| dc.description.abstract | The study records the levels of agro-chemical use on
mature coffee, viz : nitrogen, copper and captafol fungicides;
and their frequencies of application, exhibited by a sample
of small-holders in the Northern division of Machakos district
in the 1977/78 crop year. Measures of deviations in the 'levels
of (9ro-chemical use and frequencies of application from
official recommendations are also provided. Several hypotheses
are tested to determine whether on average, farmers deviate
from official recommendations in the application of agrochemical
inputs on the mature coffee enterprise.
To complete the study,simple correlation and multi-linear
regression models are used to identify socio-economic
variables which are associated with the levels of use of
agro-chemical inputs and frequencies of application.
Both primary and secondary data are employed in the study.
The former were oo Ll.ect.e d from a random sample of 35 farmers,
members of two contiguous co-operative societies in the
Northern division of Machakos district. Secondary data were
obtained from the Coffee Research Foundation technical circu+ars,
and the Machakos district coffee annual reports.
The three general hypotheses tested are:-
(1) That farmers on average do not deviate from
recommended levels of agro-chemical use and
frequencies of application on mature coffee;
(2) That farmer expenditures on agro-chemical
inputs for mature coffee enterprises bear
no definable simple linear relationships with
enterprise total variable costs and total costs
in the year of study; and
(3) That levels of agro-chemical use and related
frequencies of application on mature coffee enterprises
are not correlated with selected socioeconomic
variables.
(xii)
The research findings are divided into two categories,
the descriptive data analysis presented in Chapter 7 and
the results of tests of general hypotheses discussed
in Chapter 8. Included in the former are figures summarising
costs and returns for sample mature coffee enterprises.
These data are arrived at using a strictly defined residual
accounting model. In the 1977/78 crop year these data
show that the average mature coffee enterprise adequately
covered all incurred costs. Furthermore income from coffee
accounted on average for 69 per cent of total farm income.
Thereby easily making it the most important cash earning ~rop
in the Northern division of Machakos district where the
study was centered.
The descriptive data analysis also indicates that on
average small-holders deviated significantly from the
officially recommended levels of agro-chemical use and
frequencies of application on mature coffee enterprises in
the crop year 1977/78. Compared with recommendations,
nitrogen was on average over-used on a physical basis by
approximately 63 per cent, while fungicides, viz: copper,
and captafol, were under-used on average on a physical
basis by approximately 74 and 82 per cent respectively·
Using It' tests on regression coefficients for the 2nd
general hypothesis stated above null hypotheses were rejected
at the .05 level of significance, leading to sample based
inference that for the population of mature coffee enterprises
in the period studied, total enterprise expenditure on each
of the agro-chemical inputs investigated - nitrogen, copper
.and captafol fungicides (the latter two are considered together)
bore fixed proportional relationships to total varVable costs
and total costs for the same enterprises. While this result
must be interpreted cautiously it nonetheless suggests
strongly that small-holder coffee producers probably see
their expenditures on agro-chemicals in terms of proportional
relationships with cost totals rather than according to
strict interpretation of area recommendations.
,
(xiii)
Therefore this finding offers a definite reason why
coffee growers were not found to be following official
recommendations closely. More research is needed to
compare the results for other crop years in the area
studied.
If in fact there is correspondence in the picture
disclosed for other periods, then inquiry needs to
establish why expenditure should be related to cost
totals and not more strictly to recommendations.
Finally the third general hypothesis was neither
totally rejected nor totally accepted. This is because
a number of important selected socio-economic variables
were found to be significantly correlated with levels
of agro-chemicals use and frequencies of application,
while others showed no acceptable strengths of
relationship | en |
