Participatory identification and selection of collections of Cenchrus ciliaris in the Southern Rangelands of Kenya

Everlyne C Kirwa, Eliud K Ngugi¹, George N Chemining’wa¹ and William N Mnene²

Kenya Agricultural and Livestock Research Organization, Genetic Resources Research Institute, Muguga, P O Box 781-00902, Kikuyu, Kenya
eeckirwa@ yahoo.com
¹ Department of Plant Science and Crop Protection, University of Nairobi, P O Box 29053, Nairobi, Kenya
² Kenya Agricultural and Livestock Research Organization, Arid and Rangelands Research Institute, P O Box 12-90138, Makindu, Kenya

Abstract

A study was carried out through Focused group Discussions (FGDs) to evaluate the farmer knowledge and perceptions regarding ecotypes of C. ciliaris grass species. Selection of farmer preferred ecotypes was done at KALRO Kiboko research Station using the ribbon technique of participatory variety selection. Farmers selected among twelve established ecotypes following their own developed farmer criteria of a good and bad grass. Through FGDs, it was observed that farmers were knowledgeable on the existence of the ecotypes of Cenchrus ciliaris whose occurrence was similar in the three different farmers’ groups interviewed. Three main ecotypes were identified by all groups; the small type with purple colored flowers, the robust bluish type and robust green type. The small type with purple flowers was noted as the most preferred by all groups during the FGDs. The ecotype was said to be a heavy seeder dropping a lot of purple colored seeds on the ground thus allowing for its spread and establishment in different habitats. Also, the ecotype is perceived to be tolerant to droughts and heavy grazing. The criteria for selection of ecotypes varied depending on the type of utilization of the pasture. KLF1 and TVT1 ecotypes received the highest hits from the farmers due to their small stature and thus good as a grazing type and perceived to be drought tolerant.

Key words: Cenchrus ciliaris, ecotype, forage production, participatory variety selection

Introduction

Cenchrus ciliaris is a tufted perennial grass species that is widely cultivated for forage production in the tropics. It is adapted to a wide variety of soils, is drought tolerant with some ecotypes having the potential to withstand temperatures as low as -10 ^o C and as high as 46.5 ^o C (Griffa et al 2006; Arshad et al 2007). The species is able to utilize both low and high rainfall amounts by branching of the existing tillers or development of new tillers, respectively (Visser et al 2008). The grass is highly tolerant to grazing pressure which could be attributed to its varied forms of tiller developments or the extensive deep rooted system that may exceed 2 m. Wide adaptation has resulted in many ecotype variants. The species has been promoted widely for natural pasture improvement and rehabilitation programs in the arid and semi-arid lands (ASALs) of Kenya. This study describes a participatory selection experiment that was conducted on C. ciliaris ecotypes that had been established at KALRO-Kiboko to evaluate farmer preference.

Participatory variety selection (PVS) is the involvement of farmers in the selection of non-segregating, described products of plant breeding, for example in form of lines, hybrids or clones. Involvement of farmers in selection of varieties builds in a sense of ownership (Ceccarelli 2000) as it includes testing and selecting of the varieties either in farmer’s fields or on-station (Almekinders et al 2006). PVS is commonly initiated after genetic variability has been reduced in the selection materials in the first steps of the plant breeding cycle although evaluation and selection from natural populations or ecotypes in grass species has also been successfully conducted (Savidan 2001). In addition, C. ciliaris, is a non-segregating obligate apomict (Griffa et al 2006). Apomixis is common in perennial forage grasses and grass cultivars derived directly from natural populations are widely sown (Savidan 2001). For instance, the Kentuckey bluegrass has highly variable cultivars due to its apomictic mode of reproduction (Huff 2010).

The aims of the present study were to establish the knowledge and perceptions of farmers on ecotypes of C. ciliaris and test the use of participatory techniques in selection of the grass ecotypes.

Materials and methods

The study was carried out in two parts. First, the use of Focused Group Discussions as the tool to collect information from the farmers on their knowledge and perception on the ecotypes of Cenchrus ciliaris and develop the criteria for selecting grasses. Second was participatory selection of established ecotypes in the field using the ribbon technique was.

Focused group discussions (FGDs)

Three FGDs with each targeting eight farmers of mixed gender were held in late October 2014. Two of the farmer groups (group two as (GP2 and Group three as GP3) were made up of the agro-pastoral Kamba farmers from Kibwezi and Makindu Sub-counties while one had pastoral Maasai community from Mashuru Sub-county. The third group (GP3) involved farmers that were in leadership who actively participated in pasture improvement activities. Previously identified questions were used to guide the discussion (Table 1). Samples of some of the ecotypes were also availed to ensure the farmers were talking about the species of interest. Pairwise ranking was used during the discussions to rank the different criteria as well as the ecotypes of choice. At the end of the sessions, each group was also asked to identify the criteria for a good and bad grass. These were to be used in the selection of ecotypes in the field. A good grass was defined as the one that the farmer would want to take home to improve their pasture production while a bad grass was the one not to be promoted in their region.

Participatory selection of ecotypes of Cenchrus ciliaris using the ribbon technique

Eleven ecotypes of C. ciliaris were previously established at KALRO-Kiboko research station pasture plots between October 2012 and April 2014 for morphological characterization. The selected field site was similar in environmental conditions to the targeted farmer region because farmer selection is environment dependent (Ceccarelli et al 2000). The experiment was arranged in a complete randomized block design replicated three times. Subsequently, replication was ignored and only one block was used because use of replications was found to have no significant difference and farmers complained of replication being too tedious. This activity targeted the same farmer groups that attended the FGDs. Twenty four farmers (11 GP1, 8 GP2 and 5 GP3) participated in the activity. Each of the three groups selected two different colored ribbons representing a good and bad ecotype (Table 2).

The women in the groups had their ribbons tied to form knots to distinguish theirs from the men during the counting. Each farmer was given 5 ribbons for a good ecotype and 3 for a bad one. Each group went round the plots to first select the good ecotype by tying each of the 5 ribbons to ecotypes of choice. Poles were erected at the edge of each plot to facilitate the process. The same process was repeated for selection of the bad ecotype with each group going in at different times. Counting of the total number of ribbons per ecotype based on color and gender was done at the end of the exercise. The data was analysed using SPSS where mean percent ribbon numbers was generated and compared between gender per ecotype.

Results

Identification of ecotypes of Cenchrus ciliaris by farmer groups

The identified ecotypes during the FGDs are listed in Table 3. All the identified ecotypes were among the collections established at KALRO-Kiboko research station. The pastoral group (GP1) indicated that they knew three ecotypes, namely, robust green (Entomonyua), small with purple flowers (Enkamba) and the bluish type. Similar ecotypes were recorded by GP2 except with an additional black headed ecotype. The black headed ecotype was identified by one farmer who noted that it had was upright, very leafy with deep green leaves and was found along Makindu River. The small purple flowered and bluish types were also identified by GP3. However, the group indicated that they knew of two types of the robust green based on flower color, that is, the purple flowered and the green flowered ecotypes. All the groups indicated that the small purple flowered was the most preferred by animals and all its parts are palatable. The ecotype is also reducing overtime in composition and size due to overutilization, frequent droughts and land fragmentation. Based on farmer description of the ecotype, it was found that the ecotype is similar to Kilifi 1 (KLF1) collection from Kilifi County, a sea-coast ecology.

Summary of farmer description of the identified ecotypes

Small, purple flowered ecotype: It is short, about 60 -100 cm tall, prostrate with faster regrowth after rains. It is a high seeder whose seeds germinates easily and forms a good ground cover once established. The ecotype is liked by animals where all parts are utilized because stems are not tough. The animal preference does not change with growth stage or season. Other uses included good soil erosion control because its prostrate growth form. The Kamba community pound its roots and boil or directly chew as cough remedy in children. The ecotype is a problem in croplands since it spreads easily and is difficult to eradicate. Threats to its conservation included reduced presence in pasturelands due to overgrazing, land fragmentation and frequent drought. Reduction in plant size had also been observed by agro-pastoral group two.

Robust Green ecotype: Tall up to 150 cm with wide leaves and tough stems when mature. Has strong, deep rooting system and is found in valleys, near water collections or along rivers. The ecotype is persistent and re-sprouts when grazed or clipped even during the dry season, especially in black cotton soils. Animals prefer it only before seed maturity due to its tough stems. It is good for hay baling and is used by agro-pastoral community in tying other animal feeds during harvesting or when preparing for preservation. Other uses included soil erosion control, thatching houses, making nest for hens to lay eggs in and supports bee production since it is liked by bees. The ecotype, however, is liked by termites and snakes.

Bluish ecotype : Very tall to about 200 cm depending on soil type and moisture. It is upright with tough stems and rough leaves when mature. Produces many tillers and has a wide tussock (robust). Grows well in wetter areas, riverbanks and in croplands along the terraces but not common in the grazing areas. Does well in in black cotton soils. It is not preferred by animals but animals mainly select leaves and flowers before maturity. Some farmers indicated that intake at maturity was enhanced when chopped and mixed with molasses. The ecotype is good for hay making. Other uses included soil erosion, making brooms and thatching houses.

Selection of identified ecotypes of Cenchrus ciliaris for promotion in grass reseeding

All the participants in the pastoral Maasai group (GP1) selected only the small purple flowered ecotype because it is small with all parts liked by animals irrespective of the season. Tables 4 and 5 show the ecotypes selected by GP2 and GP3 for promotion. The small purple flowered ecotype was selected for promotion by the majority of participants in the other two groups. The green robust and the bluish type were not selected by GP2 and GP3, respectively.

Development of criteria for selection of ecotypes

It was difficult to agree on the criteria with GP1 because the group based their decisions mainly on the animal preference or performance such as increased milk yield or fattening of the animals. For instance, Digitaria macroblephara was said to increase milk yield and to fatten across seasons. In addition, they seemed to base their selection on seasons. For instance, during the dry season, a good grass does not fill up the animal stomach allowing the animal to remain light (fit) and thus not tire easily. However, grasses that fill up the animal stomach quickly are preferred during the wet seasons since they quickly fatten the animal for the market. Wet season grasses are believed to be high herbage yielding species. Three criteria were finally agreed upon for commonly preferred good grasses. These were thin stems, short and prostrate grass for grazing. Ranking was not done since the group maintained only the three criteria and indicated that the opposite applied for a good grass under wet conditions. Consideration of the semi-nomadic lifestyle seemed to have influenced their seasonally based choices.

Table 6 shows the criteria identified for a good grass by GP2. Through pairwise ranking they identified the three most important criteria as high seed yield, many leaves and small plant size, particularly leaves.

High yield of seed was ranked highly because the farmers insisted that their main target is the business value in dealing with grass seeds. Thus, germination capacity had been listed in the criteria, but was removed after agreeing that it could not be applied in the field except if lab results are included in the target selection. The high ranking of small sized and leafy type was in line with their selection of the small purple flowered type among their ecotypes. They indicated that a tall, bushy grass is not well grazed and animals would keep moving in the field picking leaves instead of keen grazing. However, those that targeted cut and carry system preferred a robust type grass since they believed chopping would increase intake. Group two (GP2) also identified criteria for a “bad” (meaning un-preferred) grass as hairy, thick stems, spaced or few leaves (less leafy) and low amount of seed yield. This was to be done through evaluation of seed amounts on the flower heads and observed number of seeding heads.

Table 7 shows five factors identified by GP3 for a good grass. Through pairwise ranking they ranked the three most important criteria for a good grass as many leaves, soft stems and many tillers. The group also listed the criteria for a “bad” grass as rough leaves (hairy or rough to touch), hard or tough stems and low biomass through observed herbage yield in the field.

Selection of good ecotypes using the ribbon technique

Table 8 shows farmer preference and selection of C. ciliaris ecotypes at KALRO Kiboko Research Station. The smaller ecotypes received significantly higher mean percent number of ribbons for a good grass. TVT1 (90.3 %) and KLF1 (80.6 %) ecotypes had the highest percent ribbons while TVT3 with 6.1 % had the least. All men present selected TVT1 as their preferred grass ecotype. For female, KLF1 and KLF3 ranked highly each with 9 ribbons out of 10 although KLF2 and TVT1 each received 8 ribbons.

There was no significant difference (p>0.05) in preference between male and female for all ecotypes. According to field plenary discussions, TVT3 which was a bluish type, received the least ribbons because of its big and hard stems, low leaf to stem ratio (less leafy) and course leaves when touched. Although KBK1 is bluish in colour, it is leafier than the other robust ecotypes, thus it received 25 %.

Discussion

The farmers were aware of existence of ecotypes of C. ciliaris. There were similarities in three of the ecotypes identified. These were the small-purple flowered, bluish type and robust green. However, group three were able to identify two types from robust green, i.e. the purple head (due to purple awns) and pure green head with no tint of purple colour. All the mentioned ecotypes have been observed among the collections at KALRO- Kiboko centre. The small purple flowered type is actually similar to KLF1 ecotype. In a previous study, Millar et al (1997) found that farmers in Australia were knowledgeable in perennial grasses in terms of animal performance, drought and persistence.

The majority of the farmers in each FGD group nominated the smaller ecotype due to its soft stems and leaves as well as being short, prostrate and perceived preference by the animals and tolerance to grazing. Shortness and prostrate growth form confers the plants tolerance to grazing (Milchunas et al (1988). Li et al (2015) observed reduced leaf length, leaf width, stem width, stem diameter and the overall plant size of Leymus chinensis grass species due to long term defoliation by either grazing or mowing. However, those that targeted harvesting for baling preferred the robust ones. Thus the mode of utilization is an important criteria in selection for different grass ecotypes. The appropriateness of the intended use is an important consideration while planning pasture production (Barnhart 2011).

The criteria for a good grass were relatively similar among the groups. Robust related traits were not preferred especially height due to perceived poor grazing habits by animals and thick stems due to associated toughness with maturity. The pastoral group (GP1) based their criteria development on seasons. They had special reasons for dry season preferred grass. This was to allow for fitness in the animals for ease of movement. This implies that promotion of grass species for reseeding pastoral Maasai areas should consider grazing tolerant species of both small and big types.

As was shown in Table 8, KLF1 and TVT1 were the most preferred because of being small, leafy and with thin stems as indicated in selection criteria. TVT1 is medium in height and semi prostrate while KLF1 is short and prostrate. The KLF1 ecotype was the small purple flowered ecotype identified by farmers during the FGDs as shown in Table 3 while TVT1 was identified by only one farmer in GP2 as the black headed ecotype. The two ecotypes had no significant difference in regards to crude protein, crude fiber, in-vitro dry matter digestibility (INVDMD) and percent DM (Kirwa et al., 2015). However, TVT1 had the least INVDMD of 45.6 % among twelve evaluated ecotypes.

Among the robust ecotypes, the Magadi collections, MGD1 and MGD3 with 41.7 and 62.5 %, respectively, received higher number of ribbons compared to the Kiboko collections, that is, KBK1 and KBK2 with 25 % and KBK3 with 8.3 %. The preference was mainly from GP2 and GP3 farmers targeting hay production in their selection. The recorded good ribbons for MDG1 from GP1, the Maasai community, may be as a result of its stems and height that did not seem to differ with the small types. The ecotype height and stem thickness did not differ significantly to the small types during characterization. The ecotype is also leafy compared to the other robust types.

Some of the characteristics depicted by the highly selected ecotypes such as high tiller numbers, small size and prostrate growth form are associated with grazing effects. The grass plant developed some of these traits due to the long term interaction with the grazer as a way of adapting to grazing (Sala 1988). N’Guessan and Hartnett (2011) observed increased tiller recruitment and development of meristems in different positions of the little bluestem grass leading to a more prostrate position as a response to frequent grazing. High seed production is also a survival trait in the grass family allowing it to maintain persistence in competitive environments, including croplands (Simpson 1990). This was indicated in Table 5 that the small purple flowered ecotype was noted to be common in agro-pastoral areas since it produces a lot of seeds and was a problem in croplands. The high seed yield trait was also expressed by KLF1 ecotype.

Conclusions

• The farmers were knowledgeable on the existence of varied ecotypes of C. ciliaris in their environs. They were able to give details on their preferences as well as their reasons for the same. This information should be synthesized and used in making additional collections. This is in particular reference to the pastoral Maasai areas where a collection of their small purple-flowered ecotype should be made and evaluated for nutritive value.
  
  
• The criteria for selection of ecotypes varied depending on the type of utilization of the pasture. Thus the small ecotypes received the highest percent ribbons as good grass for grazing. This means cut and carry or baling of hay as techniques for utilization of pastures are yet to be fully embraced. Therefore, successful development and promotion of grass varieties needs to consider the mode of utilization by the target group.
  
  
• TVT1 and KLF1 were highly preferred by the farmers. Thus, the two ecotypes should be selected for evaluation on performance and adaptability in the farmer target areas before their promotion. However, the low INVDMD particularly for TVT1 may require further improvement.
  
  
• The study indicates that participatory selection through the ribbon technique can be used in selecting grass types for promotion in rangeland reseeding.

Acknowledgment

The authors acknowledge funding support from the World Bank through EAAPP Programme and the EU through ASAL-APRP without which this work would not have been achieved. Thanks also the Director General KALRO, the management and the staff of KALRO Kiboko for their technical and logistical support.

References

Almekinders C and Hardon J 2006 Bringing Farmers Back into Breeding. Experiences with Participatory Plant Breeding and Challenges for Institutionalization. Agromisa Special 5, Agromisa, Wageningen, pp. 135.

Arshad M, Ashraf M Y, Ahamad M and Zaman F 2007 Morpho-genetic variability potential of Cenchrus ciliaris L. from Cholistan desert, Pakistan. Pak. J. Bot. 39: 1481-1488.

Barnhart S K 2011 Selecting forage species. University Extension publications, Iowa State University, PM1792.

Ceccarelli S 2000 Selection methods Part 1: Organizational aspects of a plant breeding programme. In Plant breeding and farmer participation, pp. 195 – 222.

Griffa S, Díaz D, Ribotta A, Lanza S, Muñoz N, López E, Luna C, Grunberg K and Biderbost E 2006 Molecular genetic discrimination of Buffel grass genotypes and F1 hybrids for breeding purposes using amplified fragment length polymorphism analyses, Grass Forage Science, 61: 454–458. doi: 10.1111/j.1365-2494.2006.00546.x

Huff D R 2010 Bluegrasses. In: B. Boller, Posselt, U.K. and Veronesi, F. (eds). Fodder crops and amenity grass. Hand book of plant breeding. Chapter 15 volume 5. Springer Science, ISBN 978-1-4419-0759-2, pp. 345–380

Kirwa E C, Kiarie N, Chemining’wa G and Mnene W N 2015 Nutritive composition of Eragrostis superba Peyr and Cenchrus ciliaris L. collections from the ASALs of Kenya. .Livestock Research for Rural Development, 27 (8).

Li X, Wu Z, Liu Z, Hou X, Badgery W and Guo H 2015 Contrasting effect of long term grazing and clipping on plant morphological plasticity: Evidence from a rhizomatous grass. PLoS ONE, 10(10): 1-19. Doi: 10.1371/journal.pone.0141055

Milchunas D G, Sala O E and Lauenroth W K 1988 A generalized model of the effects of grazing by large herbivores on grassland community structure. The American Nature, 132: 87-106.

Millar J, Allan-Curtis A and Terry De Lacy T 1997 Perennial Grasses in Australia: The Place for Local Knowledge. Rangelands , 19: 6-11.

N’Guessan M and Hartnett C D 2011 Differential responses to defoliation frequency in little bluestem ( Schizachyrium scoparium) in tallgrass prairie: implications for herbivory tolerance and avoidance. Plant Ecology, 212: 1275-1285.

Sala O E 1988 The effect of herbivory on vegetation structure. In: M.J.A. Werger, Van der Aart, P.J.M., During, H.J. and Verboeven, (eds). Plant form and vegetative structure. Academic publishing, The Hague, The Netherlands, pp. 317-330

Savidan Y, Carman J G and Dresselhaus T 2001 The Flowering of Apomixis: From Mechanisms to Genetic Engineering. Mexico, D.E: CIMMYT, IRD, European Commission DC VI (FAIR)

Simpson 1990 Seed dormancy in grasses. ISBN 0521372887, Cambridge University Press, digital 2007.

Visser M, M’seddi K, Chaïeb M and Neffati M 2008 Assessing yield and yield stability of remnant populations of Cenchrus ciliaris L. in arid Tunisia: developing a blueprint for initiating native seed production. Grass forage science, 63: 301-313

Received 22 August 2018; Accepted 24 September 2018; Published 1 November 2018

Go to top