Weed Control in Small Farm Systems

 John L. Hammerton
 Weed Scientist
 Caribbean Agricultural Research and Development Institute (CARDI),
 P.O. Box 971, Castries, St. Lucia
 An Output of the CARDI/USAID FSR/D Project No. 538-0099


 Small farmers in the East Caribbean spend about 30% of total crop
labour on land preparation, and another 30% on post-planting weed
control depending on the cropping system. Methods rely largely on hand
labour, using cutlasses, hoes, forks and hand pulling of weeds. The only
herbicide used in any quantity is paraquat. Constraints to the adoption of
improved technologies include cash availability, the practice of mixed- or


inter-cropping, and the complexity of many farm systems. Many her-
bicides are susceptible to rain, and their species spectrum is often limited.
A holistic approach is necessary to ensure integration of weed control into
the entire cropping system. Guidelines for the design of improved weed
control systems are discussed and CARDI's programme in weed control in
the FSR/D project is outlined.


 In a study of small scale farming in the English-speaking coun-
tries of the East Caribbean (from north to south, these countries
are: St. Christopher and Nevis, Antigua and Barbuda, Montser-
rat, Dominica, St. Lucia, St. Vincent and Grenada), carried out
between 1978 and 1982, weed control was identified as a major
constraint to crop production and to profitability (CARDI, 1983,
a,b,c,). This paper presents some of the data collected on labour
use for weed control and on the economics of small scale farming,
describes the traditional methods of controlling weeds, and
outlines CARDI's approach to developing improved weed control
systems. Since land preparation almost invariably involves the
destruction of weeds, this activity is also considered along with
"post-planting" or "in-crop" weed control.

Labour Use
 Table 1 summarizes some farm characterisation data collected
over 12-month periods during 1980-81, in five countries (Ham-
merton, 1984a). Thirty-three farmers, out of 79, devoted be-
tween 20 and 40% of total crop labour time to post-planting
weed control, and eight farmers spent more than 40% on this ac-
tivity. The St. Lucia data may be atypical as Hurricane Allan


struck partway through the characterisation period. Forty farmers
used more than 20% of total crop labour for land preparation, 13
of these using over 40%.
 One root crop farmer in Nevis devoted 85 % of crop labour to
land preparation, and a tree crop farmer in Grenada spent 81%
of crop labour time on post-planting weed control. At the other
extreme, there were farmers devoting 5 % or less of crop labour to
land preparation or weed control. These were generally livestock
farmers or farmers with established tree crops.
 On 36 of the 79 farms, the farmer him- (or her-) self was the
sole or major source of labour for crop production activities. On
only 11 farms, mainly in Grenada, was hired labour the major
source of crop labour. The remaining farms used a mixture of
farmer, family and hired labour, in varying proportions (Ham-
merton, 1984a).

Economics
 Analysis of the economic data from these characterisation
studies (CARDI, 1983b), spanning 87 farms in the same five
countries listed in Table 1, shows wide variations in output and
input levels both between and within countries. Total farm pro-


 < 20
20 40


TABLE 1. The number of farms, in five countries, and the total number of farms, devoting differeing percentage ranges of total
crop labour' to postplanting weed control and to land preparation.


Activity, and % of
total crop labour


Weed
Control


 > 40

Land < 20
Preparation 20 40
 > 40


*Labour days were weighted at 0.75 for women and 0.50 for children.
Source: CARDI (1983b), based on data from CARDI's farm characterisation studies carried out over 12-month periods during 1980-81.


VOL. XX-PROCEEDINGS of the CARIBBEAN FOOD CROPS SOCIETY


133