Creating Opportunities for Domesticating and Commercializing Miombo Indigenous Fruit Trees in Southern Africa - Indigenous Fruit Trees in the Tropics: Domestication, Utilization and Commercialization - page 148

Chipata, Zambia was opposite. Tree survival was low at 30 MAE (25-64%),
and only two of the 12 provenances exceeded 40% at 54 MAE. Tree height in
Zambia also ranged from 0.47 to 0.78 m at the Chipata site. The pooled
analysis of variance showed there was significant interaction between tree age
and site ( P < 0.01). There was strong correlation of the latitude at the site of
origin with survival ( r = 0.96), root collar ( r = 0.96) and tree age

( r = 0.93). This implies that the nearer the germplasm is planted to the

mother tree (site of origin) the better the performance at both sites. In Tanzania,
Malawi provenances from Mangochi outperformed all the 16 provenances
planted at Tumbi, with 1.37 m height at 12 months compared with 0.6 m for
Mpandamatenga provenance from Botswana.
Tree survival showed negative correlation with rainfall level, indicating that
germplasm collected from sites having low rainfall adapted better than that
from wetter sites (Akinnifesi et al. , 2004b). Poor performance and adaptability
of U. kirkiana provenances established at Chipata signals the possibility of a
site effect on tree growth. Site conditions at Makoka may offer a better potential
for orchard management than those at Chipata in Zambia or Domboshawa in
Zimbabwe. Zimbabwe faced challenges of frost damage, and U. kirkiana trees
established at the Tabora site were wiped out early due to flooding (Akinnifesi
et al. , 2004b). In addition, trees of Uapaca planted at the lower position in the
soil catena seem to have performed poorly compared with those on the middle
or upper slopes, suggesting that Uapaca is better adapted to well-drained soils
(Akinnifesi et al. , 2004b). This explains why 93% of U. kirkiana trees are found
in the mountains and hills (Malembo et al ., 1998). It seems that latitude and
climate may have an overriding influence on tree growth (Akinnifesi et al .,
2004b). Whether this would translate to improved yields or fruit quality is not
yet certain.
The provenance trial showed that most of the priority indigenous fruit trees
showed considerable phenotypic variation across southern Africa (Akinnifesi et
al. , 2004b). Other workers have shown that U. kirkiana exhibits variation in
growth, flowering and fruiting that could be manipulated by management and
improvement (Mwamba, 1995a; Ngulube et al. , 1995, 1998). Akinnifesi et al.
(2006) reported that a few individuals from the Mapanzure and Serenje
provenances from Zimbabwe and Zambia, respectively, fruited at Makoka,
Malawi for the first time after 8 years. The percentage of trees fruiting was less
than 5% of the provenance, and fruiting was not observed in the next 2 years
for all trees in the trial. This agrees with the earlier assertion that fruiting in
U. kirkiana may occur more than 10 years after seedling establishment.

8.3.2 Participatory selection

Genetic gain and diversity in the domestication of miombo fruit trees

The high genetic diversity of selected Uapaca populations in Malawi compared
with the mother source and their progenies confirmed that the extensive
participatory selection used to capture superior individuals was adequate