DR. A. HANSEN
FARMING SYSTEMS RESEARCH
CHITEDZE. P.O. BOX 158
LILONGWE, MALAWI Dr. A. Hansen
July 22, 1981
LADD LOCAL MAIZE TRIALS
One of the recommendations arising from the April 1981 farming
systems survey of Lilongwe Rural Development Project (LRDP) was
that local maize should be tested for fertilizer responsiveness.
Local maize is the term applied to any maize that is not a recognized
variety (MH12, MH13, UCA, CCA, etc.) obtained from ADMARC and, there-
fore, local maize covers a wide range of varieties where the seed is
saved by the farmer from previous harvests or purchased from other
farmers or on local markets.
Local maize is the most important and highly valued food crop
grown in LRDP and in Malawi. Every farmer plants local maize; almost
all farmers have more acreage in local maize than in any other single
crop, and most farmers have more acreage in local maize than in all
other crops combined. Farmers almost always prepare local maize fields
before other fields and plant local maize first before planting any
other crop at the beginning of the rains. Even when farmers plant
governmentally-sponsored varieties of maize such as UCA or MH12, they
continue to grow local varieties, to trust local varieties more than
the sponsored ones for home consumption purposes, and usually to grow
more acres of local than sponsored.
Manure when used, is almost totally used on maize fields. As
fertilizer use has become more common, farmers have been applying it
more and more to local maizes as well as to the sponsored varieties.
Sometimes this fertilizer used on local varieties has been purchased
and other times obtained on credit, usually as part of a package with
a sponsored variety of maize.
Although farmers are choosing to apply capital to this high
priority food crop, research and extension are unable to help small-
holders make the most effective decisions about this input because
almost nothing is known about the responsiveness of local maize
varieties to fertilizer. There is an existing nationwide blanket
recommendation for 1i bags of S/A or JAN per acre, but there are no
research records indicating the background or rationale for that
The maize agronomy section of D, sh :, noted, did
conduct fertilizer responsiveness trial oI loc., maizes on 3 farmers'
fields in Dowa West during the 1980/81 season and plans to continue
those trials. Two of the 3 trials were completed satisfactorily,
revealing in each instance a linear relationship between
application of nitrogen (up to 80 kilograms per hectare, which was the
highest level in the trials) and yield.- The following conclusions are
tentative and partial, based as they are on only two trials. Manage-
ment of the trials was divided between research personnel (planting
and fertilizer applications) and the farmers (all other labour and
decisions). In one trial the rate of increase in yield was low enough
so the farmer would not gain any profit from the higher applications
of fertilizer, while the higher rate of increase in the other trial
meant the farmer would have found it profitable.
During the 1981/82 cropping season Chitedze Research Station will
sponsor on-station and on-farm research trials of the fertilizer res-
ponsiveness of local maizes in LRDP, Thiwi Lifidzi and Ntcheu (all
within Lilongwe ADD) and Dowa West (of Kasungu ADD).
The on-station trials will test responsiveness to three levels
of nitrogen: 0,40 and 80 kilograms per hectare (40 kg/ha approximates
the present recommendation.) Local maize samples (15 kg each) will be
collected from 12 farmers in LRDP, 2 in Thiwi Lifidzi, 7 in Ntcheu,
and 3 in Dowa West. (Each farmer's local maize will be treated as if
it were a separate variety.) UCA will be run as a control, so there
will be 25 "varieties" in all with 3 fertilizer treatments and 4
Each local "variety" will be tested on-station and on the field
of the farmer from which it was taken. The on-farm trials will follow
the pattern established for the Dowa West 1980/81 local maize trials.
There will be 5 treatments of nitrogen (0,20,40,60 and 80 kg/ha) and
3 treatments of phosphorous (0,20 and 40 kg/ha of P200). The 5 by 3
factorial will be a random bloc design with 2 replicates.
The three farmers for Dowa West will be the same ones who were
used in the 1980/81 trials. The 21 farmers in LADD will be selected
randomly from strata selected from the existing National Sample Survey
of Agriculture (NSSA) households. NSSA is itself a stratified random
sample. Because the 21 LADD farmers will be NSSA farmers, the NSSA
will provide background data on the farmers, and the evaluation department
of LADD will be in constant contact with them and will provide
continuing liaison between the farmers and the project. The extent
to which the 21 farmers are representative of their areas may be
evaluated through the NSSA, and the yields which the 21 farmers obtain
on their own local maize fields (measured as part of the NSSA) may be
compared with the yields of those varieties when fertilized on the
farms and on the station.
The testing will provide, therefore, a number of comparisons.
1. The Chitedze trial will test varietal responses to
several levels of nitrogen, while holding constant the location
(at least within each replication) and manru.rnent. Random selec-
tion of the local maizes will in ,i. e the 1 robability that the
results are generally representative of i1 'i maize in the ADD.
As a consequence, research and extension will be better able to
assess the recommendations that may be given to farmers and to
project management about the advisability of fertilizing local
2. Each local variety will be generally measured under several
managements (professional at Chitedze, farmer and professional in
the on-farm trials, and unaided farmer in the farmers' other fields),
a range of fertilizer levels, and several locations (Chitedze and
various micro-niches on each farm).
3. Farmer management may be compared on the basis of NSSA yield
data from farmer-managed local maize fields, using as controls the
on-farm research plots and each variety's performance at Chitedze.