Evaluation of Grass Ecotypes for Potential Use in Reseeding of Pastoral Fields in the Arid and Semi-arid Lands of Kenya

Abstract

Cenchrus ciliaris L. (African foxtail grass) and Eragrostis superba Peyr. (Maasai lovegrass) grass species that are native to the rangelands have been promoted for rehabilitation of degraded areas and improve forage production. A major challenge to successful reseeding of the rangelands has been lack of certified seeds of these species in the formal seed systems except through collections from the wild. The main objective of the study was to contribute to improved livestock production in the ASALs through characterization, evaluation and identification of higher yielding ecotypes of C. ciliaris and E. superba that can be submitted for certification process. Germplasm was collected from the wild in four agro-ecological zones (AEZ), represented by Kilifi, Taita Taveta, Makueni and Kajiado Counties for AEZ III, IV, V and VI, respectively. Seeds of eleven ecotypes for C. ciliaris and nine for E. superba were processed and planted in five rows of four metres long in randomized complete block design with three replicates at KALRO Kiboko Research Centre. Data were collected for stem, leaf and flowering traits while plant samples were analysed for crude protein, crude fibre, ash, percent dry matter, and in-vitro digestibility of dry matter. The levels of correlation and relatedness among the ecotypes was determined. Two clusters of robust and small sized types were formed using plant height, stem thickness, leaf length and leaf width. The small sized ecotypes were clustered as early flowering while the robust types were late flowering indicating presence of early and late maturing ecotypes among the C. ciliaris collection. MGD1 ecotype was found to be different from the rest due to clustering as a robust and early flowering type. The recorded correlation between stem and leaf traits and nutritive components, CP and INVDMD, in C. ciliaris ecotypes could be used to select for higher yielding plants for the target nutritive values within the ecotypes. Clustering patterns for E. superba

ecotypes remained the same using either 16 different morphological traits, selected robustness traits, namely, plant height, stem thickness, leaf length and leaf width, or seed yield traits. KBK1 and KBK2 ecotypes of E. superba remained clustered together in one group as robust types against the rest of the seven ecotypes. Effect of the environment of collection including climate and grazingland management may have influenced the plant size and maturity time among the C. ciliaris ecotypes. The identified different clusters of C. ciliaris ecotypes allows for selection along maturity time such as within the early flowering types, late flowering types and MGD1 as an early maturing and robust type. Significant genetic differences (<0.01) was recorded among ecotypes of C. ciliaris where Kajiado population recorded the highest diversity indices while Kilifi and Narok collections were the most distant populations. High genetic differentiation between populations of E. superba was recorded with Fst=0.237, Gst= 0.534, mean Shannon diversity index (I=0.357) and Nei’s genetic diversity index (h=0.223) among populations. There was possible exchange of genetic materials between ecotypes of E. superba conserved in common gardens. Evaluation of biomass yield among C. ciliaris ecotypes was done in three sites Kiboko, Buchuma and Mtwapa KALRO Centres. Plots were established in three replicates of five rows each measuring four metres long. Dry matter yield data was collected for two seasons and analysed using AMMI stability value (ASV) and Yield stability index (YSI). Mean dry matter yield (DM) of the C. ciliaris ecotypes across three sites ranged from 3986 to 11,792 kgha-1 where the small sized types had the lowest yield. KBK3 ecotype was ranked the most stable across sites with ASV and the highest yielder with YSI. Ecotype KBK1 was the most suitable for Kiboko and Mtwapa sites and MGD3 for Buchuma.

Evaluation of farmer knowledge and perceptions on ecotypes of C. ciliaris grass species was carried out using Focused group Discussions (FGDs) where farmers developed criteria for preferred grass types. The farmers knew of the existence of various ecotypes of C. ciliaris and had varied perceived preferences on them. TVT1 and KLF1 ecotypes were selected by over 80 % of the participants due to their perceived tolerance to droughts and heavy grazing. The criteria for selection of ecotypes by farmers varied depending on the type of utilization of the grass. Successful development and promotion of grass varieties should consider the mode of utilization by the target farmer group