The genotypes labelled =wild types, whose plants do not have
agriculturally favorable characteristics for flower type and color,
had an average genetic similarity index superior in comparison to
the one found among the commercial cultivars, namely, 0.59
against 0.54, respectively, demonstrating that even among
cultivated genotypes significant genetic diversity can be found
and employed in genetic breeding programs.
Similar genetic diversity has been reported in French bean
cultivars using RAPD markers by Chiorato et al. (2007) who
analyzed 220 accessions representing the two most important
genetic diversity centers of the species and found ı´ndices ranging
from 0.54 to 0.82. Therefore, the magnitude of the genetic diversity
obtained in gerbera confirms the expressive genetic variation
found in the species.
The smaller similarity indices were found for the Gunda
genotype although its commercial characteristics do not strikingly
differ from the other accessions (Table 1). Considering, for instance,
the genotype Brasil, it shares three commercial features with
Gunda (flower type, flower stem and inflorescence color) among
the four characteristics analyzed, however its similarity coefficient
is of 0.48. This information along with the coefficients between
Feria and Brasil (0.82), which have two characteristics in common,
and Amazone and Avana (0.59) that share all the agricultural
characteristics with the other accessions, indicate that the
evaluated commercial characteristics do not directly relate to
the genetic similarity coefficients, leading to the conclusion that
the highest genetic similarity coefficients not always indicate a
higher number of shared characteristics. Similarly, it was not
possible to associate the phenotypic characteristics with the
genotype clusters.
In order to evaluate the potential of RAPD markers to
characterize the genetic diversity of gerbera accessions in
comparison to the main horticultural characteristics employed,
comparative analyses of genetic diversity were performed using
the information from the afore mentioned markers and the
botanical descriptions of 28 accessions with available information
(Table 1). Analyses were carried out by transformation of the
botanical data into a disjunctive table (presence or absence of the
class for each characteristic) to obtain the genetic similarity indices
for both data sets (=Jaccard). Despite the botanical data are only
from four characteristics, the average value of genetic similarity
found for RAPD markers was of 0.58 against 0.23 for the botanical
descriptions, indicating a greater ability of RAPD markers to reveal
the genetic divergence among the accessions investigated in this
work than the botanical description. Moreover, the botanical
descriptors were not able to discriminate between the accessions
Avana and Amazone, whereas RAPD markers allowed the
distinction among all 42 accessions investigated in the comparison
between the datasets. In contrast, null similarity values were
absent from the analysis of RAPD markers whereas the botanical
descriptions have provided several comparisons of total diver