International Journal of Horticulture, 2016, Vol.6, No.2, 1-10
7
the bottom most cluster in the present set of hierarchical clustering pattern, which clubbed all parental test
genotypes except M 44/3. This clearly suggests that the recombination breeding programme in the present pursuit
has unequivocally resulted wide genetic diversity (genetic variation) exceeding the parental gene pool in terms of
heterotic hybrids. Among the parents, M 44/3 was genetically far distant followed by RP 1 which constituted a
mono-genotypic group ‘Cluster IVB’. Thus, RAPD markers were potent enough to discriminate all test genotypes
beyond 71% phenon level. Neto Silva et al. (1995) used RAPD markers in cashew to distinguish four dwarf
cashew seedlings in Brazil. Similarly, Thimmappaiah et al. (2009) reported clear distinction of 67 accessions out
of which NRC-142 and NRC-12 emerged as highly divergent while, NRC-231 and NRC-232 were genetically
similar among 100 accessions of cashew. In an another study, Thimmappaiah (2011) reported genotype
combinations of NRC 335 with NRC 338; and NRC 362 with NRC 388 to be highly divergent. Mneney et al.
(2001) used RAPD to determine the genetic diversity within and between populations of cashew (
Anacardium
occidentale
L.).
RAPD technique together with study of morphological traits has proved useful for assessment of genetic diversity
(Ramessur and Ranghoo-Sanmukhiya, 2011). Therefore, an effort was made to note agronomic performance of the
test genotypes in relation to their distribution in the dendrogram clusters using RAPD markers. The order of
occurrence of the genotypes in the above RAPD based clusters was used as a reference for arranging performance
values of the accessions for 14 agro-morphological traits that were observed in the field over two consecutive
seasons laid out in RBD. The average values of these traits are given in Table 2. For better clarity, the significantly
lower/higher performance values for plant height and for rest of the morpho-economic traits significantly higher
mean values compared to the experimental grand mean have been marked with an asterisk (*), are plant types of
special consideration for cashew genetic improvement. In fact, genotypes with significantly higher values than the
grand mean were identified to assess relation (if any) of genotypic performance for yield and yield attributing
traits with RAPD marker based clustering. The table values arranged cluster-wise allow a simultaneous
comparison of several quantitative traits in the test genotypes with reference to the position of each genotype in
the dendrogram. It is observed that the genotypes in the same cluster based on DNA profiling have some common
phenotypic performance and such clusters with unique phenotypic performance could be identified for genetic
improvement in cashew nut.
Dwarf plant types are of special consideration to restructure plant ideotype. Cluster IVB and Cluster IIB had
shown characteristically dwarf plant types. Among the genotypes under these genetic groups; M 44/3 in Cluster
IIB and RP 1 comprising Clusters IVB, exhibited significantly dwarf plant height. Such dwarf type cashew
genotypes have been purposefully used as parents for development of heterotic cashew hybrids. For instance,
Cluster IC and Cluster III are characterized by tall plant types. B 27 included in Cluster IC and B-5 in Cluster III
had shown significantly tall plant type among the selected 20 hybrids under study. It is worth to note that the
dwarf type parent RP 1 was one of the parents in the heterotic hybrids B-27 and B-5. Besides, cashew hybrids
D-19 and H-6 constituting the most divergent heterotic genetic group ‘Cluster IA’ have M 44/3(dwarf plant type)
as one of the parent. Barros et al. (2002) reported that clones of early dwarf cashew have greatly impacted cashew
cultivation as they are more productive, early, short, easily harvestable, and have uniform nuts and apples.
In cashew nut, trunk girth, canopy spread (N-S), number of flowering laterals/m
2
, number of perfect flowers, sex
ratio, nuts/panicle have considerable role in determining nut yield. Cluster IA was unique in significantly higher
trunk girth while Cluster IC had highest mean values for canopy spread (N-S) and number of flowering laterals/m
2
which has direct bearing on productivity. In this context, Cluster IB and Cluster IIB had shown characteristically
higher mean values for number of perfect flowers, sex ratio and nuts/panicle.
RAPD based clustering differentiated all parental genotypes except M 44/3 into the Cluster IV which revealed
appreciably lower mean values for almost all agro-economic traits including nut yield. This is indicative of the
fact that the 20 selected experimental hybrids developed from different cross combinations have expressed
