mean temperature during the growing period in the area ranges from 20.1 to 22.5°C. Target areas were characterized to the warm semi-arid to cool and humid agro-climatic zone. The altitude is between 500 and 2500 m.a.s.l. The average annual temperature ranges from 27.5-18°C, the average annual rainfall ranges from 200 to 1400 mm, and the potential evapo-transpiration were estimated to be from 1438 to 2099 mm (Oromia Water Works Design and Supervision Enterprise - OWWDSE, 2012).

Sampling procedure

Multi-stage purposive random sampling procedure was followed from higher to lower administrative levels, with agro-pastoralists being sampling units. Eight sorghum producing villages and 16 farmers/agro-pastoralists were selected for interview. The interview was conducted between August to October 2014 using pretested semi structured questionnaires administered by trained enumerators and researchers. Researchers and agro-pastoralists observed sorghum field and assessed the sorghum plant selection criteria, particularly for seed purposes.

A four-stage sampling procedure has been used to select potential areas of sorghum production Somali region. It involves the selection of administrative regions, zones, districts and villages.

First stage: From Fafen Zone, two sorghum producing districts (Jijiga and Gursum) were purposively selected.

Second  stage:   From    selected    districts,    two    villages   were purposively selected based on the area allocated for sorghum crop production.

Third stage: Within each of the two selected villages, at least two major sorghum producing villages were selected randomly by considering the proportional area planted by sorghum in the villages.

Fourth stage: From selected villages, sixteen experienced sorghum producer agro-pastoralist's were selected.

Seed quality analysis

Sorghum stored for seed propose was collected from local stores and analyzed for physical and physiological quality parameters. One kilogram of sorghum seed was collected from farmers/agro-pastoralists stores using sample bag. The seed is submitted to Jigjiga University Dryland Crop Science Laboratory for physical and physiological evaluation.

Physical purity test

Each working sample was divided into two 350 g portions for physical purity analysis. The components were separated into pure seed and inert matter and each component was weighed using analytical balance. Finally, the percentage composition of the seed lot was calculated based on the weight of each component.

Physiological quality test

For germination test, forty-eight seeds were placed on moistened filter paper in the Petri-dishes. Seeds in Petri dishes were placed on working board at room temperature. First and final counts were made on 5 and 9 days after planting.

Statistical analysis

The average value of panicle length, panicle weight, seed physical purity and physiological quality test were calculated and illustrated in table form.

Indigenous folk names associated with sorghum seed selection

In the target study areas, sorghum is known by the collective folk name ‘harur’. This study has identified six sorghum landraces predominantly produced in the Jijiga and Gursum district of the Somali region (Table 1). Each local landrace are their own specific varietal names and help the farmers to describe, identify and distinguish. The agro-pastroalists have particular names and identify the cultivars with  local names such as ‘Eilmi-jama’, ‘Wogera’, 

‘Asse’, ‘Dongae’ and ‘Ahmednasir’. Sorghum field observation with the key informants and researchers identified that the landraces folk names given were linked to distinct morpho agronomic attributes such as plant height, days to maturity, seed color and bird resistance. Furthermore, the folk names were related to the original source of the material, morphology, end use and name of the person who introduced it to the particular location.

Nevertheless, vernacular names may not always correspond to botanical distinctiveness, although they are quite often descriptors used for variety identification. This indicates that the farmers understand crop genetic diversity on the farm and the value associated to it, and such folk names are extremely useful (Liu, 2013). In our study, the agro-pastoralists classified the sorghum local varieties by seed color, grain head, panicle type, bird resistance and days to maturity. The target communities’ practices agro-pastoral production system which relies both on livestock and crop farming, and the sorghum dynamic classification is an indicator for recognition that the significant contribution of the crop is as human food and livestock feed security. Sorghum is widely grown in the Fafen valley of of Gursum districts; particularly in Kumijaro area. This study covers the altitude range of 1517-1842 m.a.s.l (Table 1). Despite the variation in altitude and sorghum cultivars and agronomic practices, six varieties were showed wide adaptation across the elevations studied.

Crop and varietal mixture (poly-cultivar)

In our study, we identified that the agro-pastoral communities have special planting calendar for sorghum landraces. Through field observation and key informants’ interview, we noted that there was special linkage between sorghum seed maintenance and planting. For instance, every agro-pastoralists were maintaining an average of 2-5 sorghum cultivars per household.  In our view, these practices can be considered as climate change coping strategies practiced at small-scale level. Similarly, Yemane et al.  (2009)  reported  that  the  North Ethiopia farmers manage different crop landraces as a climate change coping mechanism. In this study, we noted that the indigenous seed selection was complimented with the need to mitigate the risks of drought by planting diversity of sorghum on a single farm. Using this kind of indigenous knowledge, the agro-pastoralists minimize on-farm micro-environment associated risk such as insect damage, diseases protection, and drought, and hence obtain biomass and biological yield, at the same time prolong varietal stability (Altieri and Merrick, 1987). The sorghum farming communities had established varietal mixtures sorghum production strategy.

The varietal mixture prepared based on different days to maturity was a key indigenous strategy implemented by the agro-pastoralists to cope with the biotic and abiotic stress existing in the study areas.

On the other hand, the reasons for using intra-specific varietal mixture were agronomic and gastronomic followed by different secondary uses (feed, fuel wood). Such strategy was commonly practiced in Sub-Saharan Africa and particularly in Ethiopia for other crops such as legumes (Ruelle et al., 2019). Varietal mixture composition was made in order to meet the differential food grain preferences of household members. The agronomic reasons for the use of varietal mixture include stable yield, reduced lodging, diverse maturity groups for minimization of moisture stress risk, confusing birds attack and improved soil fertility utilization. The gastronomic reasons were taste, baking quality, digestibility and diversity of diet. The different secondary uses were feed, fuel wood and construction materials. Several previous studies have shown that many farmers prefer sorghum due to its cultural practices and end uses (Barnaud et al., 2007; Missihoun et al., 2012; Muui et al., 2013; Mekbib et al., 2009).

Dynamics of on-farm seed selection

In  each   cropping  season,  the  household  head  is  the decision-making unit (mainly the household head) a cultivar which type and how much seeds of a given variety apply to a plant. Seed selection take place at different times, for instance, prior to the harvest, at the time of harvest, after the harvest, at different places e.g. in the field, at a drying or storage facility, or in the home. Table 2 illustrates the farmers’ on-farm seed selection parameters for each local variety. Likewise, Bellon (1996) reported how farmers decided to maintain a pool of genetic resources since the beginning of crop domestication. The quality parameters such as seed color, bird resistance and days to maturity were the main on-farm seed selection criteria. On top of this, the on-farm selection practices of agro-pastoralists were found to include leaf main vein colors (white, yellow, reddish brown), ever green stalk indicator for drought tolerance and uses as fodder, bent panicle indicator for bird resistance, healthy stalk indicator for tolerance to stalk borer. Interestingly, agro-pastoralists in the study areas have had enormous indigenous experience on diversity of sorghum seed selection and promoting local sorghum varieties for a decade. The farmers were found to be practicing several crop improvements approaches such as introduction, simple mass selection, modified mass selection, modified bulk selection, and pure line selection. This indicates that, the sorghum agro-biodiversity in the study area is highly dynamic and strongly associated with livestock production. This revealed the significance of local knowledge and suggests the consideration of small scale-farmers interests in seed system improvement (Berg,1993; Seboka and Hintum van, 2006).

The seed selection diversity criteria detected in these study villages has paramount significance for the establishment  of    on-farm    conservation  programmes; however, it should be well analyzed using biodiversity indices. Similarly, diversity of local variety maintenance practices was reported in Alamata and Raya-Azebo woredas at nothern Ethiopia by Yemane et al. (2009) and in eastern Ethiopia by Mekbib et al. (2009). Teshome et al. (2007) reported that farmers in south Welo were using intra- and inter-specific crop diversities in their field, and as a result, over 30 different sorghum landraces were found in a single field. This practices could also allow the farmers to exploit different microclimates and derive multiple nutritional values and harvest security in times of unpredictable environmental stress. This finding was in agreement with previous report of Brush and Meng (1998), in which crop mixture is purposively practiced for cultural reasons, tastes, gifts, local identity, and for market preference.

Sorghum yield indicators used for seed selection

It was known that the selection of local varieties for seed is usually based on several criteria. Among these, panicle length and panicle weight are predominantly used in our study area. Panicle length and panicle weight were sorghum yield components strongly correlated with sorghum yield. It was obvious that the agro-pastoralists recognized the importance of these traits and included it in their seed selection criteria. The highest panicle length was observed for ‘Elmi-jama’ grown in Kuramatana, whereas that of panicle weight was observed for ‘Adengab’ grown at 2^nd Turuad (Table 3). Both local varieties have a characteristic of long days to maturity with availability of soil moisture (Table 2).  The majority of respondent farmers had begun selection before harvesting  for  a  number  of  traits  such  as big and long panicles, early maturing types, straw quality, disease free and good tiller capacity. They also exercise based on indicators selection of morpho-types from other farmers’ field through careful day-to-day observation of plant morphology at farmers' field and obtain access by making an agreement before harvesting. A number of characters are also mentioned upon which farmers focus when selecting individual plants after harvesting and during storage, such as yield, bigger seed size and colour.

Sorghum seed system in the target area

This study reveals that the informal seed system is dominantly operated in the area. The sorghum seed used for planting could be either own saved or seed obtained from relatives and from other informal networks. The informal seed exchange system enhances the resilience of sorghum production to unpredicted weather such as increased temperatures and diminished rainfall occurring in the areas which resulted in crop failures. In the study areas, seed production was observed as an integral part of grain production. Seed quality controls were also purely informal and solely based on belief and trust between farmers.

The farmers and agro-pastoralists in the study area also obtained seeds from various sources, such as exchange, neighbours and market, through payment in cash. The flow of seeds of named varieties from nearby spatial scales and villages was evident in this study. For instance, there was much similarity between the names of sorghum varieties grown in both Jijiga and Gursum districts and even eastern Ethiopia (Mekbib et al.,  2009), and as the areas are close to one another, the informal seed exchange as part of the local network might cause similarity of sorghum local names. For instance, in bad seasons where drought and birds attack led to crop damage, seed saving may totally fail and seeds may be obtained through market networks with reliable source.

Farmers/agro-pastoralists in the study area use their own farmer-saved seeds (unless unpredicted factors such as drought cause complete crop failure) although they may obtain seeds through exchange, gift or purchase. On the other hand, farmers’ practice of exchanging seed lots for the same named varieties across different large spatial scales has been noted in a number of studies (Almekinders et al.,  1994),  though  an assessment of the direction of the seed flow, the structure of the genetic diversity within the landraces as well as the ethno-botanical knowledge associated with the landraces at respective sites is deemed important for better understanding and utilization of the landraces. Moving landraces within and/or across similar ecological zones could be a powerful way of improving yield production stability. Seed exchange network can be established as a means of facilitating access to locally adapted sorghum genetic resources. However, this may require the establishment of mini-community seed banks (satellite seed banks) designed and managed by farmers themselves. The satellite community seed banks could facilitate mobilizing masses of useful local germplasms at faster rate and can reduce transaction costs.

 In our study, majority of the farmers/agro-pastoralists responded to retain their produce and depend on their own seeds; however, during bad season, exchange seeds and sources of named varieties in kinds, either from neighbours, relatives and markets exists along with seed aid. This was probably due to lack of modern varieties that suits their local environment or farmers criteria and that were the main reasons causing reluctance to use and rather stick to their own genetic resources adapted to their locality.

From the focus group discussion, it was noted that women farmers play the leading role in cleaning of stored seeds. The main storage systems used were plastic bags, storage jars and underground pits. However, underground pit storage is mainly used for grains, even though few farmers are storing part of their seed with farmers witnessing severity of mold damage. In our previous study, we identified that the underground storage pit affected the seed quality attributes including germination and market preference (Mulu and Belayneh, 2016). Interesting seed storage practices explored in our study includes mixing sorghum seed with common bean seed to confuse the pests and minimize the damage from storage pest.

Physical seed purity analysis

The physical purity and hundred seed weight of local varieties seeds obtained from storage were illustrated in Table 4. The physical purity of seeds obtained from Jigjiga district (89.92%) was higher than the one collected from Gursum district (84.58%). Conversely, higher hundred seed weight was observed for Gursum districts.

The physical purity analysis indicated that the contamination with panicle residue appeared to be the major impurity observed in the study samples.

Physiological quality

Seed germination of seeds varied across the districts. Seed samples collected from Gursum exhibited relatively better germination (86.99%) than those obtained from Jigjiga districts (35.77%) (Table 5). However, germination percentage of seeds produced in Jijiga district systems did not fulfil the national standard set (85%) for sorghum seed in Ethiopia. Germination percentages of seed samples collected from different villages were considered; those from Adade (89.58%), Adade (87.5%) and Kuramatana (85.42%) fulfilled the national seed standard whereas none of those seed samples were collected from Jijiga districts (35.77%). Seed samples that had not shown germination percentage was 90%. About 66% of the seed samples had a germination capacity of within 80% for Gursum district. Nonetheless, if more than one quarter of seeds planted by farmers is not viable, then this poses a threat to food security in the region and should be an area of concern. More so, it is not usually true that all seed lots exceeding “the critical minimum germination” are equally good in the eventual test of quality and emergence in the field. This is because, the lower the germination, the poorer the performance since deterioration has occurred even though all lots met the minimum germination recommendation. While it is sometimes possible to compensate for reduced germination by increasing sowing rates so as to achieve a desired population as practiced by many farmers in the district, a point is reached where there can be deleterious effects on yield and quality.

The results of seed quality test reveal that more than one quarter of sorghum seeds used for planting were inadequate in quality parameters including physical seed quality and physiological germination test. This proportion of seed with inadequate quality can lead to poor crop stand and generation of inferior seed quality.

Indigenous seed selection practiced in Fafen Zone has been contributing to enhancement of sorghum genetic resources pool. The contributions of farmers/agro pastoralists are critical to the conservation, use and enhancement of biodiversity. The two districts farmers managed to maintain more than six locally named varieties. Many of the sorghum local varieties are exquisitely fitted to the specific niches and highly associated with the livelihood of the crop livestock farming systems. However, the seed quality is very poor and below national standard. Furthermore, seed quality comprised of more quality aspects and, therefore, it is recommended that further research has to be conducted on storage, seed health and genetic aspects to enable the clarification of the extent to which seed quality affects crop production. In order to make indigenous sorghum seed selection sustainable, it is imperative to take farmer seed selectors as partners in the research extension programs and improve the seed storage mechanisms.

The authors have not declared any conflict of interests.