Fertilizer Research 29: 45-54, 1991. © 1991 Kluwer Academic Publishers. Printed in the Netherlands.

45

Overcoming soil nutrient constraints to crop production in West Africa: Importance of fertilizers and priorities in soil fertility research M.T.F. Wong t, A. Wild 1 & A.U. Mokwunye;

1 Department of Soil Science, The University, Reading, UK 2 International Fertilizer Development Center-Afrique, Lomb, Togo Key words:

West Africa, soil fertility, organic matter, nutrient dynamics, water/crop/fertilizer

interactions Abstract

The Alfisols, Oxisols, Ultisols and Inceptisols which dominate the soils of West Africa have sustained crop growth for a very long time. As a consequence, their fertility has become perilously low and the task of increasing or even maintaining the productive capacity of these soils has become perhaps the greatest challenge to agricultural scientists in this latter half of the 20th century. Water is useful not only for the growth of plants but also for the efficient use of costly inputs such as fertilizers. On the other hand, fertilizers increase the water-use efficiency. Such interactions must be closely studied so as to maximize the impact of inputs of agricultural production. The nutrients in the soil are always in a state of flux, with additions and subtractions. Monitoring the dynamics of the nutrients would promote their efficient use by crops and prolong the productive life of the soils. Several models currently exist for the study of organic matter dynamics in soils. These models should be applied to the West African situation since it is vital to develop management practices that can promote efficient use of nutrients released during mineralisation of soil organic matter. Judicious fertilizer use in West Africa should be promoted as this practice will enhance agricultural production while protecting the fragile environment.

Introduction

The key to economic development in West Africa lies in the development of the agricultural sector. With a population growth rate in excess of 3% per annum, the task facing the agricultural scientist is the promotion of agricultural production while protecting the natural resource base. As a continent that was not submerged and glaciated, Africa contains some of the oldest and most exhausted soils. The Alfisols, Oxisols, Ultisols and Inceptisols which dominate the soils of West Africa have been sustaining agricultural production for a very long time. As a consequence, their native fertility has become perilously low. In addition, they

are easily prone to erosion. The task of increasing and sustaining crop production in such soils poses what is perhaps the greatest challenge to agricultural scientists in the latter quarter of this century. With finite limits in arable land, the problem of food shortages can only be overcome by yield increases through intensive agriculture. Fertilizers will play a key role in this strategy. In other regions of the world where the so-called 'green revolution' has taken place, several factors such as improved crop varieties, expanded irrigation, better farm practices and increased use of fertilizers have contributed to the phenomenal growth in food production. According to the World Commission on Environment and De-

46 velopment [1987] increased use of fertilizers contributed the most to such increases. Similar 'miracles' can occur in West Africa. In the first symposium of these series, several presentations demonstrated the good response that can be obtained from the application of nitrogen and phosphorus fertilizers. Economic analysis of experimental results showed that the agronomic responses also translated into increased profit for the farmer [1]. The central role of fertilizers therefore is to build up the soil fertility and provide optimum conditions for the growth of the crop. Given this central role, it is important that the physical, chemical and biological parameters and processes which influence the fertility of and therefore the effectiveness of the nutrients added to the soil be better understood. In the remaining pages of this paper, we will attempt to discuss the research priorities in any efforts to overcome the physical and chemical constraints to crop production in West Africa.

Water/Crop variety/Fertilizer interactions An important physical constraint in most of West Africa but especially in the savanna and the sahel is the supply of water from rainfall, which can vary greatly between sites and years, and in its distribution within the growing season. Management of water involves efforts to improve infiltration into the soil, to preserve the soil water in usable form and to facilitate the use of the stored water. Most tillage and soil conservation practices are relatively efficient in improving infiltration and preventing run-off. But unless other factors that influence root development and plant nutrition are taken care of, the effective and efficient use of the conserved water is not possible. Relatively few studies have been done in West Africa on how to take advantage of the relationships between water management and fertility management. In Burkina Faso, Ohm et al. [26] found that tied-ridging improved crop yields but the yield increases were greatest when soil fertility was not limiting. Some of the most thorough work on the interaction of crop variety, water supply and fertilizer application in the developing world has been undertaken in

Syria. Although it is a cool-season (winter) rainfall region, the work is sufficiently relevant to West Africa to be worth describing here. In the eastern Mediterranean region, rainfed farming is practised with an expected annual rainfall of between 200 and 600 ram, most of which falls in the period October-May when the monthly mean temperature is below 20°C and in the middle of this period is below 10°C. The experiments to be referred to relate to barley grown in Syria. Figure 1 shows the relationship between grain yield of barley and seasonal rainfall at three sites for three cropping seasons [31]. Fertilizer nitrogen was applied as ammonium nitrate at 0, 60, 90 or 120 kg N/ha, according to the site, and phosphate was applied as triple superphosphate at 0, 26.4, 39.6 or 52.8 kg P/ha; the fertilizer treatments are grouped together in Figure 1. These results show a positive relation between rainfall and the response to fertilizer, which might be expected. Measurements also showed that application of fertilizer did not substantially increase the amounts of water lost by evapotranspiration. This was mainly achieved by reducing soil evaporation to compensate for the increase in water uptake by the crop. Because grain yields were increased, the water use efficiency (WUE), expressed as kg grain ha-1 mm-1 evapotranspiration was therefore in-

~

4

•

_.N -~

•

©

2

o 0 200

I 250

I 300

I 350

I 400

I 450

500

Total seasonal rainfall ( m m )

Fig. 1. Relationshipbetweenseasonalrainfalland grain yield

of barley. Data from Shepherd et al. [31]. Regressionequations: • plus fertilizer y = 0.015x- 2.4 (r2 = 0.94) o no fertilizer y = 0.007x- 0.86 (r2= 0.81)

47 creased. The average W U E was 5.6 without fertilizer and 9.7 with fertilizer. The difference was similar when expressed per mm rainfall and similar results have been reported by Cooper et al. [7]. For example, in a five year experiment at Breda in NW Syria (average rainfall 278 mm) application of fertilizer (26.4 kg P and 20 kg N per ha) increased the average grain yield of barley by 880kg ha -1 whilst water use only changed by - 4 to +8 mm per season. Comparable results have been reported from three sites in Niger as shown in Table 1. The pearl millet (Pennisetum typhoides) in these experiments was grown with fertilizer (45, 20 and 25 k g / h a of N, P and K respectively) or with no fertilizer. Use of fertilizer increased grain yields, but had little effect on water use, and water use efficiency was therefore increased. The mechanisms leading to the improvements in water use efficiency were further investigated in the experiments in Syria. It was found that fertilizer increased the length of barley roots in the top 60cm of the soil profile and also increased water use during the vegetative period of growth [6]. The distribution of roots differed between crop genotypes and these differences were associated with different timing of water use [2]. It appears that greater water use in the vegetative phase was related to increases in grain yields. Another important observation is the pronounced effect of phosphate on crop yields. The field observation is that phosphate applications increase the rate of development of the barley crop and can advance maturity by up to 14 days [4]. Where water is in short supply towards the end of the growing season early maturity provides a drought escape mechanism, as described by Cooper. Experience in Syria is that the relative response to phosphate applications increases as rainfall decreases [21].

The work at I C A R D A has highlighted several routes to increased crop yields in regions where water is the main limitation to crop growth and yields. These include choosing genotypes suited to the particular climatic conditions, application of fertilizer, weed control and application of crop residues as mulches where available. These routes lead to improved water use efficiency. A conclusion that is relevant to West Africa is that it is a myth that fertilizer use should not be advocated in areas with unsure rainfall such as the sahel region. The use of a purchased input such as fertilizers is highly risk-prone under rainfed conditions. Studies geared towards a more efficient utilization of fertilizers are therefore critical. Such studies must define the right types and correct quantities of fertilizers for each crop and cropping system. In addition, they must address issues related to management practices that must be adopted, including soil conservation measures to maximize efficiency.

Nutrient dynamics in West African soils The second concept we want to emphasize is that the soil is not a static system. The nutrients contained in soil are in a state of flux: there are additions, losses, crop removal, and internal cycling processes. These processes are often driven by temperature, rainfall, and additions of organic matter to the soil, any or all of which can be extreme in the tropical environments of West Africa. It is therefore in such regions, whether humid or seasonally arid, where it is most important that the dynamics of soil nutrients be understood. This applies particularly to nitrogen, but as referred to later, it applies also to the other crop nutrients. Before developing this

Table I. Water use (WU), grain yield (Y), and efficiencyof water use (WUE) of pearl millet with and without fertilizer application at three sites in Niger during the rainy season of 1985. From ICRISAT [10] Sador6

Fertilizer No fertilizer SE +-

WU (mm) 382 373 3.7

Dosso Y (kg/ha) 1570 460 162

WUE (kg/ha/mm) 4.14 1.24 0.44

WU (mm) 400 381 3.0

Bengou Y (kg/ha) 1700 780 103

WUE (kg/ha/mm) 4.25 2.04 0.26

WU Y (ram) (kg/ha) 476 2230 467 1440 15.2 126

WUE (kg/ha/mm) 4.68 3.08 0.22

48 main theme some background information will be given. The amounts of nutrient elements in soils of the savanna are usually low and are sometimes very low [18]. The amounts are often higher in the forested region but work at the International Institute of Tropical Agriculture (IITA) and elsewhere has shown that responses to N, P and S-containing fertilizers can occur within 2 or 3 years of bringing land into cultivation. We know that the nutrients which accumulate in the soil and vegetation under a bush or forest fallow [18, 19, 25] can meet crop requirements when the land is first cultivated. What we have failed to do is to investigate the processes of nutrient accumulation when land is under fallow and the processes of nutrient loss when land is under cultivation. Is the nitrogen accumulation through symbiotic fixation, and if so by what plant and microbial species? An alternative mechanism for which there is some evidence is that nutrients (N, P, S, K etc) are captured by trees from rainfall and dry deposition, although in only small amounts annually [19]. Another common explanation is that deep-rooted trees transfer nutrients to the soil surface through leaf fall, In view of the importance of the bush/forest fallow in providing a successful means of land management for many generations of farmers it is surely time for mechanisms of soil fertility regeneration to be known.

Nitrogen On a regional basis Robertson and Rosswall [29], in their comprehensive review of the nitrogen cycle in West Africa, reported amounts of biologically fixed nitrogen that ranged from about 100 kg ha-1 a-1 in early successional forest to less than I kg ha -~ a -~ in unused but grazed desert. They reported that the biggest loss of N is caused by fire, but a gap in the information is how much of this nitrogen (NH 3 and NOx) returns elsewhere as wet or dry deposition. At the field scale there is very little information on the extent of gains and losses of nitrogen by the various routes. For West Africa, mineralisation of soil organic nitrogen when land is brought into cultivation

has been described by Greenland [8]. The paper discusses the rate of decrease of the organic N left in soil, but at the time the data were collected there appears to have been no informations on how much of the mineralised nitrogen was used by crop plants. This kind of information was the subject of a study conducted at IITA, Ibadan, Nigeria [24]. The experiment was conducted on a flat site which had been under secondary forest for about 15 years and was cleared by hand before the first growing season in 1979. The soil was sampled before the trees were removed and then at intervals during the following 22 months during which time, three crops of maize and one crop of soybean were grown. No fertilizers were applied and the crops were grown under no tillage conditions. On each sampling occasion 64 soil samples were analyzed, each sample containing 8 bulked soil cores. The initial soil samples were taken to 10 cm depth and because of the increase in bulk density later samples were taken to 8 cm. They were analyzed for total N, P and S and for organic C and P. During the 22 month cropping period each loss, expressed as a per cent of the amount initially present at 0-10 cm, was organic P 25, total N 32 and total S 44. In the first 15 months the two crop harvests removed 276 kg N ha -~ compared with 509kg N ha -~ that was mineralised; for P the amounts were 27 and 29 kg ha 1 respectively. It follows from these analyses that 233 kg N, the difference between the amount mineralised and the amount in the crops, was lost per hectare from the top 10 cm of the soil. As the annual rainfall was about 1500 mm it is assumed that much of the nitrogen was lost as nitrate by leaching. This is a serious economic loss, especially as valuable cations will also have been leached with the nitrate. The loss would probably have been even higher if normal cultivation practices had been used. In another experiment, twelve monolith lysimeters were installed at Onne, near Port Harcourt [35]. Four of the lysimeters were uncropped in the first year, two received urea labelled ~SN and two received no nitrogen fertilizer. Drainage water was collected from each lysimeter over the two-year period of the experiment and analyzed for nitrate and ammonium. The urea was quickly hydrolysed, and as only

49 traces of ammonium were found the results are essentially for nitrate leaching. Because the soil that was used for the lysimeters was a coarse textured Ultisol and average annual rainfall was 2420 mm, it was expected that the nitrate would be quickly leached below 1.35 m, the depth of the lysimeters. The simplest assumption made was that net rainfall, equal in volume to that contained in the soil of the lysimeters at field capacity, would be just sufficient to displace the nitrate from the lysimeter. Net rainfall was calculated as rainfall minus pan evaporation. The expected volume for the peak concentration of nitrate was 160 dm 3 whereas the actual volume was 4 0 0 d m 3. The ratio 400/ 160 = 2.5 pore volumes means that 2.5 times as much rain was needed to displace the nitrate than was expected. Laboratory experiments showed that the soil adsorbed nitrate. Measurements of nitrate adsorption have since been extended to a wide range of soils from other countries (Wong et al., unpublished). Nitrate adsorption has been reported in soils from other regions of the world, but as far as we know, this is the first observation of adsorption in West African soils. It should be noted that the soil at O n n e is acid (pH 4.3-4.5 in water), and if the p H had been higher, nitrate adsorption might be less. One advantage in using monolith lysimeters for research on the fate of fertilizer nitrogen is that, as illustrated above, nitrate leaching below the root zone can be quantified because drainage water is collected from a defined soil column and can be analyzed. Because the mass of soil in the lysimeter is known another advantage is that the amount of the label from ~5N labeled fertilizer retained in the soil can be measured with a high degree of confidence.

Using the Onne lysimeters the recovery of the label from 15N labelled urea (10.35% atom excess *SN) was measured in drainage water, crops (maize and rice grown in each of two years) and, at the end of the experiment, in the soil plus weeds. Details are given elsewhere [321 but by way of summary, the recoveries are given in Table 2. The urea was applied at 138 kg N / h a to maize and 92 kg N / h a to rice. The average grain yields for maize and rice were 2500 and 2300 kg ha -1 respectively. The maize yields were considered normal under experimental conditions in this region, but rice yields were low. The points which are important are the large losses in the drainage water, retention of the label in the soil, and total recoveries of between 70 and 91% implying denitrification losses. Recovery of the label in the crops was also low, but this is not a true measure of the effect of the fertilizer. As is now well established, 15N usually undergoes exchange with soil nitrogen due to mineralisation/immobilization turnover (MIT) so that some of the ~SN is retained in soil and some soil nitrogen is taken up by the crop [11, 14]. Judging from the 15N remaining in the soil at the end of the experiment, MIT will have occurred in the lysimeters, but this could not be checked because the experimental design did not include a cropped, zero nitrogen treatment. Analysis of the drainage water also showed that, apart from the loss of I5N, there was a large leaching loss of unlabelled nitrogen. Over the two years the loss was 270, 324 and 243 kg ha-1 from treatments A, B and C / D , respectively [32]. This is loss of a valuable nutrient and, as the drainage water analyses showed, there was also a corresponding loss of cations. We are not aware of any direct measurements of denitrification in West Africa and the direct

Table 2. Recovery of ~SN from labelled urea in crops + weeds, drainage and soil in a lysimeter experiment at Onne, near Port Harcourt, Nigeria

Treatment A. 15Napplied to 1st crop, (maize)in 1st year B. 15Napplied to 2nd crop, (rice) in 1st year C/D. 15Napplied in 1st year, uncropped, and cropped in 2nd year

percentage of recovery over 2 years Crops + weeds

Drainage

Soil

Total

35 19 2

29 22 65

27 29 20

91 (82*) 70 87

* True recovery is 82% because crop residues containing 9% of the added ~SNwere returned to the soil.

50 measurements of denitrification in tropical rainforests have only recently been reported [30]. These were made in Costa Rica and indicated that the highest gaseous losses of N occurred in primary forest (13 kg h a -1 a -I) and early successional sites (19kg ha -~ a -1) and least in midsuccessional forest (4 kg ha -1 a-l). The relative amounts are probably related to the amounts of nitrate in the soils. In the Sahelien region of West Africa where millet is the principal crop, few trials using 15N have been conducted. In a series of lysimeter studies in Senegal, Ganry and Guiraud [20] found that only 34% of the urea applied at 150 kg N/ha was taken up by the millet crop. 13.7% of the N was found in the grain and 16% was in the soil (primarily in the top 10 cm layer). Gaseous losses were estimated at 47% of the applied N. Leaching was not a problem since only 1% of the applied N was found in the lysimeter percolates. Monitoring what happens to applied N and developing technologies to arrest losses have occupied priority attention in IFDC's N research in West Africa. Since 1982, field studies have been conducted in the different ecologies using ~SN labelled fertilizers to study the fate and efficiency of applied N. A summary of the results obtained to-date is presented elsewhere in this volume. It is important to recognize the high losses (sometimes in excess of 50%) of applied N in the drier and sandy environments of the Sahel. The conjecture is that the primary loss mechanism is ammonia volatilisation. It is of prime importance that the exact loss mechanism be ascertained. ~SN-labelled fertilizers have also been used in trials to develop management practices for a more efficient use of N fertilizers. Such studies have shown that crops utilize the second dose of N more efficiently when the fertilizer is split-applied. Studies on integrated soil fertility management including the use of legumes can benefit from the use of 15N-labelled fertilizers. These studies help scientists to design farming systems and farm management practices that can promote the efficient use of inputs.

Phosphorus and sulphur There are models for the cycling of phosphorus

in natural grasslands [3] and in forests [9] which incorporate mineralisation and immobilization turnover, as referred to in the case of nitrogen. Net mineralisation occurs when a forest or bush fallow is brought into cultivation and in subsistence farming this process provides phosphate and sulphate, in addition to ammonium and nitrate, for crop growth. In the work at IITA, Ibadan, referred to earlier [24] the amounts of organic phosphorus and sulphur mineralised in the top 10 cm soil during a 22-month period were 53 and 80 kg ha -~ respectively. The amount of phosphorus mineralised was approximately equal to crop uptake. The crops were not analysed for sulfur but it seems likely that the amount mineralised equalled or exceeded crop uptake. It is not known whether the rates of mineralisation, on say, a weekly basis, matched the rates of uptake by the crops. But this matters less for phosphorus and sulfur than for nitrogen because their supply in soil is better buffered. The mineralisation ratios for N : P : S were 12:1 : 1.5. Table 3 gives the calculated amounts mineralised assuming that these ratios apply generally to soil organic matter, and that the annual mineralisation rate, apart from the first two years after a bush/grass fallow is either 2% or 6% [8]. Although the calculated amounts will apply only approximately to a particular site, they show that soil organic matter is an important source of phosphate and sulphate for plants, thus supporting earlier work [23]. Plant residues, whether leaf litter under forests, stubble and roots left by a crop, or an organic mulch, also provide phosphate [19] and other nutrients. It should, however, be noted that phosphate and sulphate are subject to mineralisation/immobilization turnover, as referred to earlier for nitrogen. For example, in experiments recently reported from South Australia [22], a large proportion of the P-33 from labelled residues of medic (Medicago truncatula) was incorporated into soil microbial biomass and very little was taken up by the test plant (wheat). Better understanding of internal cycling in the soil of phosphorus and sulphur is needed. Because of widespread deficiencies in West African soils, especially of phosphate, fertilizer additions are essential for improved crop yields.

51

Organic matter dynamics

CO~

The two field experiments in Nigeria referred to in the section on nitrogen show the important contribution of mineralisation in supplying the mineral nitrogen pool. The calculations that are given in Table 3 need to be developed to include additions, especially of fertilizers, organic manures and plant residues and their C : N ratios; gaseous and leaching losses of nitrogen; measurements of mineralisation/immobilisation. An important requirement is, however, to be able to predict the rate of mineralisation of nitrogen held in soil organic matter. Except for the initial stages of decomposition, when the C / N ratio is high, there is a relationship between the rates of carbon and nitrogen mineralisation. Modelling of carbon mineralisation is therefore useful in this context. Several models have been developed to describe the loss of carbon from organic matter. To be realistic, these models must have several compartments to cater for differences in the decomposition rates of the various organic matter fractions and of the physical and chemical nature of the plant residues. The simple two compartment model shown in Figure 2, illustrates the concept used to separate the decomposition process into a fast and a slow part [12]. In this model, compartment P, containing undecomposed plant material and animal remains, feeds compartment Q containing the remainder of the soil organic matter. A fraction, f, of the annual input to P is transferred to compartment Q and r is the fraction of soil C decomposed per year. The carbon contained in each compartment is assumed to decompose uniformly. The rate of change of soil organic C is given by:

Annualinput( A)~I'~

H

CO~

I~

FA

~

Compartment P (undecomposed plant and animal remains)

C

__

Compartment Q (all organic matter in soil, except that in P)

Fig. 2. Two-compartment model for the turnover of organic carbon in soil. The fraction of A entering compartment Q each year isf. H and C are the quantities of organic carbon in compartments P and Q respectively and r is the fraction of C decomposed each year from Jenkinson [15].

dC dt

fA-

rC

On integration this becomes

where t is time in years and Co is the initial carbon content in Q, which is effectively the initial carbon content of the soil. The two compartment model effectively simulates the rapid initial loss of carbon observed during decomposition [12]. During that phase there is no parallel decrease in soil organic N and a narrowing of C : N ratio occurs. Such reasoning had guided Jenny [17] in using a single compartment model to describe the turnover of organic N in soils. Greenland [8] and Mueller-Harvey et al. [24] used similar models to describe the mineralisation of N in West African soils. The simple models are useful but lack accuracy. More sophisticated models have therefore been developed, one of which is that of Jenkinson and Rayner [16]. They partitioned the in-

Table 3. Calculated amounts of N, P and S mineralised annually (kg/ha) at 0-10 cm from soil organic matter, assuming ratios of N:P:S of 12:1:1.5, at four levels of soil total N and two annual rates of mineralisation

Soil N

Annual mineralisationrate

(%)

2%

0.02 0.05 0.10 0.20

6%

N

P

S

N

P

S

5 13 26 52

0.4 1.1 2.2 4.4

0.7 1.7 3.3 6.6

16 39 78 156

1.3 3.3 6.6 13

2.0 5.0 10 20

52 coming plant material into two compartments called decomposable plant material (DPM) and resistant plant material (RPM). Decomposition of the plant material produces CO 2 and organic products in the form of microbial biomass, physically stabilized organic matter and chemically stabilized organic matter each decomposing exponentially at its characteristic rates. Van Faassen and Smilde [33] used this model to calculate the rate of carbon loss from sorghum and cotton residues and assumed the rate constants to be four times those observed for temperate soils. They calculated the assumed C : N ratio of each compartment, the time when net mineralisation of N will occur and the rate. In view of the importance of plant residues especially in the sub-humid and dry parts of West Africa, priority should be given to testing such models. We will not review here the various models of soil organic matter decomposition that have been devised, but mention should be made of those of Van Veen and Paul [34] and Parton et al. [27]. The latter includes simulation of primary production and nitrogen flow. A new version also includes the effects of soil texture and grazing intensity on the level of soil organic matter [28]. Its prediction was good for a wide range of grassland soils in the Great Plains of the USA. The model of Jenkinson and Rayner [16] has since been modified (Fig. 3) for wider application [15]. If a compartment contains y kg C/ha in a given layer, it declines to y e - r t after a specified time t (usually a month) and r is the rate constant. The latter takes account of monthly tem-

Organic . 14 input " <~.

n~D--~p M ] Decay// ]proces~ A ~ I

perature, soil moisture and crop cover. The soil texture determines the fraction of the organic carbon lost as CO2; soil pH is not taken into account. Plant materials added monthly are partitioned into DPM and RPM which are decomposed by process A to give CO2, zymogenous microbial biomass (BIOZ) and humified organic matter (HUM). The H U M is decomposed by process B to produce CO2, autochthonous microbial biomass BIOA, and more humus. The model accurately simulated the decomposition of 14C labelled ryegrass at Rothamsted, and also at Ibadan where the rate of decomposition was four times faster [13]. It also predicted the effect of different agronomic practices on the level of soil organic matter. Modelling techniques are useful in determining the most appropriate strategy for maintaining the soil organic matter content at a prescribed level. Good models have the resolution required to calculate net immobilisation and mineralisation rates on a monthly basis. They are easier to develop and validate for C than for N (or probably also for P and S), and they show promise for predicting the rate at which nutrients held in soil organic matter become available to crop plants.

Conclusions

Many experiments conducted for several years show that cost/benefit ratios justify use of fertilizers in West Africa. To maximise the benefit,

¢01)

* " -I" f Decay ." . B~io--~Decay / ' " pro~ess p r~c e s ~-~.~

~1 BIOAIproAceSS~. ~.

'r-------7Decay / ~[ H U M ] p r i c e s s

r ~0~

f ~

Decay / _

~0-a-q De ca~ < / _ '

"price s s ~ . ~

Fig. 3. Flow of C through the model. DPM is decomposable plant material, RPM is resistant plant material, BIOZ is zymogennus soil microbial biomass, and BIOA is autochthonous biomass: all these compartments decay by process A. H U M is humified organic matter which decays by process B. These processes differ only in the proportions of CO2, biomass, and H U M formed. IOM is biologically inert organic matter. From Jenkinson [12].

53 r e s e a r c h is n e e d e d : (i) O n t h e i n t e r a c t i o n o f f e r t i l i z e r ments, water supply, crop variety ronomic practices; (ii) o n t h e i n p u t s , l o s s e s a n d i n t e r n a l c y c l i n g in soil; (iii) o n t h e d y n a m i c s o f o r g a n i c m a t t e r o f its i m p o r t a n c e as a s o u r c e o f a p a r t f r o m its i m p o r t a n c e in o t h e r

11. requirea n d ag12. nutrient because nutrient, respects.

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

Baanante CA (1986) Economic evaluation of alternative fertilizer technologies for tropical African agriculture. In Mokwunye A U and Vlek PLG (eds) Management of nitrogen and phosphorus fertilizers in subSaharan Africa, pp 319-362. Dordrecht, Martinus Nijhoff Brown SC, Keatinge JDH, Gregory PJ and Cooper PJM (1987) Effects of fetilizer, variety and location on barley production under rainfed conditions in northern Syria. 1. Root and shoot growth. Field Crops Res 16: 53-66 Cole CV, Innis GI and Stewart JWB (1977) Simulation of phosphorus cycling in semi-arid grasslands. Ecol 58: 1-15 Cooper PJM (1983) Crop management in rainfed agriculture with special reference to water use efficiency. In Nutrient balances and the need for fertilizers in semi-arid and arid regions. Bern, International Potash Institute Cooper PJM (1984) The effects of fertilizer on water use and water use efficiency of barley in dryland regions. In Proceedings of the soils directorate/ICARDA workshop on fertilizer use in the dry areas. Aleppo, Syria: ICARDA Cooper PJM, Gregory PJ, Keatinge JDH and Brown SC (1987) Effects of fertilizer, variety and location on barley production under rainfed conditions in northern Syria. 2. Soil water dynamics and crop water use. Field Crops Res 16:67-84 Cooper PJM, Gregory P J, Tully D and Harris HC (1987) Improving water use efficiency of annual crops in the rainfed farming systems of West Asia and north Africa. Expt Agric 23:113-158 Greenland DJ (1980) The nitrogen cycle in West Africa- agronomic considerations. In Rosswall T (ed) Nitrogen cycling in West African ecosystems pp 73-81. Sweden, SCOPE/UNEP Harrison AF (1978) Phosphorus cycles of forest and upland systems and some effects of land management practices. In Phosphorus in the environment: its chemistry and biochemistry. CIBA Foundation Symp 57 pp 175-195. Amsterdam, Elsevier 1CISAT (1985) ICRISAT Sahelian Centre Annual Report pp 40-41, Niamey, Niger

13.

14.

15.

16.

17. 18.

19. 20.

21.

22.

23.

24.

25.

26.

27.

Jansson SL and Persson J (1982) Mineralisation and immobilization of soil nitrogen. In Stevenson FI (ed) Nitrogen In Agricultural Soils. pp 229-252, American Society of Agronomy, Madison, Wisconsin, USA Jenkinson DS (1988) Soil organic matter and its dynamics. In Wild A (ed) Russell's Soil conditions and plant growth llth ed. Longman Scientific and Technical Jenkinson DS and Ayanaba (1977) Decomposition of carbon-14 labelled plant material under tropical conditions. Soil Sci Soc Am J 41:912-915 Jenkinson DS, Fox RH and Rayner JH (1985) Interaction between fertilizer nitrogen and soil nitrogen- the so called priming effect. J Soil Sci 36:425-444 Jenkinson DS, Hart PBS, Rayner JH, Parry LC (1987) Modelling the turnover of organic matter in long term experiments at Rothamsted. Intecol Bulletin 1 5 : 1 - 8 Jenkinson DS and Rayuer JH (1977) The turnover of soil organic matter is some of the Rothamsted classical experiments. Soil Sci 123:298-305 Jenny H (1941) Factors of soil formation. McGraw-Hill, New York Jones MJ and Wild A (1975) Soils of the West African savanna. Commonwealth Bureau of Soils Tech Comm 55. Harpenden: Commonwealth Agriculture Bureaux Jordan C (1985) Nutrient cycling in tropical forest ecosystems. Chichester: Wiley Ganry F and Guiraud G (1979) Mode d'application du fumier et bilan azote dans un system rail-sol sableux du S6n6gal: Etude au moyer de l'azote-15. In Isotopes and radiation in research on soil plant relationships. International Atomic Energy Agency, Vienna Matar AE (1977) Yield and response of cereal crops to phosphorus fertilization under changing rainfall conditions. Agron J 69:879-881 McLaughlin MJ, Alston AM and Martin JK (1988) Phosphorus cycling in wheat-pasture rotations. I. The source of phosphorus taken up by wheat. Aust J Soil Res 26:323-331 Mokwunye AU, Chien SH and Rhodes ER (1986) Phosphate reactions with tropical African soils. In Mokwunye AU and Vlek PLG (eds) Management of nitrogen and phosphorus fertilizers in sub-Saharan Africa pp 253-281. Dordrecht, Martinus Nijhoff Mueller-Harvey I, Juo ASR and Wild A (1985) Soil organic C, N, S and P after forest clearance in Nigeria: mineralisation rates and spatial variability. J Soil Sci 36: 585-591 Nye PH and Greenland DJ (1960) The soil under shifting cultivation. Techn Comm 51. Harpenden: Commonwealth Bureau of Soils Ohm HW, Nagy JG and Sawadogo S (1985) Complementary effects of tied ridging and fertilization with cultivation by manual, donkey and ox traction. In Ohm HW and Nagy JG (eds) Appropriate technologies for farmers in semi-arid West Africa. International Programs in Agriculture, Purdue University, West Lafayette, Indiana Parton WJ, Anderson DW, Cole CV and Stewart JWB (1983) Simulation of soil organic matter formations and mineralisation in semi arid agro ecosystems. In Lawr-

54

28.

29. 30.

31.

ence RR, Told RL, Asmussen LA and Leonard RA (eds) Nutrient cycling in agricultural ecosystems. The University of Georgia Special Publication No 23 pp 533-550 Parton WJ, Schimel DS, Cole CV and Ojima DS (1987) Analysis of factors controlling soil organic matter levels in Great Plains grasslands. Soil Sci Soc Am J 51: 1173-1179 Robertson GP and Rosswall T (1986) Nitrogen in West Africa: the regional cycle. Ecol Monog 56:43-72 Robertson GP and Tiedje JM (1988) Deforestation alters denitrification in a lowland tropical forest. Nature 336:756-759 Shepherd KD, Cooper PJM, Allan AY, Drennan DSH and Keatinge JDH (1987) Growth, water use and yield of barley in Mediterranean-type environments. J Agric Sci Camb 108:365-378

32.

33.

34.

35.

Van der Kruijs ACBM, Wong MTF, Juo ASR and Wild A (1988) Recovery of tSN-labelled fertilizer in crops, drainage water and soil using monolith lysimeters in south-east Nigeria. J Soil Sci 39:483-492 Van Faassen BG and Smilde KW (1985) Organic matter and its turnover in soils. In Kang BT and van der Heide J (eds) Nitrogen management in farming systems in humid and sub-humid tropics. Inst Soil Fertility Netherlands and Int Inst Trop Agric Ibadan Nigeria pp 39-55 van Veen JA and Paul EA (1981) Organic carbondynamics in grassland soils. 1. Background information and computer simulation. Can J Soil Sci 61:185-201 Wong MFT, Wild A and Juo ASR (1987) Retarded leaching of nitrate measured in monolith lysimeters in south-east Nigeria. J Soil Sci 38:511-518