| dc.description.abstract | Banana (Musa spp.) is a crop grown globally and it is of major economic importance. Banana is the fourth most important crop of the food market after rice, wheat, and maize (Gebre et al., 2020). Banana has become a major crop cultivated worldwide with over 1000 varieties being produced. The study focused on estimation of banana mat densities in farmer’s fields for yield evaluation in south western Uganda in Isingiro and Kabarole districts , the data used in estimating the banana yield in the model was observed density, predicted density and intermediate density as from the study.
There is limited information on banana mat densities existing in Uganda and their use as an estimate of banana yields is lacking. In Uganda, banana yield estimates are often done based on whole field plant density and bunch qualities (Mukasa et al., 2005). This undermines variations that exist within the plantation such as the difference in soil conditions, age of the mats and densities of the mats that are critical in determining the productivity of the plantation. Also, in sub-Saharan Africa, bananas are often intercropped in heterogeneous planting arrangements, especially in smallholder farming systems that dominate the region (Hauser and van Asten, 2010). Such differences in planting arrangements affect the inter mat distances that could influence banana yields. This study was timely to investigate on estimation of banana mat densities in farmer’s fields for yield evaluation.
A paired t-Test and linear regression analysis was done to confirm hypothesis, in which the slope of the linear regression line was compared against 1 and the intercept against 0 was also run as an additional Test for Hypothesis. The study findings confirmed that there was a strongly positive relationship between the observed density (mat/ha) and the predicted density (mat/ha) in Isingiro and Kabarole district. Regardless of the location, the regression model also showed a strong relationship between observed density and predicted density. The model indicated that there was increase in predicted density for a unit increase in observed density.
Regardless of the location, the regression model 1, the probability level of significance showed a strong relationship (p<0.001, t=9.874) between observed density (mats/ha) and predicted density 1(mats/ha). The model indicates that there is an increase in predicted density (mats/ha) of 264.4 for a unit increase in observed density. This model explains the relationship between observed density (mats/ha) and predicted density (mats/ha) by 77.8%with 22.2% being random variation hence observed density had a strong influence on predicted density. For regression model 2, the probability level of significance showed a strong relationship (p<0.001, t=8.663) between observed density (mats/ha) and predicted density 2(mats/ha). The model indicates that there is an increase in predicted density (mats/ha) of 333.1 for a unit increase in observed density. this model explains the strong relationship between observed density(mats/ha) and predicted density(mats/ha) by73.3%with 16.7% being random factors/variations showing that observed density had a strong influence on predicted density.
The study found a strong relationship between adjusted density(mats/ha) and observed density regardless of the area of study and both models predicted the relationship with higher prediction level and the less percentage was accounting for random factors/variations. According to the dataset from Wytze and Martin’s survey of observed banana mat densities on 11 farms in western Uganda showed that model predicted the positive relationship between adjusted density and observed density and the study was in line with what Wytze and Martin‘s survey observed implying our model is accepted and being explained by 83% and can be an alternative tool in estimating banana yield by farmers. This model can be used in estimation of banana mat densities in farmer’s fields for yield evaluation.
More research is needed to develop models for estimating yield in all food crops and cash crops in Uganda like coffee, cassava among others. More research is needed on climate of the study area, soil fertility status, agronomic practices, production cycle and be incorporated in the model for high efficiency in predicting banana yield | en_US |
