Agroforestry Systems 51: 239–251, 2001. 2001 Kluwer Academic Publishers. Printed in the Netherlands.
Litter and biomass production from planted and natural fallows on a degraded soil in southwestern Nigeria F. K. Salako & G. Tian* Soil Fertility Unit, Resource and Crop Management Division, International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria, c/o L.W. Lambourn & Co., 26 Dingwall Road, Croydon CR9 3EE, England (*E-Mail:
[email protected]) Received 12 July 1999; accepted in revised form 17 August 2000
Key words: leguminous trees, litterfall, planted and natural fallows, southwestern Nigeria
Abstract To rehabilitate a degraded Alfisol at Ibadan, southwestern Nigeria, Senna siamea (non-N2-fixing legume tree), Leucaena leucocephala, and Acacia leptocarpa (N2-fixing legume trees) were planted in 1989, and Acacia auriculiformis (N2-fixing legume tree) in 1990. Pueraria phaseoloides (a cover crop) and natural fallow were included as treatments. Litterfall and climatic variables were measured in 1992/1993 and 1996/1997 while biomass production and nutrient concentrations were measured in 1993 and 1995. Total litter production from the natural and planted fallows was similar, with means ranging from 10.0 (L. leucocephala) to 13.6 t ha–1 y–1 (natural fallow) during the 1996/1997 collection. Leaves constituted 73% (L. leucocephala) to 96% (A. auriculiformis) of total litterfall. Acacia auriculiformis grew most quickly but S. siamea produced the highest aboveground biomass which was 127 t ha–1 accumulated over four years, and 156 t ha–1 accumulated over six years of establishment. The aboveground biomass of P. phaseoloides and natural fallow was only 6 to 9 t ha–1 at six years after planting. Nitrogen concentration in the leaves/twigs of was 2.5% for L. leucocephala, and 2% for other planted species and natural fallow. Pueraria phaseoloides had concentrations of P, K, Ca and Mg comparable to levels in the leaves/twigs of the tree species. Through PATH analysis, it was found that maximum temperature and minimum relative humidity had pronounced direct and indirect effects on litterfall. The effects of these climatic variables in triggering litterfall were enhanced by other variables, such as evaporation, wind, radiation, and minimum temperature. Improvement in chemical properties by fallows was observed in the degraded soil.
Introduction Soil regeneration, in the context of this study, is used to describe efforts aimed at reclaiming previously degraded agricultural land for further agricultural use. Terms commonly used in the literature for such research include soil restoration and soil rehabilitation. Johnson and Bradshaw (1979) observed that soil restoration might not be a complete reinstatement of the original environment but the restoration of a properly functioning
soil/plant ecosystem. The objectives for reclaiming a soil need to be properly articulated before decision implementation. A key component to soil regeneration for agricultural land use is the accumulation of organic matter and its conversion to humus. Litter production from plants, particularly trees, is a major source of organic matter in soils under restoration for agricultural purposes. Both the species and age of plantations largely determine the amount of litter production of trees (Lugo and
240 Brown 1993). The efficient conversion of this litter to soil humus by soil organisms (Johnson and Bradshaw 1979) depends largely on climatic factors, and the selected species (i.e., litter quality). The peak of litterfall in the tropics coincides with the dry season (John 1973). Apart from this broad observation, there have been efforts to find the relationship between climatic variables and litter production at regional levels (Vitousek 1984; Vogt et al. 1986). Vitousek (1984) reported that the best single predictor of fine litterfall dry mass was a temperature-precipitation interaction. Williams and Tolome (1996) reported that some tropical tree species showed a positive correlation of litterfall with maximum temperature. Williams et al. (1997) observed that leaffall coincided with the attainment of the seasonal minimal level in leaf water potential. There is a need for a narrower, location-specific investigation of the interaction of climatic variables and litter production if the processes involved are to be well understood. There are potential extra benefits from trees planted for soil regeneration in the humid tropics; timber, fuel-wood, and fodder production. Leguminous trees such as Leucaena leucocephala, Senna siamea and Acacias (e.g., Acacia leptocarpa and Acacia auriculiformis) are widely used in agroforestry systems (Buresh and Tian 1998). Both the N2-fixing species such as L. leucocephala and the Acacias, and non-N2-fixing legumes such as S. siamea have potential for the restoration of soil fertility (Giller 1998; Buresh and Tian 1998). The objective of this study was to evaluate litter and biomass production by different leguminous fallow species planted for the regeneration of a degraded soil; the relationship between climatic variables and litter production was also evaluated. This paper, one of a series addressing the biological restoration of a degraded tropical Alfisol, aims to supplement the previous studies on the responses of soil properties and crop yields to the fallows (Kang et al. 1997), and the restoration of microarthropods (Adejuyigbe et al. 1999) and earthworms (Tian et al. 2000).
Materials and methods Study site and experimental design This study was carried out between 1992 and 1997 at the International Institute of Tropical Agriculture (IITA), Ibadan (latitude 7°30′ N and longitude 3°54′ E), southwestern Nigeria, on an experiment established in 1989 to regenerate a degraded soil by planting woody and herbaceous fallow species. Mean annual rainfall at Ibadan is 1300 mm. The primary rainy season is from April to August and the minor rainy season is from August to October, followed by a dry season from November to March. Mean monthly rainfall ranged from 0 (January) to 234 mm (September) between 1992 and 1997. The soil, an Oxic Kandiustalf (sandy loam or loamy sand topsoil) (Moormann et al. 1975), was degraded by compaction from land clearing in 1979, and soil erosion during 10 years of cropping (Lal 1992). The surface (0–10 cm) soil had 652 g kg–1 sand, 100 g kg–1 silt and 248 g kg–1 clay in 1989. Bulk density was 1.45 Mg m–3 before fallow establishment in 1989 (Hulugalle 1992). Means of prefallow soil chemical characteristics (0–10 cm) at the same depth were 5.4 for pH, 11 g kg–1 for organic carbon, and 6.5 cmol (+) kg–1 for cation exchange capacity. Part of the upper and mid-slope (8 to 10%) of the degraded site was demarcated into experimental plots of 12 × 24 m each in a randomized complete block design with four blocks. Details of the 15 woody and herbaceous fallow species which were planted as treatments in 1989 in addition to natural fallow and continuous cropping with maize/cassava intercrops were described by Hulugalle (1992). Woody species were planted at a spacing of 4 m (between rows) and 0.5 m (within rows), and herbaceous species at a spacing of 1 × 0.5 m. Acacia auriculiformis was planted in 1990, while other fallow species were planted in 1989. Four woody species (S. siamea, L. leucocephala, A. leptocarpa, and A. auriculiformis) and two herbaceous species (Psophocarpus palustris and P. phaseoloides) had survived by 1993 (Kang et al. 1997). One-third of each plot (12 × 8 m) was cleared in 1993 and in 1995 and the other onethird remained under fallow. Psophocarpus palus-
241 tris had been overgrown by the natural fallow vegetation by 1995. Litterfall dynamics Litterfall was measured between March 1992 and February 1993, and between April 1996 and March 1997 on the woody species and natural fallow plots, using litter traps 50 × 50 cm wide and 15 cm deep with 1 mm mesh size. In 1992/1993, litterfall was collected at two-week intervals. During this period, four litter traps were placed in each plot, two close to tree rows (within 50 cm) and two in central positions between two hedgerows. This spatial arrangement was not applicable to the natural fallow plots where all the traps were considered to be centrally located. By 1996/1997, there were only two rows of trees on a 12 × 8 m plot for litterfall collection. Therefore, the number of litter traps was reduced to one for each position. The 1996/1997 measurement intervals were disrupted by national socio-political crises causing them to range from 14 to 48 days. Litterfall was collected, placed in a paper bag, and dried in an oven at 65 °C before being separated into pod, wood (very small branches of about 2 to 4 cm circumference) and foliage components. The components were weighed separately and total litterfall was obtained by adding their weights. Climatic variable measurements The measurements of different climatic variables were carried out in the open at the central weather station of the IITA. Similar instruments used for the measurements are described by Lee (1978) and Rosenberg et al. (1983). Aboveground biomass characteristics Breast height (1.3 m above ground level) circumference of the trees was measured in March 1992, January 1993, and October 1996 using flexible graduated tapes. In 1992 and 1993, all the trees within the inner 10 × 20 m of each plot were measured while 12 trees from each plot were measured in 1996.
In 1993 and 1995, aboveground biomass and nutrient concentrations were measured just before the clearing of segments (12 × 8 m each year) of the whole plot for phased cultivation. For each of the tree species, 30% of the stand density in the 12 × 8 m area for cultivation in 1995 was cut to the ground level. The trees were chopped into parts, tied, and weighed for the determination of fresh biomass of components (leaves/twigs [including pods] and stems/main branches). Hereafter, the stems/main branches are referred to as wood. Subsamples of the leaves/twigs and wood were further chopped to pieces for ovendrying at 65 °C. The aboveground biomass of P. phaseoloides and natural fallow were sampled from two 4 × 4 m quadrats. There was no component partitioning for P. phaseoloides and natural regrowth vegetation (mainly Chromolaena odorata in 1993 but with some interspersing shrubs and woody species in 1995). The biomass of the undergrowth (mainly Chromolaena odorata) in each tree plot was also determined by clearing two 4 × 4 m quadrats in each plot. After subsampling and subsequent removal of the large tree trunks, the remaining vegetation was burned. The dry subsamples of various kinds of vegetation were ground for analysis of N, P, K, Ca and Mg concentrations using routine procedures described in IITA (1979). Composite soil samples at 0–15 cm depth were collected before clearing (9th March) in 1995 and after burning (mid-April in 1995). The samples were analyzed for pH-H2O (1:1), organic carbon (Org. C), K+, Ca2+ and Mg2+ and extractable Bray-1 P following routine procedures as described in IITA (1979). Data calculation The litterfall measurements (i.e., the different components and total litterfall) were converted to amounts per area per year (kg ha–1 y–1). The means and standard errors were calculated for each species (S. siamea, L. leucocephala, A. leptocarpa, A. auriculiformis, and natural fallow) and locations (hedgerow and center). Analysis of covariance was carried out using the fallow species as treatments, and locations and cumulative days of litterfall collection as covariates. Contrasts were
242 then calculated to compare the treatments (Hinkelmann and Kempthorne 1994). The descriptive statistics, analysis of covariance and contrasts were calculated with STATISTIX (Analytical Software 1998). PATH analysis (Li 1975) was carried out to determine the relationship between litterfall components (podfall, woodfall, leaffall, and total litterfall) for each species (dependent variable) and climatic variables (independent variables). The 1992/1993 and 1996/1997 data were combined for this analysis, resulting in 23 data points for each fallow species. The independent variables were (i) cumulative days of litter collection, (ii) rainfall amount (mm), (iii) Class A pan evaporation (mm), (iv) wind speed (km h–1), (v) radiation (cal m–2 day–1), (vi) maximum temperature (°C), (vii) minimum temperature (°C), (viii) maximum humidity (%), and (ix) minimum humidity (%). Daily records of climatic variables were reconciled with appropriate intervals of litter collection, and the average values of these variables for each
interval were obtained for pairing with litterfall data. First regression and correlation analyses were carried out using SAS procedures (SAS Institute Inc. 1990) to obtain standardized estimates and correlation matrix. These data were then subjected to PATH analysis with a BASIC program. Analysis of variance (ANOVA) was also carried out using SAS procedures. Least significant difference was calculated at 5% probability level (LSD0.05). Results Litter production The total litterfall for S. siamea (non-N2-fixing legume) was significantly higher than that of any other species (N2-fixing legumes) in 1992/1993 but not in 1996/1997 (Tables 1 and 2). Leucaena leucocephala had a significantly higher total litterfall than A. leptocarpa and A. auriculiformis in
Table 1. Litterfall (mean ± standard error), and contrast information for species comparison during the 1992/1993 period on the degraded Alfisol in Ibadan, southwestern Nigeria. Litterfall (kg ha–1 y–1)
Fallow species (n = 304 for each)
Location
S. siamea
L. leucocephala
A. leptocarpa
A. auriculiformis
Natural fallow
Hedgerow (n = 608)
Center (n = 912)
Pod Wood Leaf Leaffall % of total
0083 ± 21 0578 ± 117 7118 ± 221 0091.5
1011 ± 136 1274 ± 326 6499 ± 198 0074.0
1141 ± 226 0126 ± 31 6335 ± 333 0083.3
0155 ± 58 0114 ± 43 6650 ± 390 0096.1
0553 ± 126 0401 ± 70 6760 ± 287 0087.6
0617 ± 99 0629 ± 151 7004 ± 214
0570 ± 76 0412 ± 65 6451 ± 166
Total
7779
8783
7602
6919
7714
8250
7433
Contrast information Probability values for significance Pod
Wood
Leaf
Total
Contrasts Natural fallow versus planted fallows Acacia species versus planted non-Acacias Non-nitrogen-fixing legume (S. siamea) versus nitrogen-fixing legumes (other planted fallows)
NS NS
NS 0.000
NS NS
NS 0.000
0.000
NS
NS
NS
Covariates Location Cumulative days
NS NS
NS NS
0.010 0.000
0.003 0.000
NS not significant at P = 0.05.
243 Table 2. Litterfall (mean ± standard error), and contrast information for species comparison during the 1996/1997 period on the degraded Alfisol in Ibadan, southwestern Nigeria. Litterfall (kg ha–1 y–1)
Fallow species (n = 98 for each)
Location
S. siamea
L. leucocephala
A. leptocarpa
A. auriculiformis
Natural fallow
Hedgerow (n = 176)
Pod Wood Leaf Leaffall % of total
00546 ± 83 01306 ± 169 08516 ± 751
00511 ± 105 01154 ± 152 08384 ± 813
00058 ± 21 00934 ± 177 09673 ± 996
00888 ± 212 00693 ± 194 10526 ± 1281
00122 ± 40 00524 ± 96 02056 ± 234 01213 ± 129 11434 ± 1105 09583 ± 688
00082.1
00083.4
00090.7
00086.9
00084.0
Total
10368 ± 726
10049 ± 791
10665 ± 995
12107 ± 1361
13612 ± 1129 11320 ± 704
Center (n = 264) 00359 ± 60 01213 ± 129 09789 ± 599
11388 ± 611
Contrast information Probability values for significance Pod
Wood
Leaf
Total
Contrasts Natural fallow versus planted fallows Acacia species versus planted non-Acacias Non-nitrogen-fixing legume (S. siamea) versus nitrogen-fixing legumes (other planted fallows)
0.004 NS
0.000 0.017
NS NS
0.015 NS
NS
NS
NS
NS
Covariates Location Cumulative days
NS 0.000
0.041 NS
NS 0.000
NS 0.000
NS not significant at P = 0.05.
1992/1993. Total litterfall was significantly different between the Acacia species (A. leptocarpa and A. auriculiformis) and non-Acacias (S. siamea and L. leucocephala) in 1992/1993 but not in 1996/1997. Total litter production increased with age, but this trend was mainly reflected by the leaffall component, which was less variable within each period than podfall and woodfall. Leaffall generally constituted more than 80% of total litterfall at the site (Tables 1 and 2). Woodfall from the Acacia species was significantly less than that from the two non-Acacia woody species in both periods. Woodfall amounts were similar among S. siamea and the other three woody species in 1992/1993 and 1996/1997. Podfall was significantly lower for the S. siamea (non-N2-fixing) than for L. leucocephala, A. leptocarpa, and A. auriculiformis (N2-fixing species) in 1992/1993, and was similar in 1996/1997 (Tables 1 and 2). The effects of litter trap location on litterfall
amount were different in 1992/1993 and 1996/1997 (Tables 1 and 2). Variations of litterfall variables (podfall, leaffall, and total litterfall) were strongly influenced by seasonal changes (i.e., cumulative days). For natural and planted fallow, litterfalls in both 1992/1993 and 1996/1997 were highest between November and December. A tendency for peak total litterfall between January and March was particularly exhibited by A. auriculiformis and natural fallow. Peak podfall occurred in January for A. auriculiformis, in February for S. siamea, in March for L. leucocephala and natural fallow, and in April for A. leptocarpa. Multiple peaks were observed for woodfall in January, February, June, July, October, November, and December. The seasonal pattern for the mean total litterfall (kg ha–1 y–1) (variable ‘Y’) for each month (January = 1 to December = 12) (variable ‘X’) is explained for each seven- to eight-year-old species with the following equations:
244 S. siamea: L. leucocephala: A. leptocarpa: A. auriculiformis: Natural fallow:
Y = 436X2 – 5441X + 21863 R2 = 0.91 Y = 160X2 – 1801X + 13078 R2 = 0.50 Y = 338X2 – 3952X + 17136 R2 = 0.55 Y = 485X2 – 5753X + 20752 R2 = 0.73 Y = 321X2 – 4252X + 20358 R2 = 0.77
where R2 is coefficient of determination. The multiple peaks of woodfall in a year were a source of high variability in total litterfall. Effects of climatic variables Leaffall and total litterfall were largely controlled by minimum relative humidity and maximum temperature (Tables 3 and 4). The trends observed for
the leaffall of S. siamea, A. auriculiformis, and natural fallow were observed for their total litterfall. Evaporation enhanced effects of temperature and humidity on leaffall and total litterfall from L. leucocephala and A. leptocarpa. Podfalls from S. siamea, L. leucocephala, and natural fallow were substantially influenced by maximum temperature and minimum relative humidity. Also, temperature and radiation had a significant influence on podfall from the Acacia species. The overall influence of these climate variables on podfall was enhanced by wind and evaporation. Minimum relative humidity was considered an important climatic variable influencing woodfall from all the species. It had a significant effect on woodfall from S. siamea and considerably enhanced the influence of maximum temperature on woodfall from L. leucocephala and A. auriculiformis. The direct effect of wind on woodfall from A. leptocarpa was pronounced.
Table 3. Pronounced* direct and indirect effects of climatic variables on leaffall as obtained through PATH analysis of the 1992/1993 and 1996/1997 data from Ibadan, southwestern Nigeria. Effects
Estimation of effect size
Senna siamea Direct effect of minimum relative humidity Indirect effect of minimum relative humidity via maximum temperature
–1.69 –0.95
L. leucocephala Direct effect minimum relative humidity Indirect effect of minimum relative humidity via evaporation Indirect effect of minimum relative humidity via radiation Indirect effect of minimum relative humidity via maximum temperature
–1.15 –0.62 –0.31 –0.39
A. leptocarpa Direct effect of maximum temperature Indirect effect of maximum temperature via evaporation Indirect effect of maximum temperature via minimum temperature Indirect effect of maximum temperature via minimum relative humidity
–2.37 –0.33 –0.29 –2.10
A. auriculiformis Direct effect of maximum temperature Indirect effect maximum temperature via evaporation Indirect effect maximum temperature via minimum relative humidity
–1.86 –0.38 –1.70
Natural fallow Direct effect of minimum relative humidity Indirect effect of minimum relative humidity via maximum temperature
–1.55 –0.92
* The figures in the table represent comparatively high numerical values among all the nine variables considered when the mathematical signs are ignored. This implies that these variables are the main ones influencing the variation in the litterfall component while the others can be ignored without losing much information.
245 Table 4. Pronounced* direct and indirect effects of climatic variables on total litterfall as obtained through PATH analysis of the 1992/1993 and 1996/1997 data from Ibadan, southwestern Nigeria. Effects
Estimation of effect size
S. siamea Direct effect of minimum relative humidity Indirect effect of minimum relative humidity via maximum temperature
–1.81 –0.96
L. leucocephala Direct effect of maximum temperature Indirect effect of maximum temperature via evaporation Indirect effect of maximum temperature via minimum relative humidity
–1.04 –0.45 –0.34
A. leptocarpa Direct effect of minimum relative humidity Indirect effect of minimum relative humidity via evaporation Indirect effect of minimum relative humidity via maximum temperature
–2.33 –0.39 –1.83
A. auriculiformis Direct effect of maximum temperature Indirect effect of maximum temperature via evaporation Indirect effect of maximum temperature via minimum relative humidity
–2.00 –0.42 –1.76
Natural fallow Direct effect of minimum relative humidity Indirect effect of minimum relative humidity via maximum temperature
–1.64 –0.99
* See the footnote for Table 3.
82% for A. leptocarpa and 215% for A. auriculiformis. The biomass of leaves/twigs was low for L. leucocephala in 1993 and 1995 (Figure 1). In 1995, the biomass for these components of trees ranged from 3.5 t ha–1 (L. leucocephala) to 19.7 t ha–1 (A. auriculiformis) whereas the biomass of P. phaseoloides was 6.3 t ha–1. Natural fallow had a biomass of 9.4 t ha–1 in 1995. The leaves/twigs biomass of A. auriculiformis was significantly higher than the leaves/twigs biomass of L. leuco-
Biomass production Stand density of trees varied among species as the trees matured (Table 5). The overall mean stand density of 4505 ha–1 cleared in 1993 was significantly higher than that of 2897 ha–1 cleared in 1995 (LSD0.05 = 362). Senna siamea was the tallest tree in 1993 and A. leptocarpa in 1995 (Table 5). The data show that the increments in diameter at 1.3 m above ground level from 1992 to 1996 were 71% for S. siamea, 106% for L. leucocephala,
Table 5. Growth characteristics after clearing* the four-year (1993) and six-year old (1995) trees on the degraded Alfisol in Ibadan, southwestern Nigeria. Fallow species
S. siamea L. leucocephala A. leptocarpa A. auriculiformis LSD0.05
Stand density (tree ha–1)
Height (m)
1993
1995
1993
1995
1992
1993
1996
4479 6537 3855 3151 1246
2786 4245 1823 2734 1146
9.2 7.0 7.5 4.7 1.5
11.7 10.0 14.1 12.5 –2.2
7.6 5.1 7.0 4.1 1.0
9.1 5.9 9.6 6.5 0.8
13.1 10.5 12.7 13.1 –1.9
Diameter at breast height (cm)
* The 1993 data were from Kang et al. (1997). Also, A. auriculiformis was a year younger than other trees while the 1996 data were included for diameter at breast height.
246
LSD bars (0.05) are shown while NS indicates non-significance in differences and ND indicates means which were not determined. PP = P. phaseoloides; SS = S. siamea; LL = L. leucocephala; AL = A. leptocarpa; AA = A. auriculifomis; and NF = natural fallow. Figure 1. Dry matter of leaves/twigs, wood and undergrowth of planted fallows in 1993 and 1995 on a degraded Alfisol in Ibadan, southwestern Nigeria.
cephala, and the biomass of P. phaseoloides, and natural fallow which were all ≤ 9.4 t ha–1. However, when the vegetation was grouped by type, there was no significant difference in overall mean biomass (11.6 t ha–1) of tree leaves/twigs, herbaceous type (P. phaseoloides) and natural fallow. Senna siamea and A. leptocarpa had significantly higher wood biomass than A. auriculiformis
in 1993. By 1995, all the trees had similar wood biomass. As expected, the trees had significantly higher biomass (leaves/twigs + wood) than P. phaseoloides and natural regrowth (LSD0.05 = 68). The biomass of undergrowth was similar for L. leucocephala and A. leptocarpa, and for S. siamea, A. leptocarpa, and A. auriculiformis (Figure 1). Overall, the wood accounted for more than 72% of the dry aboveground biomass.
247 Nutrient concentrations The N concentration in leaves/twigs was significantly higher for L. leucocephala than for A. leptocarpa in 1993 and 1995 (Table 6). The overall means for 1993 and 1995 were similar. Nutrient concentrations of wood (Table 7) decreased from1993 to 1995 for all species except for K and Mg in A. leptocarpa with an increase of 89%. In 1993, wood N concentrations were
similar among the species. However, in 1995, A. auriculiformis showed a higher level of wood than S. siamea and had a levelve similar to that of L. leucocephala and A. leptocarpa. The mean wood N concentrations for 1993 and 1995 were significantly different. Phosphorus concentrations in leaves/twigs were similar among the tree species in 1993 (Table 6). In 1995, L. leucocephala had a significantly higher P concentration than the two Acacias but similar
Table 6. Nutrient concentrations (%) in leaves/twigs of trees, P. phaseoloides and natural fallow* which were planted on a degraded Alfisol in Ibadan, southwestern Nigeria. Parameter
Year
Tree species*
Non-tree
S. siamea
L. leucocephala
A. leptocarpa
A. auriculiformis
P. phaseoloides
Natural fallow
LSD0.05 species
Year
N
1993 1995
2.07 2.06
2.22 2.46
1.82 1.75
2.06 2.04
1.62 1.87
1.03 2.18
0.35 0.31
NS
P
1993 1995
0.08 0.06
0.06 0.07
0.06 0.05
0.06 0.04
0.04 0.06
0.04 0.10
NS 0.02
NS
K
1993 1995
1.03 1.14
1.32 1.57
0.79 0.92
1.02 1.40
0.56 1.92
0.51 1.77
NS NS
NS
Ca
1993 1995
1.61 1.91
1.31 2.59
1.40 1.28
1.43 1.41
1.27 1.66
1.29 1.85
NS 0.55
0.32
Mg
1993 1995
0.32 0.38
0.35 ND
0.25 0.33
0.23 0.40
0.20 0.56
0.04 0.58
0.09 NS
ND
* ANOVA was carried out for the leaves/twigs of trees only. NS = No significant differences among the means. ND = Not determined. Table 7. Nutrient concentrations in wood of trees which were planted on a degraded Alfisol in Ibadan, southwestern Nigeria. Parameter
Year
Tree species S. siamea
L. leucocephala
A. leptocarpa
A. auriculiformis
LSD0.05 species
Year
N
1993 1995
0.65 0.24
0.51 0.43
0.66 0.51
0.71 0.51
NS 0.25
0.14
P
1993 1995
0.03 0.01
0.04 0.02
0.02 0.02
0.03 0.01
NS NS
0.01
K
1993 1995
0.77 0.42
0.62 0.26
0.22 0.42
0.35 0.16
0.39 NS
NS
Ca
1993 1995
0.98 0.45
0.93 0.40
1.07 0.60
1.10 0.53
NS NS
0.33
Mg
1993 1995
0.15 0.04
0.18 0.12
0.06 0.10
0.10 0.06
NS NS
NS
NS = No significant differences among the means. ND = Not determined.
248 icantly higher in the leaves/twigs than in wood (Tables 6 and 7). In 1993 the leaves/twigs had three times more N concentration than the wood, and about twice as much P, K, and Mg. Calcium concentration in leaves/twigs (1.4%) was close to that in wood (1.02%). In 1995, it was observed that the leaves/twigs had six times more N and Mg concentrations, and four times more P, K and Ca concentrations than the wood.
to that of S. siamea. Nonetheless, S. siamea and the two Acacias had similar P concentrations in 1995. The overall means of P concentrations were similar between 1993 and 1995. Phosphorus concentration in wood decreased by 6% (A. leptocarpa) to 72% (S. siamea) from 1993 to 1995 (Table 7). The concentrations of K in leaves/twigs were similar among species or between years (Table 6). In 1993, the K concentration in S. siamea wood (Table 7) was significantly higher than A. leptocarpa and A. auriculiformis and similar to that in L. leucocephala. Calcium concentrations in the leaves/twigs (Table 6) of the species were similar in 1993. In 1995, L. leucocephala had a significantly higher Ca concentration in its leaves/twigs than other species and S. siamea had a significantly higher Ca concentration in the leaves/twigs than A. leptocarpa. The wood Ca concentrations (Table 7) decreased by 44% (A. leptocarpa) to 57% (L. leucocephala) from 1993 to 1995. In 1993, Mg concentration was significantly higher in the leaves/twigs of L. leucocephala than for A. leptocarpa and A. auriculiformis (Table 6). The concentrations in leaves/twigs were similar among the species in 1995. The wood Mg concentrations (Table 7) decreased by 38% to 71% for S. siamea, L. leucocephala, and A. auriculiformis but increased by 89% for A. leptocarpa. The concentrations of all nutrients were signif-
Effects of fallow species on soil chemical characteristics The chemical properties of surface soil before clearing/burning were improved with six-year fallow (Table 8). As compared with the results of continuous cropping, S. siamea and natural fallow significantly increased soil pH, A. auriculiformis increased soil extractable P, and L. leucocephala and natural fallow increased exchangeable Mg. Burning of the cleared biomass caused a significant increase in soil pH, organic carbon, available P, and exchangeable cations (Table 8). Thre was a significant improvement in soil organic carbon due to fallow after burning.
Discussion Litter production at the site was considered high (Table 2) compared with reported values for other
Table 8. Surface soil (0–15 cm depth) characteristics before and after burning biomass after a six-year fallow on a degraded Alfisol in Ibadan, southwestern Nigeria. Treatment
pH (1:1 H2O)
Org. C (g kg–1)
Extr. P (mg kg–1)
Exchangeable cations [cmol (+) kg–1] K
S. siamea L. leucocephala A. leptocarpa A. auriculiformis P. phaseoloides Natural fallow Continuous cropping LSD0.05
Ca
Mg
BB
AB
BB
AB
BB
AB
BB
AB
BB
AB
BB
AB
6.0 5.5 5.1 5.5 5.1 5.9 5.0 0.8
7.1 6.4 6.4 6.7 6.1 6.2 5.0 0.9
13.7 13.2 12.5 12.6 14.0 14.3 10.6 NS
16.9 19.0 15.7 15.7 17.9 15.6 10.6 6.5
13.9 –8.6 15.5 19.4 –6.9 –5.6 –6.1 10.3
42.1 24.8 23.6 40.1 23.4 29.4 –6.1 19.2
0.35 0.33 0.30 0.22 0.36 0.54 0.22 NS
2.8 1.2 1.3 1.8 0.9 1.8 0.2 2.0
3.7 3.3 3.4 2.3 3.0 3.6 2.8 NS
10.1 –6.6 10.7 10.4 –8.2 –7.5 –2.8 –5.7
1.0 1.2 0.9 0.8 0.9 1.2 0.7 0.4
2.5 2.3 2.0 2.5 2.1 3.1 0.7 1.5
NS = No significant differences among the means. BB = Before burning. AB = After burning.
249 types of vegetation in southwestern Nigeria (Nwoboshi, 1981; Vitousek, 1984; Vogt et al., 1986). A greater production of litter in plantations with high tree-stand densities can be obtained. Litter production at the site would increase for planted tree species as fallow length increased up to eight years. There may be a need to ascertain the fallow length required for this production to stabilize but the observations in this study, particularly with regard to a degree of selfthinning and/or self-pruning, suggest that this level was being approached. This is based on the assumption that the soil was being transformed from a degraded state to an undegraded state with time. Invariably organic-matter content of the soil would increase with time (Kang et al. 1997). Nwoboshi (1981) reported that litterfall tended to increase with age for exotic trees in Nigeria. The high proportion of leaffall in total litterfall is widely reported for different vegetation ecosystems (John 1973; Lowman 1988; Nwoboshi 1981; Kang et al. 1997). The observation that the amount of total litterfall collected was not influenced by the proximity of litter traps to trees in 1996/97 (Table 2) is consistent with the finding by Lowman (1988). He reported that the traps nearest to tree trunks in Australian rainforests did not accumulate higher litter weights from the dense canopy above, and litter traps situated on the different sides (north, south, east and west) of Dendrocnide excelsa collected similar amounts of litter. In the present study, it was established that the climatic variables which influenced litterfall were maximum temperature and minimum humidity (Tables 3 and 4). Other climatic variables which were relevant in enhancing these variables were evaporation, wind, radiation, and minimum temperature. Variables considered but which were not shown by PATH analysis to have pronounced effects on litterfall, were cumulative days of litter collection (representing a gradual shift from period to period) and rainfall. Therefore, this study showed that the change in litterfall on a seasonal basis was not gradual but an erratic process with the dominating climatic variables asserting their role more in the dry season than in the rainy season. It was not the shift from one season to the other that made the difference in quantity of litter produced, but the interactions of relevant climatic
variables within seasons, which also had differing effects depending on the phenological characteristics of the fallow species (Tables 3 and 4). The positive and negative signs ascribed to different climatic effects showed that there were counteracting effects of these variables for the enhancement of litterfall. Agbim (1987) found a significant negative correlation between Chromolaena odorata litterfall and atmospheric relative humidity in a study conducted in the dry season in southeastern Nigeria. Vogt et al. (1986) found that climatic variables (latitude, temperature, and precipitation) could explain 50% of the variation in aboveground litterfall for different broadleaved forests in the world, but they were poorly related to data analyzed for needleleaved forests. Vitousek (1984) also reported that the interaction of temperature and precipitation was significant for litter production in the tropics. From the data (Tables 3 and 4), it was only the natural fallow that followed the stereotypic observation of high litterfall in the dry season and low litterfall in the wet season. Nwoboshi (1981) observed that rapid shedding of litter by teak in southwestern Nigeria in the dry season was a mechanism for combating soil-water stress. The differences in the seasonal litterfall pattern between the planted and natural fallows suggest that the introduction of these planted fallow species for soil rehabilitation will require that the native soil fauna to adjust to the different litterfall patterns exhibited by the planted species (Adejuyigbe et al. 1999; Tian et al. 2000). Therefore, it can be deduced that the differences exhibited would alter the rates of litter decomposition at the site. Judging from the leaf components, litter from S. siamea and A. auriculiformis seemed to require similar climatic conditions to trigger litter production. Also, the environmental triggers of litterfall were similar for L. leucocephala and A. leptocarpa. Stand density and diameter at breast height (dbh or 1.3 m above ground level) suggest that after land clearing, the stumps of the trees were specifically occupying 11% (A. auriculiformis) to 30 % (S. siamea) of the plots in 1993, and 24% (A. leptocarpa) to 39% (S. siamea) in 1995 (Table 5). Thus, the least cultivable area was 70% of the cleared area in 1993, and 61% in 1995. Acacia
250 auriculiformis among the tree species grew most quickly as shown by the dbh increments (Table 5), and the comparison of heights and biomass (Table 5 and Figure 1) with those of other species which were a year older. Also, the N2-fixing trees (L. leucocephala, A. leptocarpa, and A. auriculiformis) had higher dbh increments than the non-N2-fixing S. siamea. The growth of the trees was non-linear. Senna siamea had the highest aboveground biomass (Figure 1) despite its comparatively slow growth rate (Table 5). The wood component constituted 72 to 95% of the total tree biomass. Therefore, a very substantial part of the tree biomass was useful as fuel-wood or construction materials, or wood materials for commercial purposes. The stand density of the trees (Table 5) and the biomass of the leaves/twigs (Figure 1) indicate that high leaves/twigs biomass was associated with comparatively low stand density and vice versa. On Alfisols and Ultisols in southern Nigeria, the mean heights of five- to 16-year old S. siamea ranged from 10 to 18 m while their mean breast height diameters ranged from 9 to 20 cm (Akinnifesi and Salako 1997). The data suggest that undergrowth presence would be minimal with high tree biomass (Figure 1). However, other factors such as canopy architecture, and soil and environmental conditions would also influence undergrowth. The possible influence of canopy architecture on undergrowth development is reflected by the analysis of variance with undergrowth biomass data vis-à-vis the biomass of leaves/twigs for individual trees (Figure 1). For instance, S. siamea had a leaf/twig biomass statistically similar to L. leucocephala, but it had a significantly lower undergrowth biomass. Implications of the presence of the undergrowth are diverse, e.g., weed control in agroforestry (MacDicken et al. 1997), and soil erosion under canopies (Ruangpanit 1985; Akinnifesi and Salako 1997). The litterfall from the fallow species has resulted in an improvement in the quality of the degraded soil, depending on the parameters. High levels of microarthropod populations were observed by Adejuyigbe et al. (1999) and of earthworms by Tian et al. (2000), largely due to the contribution of litterfall to the improvement of microclimate and substrates. The little change in
soil physical properties as reported by Kang et al. (1997) was probably due to the high decomposition rate of the litterfall in the tropics (Agbim 1987). As the level of soil improvement by tree fallow was to that by P. phaseoloides (a herbaceous legume) and natural fallows, the use of the trees for soil rehabilitation would be justifiable because of the additional advantage in wood production.
Conclusions Total litter production by S. siamea, L. leucococephala, A. leptocarpa, and A. auriculiformis was similar (10–12 t ha–1 y–1) when the trees were seven to eight years old in southwestern Nigeria, and this was slightly less than that from natural fallow (13.6 t ha–1 y–1). Leaffall constituted 73–96% of total litterfall. Overall, litter production was higher in the dry season than in the wet season, though minor components (podfall and woodfall) from planted trees may not follow this trend. Peaks for podfall were observed both in the dry season and at the onset of rains, while woodfall peaks were observed in both dry and wet seasons. The effects of maximum temperature and minimum relative humidity on litterfall were pronounced. Senna siamea, L. leucocephala, A. leptocarpa, A. auriculiformis (leguminous trees) and P. phaseoloides (herbaceous legume) adapted well as planted fallows on a degraded Alfisol in southwestern Nigeria. The trunks of the trees were developed well enough after four years of establishment for economic benefits to be derived through timber production from them, but better benefits would accrue if the trees were left to grow for up to seven years.
Acknowledgements The authors gratefully acknowledge the contributions of Messrs. S. Ojo, Yekini Olaide and Bayo Oke of the Resource and Crop Management Division (RCMD), IITA, Ibadan, Nigeria to data collection. The Crop Modeling Unit collected the climatic data while the PATH analysis program was written by the Biometrics Unit. IITA manuscript number: IITA/99/JA/25.
251 References Adejuyigbe CO, Tian G and Adeoye GO (1999) Soil microarthropod populations under natural and planted fallows in southwestern Nigeria. Agrofor Syst 47: 263–272 Agbim NN (1987) Carbon cycling under Chromolaena odorata (L.) K. & R canopy. Biol Agric Hort 4: 203–212 Akinnifesi FK and Salako FK (1997) Growth of Gmelina arborea and Senna siamea and their effects on microclimate in southern Nigeria. Trop Ecol 38: 55–63 Analytical Software (1998) Statistix for Windows. Version 2. Analytical Software, Florida, USA Buresh RJ and Tian G (1998) Soil improvement by trees in sub-Saharan Africa. Agrofor Syst 28: 51–76 Giller KE (1998) Tropical legumes: providers and plunderers of nitrogen. In: Bergström L and Kirchmann H (eds) Carbon and Nutrient Dynamics in Natural and Tropical Ecosystems, pp 33–45. CAB International, Wallingford, Oxon, UK Hinkelmann K and Kempthorne O (1994) Design and Analysis of Experiments Volume I: Introduction to Experimental Design. John Wiley & Sons, New York, USA Hulugalle NR (1992) Amelioration of highly degraded tropical Alfisol. I. Changes in soil physical and chemical properties 1989–91. Land Degrad Rehab 3: 141–152 IITA (International Institute of Tropical Agriculture) (1979) Selected methods for soil and plant analysis. Manual Series No. 1. IITA, Ibadan, Nigeria John DM (1973) Accumulation and decay of litter and net production of forest in tropical West Africa. OIKOS 24: 430–435. Johnson MS and Bradshaw AD (1979) Ecological principles for the restoration of disturbed and degraded land. Appl Biol IV: 141–200 Kang BT, Salako FK, Akobundu IO, Pleysier JL and Chianu JN (1997) Amelioration of a degraded Oxic Paleustalf by leguminous and natural fallows. Soil Use Manage 13: 130–136 Lal R (1992) Tropical agricultural hydrology and sustainability of agricultural systems. The West Bank Project (1978–1988). Ohio State Univ, Ohio, USA/IITA, Ibadan, Nigeria
Lee R (1978) Forest microclimatology. Columbia University, New York, USA Li CC (1975) Path analysis – a primer. The Boxwood Press, Pacific Grove, California, USA Lowman MD (1988) Litterfall and leaf decay in three Australian rainforest formations. J Ecol 76: 451–465 Lugo AE and Brown S (1993) Management of tropical soils as sinks or sources of atmospheric carbon. Plant Soil 149: 27–41 MacDicken KG, Hairiah K, Otsama A, Duguma B and Majid NM (1997) Shade-based control of Imperata cylindrica: tree fallows and cover crops. Agrofor Syst 36: 131–149 Moormann FR, Lal R and Juo ASR (1975) The soils of IITA. IITA Technical Bulletin No. 3. IITA, Ibadan, Nigeria Nwoboshi LC (1981) Nutrient cycling in a managed teak plantation ecosystem. II. Litterfall and macronutrient return to the forest floor. Niger J Agric Sci 3: 9–14 Rosenberg NJ, Blad BL and Verma SB (1983) Microclimate: the biological environment 2nd edition. John Wiley and Sons, New York, USA Ruangpanit N (1985) Percent crown cover related to water and soil losses in mountainous forest in Thailand. In: El-Swaify SA, Moldenhauer WC and Lo A (eds) Soil Erosion and Conservation, pp 462–471. SCSA, Iowa, USA SAS Institute Inc. (1990) SAS/STAT User’s Guide, Version 6, 4th edition, Volume 1. SAS Institute Inc, NC, USA Tian G, Olimah OA, Adeoye GO and Kang BT (2000) Regeneration of earthworm population in a degraded soil by natural and planted fallows under humid tropical conditions. Soil Sci Soc Am J 64: 222–228 Vitousek PM (1984) Litterfall, nutrient cycling, and nutrient limitation in tropical forests. Ecol 65: 285–298 Vogt KA, Grier CC and Vogt DJ (1986) Production, turnover, and nutrient dynamics of above- and belowground detritus of world forests. Adv Ecol Res 15: 303–377 Williams LG and Tolome J (1996) Litterfall, temperature and tropical dominant trees, and climate in a Mexican lower montane forest. Biotropica 28: 649–656 Williams RJ, Myers BA, Muller WJ, Duff GA and Eamus D (1997) Leaf phenology of woody species in a north Australian tropical savanna. Ecol 78: 2542–2558
