PhD Thesis
Cotton (Gossypium hirsutum L.) is among the most important export crops in Tanzania,
contributing directly to peoples‘ livelihood and economic development. Cotton is grown
under both conventional and organic farming systems and both farming systems are
characterised by low cotton yield. Low fertilizer input use and low fertility contributes to
such low yields. A quick solution to address the low cotton yield seems to be enhanced use
of fertilizer and pesticide, which is also challenging on how to simultaneously sustain
natural resources conservation and the supply of cotton fibre. Inappropriate use of fertilizer
is linked to increased greenhouse gases (GHG) emissions. One of these gases is nitrous
oxide (N 2 O), and nearly 2/3 of the anthropogenic N 2 O emission originates from
agricultural activities. However, there are limited studies involving in-situ measurement of
N 2 O in smllhoder farming systems in Sub-Saharan Africa. The use of organic and
synthetic fertilizers and pesticides can also change the soil environment and hence
microbial activity and soil quality. To address the problem of low cotton productivity,
adequate information on the soil status and its suitability for cotton production is
important. However, availability and acquisition of such information related to soil
qualities in the study area is limited.
This study was undertaken for two growing seasons in semi arid cotton growing areas in
Meatu, Simiyu region, Tanzania. This study aimed to generate soil information,
characterise and evaluate suitability of soils in these areas in terms of their qualities for
cotton cultivation; compare yield and economic performance of cotton in response to
different soil fertility and pest management practices under low input small holder
conventional and organic production systems; compare microbial activity as a proxy for soil quality and quantifying N 2 O emission from soil under different smallholder organic
and conventional cotton production practices.
To characterise the soils, four pedons were characterized namely Biore-P1 in Mwamishali
village, Ng'ho-P1 in Nghoboko village, IT-P1 in Sanga Itinje village and MWAB-P1 in
Mwabagalu village. Soil samples from the pedons' horizons were analyzed for physico-
chemical properties and were classified according to the USDA Soil Taxonomy and the
World Reference Base for Soil Resources. The soil suitability for cotton production was
evaluated by using most limiting criteria approach.
A field experiment was set to test performance indicators (yield, economic benefit
microbial activity and GHG emission) for fertility and pest management practice in organic
and conventional farming systems. For fertility management, the recommended and current
farming practices (30 kg N ha -1 and 3 Mg farm yard manura (FYM) ha -1 ) were tested
against high input scenario (60 kg N ha -1 and 5 Mg FYM ha -1 ) and alternative practices (30
kg N ha -1 + 3 Mg ha -1 FMY and 3 Mg ha -1 FMY + legume intercrop) for conventional and
organic cotton production, respectively. For pest management in conventional and organic
treatments, respectively, the current practice (3 sprays of synthetic pesticide and pyrthrum
spray as needed after scouting), higher rate (6 sprays of synthetic pesticide and pyrethrum
spray as needed after scouting) and alternative practice (3 sprays neem + cow urine and
spray of Neem + cow urine as needed after scouting) were tested.
Seed cotton yield, cost, revenue and gross margin for each treatment was measured to
estimate yield and economic performance. For GHG emission, in-situ measurement of N 2 O
emission from soil was done using static chambers technique and gas analysis by gas
chromatography (GC). For microbial activity, enzyme activity (arylsalfatase), potential nitrification, basal respiration and birch effect for plots under current practice,
recommended practice, high input scenario and alternative practices were tested against a
no fertilizer and no pesticide control treatment.
The results on soil site suitability show that soil moisture and temperature regimes in all
study areas were ustic and isohyperthermic respectively. Except for Mwabagalu, which
had very deep pedon (> 150 cm), all the other pedons had moderate deep and well drained
soils. For Mwamishali and Ng'hoboko CaCO 3 concretions were observed, which is a
characteristic of calcic horizon in subsoil. Soil texture was sandy clay throughout all
pedons with top soil pH ranging from neutral to strongly alkaline (pH 7.05 - 9.54). In all
areas pH increased with depth from topsoil to subsoil. The organic carbon content was low
to medium while total N concentrations were very low; CEC and exchangeable bases
ranged from low (19 and 10.1 cmol
(+)
kg -1 ) to medium (37 and 38.4 cmol
(+)
kg -1 )
repectively. According to the USDA Soil Taxonomy the soils at Mwamishali and
Ng'hoboko were classified as Pachic Calciustolls, while that of Sanga-Itinje and
Mwabagalu were Typic Rhodustalfs and Sodic Haplusterts, respectively. All the soils were
rated as marginally suitable for cotton production due to soil fertility limitations and
therfore, sustainable cotton production in these areas would need interventions for soil
fertility and soil moisture conservation improvement.
The results on yield and economic performance showed that, for the current farming
practice, organic and conventional farming practice had no significant difference in yield.
The seed cotton yield in current organic practice was 1.4 and 0.64 Mg ha -1 compared to the
conventional practice with yield of 1.3 and 0.37 Mg ha -1 for season 1 and 2 respectively.
The current organic farming practice had higher gross margin of USD 527 and 137 ha -1
compared to the conventional practice with gross margin of USD 321.2 and -74.9 ha -1 for season 1 and 2 respectively. However, at higher input rate (fertilizer and pesticide)
conventional cotton had significantly higher yield of 1.8 and 0.51 Mg ha -1 ) than organic
practice with seed cotton yield of 1.4 and 0.39 Mg ha -1 for season 1 and 2, respectively.
The high input conventional cotton had relatively lower gross margin of USD 463 and -
85.9 ha -1 compared to the organic practice with gross margin of USD 477 and -81.3 ha -1
for season 1 and 2, respectively. For organic cotton practice the alternative practices
(intercropping cotton with green gram) had lower cotton yield of 1.3 Mg ha -1 in season 1
than the alternative conventional (1.7 Mg ha -1 ) but had higher gross margin (USD 616.2
ha -1 ) than conventional (USD 476 ha -1 ) as a result of additional revenue from green gram.
For conventional cotton the alternative practice of combining inorganic fertilizer and
manure outperformed the current practice by having both better yield and gross margin. In
the season with less rainfall (season 2) the control outperformed all other treatments in
terms of economic return indicating the rationale for farmers reluctance in investing on
inputs under such conditions.
The results on N 2 O emission show that the current organic and conventional cotton
farming practices had similar (p < 0.05) cumulative area-scaled N 2 O emission. However,
yield-scaled emissions were significantly higher in conventional than organic farming
systems. For the high input scenario conventional cotton had higher area-scaled and yield-
scaled N 2 O emission than organic cotton in season 1 which received higher rainfall (759
mm) but not in season 2 which had less rainfall (522 mm). A combination of manure and
inorganic fertilizer as alternative practice reduced yield-scale N 2 O emission by 17% from
inorganic fertilizer. In season 1, intercropping cotton with legumes reduced area-scaled
emission by 27%. The emission factor for both conventional and organic systems were <
1% of applied total N. Although the cummulative N 2 O emission varied between season 1
(0.24 and 0.31 kg N 2 O-N ha -1 ) and season 2 (0.52 and 0.60 kg N 2 O-N ha -1 ), results show that the current smallholder organic and conventional cotton farming practices had similar
soil N 2 O emission, which is very low compared to reported emissions from cotton fields in
high input farming system (0.78 - 10.6 kg N 2 O-N ha -1 ).
The results also revealed no significant (P<0.05) difference in microbial activity between
organic and conventional cotton production practices using the current levels of fertilizer
and pesticide. However, increasing N level as synthetic fertilizer or manure increased
microbial activity. Manure fertilizer combination as an integrated practice increased
arylsulfatase activity and potential nitrification but relatively reduced basal respiration in
conventional farming in both seasons. Intercropping green gram increased microbial
activity (arylsulfatase) but reduced potential nitrification and basal respiration and hence
had no clear trend on improving soil microbial activity. Pest management had no effect in
microbial activity in both organic and conventional farming practice and there was no
significant difference between organic and conventional.
The major conclusions drawn from this study are that, all the soils were rated as marginally
suitable for cotton production due to soil fertility limitations and hence sustainable cotton
production in these areas would need interventions for soil fertility improvement. For the
existing smallholder farming system in Meatu, Tanzania, organic and conventional cotton
system have similar agronomic and economic perfromance. Increasing input level in
conventional cotton would have higher yield with low economic return than organic.
Combining manure and inorganic fertilizer as well as intercropping cotton with grain
legumes has potential for increasing yields while reducing the risk of increased N 2 O
emission from cotton fields and improving microbial activity and hence soil quality.
Under the prevailing semi-arid conditions, smallholder farmers are rational on limited use
of fertilizer and pesticide Based on these conclusions it is recommended that alternative practice of cotton-legume
intercrop, fertilizer-manure combination and pest control by neem and cow urine mixture
are viable farming practices for smallholder cotton farmers. Further studies on the effect of
cotton-legume intercrop on soil microbial activity that include other legume species are
recommended. Further studies that include higher input levels and higher sampling
frequencies for GHG are also recommended.