SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (26)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Buerkert, A.
Right arrow Articles by Dossa, K.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Buerkert, A.
Right arrow Articles by Dossa, K.
Agricola
Right arrow Articles by Buerkert, A.
Right arrow Articles by Dossa, K.
Soil Science Society of America Journal 64:346-358 (2000)
© 2000 Soil Science Society of America

DIVISION S-6-SOIL & WATER MANAGEMENT & CONSERVATION

Mechanisms of Residue Mulch-Induced Cereal Growth Increases in West Africa1

Andreas Buerkerta, André Bationob and Komlan Dossac

a Institute of Crop Science, Univ. of Kassel, Steinstr. 19, D-37213 Witzenhausen, Germany
b International Fertilizer Development Center (IFDC) at International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) Sahelian Center, B. P. 12404, Niamey, Niger
c IFDC-Africa, B.P. 4483, Lomé, Togo

buerkert{at}wiz.uni-kassel.de


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
The use of crop residues (CR) has been widely reported as a means of increasing crop yields across West Africa. However, little has been done to compare the magnitude and mechanisms of CR effects systematically in the different agro-ecological zones of the region. To this end, a series of field trials with millet (Pennisetum glaucum L.), sorghum [Sorghum bicolor (L.) Moench], and maize (Zea mays L.) was conducted over a 4-yr period in the Sahelian, Sudanian, and Guinean zones of West Africa. Soils ranged in pH from 4.1 to 5.4 along a rainfall gradient from 510 to 1300 mm. Treatments in the factorial experiments were three CR rates (0, 500, and 2000 kg ha-1) and several levels of phosphorus and nitrogen. The results showed CR-induced total dry matter (TDM) increases in cereals up to 73% for the Sahel compared with a maximum of 16% in the wetter Sudanian and Guinean zones. Residue effects on weakly buffered Sahelian soils were due to improved P availability and to a protection of seedlings against wind erosion. Additional effects of CR mulching on topsoil properties in the Sahel were a decrease in peak temperatures by 4 °C and increased water availability. These mulch effects on soil chemical and physical properties strongly decreased from North to South. Likely explanations for this decrease are the decline of dust deposition and wind erosion hazards, the higher soil clay content, lower air temperature, and a faster decomposition rate of mulch material with increasing rainfall from the Sahel to the Sudanian and Guinean zones.

Abbreviations: CAN, calcium ammonium nitrate • CEC, effective cation exchange capacity • CR, crop residues as cereal stover • DAS, days after sowing • NPK, 15-15-15 mineral fertilizer with 15% N, 15% P2O5, and 15% K2O • SSP, single superphosphate • TDM, total dry matter • TRP, Tahoua rock phosphate


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
LOW AVAILABILITY of mineral nutrients severely limits primary production and particularly crop growth on acid, sandy soils in sub-Saharan West Africa above 300 mm of annual rainfall (Penning de Vries and van Keulen, 1982; Bationo and Mokwunye, 1991). Given the low clay contents in West African soils and that kaolinite is the dominant clay mineral, soil cation exchange capacity (CEC) mainly depends on the organic carbon content of the topsoil. The current fertility status of these soils may be explained by a number of factors: age and origin of the soils, leaching, soil erosion by wind and water, short fallow periods, and continued nutrient mining of cropped fields (Stoorvogel and Smaling, 1994).

Under these conditions, application of mineral N and P fertilizers and even crop residues (CR) surface applied at 2000 kg ha-1 yr-1 have been reported to cause large yield increases in pearl millet across the southern Sahel (Bationo and Mokwunye, 1991; Bationo et al., 1992, 1993, 1995). For the sub-humid rainforest zones of West Africa with a bimodal annual precipitation of up to 1600 mm, many studies have documented changes in physical and chemical soil parameters as causes for mulch-induced crop growth increases (De Vleeschauwer et al., 1978, 1980; Lal et al., 1980; Maurya and Lal, 1981). However, most of this research was conducted in western Nigeria on moderately sloping Paleustalfs or Luvisols of pH > 6. Furthermore, excessive rates of mulch application of up to 24 Mg ha-1 yr-1 were used. For the inner part of West Africa, the northern Guinean, the Sudanian, and the southern Sahelian zones, where annual average rainfall declines from 1300 to 300 mm and total biomass production is much lower, the causes of CR effects on crop growth and their declining magnitude from North to South are still poorly understood (Bationo et al., 1995). For the Sahel, with its many crust prone sandy soils (Hoogmoed and Stroosnijder, 1984; Valentin and Bresson, 1992), most reported data come from a very limited zone with rainfall ranging between 500 and 600 mm. For this area, mulch effects have been attributed to increased P availability (Kretzschmar et al., 1991), more vigorous root development (Hafner et al., 1993b), enhanced potassium (K) nutrition (Rebafka et al., 1994), protection of young seedlings against soil coverage during sand storms (Michels et al., 1995b), and a decrease in the penetration resistance of the soil surface that affect emergence and root growth of seedlings (Buerkert and Stern, 1995). Given the urgent need to stabilize agricultural production in the region, a better understanding and subsequent prediction of CR effects on a regional scale is important. The purpose of this research was therefore (i) to verify the mechanisms of CR response at a number of sites in the Sahel and (ii) to measure with an identical experimental setup CR effects under the higher rainfall conditions of the Sudanian and northern Guinean zones.


   Materials and methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
Results
 Discussion
 Conclusions
 REFERENCES
 
The data presented in this paper are based on three rainfed experiments. Two experiments were conducted from 1991 to 1994 at ICRISAT Sahelian Center, Sadoré, Niger (13°14'N latitude, 2°17'E longitude, 235 m altitude) on a siliceous, isohyperthermic Psammentic Paleustalf (West et al., 1984). Total annual rainfall was 592 mm in 1991, 511 mm in 1992, 533 mm in 1993, and 794 mm in 1994 compared with a long-term average of 560 mm. The experiments at ICRISAT were designed to clarify cause-effect relationships involved in CR-induced increases of cereal growth in the Sahel.

The purpose of the third experiment was to compare CR effects from the Sahel with those from higher rainfall areas. This experiment consisted of a series of multi-factorial trials conducted from 1995 to 1998 at eight sites selected to provide a gradient of annual rainfall from 510 to 1300 mm and corresponding increases in clay contents from 2 to 16%. The soil types at the chosen sites were in the Sahel Psammentic Paleustalfs at Banizoumbou (13°31'N, 2°39'E; Niger), Sadoré (13°14'N, 2°17'E; Niger), Kara Bedji (13°15'N, 2°32'E; Niger), and Goberi (12°58'N, 2°50'E; Niger); in the Sudanian zone an Arenic Kandiustalf at Gaya-Bengou (11°59'N, 3°32'E; Niger) and a Haplustalf at Fada-Kouaré (11°59'N, 0°19'E; Burkina Faso); and in the Northern Guinean zone an isohyperthermic Plinthic Kanhaplustult at Koukombo (10°17'N, 0°23'E; Togo) and an Isohyperthermic Plinthustalf at Kaboli (8°45'N, 1°35'E; Togo). The sites reflected the typical range of total rainfall in the region with a single unimodal duration of 4 to 6 mo. Topsoil values for organic carbon (C org), P-water, CEC, and base saturation were also typical for the region (Table 1) . All eight sites, except Kara Bedji, which had a slope of about 2%, were completely flat and did not show any sign of water erosion over the 4 yr of the experiment.


View this table:
[in this window]
[in a new window]
  		Table 1 Mean annual precipitation and initial soil chemical parameters at 0- to 0.2-m depth for eight sites in West Africa in May 1995

 
Experiment 1
To study CR effects on the chemical and physical properties of the topsoil, data were collected from a soil fertility trial conducted from 1991 to 1993. The site was cleared after 8-yr growth of a typical bush fallow with the shrub Guiera senegalensis J.F. Gmel. then cultivated with millet at a rate of 10 000 planting holes (pockets) ha-1 in the first rains in May 1991. The 48 factorial treatments assigned to 10- by 10-m plots in a completely randomized design with two replications consisted of (i) four millet genotypes, (ii) three levels of millet stalks (CR), (iii) P fertilizer broadcast at a rate of 0 and 13 kg P ha-1 as single superphosphate (SSP) shortly before sowing, and (iv) N fertilizer applied as calcium ammonium nitrate (CAN) at 0 and 30 kg N ha-1. Crop residues harvested from the corresponding plots were broadcast in early May 1991 and again in November 1992 and 1993 on the soil surface as stalks at rates of 500 and 2000 kg dry matter ha-1. A third mode of application was the spreading of ash from 2000 kg burned CR ha-1 prior to planting that was surface-incorporated with a rag ("unmulched"). Ash plots received a supplement of elemental sulfur and N as CAN to account for volatilization losses at CR burning.

Plots with 500 and 2000 kg ha-1 of CR mulch were split in two parts at the end of 1993, to study the short-term effects of CR application on soil properties and millet growth. One part received an additional (fourth) CR application after millet harvest in November 1993, whereas on the other part residues were withheld until April 1994. Average total mineral nutrient concentrations in the mulched millet stalks were 0.81% N, 0.05% P, and 2.2% K. Details on yearly rainfall, management operations, plant harvest, and pest protection have been reported previously (Buerkert and Stern, 1995).

Plant Sampling
At final harvest, four central rows of millet were cut, shoots and heads separated and dried to constant weight at 65°C, and shoot total dry matter (TDM) was determined. For determination of mineral nutrient concentrations, all plant material was ground to pass through a 2-mm sieve. Total N was determined with a Macro-N-Analyzer (Heraeus, Bremen, Germany). For P and K determination, samples were ashed for 4 h at 500°C in a muffle furnace and the ash dissolved in 1:30 (v/v) diluted HCl. Potassium was analyzed by flame-emission photometry (Eppendorf, Elex 6361, Ismaning, Germany) and total P colorimetrically (Hitachi U-3300 spectrophotometer) according to the vanado-molybdate method (Gericke and Kurmies, 1952). To determine treatment effects on input-output balances of N, P, and K, the amount of these nutrients taken up in the stover and grain (plant nutrient output) was deducted from the amount of nutrients applied with the respective levels of mulched stalks and mineral P fertilizers (nutrient input).

Soil Sampling
To measure treatment effects on chemical properties of the topsoil, initial samples were taken before the onset of the experiment in May 1991 from 0 to 0.2 m at four locations in each plot, bulked, air-dried, and sieved to 2 mm. In May 1994, plots were sampled again at 0 to 0.1 and 0.1 to 0.2 m. Data from the initial soil sampling were used as covariates in the analysis of variance of the results from the second sampling to reduce effects of initial short-distance spatial variability on treatment effects. Separate samples were taken in both parts of the plots split in November 1993 to determine short-term mulch effects on chemical properties of the topsoil. In May 1994 additional samples were taken from 0.2- to 0.4-, 0.4- to 0.6-m, and 0.6- to 1.0-m depth in 12 plots to measure CR effects on soil chemical properties at greater depth. These plots came from both replications of all three CR levels with and without broadcast P and were all planted to the millet landrace Sadoré Local.

All samples were analyzed for pH (1:2.5 0.01 M KCl), C org (Walkley and Black, 1934), Bray-P1 (P-Bray; Olsen and Sommers, 1982), and exchangeable Al and total acidity (McLean, 1982). Exchangeable bases were extracted with 1 M NH4-acetate and Ca and Mg were determined by atomic absorption spectrophotometry, whereas K and Na were determined by flame emission spectrophotometry. Water soluble P was determined according to Sissingh (1971) but with incubation time reduced to 5 min at a soil:water ratio of 1:10 in order to obtain only the most easily available P fractions.

To examine CR effects on soil surface crusts, bulked samples were taken in May 1994 from 16 plots with a stainless steel spoon at 0- to 10-, 10- to 20-, and 20- to 50-mm depth in plots with 2000 kg CR ha-1 applied as mulch or ash, with and without broadcast P. The samples were analyzed for soil chemical parameters and soil particle size distribution using the pipette method (Gee and Bauder, 1986).

To examine treatment effects on surface crusting, penetrometer measurements were collected before the onset of the rains in late April 1994. At 20 randomly chosen locations in each plot or split-plot, measurements were taken independently at 0 to 0.02 and 0 to 0.05 m. A hand-held penetrometer (Eijkelkamp, the Netherlands) was used with tips of 35-mm diam. for the upper depth and 15-mm diam. for the lower depth.

Experiment 2
Measurements were taken in a mulching experiment with millet planted at 10 000 pockets ha-1 from 1992 to 1994 (Buerkert et al., 1996b) to study CR effects on soil surface temperature and soil water content during the growing season. In one of the three replicates of the trial, soil surface temperature was recorded with three temperature probes (type 108, Campbell Scientific, Logan, UT) randomly placed at the 0.01-m depth in a unmulched (control) plot and in a plot mulched with 2000 kg surface applied CR ha-1. Given lower variation in temperature with increasing depth, only a single copper-constantan thermocouple wire was used to measure soil temperatures at 0.05 m in each of the two plots. The differential measurements of all sensors were recorded at intervals of 20 s, averaged every 30 min and stored with a CR10 datalogger (Campbell Scientific). In all three replicates, soil water content was monitored regularly during 1993 and 1994 with a neutron probe (Troxler Int. Ltd, Research Triangle Park, NC) in three access tubes per plot at 0.1, 0.2, 0.3, 0.5, 0.7, 1.0, 1.3, and 1.6 m. The probe had been previously calibrated for this site at different depth intervals and the respective regression equations were used to calculate water contents from measured counts.

Experiment 3
The 40 mainplot treatments attributed to plots of 10 by 10 m and replicated twice at each site were factorial combinations of three factors: CR, N, and P (Table 2) . Crop residues were broadcast at 500 or 2000 kg CR ha-1 in January of each year, the middle of the dry season. For the sites in the Sahelian and Sudanian zone, CAN was broadcast at 0, 30, 60, and 90 kg N ha-1 and for the sites in the Guinean zone at 0, 60, 90, and 120 kg N ha-1. All N rates were split into three equal applications which were made after emergence (10 d after sowing, DAS), at thinning (25 DAS), and at booting (50 DAS). The eight P levels were (i) a control without P, (ii) annual P placement with the seed at a rate of 4 kg P ha-1 (P4) as single superphosphate (SSP) in 1995 and 1996 and as NPK in 1997, (iii) annual broadcast P at 13 kg P ha-1 as SSP (SSP13), broadcast P as `soft' Tahoua rock phosphate (TRP) at a 3-yr rate of 39 kg P ha-1 (iv) with and (v) without seed placement of SSP (TRP39 and TRP39+P4), (vi) as "soft" rock phosphate at a 10-yr rate of 130 kg P ha-1 (TRP130) or as "hard" rock phosphate at a 10-yr rate of 130 kg P ha-1 (vii) with and (viii) without seed placement of SSP (RP130 and RP130+P4). At the Burkina site, the hard rock phosphate was from Kodjari and in Togo from Hahotoe. Despite generally higher TDM production in the wetter zones and differences in the cereal species sown, CR mulch and P levels were kept constant to facilitate the comparison of treatment effects on crop growth and soil properties across environments. In Niger (Kara Bedji, Banizoumbou, Sadoré, Goberi and Gaya), the crop sown was millet at 10000 pockets ha-1. In Burkina Faso (Kouaré) sorghum was sown at 40 000 pockets ha-1, and in Togo (Koukombo and Kaboli) maize at 53 330 pockets ha-1.


View this table:
[in this window]
[in a new window]
  		Table 2 Treatment combinations in a multi-factorial experiment with crop residue mulch and mineral fertilizer application from 1995 to 1997 in West Africa

 
Plant and Soil Sampling
At harvest, the pockets of each plot, with the exception of border rows, were cut as stated earlier and the dry weight of grain, heads, and remaining straw determined to compute TDM. At two sites in each of the three climatic zones (Kara Bedji, Goberi, Gaya, Fada, Koukombo, and Kaboli), the topsoil was sampled before the onset of the third rainy season. Five subsamples were taken in both replicates of eight selected treatments at depths of 0 to 0.1 and 0.1 to 0.2 m. Plots sampled comprised the two CR levels (500 and 2000 kg ha-1) without N application at three levels of P (control, SSP13, and TRP39) and two additional P levels (SSP4 and TRP130) at 500 kg CR ha-1. Subsamples were bulked, dried, sieved, and analyzed for soil chemical parameters as stated above.

Statistical Analysis
GENSTAT (Lawes Agricultural Trust, 1993) was used to compute F-statistics with analyses of variance and standard errors of the difference. Wherever applicable, depth intervals were treated as split-plots thereby taking into account their spatial dependence. Some of the plant dry matter data were examined for normal distribution of residuals with SAS version 6.06 (SAS Institute, 1991). Occasionally minor deviations were found but data transformations had very little influence on the F-values. For consistency, it seemed appropriate to present statistics of untransformed data while being aware that true probabilities may slightly differ from those shown.


   Results
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
Discussion
 Conclusions
 REFERENCES
 
Experiment 1
Plant Growth
Depending on the P level, mulch-induced effects on millet TDM were negligible or even slightly negative in the first 2 yr after the prolonged period of fallow at this site but increased in 1993 across P levels leading to almost twice the TDM compared with unmulched (ash) plots (Fig. 1) . Plots with ash application had their highest yields in the first year after the prolonged fallow and were still more productive than plots with 500 kg ha-1 of CR mulch in 1992. In 1993, however, TDM yields in ashed plots collapsed regardless of the P level applied. The dynamics of residue-induced increases in millet growth after repeated cultivation became obvious with the split-plot application in 1994. Across P levels, the early fourth mulch application led to millet TDM increases of 430 kg ha-1 and 1010 kg ha-1 with the low and high CR mulch rate, respectively, as compared with the delayed mulch application in April (Muehlig-Versen, 1994, unpublished data). In contrast to the time-dependent cumulative effects of mulching on crop growth, the impact of SSP application at 13 kg P ha-1 was much more immediate, leading to TDM increases of 28% in the first, 51% in the second, and 43% in the third year of application compared with unfertilized plots (Fig. 1).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 1 Millet total dry matter production at Sadoré, Niger, as affected by phosphorus (P) applied at 0 and 13 kg P ha-1 as SSP and surface broadcast millet crop residues (CR) at annual rates of 500 and 2000 kg dry matter ha-1 or as ash (unmulched) at the rate of 2000 kg burned CR ha-1. Total annual rainfall was 592 mm in 1991, 511 mm in 1992, and 533 mm in 1993. Data are means of four millet genotypes; vertical bars represent one standard error of the difference

 
Nutrient Balances
In each year, nutrient removal strongly exceeded inputs for all mineral elements analyzed except P after SSP application and K in unmulched plots without SSP in 1993 (Fig. 2) . Losses were largest for N and K, particularly after P application. Mulch application at 2000 kg CR ha-1 decreased net N and K losses in 1991 and 1992. However, in 1993 the reverse was true when N and K balances were most negative at the high rate of mulch (Fig. 2).



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 2 Nutrient balances for nitrogen (N), phosphorus (P), and potassium (K) as affected by P applied at 0 and 13 kg ha-1 as SSP and surface broadcast millet crop residues (CR) at annual rates of 500 and 2000 kg dry matter ha-1 or as ash (unmulched) at the rate of 2000 kg burned CR ha-1. Data are means of 16 replicates. Sadoré, Niger, 1991 to 1993

 
Near-Surface Soil Properties
Overall, the three uppermost surface soil layers at 0 to 10, 10 to 20, and 20 to 50 mm had higher pH, available P, exchangeable K, and base saturation than the topsoil at the 0- to 100-mm depth (Table 3 and 4) . Four successive applications of CR mulch had significantly affected most of the measured chemical properties. Compared with ash application, mulching at 2000 kg CR ha-1 increased pH by about 0.7 units throughout the three surface layers regardless of the level of applied P. The mulch-induced increases in P-water (P < 0.001) averaged 51% or 256 µg P kg-1 across P levels and depth (Table 3). Increases in exchangeable K were large in absolute terms but statistically not significant.


View this table:
[in this window]
[in a new window]
  		Table 3 Effects of broadcast phosphorus (P) fertilizer and crop residue (CR) application on chemical properties of the surface soil crust at 0- to 50-mm depth after 3 yr of treatment

 

View this table:
[in this window]
[in a new window]
  		Table 4 Effects of 3 yr of surface mulched crop residue (CR) application at 0 (unmulched with ash application), 500 and 2000 kg CR ha-1 with and without phosphorus (P) as broadcast single superphosphate at 13 kg P ha-1 on soil pH (in 0.01 M KCl), water soluble phosphorus (P-water), P-Bray, exchangeable potassium (K), base saturation, and organic carbon (C org) at two depths. Sadoré, Niger, 1991–1994

 
Mulching also indirectly affected the particle size distribution in the surface soil leading to significantly higher proportions of clay particles (< 2 µm) and coarse sand (< 1000 µm) in mulched plots than in unmulched plots. The proportion of medium size particles (< 250 µm) tended to decrease with CR application (Table 5) .


View this table:
[in this window]
[in a new window]
  		Table 5 Effects of crop residue application as ash (CR0) or mulch at 2000 kg ha-1 (CR2000) on the texture of the soil surface crust at 0– to 50-mm depth after 3 yr of treatment. Sadoré, Niger 1991 to 1994

 
Topsoil Properties
After three consecutive CR treatment applications, both mulch rates had significantly (P < 0.001) affected all measured chemical properties of the topsoil at both P levels (Table 4). For P (water and Bray) and K the mulch-induced increases were particularly large in the 0- to 0.1-m layer but remained noticeable even at the 0.1- to 0.2-m depth. Phosphorus application only increased P-Bray and base saturation but did not affect P-water (Table 4). Organic carbon concentrations as a function of depth and P level were significantly higher with mulching than in unmulched plots. However, with 2.0 g C kg-1 in plots with 2000 kg CR ha-1 average C org levels after 4 yr of cultivation were across both P levels 11% lower than the initial C org concentration measured after 8 yr of bush fallow (Table 1 and 4).

The early fourth CR application in November 1993 led to significant increases of all measured soil chemical properties at the 0- to 0.1-m depth, except for P-Bray and C org, compared with the application in April 1994 (data not shown). The early CR application increased P-water by 9%, exchangeable K by 13%, and base saturation by 3% in the upper topsoil layer averaged across the two rates as compared with the late application.

Visual observations showed a strong increase in termite activity with the rate of applied mulch as evidenced by "termite roads" across plots and partial coverage of mulched stalks with soil hulls. Three consecutive CR mulch applications significantly increased surface soil penetrability at both measured depths (P < 0.001). At 0- to 0.02-m mechanical resistance was 320 kN m-2 in unmulched plots, 210 kN m-2 with the low mulch rate, and 200 kN m-2 for plots with 2000 kg CR ha-1, whereas at 0 to 0.05 m, resistance values were 2950, 2570, and 2110 kN m-2, respectively. A fourth mulch application right after millet harvest further decreased soil resistance (P < 0.001) at 0 to 0.02 m to 190 kN m-2 with 500 kg CR ha-1 and to 90 kN m-2 with 2000 kg CR ha-1; at 0- to 0.05-m depth, the respective resistance values were 2260 and 1500 kN m-2.

Chemical Properites of the Subsoil
After 3 yr of CR application, the analysis showed mulching effects to 0.3 m across P levels. The relative increases in P-water, C org, and base saturation in plots with the high mulch rate compared with unmulched plots with ash application were particularly large (Fig. 3) .



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 3 Soil pH KCl, water soluble phosphorus (P-water), base saturation, and soil organic carbon (C org) concentrations in a typical profile of a Psammentic Paleustalf after 3 yr of continuous millet cultivation with annual applications of three levels of millet crop residues (CR) broadcast at 500 (CR500) and 2000 (CR2000) kg dry matter ha-1 or as ash (unmulched) at the rate of 2000 kg burned CR ha-1. Data points are averages of four plots sown with the millet cultivar Sadoré Local without and with yearly application of SSP at 13 kg P ha-1 at Sadoré, Niger, from 1991 to 1993. Horizontal bars represent one standard error of the difference

 
Experiment 2
Surface Temperature
On clear days, mulching decreased soil surface temperatures from 0900 to 1800 h. In the early rainy season, at 1430 h temperatures at the 0.01-m depth in unmulched plots reached 44°C, whereas temperatures in mulched plots at 0.01 and 0.05 m were 4°C lower. While seedlings appeared dark green and vigorous in mulched plots, many of the seedlings in unmulched plots looked severely N-deficient, developed brown leaf tips, and subsequently died.

Soil Moisture
Treatments did not affect soil moisture contents early in the season. However in 1993, with heavy rains in August, water contents in mulched plots by early September were higher than in unmulched control plots to a depth of 0.7 m (Fig. 4 a). Differences were even more marked (P < 0.10) in the wetter year of 1994 when mulching strongly increased soil water contents throughout the profile in August and to 0.3 m in September (Fig. 4 b and c).



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 4 Soil water contents from 0.1 to 1.6 m with and without surface mulched crop residues (+/- CR) at 2000 kg ha-1. Data were measured at Sadoré, Niger, in early September 1993 (A), late August 1994 (B), and mid September 1994 (C) and are means of three subsamples in each of the three replicates. Horizontal bars represent one standard error of the difference

 
Experiment 3
Plant Growth
Despite large variations in average TDM levels across sites and years (Table 6) a distinct difference in CR effects on cereal growth was observed between the Sahelian sites and the Sudano-Guinean sites. In the Sahel after the first year, CR mulch at 2000 kg ha-1 resulted in cereal TDM increases of up to 72% compared with 500 kg ha-1. In contrast, at the four sites in the Sudanian and Guinean zone TDM differences between the two mulch rates varied between -10% and +15% (Fig. 5) .


View this table:
[in this window]
[in a new window]
  		Table 6 Average total dry matter (TDM) yields of cereals with crop residue mulch applied at 500 kg CR ha-1 for 4 yr at eight sites in West Africa. Data are averages of 64 replicates across eight phosphorus treatments and four nitrogen treatments, values in brackets are standard errors of the mean

 


View larger version (19K):
[in this window]
[in a new window]
  		Fig. 5 Total dry matter yield of cereals in plots with an annual application of 2000 kg crop residues (CR) ha-1 relative to plots with 500 kg CR ha-1. Data points are means of 64 replicates across eight phosphorus and four nitrogen treatments of an experiment conducted from 1995 to 1998 at four sites in the Sahel, two sites in the Sudanian zone, and two sites in the Guinean zone of West Africa

 
Topsoil Properties
Residue effects on soil chemical properties closely mirrored those on plant growth and declined with increasing rainfall from North to South. For Kara Bedji in the Sahel, changes in soil properties were very similar to those at Sadoré obtained in experiment 1 (Table 7) . Compared with 500 kg CR ha-1, the application of 2000 kg CR ha-1 for 2 yr resulted in increases in pH, available P, and extractable K. The high mulch rate also raised the base saturation and helped maintain the initial level of C org (Table 1). For Goberi, the Sahelian site with the highest rainfall and the largest clay content, CR effects on chemical soil properties were detectable but much smaller than at Kara Bedji (Table 7). Compared with the Sahel, mulching effects on pH or base saturation were minor in the Sudanian and Guinean zone. Nevertheless, the high mulch rate caused a significant increase in C org at Gaya as compared with the low rate of 500 kg CR ha-1 (Table 8) . Residue effects on P-Bray were inconclusive but the concentration of P-water, a particularly sensitive indicator for CR effects in the Sahel, was consistently lower throughout the Sudanian and Guinean zone at the high CR mulch rate as compared with the low rate. At 0 to 0.2 m, the percentage decrease in P-water with the high compared with the low mulch rate was 33% at Gaya, 22% at Fada, 9% at Koukombo, and 27% at Kaboli (Table 8). After 2 yr of cultivation, a major decrease in base saturation was observed at Kaboli which was independent of the mulch rate (Table 1 and 8).


View this table:
[in this window]
[in a new window]
  		Table 7 Effects of millet crop residues (CR) surface mulched for 2 yr at 500 (CR500) and 2000 kg ha-1 (CR2000) on soil pH (in 0.01 M KCl), water soluble phosphorus (P-water), P-Bray, exchangeable potassium (K), base saturation, and organic carbon (C org) in two depths at two sites in the Sahelian zone of West Africa. Data are means across five P levels

 

View this table:
[in this window]
[in a new window]
  		Table 8 Effects of crop residues (CR) surface mulched for 2 yr at 500 (CR500) and 2000 kg ha-1 (CR2000) on soil pH (in 0.01 M KCl), water soluble phosphorus (P-water), P-Bray, exchangeable potassium (K), base saturation, and organic carbon (C org) in two depths at two sites in each the Sudanian (Gaya and Fada) and Guinean (Koukombo and Kaboli) zone of West Africa. Data are means across five P levels each applied to two cropping systems

 

   Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
 Discussion
Conclusions
 REFERENCES
 
Mulching Effects in the Sahelian Zone
The large increases in millet TDM following P application from the first year onward provide good evidence of the major role of this mineral nutrient in the Sahel and confirm the additive effects of CR and P application shown previously by Hafner et al. (1993b). The data also show the cumulative effects of CR application with time (Fig. 1). In the first and second year of application, residue-induced increases in millet TDM appeared to be effectively masked by the residual effects of the prolonged fallow leading to a high spatial variability in plant growth. The initially high yields in unmulched plots to which ash has been applied may also be explained by residual effects of former human settlements and trees on some of these plots (Buerkert et al., 1995). Under these typical post-fallow conditions, mulch-induced TDM increases in millet became obvious only in 1993, the third year of application, when TDM production continued to increase at the high mulch rate compared with a rapid TDM decline in unmulched plots (Fig. 1).

While P application led to significant differences in mineral nutrient concentrations within the plants (Buerkert et al., 1998), its effect on nutrient balances was governed by differences in TDM production (Fig. 1). This became particularly obvious for the negative effect of P application on the N and K balances (Fig. 2). However, actual nutrient losses at the village or the watershed level are likely much smaller given the various modes of nutrient recycling via manure in crop- and rangelands as well as wind-related nutrient depositions during fallow periods (Buerkert and Hiernaux, 1998).

The pattern of TDM effects caused by P and CR application points to the interactive mechanisms responsible for the crop growth responses to both factors on sandy Sahelian soils. Phosphorus application strongly increased soil P-Bray across CR levels (Table 4). At the same time, the application of mineral P fertilizer had only minor effects on the more immediate P availability as measured by P-water; however, surface mulched CR increased P-water between 33 and 37% and P-Bray between 20 and 48%, at 0 to 0.2 m, This was most likely due to the small but consistent increase in pH as a result of added basic cations from dust deposition (Stahr and Herrmann, 1996) that was captured in mulched plots and the mobilization of soil P through the release of organic acids from decomposing CR (Hue, 1991). These acids may have acted as anion exchangers displacing P from the soil matrix and as ligands for Al and Fe (Fig. 6) . Compared with the 29 kg P ha-1 returned annually with 16 Mg of maize stover in an experiment at IITA (Juo and Lal, 1977), mineralized P from the CR mulch in our study contributed with less than 1 kg P ha-1 little to the changes in available P with mulching. To gain a more complete understanding of the dynamic changes of chemical P availability in the different pools, more detailed investigations with isotopic exchange methods are necessary (Frossard et al., 1993).



View larger version (46K):
[in this window]
[in a new window]
  		Fig. 6 Conceptual model showing differences in the agro-ecological conditions between the Sahelian and the Sudano-Guinean zone of West Africa and the importance of possible mechanisms involved in crop residue (CR) mulch-induced cereal growth increases. Meaning of symbols are: ++ very important, + important, (+) possibly important, {circ} not important, - not relevant, and ? uncertain given insufficient data

 
Mulching is also likely to have improved spatial P availability in the topsoil through increases in root length density as reported by Hafner et al. (1993b). Several interrelated factors may be cited to account for the effects of mulching on root development. These include the combination of small differences in soil moisture leading to higher nutrient availability, a smaller physical resistance to root elongation, and a hormone driven feed-back mechanism stimulating root growth through larger bacterial populations that feed on exudates of larger plants (Buerkert et al., 1995; Hafner et al., 1993a; Kretzschmar et al., 1991; Fig. 6). In most of the trials in the Sahelian zone, mulch application visibly reduced the formation of new surface crusts because of a reduction in wind-induced soil movement (Michels et al., 1995a) and the increase in termite activity observed repeatedly with higher mulch rates.

The large effects of mulching on base saturation in the topsoil of CR plots reflect the increase in pH and cations, particularly K, as a result of released nutrients from decomposing residues and dust deposits (Table 3 and 4). Compared with the 44 kg K from 2000 kg CR ha-1, of which about 85% was decomposed after 12 mo in the Sahel, K inputs from dust deposition are much smaller. For well protected surfaces such as fallow, K inputs from Harmattan winds have been estimated at between 6 and 15 kg K ha-1 yr-1 (Herrmann et al., 1994).

It may be argued that these mulch-induced changes in nutrient concentration of the topsoil are too small to explain the observed effects on cereal growth. However, the changes are large in relative terms given the fact that the acid Sahelian soils in this study are nothing more than weakly buffered sand cultures. This is supported by Wendt et al. (1993), who found that minor changes in soil parameters, particularly in P availability and in surface crusting, could explain striking differences in plant growth over short distances often referred to as "microvariability" (Brouwer et al., 1993; Buerkert et al., 1996a). The sandy nature of these soils may also explain why CR effects on soil chemical parameters were detected to the 0.8-m depth. The increased C org concentrations in the subsoil and the unexpected changes in P availability below 0.5 m might be caused by increased root growth and leaching of short chain organic molecules from decomposing CR through the profile. This has been hypothesized by Brouwer and Powell (1995), who found large losses of P from a manure trial in a nearby field at Sadoré.

With some differences in the absolute effects of mulching on crop growth and soil parameters, the pattern seems to be remarkably consistent across sites in the Sahel (Fig. 5 and Table 7). The immediate large increase of millet TDM with the high mulch rate at Kara Bedji may be due to the fact that, unlike the other sites, the chosen farmer's field had not been left fallow before the establishment of the trial in 1995 but was sown to millet for over five consecutive years.

The significant decrease in soil surface temperature measured in mulched compared with unmulched plots was unexpected given that a mulch rate of 2000 kg ha-1 is equivalent to a surface coverage of barely 10%. However, similar temperature differences between unmulched (48.2°C) and mulched plots (41.7°C) have also been measured in an on-farm trial at Maradi, south-western Niger in 1994 (Buerkert, 1994, unpublished data). The lower soil temperature with mulching may be caused by the combined effects of changes in albedo and surface roughness, increased plant cover, and higher water content at the soil surface. The differences in topsoil temperatures are remarkably similar to those reported from western Nigeria by Lal (1974) and De Vleeschauwer et al. (1978) between unmulched plots and plots with 4 to 6 Mg rice straw ha-1.

The increased water content in mulched plots with higher observed termite activity (Fig. 4) confirms results by Mando et al. (1996) and Mando and van Rheenen (1998) on a Sahelian soil, who measured a 41% higher infiltration and up to 50 mm more stored soil water with termites compared with plots without termites. From their work in Nigeria, De Vleeschauwer et al. (1978) reported up to 11% higher soil moisture storage in the top 0.05 m of a freshly cleared Alfisol due to the application of 6 Mg rice straw ha-1. Under the more humid conditions of their site, this difference was attributed not to termites but to an earthworm-related increased volume of macro-pores and less surface crusting leading to a higher infiltration rate (De Vleeschauwer et al., 1980).

Mulching Effects in the Sudanian and Guinean Zones
Despite slightly higher C org and base saturation in plots with the high mulch rate at most sites (Table 8), the annual application of 2000 kg CR ha-1 failed to increase cereal TDM production in the Sudanian and Guinean zone. This was even true in plots that had received yearly applications of 13 kg P ha-1 as SSP and 60 or 90 kg N ha-1. To obtain significant effects of mulching on cereal TDM in the more humid zones of West Africa such as the 10 to 29% (Juo and Lal, 1977) and the 50% (Lal, 1974) reported for western Nigeria, mulch rates would likely have to be increased considerably. The latter authors had applied rice straw at rates between 4 and 10 Mg ha-1 yr-1. There is little data about residual effects of mulching on crop growth for West Africa that would justify the application of higher mulch rates at intervals. Even for the Sahel with its long dry season and relatively low mineralization rates, the few available reports indicate a substantial decline of mulch effects on millet growth after 1 yr of withdrawal (Rebafka et al., 1994). This is in sharp contrast to large residual effects of straw mulching in the U.S Midwest on corn yield, C org, and N and P availability with a half-life of up to 10 yr (Power et al., 1998).

In view of the mechanisms underlying CR responses in the Sahel (Fig. 6), the results of the soil analyses seem particularly helpful in explaining the striking differences between agro-climatological zones. The minor influence of CR on pH, P-Bray, and particularly P-water at the Sudano-Guinean sites is likely due to (i) less dust deposition, (ii) higher inherent soil fertility and clay contents, (iii) lower air and soil temperatures, (iv) higher rainfall and soil water holding capacity on the heavier soils, and (v) a much higher rate of CR decomposition (Buerkert and Hiernaux, 1998; Fig. 6). While in the Sahel between 15 and 55% of the surface applied 2000 kg CR ha-1 remained after 1 yr, in the Sudanian and Guinean zone the complete mulch cover disappeared within 9 to 12 mo. Soil preparation and cultivation methods particular to each zone may also contribute to regional differences in CR mulch effects on soil properties and plant growth. Fields in the Sudanian and Guinean zone are traditionally animal plowed with the first rains before sowing which leads to a mechanical disruption of surface crusts and a partial incorporation of mulch material. Weeding is done with a hand-held hoe which turns the soil to about 0.1 m. In the Sahel, however, millet is typically sown on the untilled soil and fields are weeded with a hand-held metal blade on a long handle that slides on the soil surface at about 0.01 m.

The data of TDM yields at 2000 kg CR ha-1 from the eight sites indicate that annual application of mineral N and P fertilizers, the on-farm availability of stover to maintain this mulch rate may be questionable given the multiple competitive uses of CR in the predominant agro-pastoral systems of the region (Table 6; Lamers and Feil, 1993).


   Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
 Discussion
 Conclusions
REFERENCES
 
The data support the view that enhanced P availability resulting from residue mulch application is the major cause of subsequent increases in cereal TDM production on weakly buffered acid sandy Sahelian soils. The results also confirm earlier data from western Nigeria that given the strong activity of the soil fauna in the region, mulching can only slow down the decline in C org commonly observed under continuous cultivation, but not increase it.

To predict CR effects on cereal growth across climatic zones, an understanding of the relative importance of the processes involved along the agro-ecological gradient from the Sahel to the Guinean zone is necessary. Nevertheless, it should be stressed that even with substantial inputs of mineral N and P fertilizers, it may remain difficult to obtain crop yields under on-farm conditions which allow the recycling of residues as surface mulch at the often recommended rate of 2000 kg ha-1.


   ACKNOWLEDGMENTS
 
The authors are grateful to A. Moussa, M. Bachir, A. Thiombiano, and I. Oumarou for their technical help at sampling and sample analysis, to H. Traoré (INERA, Burkina Faso) and ICRISAT Sahelian Center for logistical support and to F. Graef for his help with soil classification. They also thank B. Buerkert, B. Gérard, S. Alvey, V. Roemheld, N. van Duivenbooden, E.A. Kirkby, and three anonymous reviewers for their constructive comments on earlier versions of this paper.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
Joint contribution from Univ. of Hohenheim and ICRISAT Sahelian Ctr., ICRISAT journal article no. 2109. Research supported by the Deutsche Forschungsgemeinschaft (SFB 308).

1 Dedicated to Horst Marschner and his commitment to process-oriented soil fertility research in West Africa. Back

Received for publication November 9, 1998.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
 Discussion
 Conclusions
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (26)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Buerkert, A.
Right arrow Articles by Dossa, K.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Buerkert, A.
Right arrow Articles by Dossa, K.
Agricola
Right arrow Articles by Buerkert, A.
Right arrow Articles by Dossa, K.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome