MSc

The study entitled ‘Variability of mangosteen (Garcinia mangostana L.) in
Kerala’ was conducted under Department of Fruit Science, College of Agriculture,
Vellayani during the period of October 2020 to December 2021 with an objective
of assessing the variability and identification of elite clones in mangosteen grown
in Kerala.
Since Wayanad and Thrissur districts represent the major mangosteen
growing areas of Kerala, assessment of the variability of the crop was done in
different agro ecological units (AEUs) of Wayanad (AEU 15 and AEU 20) and
Thrissur districts (AEU 10). A total of 153 trees were selected for the study.
Morphological and biochemical characters, abiotic stress (sun burn) susceptibility
and incidence of pests, diseases and physiological disorders were recorded from
each tree (accession).
Variability was observed in qualitative characters like flower and fruit
clustering habit, flowering season, fruit size and fruit skin thickness. Dendrogram
based on these qualitative characters by UPGMA method grouped 153 accessions
into 6 different clusters at a Euclidean similarity distance measure of 1.5. Clusters
I, III and IV together consisted of accessions of Wayanad district (AEU 15 and
20) had small and medium sized fruits with medium fruit skin. Cluster I had
accessions possessing a combination of flower/fruit clusters which have 1, 2 and
3 flowers/ fruits per cluster. Accessions which flower either during February to
April period or April to May period were seen in this cluster. Accessions in cluster
III had a combination of flower/fruit clusters posessing either 1 or 2 flowers/fruits
per cluster. Flowering was from April to May and May to June. Cluster IV
consisted of accessions which had one flower and fruit per flower/fruit cluster.
Flowering seasons were same as that of cluster III. Cluster II, V and VI consisted
of accessions from Thrissur (AEU 10) only and these accessions had small sized
fruits with thin fruit skin. Accessions of cluster II had a combination of
flower/fruit cluster with 1 or 2 flowers/fruits per cluster. Flowering season of this
cluster was from February to April. Accessions of cluster V had flower/fruit
177
clusters which were similar to that of cluster II. Flowering season of this cluster
was from March to April. Cluster VI consisted of two types of accessions; those
which produce flower/fruit clusters of single flower/fruit only and those which
produce a combination of flower/ fruit clusters of either 1, 2 or 3 flowers/fruits.
The flowering season was the same as that of cluster V.
Dendrogram based on twenty six quantitative characters, formed using
UPGMA method, grouped the accessions into 8 different clusters at a Euclidean
similarity distance measure of 30. Principal component analysis (PCA) based on
tree and leaf characters revealed that PC1, PC2 and PC3 together contributed
70.95% of variance and had correlation with traits such as age, tree height and leaf
petiole width. The loading plot revealed that tree age and tree height; leaf blade
length and leaf blade width were the variables having strong positive correlation.
PCA bi-plot indicated that Acc.013 to Acc.020, and Acc.077 to Acc 082 were
highly diverse from other accessions and the characters contributing to their
variance were mainly age of the tree and tree height. PCA based on inflorescence
and fruit characters revealed that first six principal components accounted for
67.99% of the total variance among the accessions. Characters contributed mainly
to the variance were number of aborted seeds, TSS/acidity ratio, number of mature
seeds and acidity. Loading plot showed a strong positive correlation between fruit
length and fruit diameter. A strong negative correlation was observed between
four pairs of characters. They are number of matured seeds and number of aborted
seeds, TSS-acidity ratio and acidity, fruit length and edible portion and fruit
diameter and edible portion. PCA bi-plot showed that 30 accessions of Thrissur
district i.e., from Acc.124 to Acc.153 were diverse from other accessions and
maximum contribution to their variance was from characters like number of
aborted seeds or matured seeds and TSS/acidity ratio. The above results indicated
that considerable variability exists in qualitative and quantitative characters of
mangosteen.
Clustering of 62 identified elite clones based on fruit characters like fruit
size, edible portion, TSS and acidity using PCA score plot gave 12 different
clusters. 9 superior clones (Acc.004, Acc.042, Acc.057, Acc.104, Acc.005,
178
Acc.017, Acc.048, Acc. 121 and Acc.037.) were identified based on economical
characters like fruit weight and edible portion. Seven accessions among the
identified superior clones were located at RARS Ambalavayal (AEU 20) and two
accessions were located at the farmer’s field in Meppadi (AEU 15) of Wayanad
district. In future, these nine superior clones could be used for the multiplication of
elite clones as well as for creating a germplasm of superior clones. Also, the
accessions identified with variations should be subjected for molecular
characterization for checking whether variability exists in genetic level.
Mangosteen accessions under this study were not susceptible to sun burn.
Mild incidence of minor pests and diseases like thrips, mealy bugs, algal spot and
anthracnose were observed and were managed at the field itself. Physiological
disorders found in mangosteen accessions were gamboge and translucent flesh
disorder (TFD).