ABOVEGROUND BIOMASS PRODUCTION AND NITROGEN CONTENT IN Gliricidia sepium (JACQ.) WALP. UNDER SEVERAL PRUNING REGIMES

Isidro Melchor Marroquín, Jesús Vargas Hernández, Alejandro Velázquez Martínez and Jorge Etchevers Barra

Isidro Melchor Marroquín. Agronomist Engineer, Universidad Autónoma Chapingo (UACh), Mexico. M.Sc. and D.Sc. Colegio de Postgraduados, Mexico. Investigador, Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias, Mexico. e-mail: marroquin59@hotmail.com

Jesús Vargas Hernández. Agronomist Engineer, UACh. M.Sc., Colegio de Postgraduados, Mexico. Ph.D., Oregon State University, USA. Professor Researcher, Colegio de Postgraduados, Montecillo, Mexico. e-amail: vargashj@colpos.mx

Alejandro Velázquez Martínez. Agronomist Engineer, UACh. M.Sc., Colegio de Postgraduados, Mexico. Ph.D. Oregon State University. USA. Professor Researcher, Colegio de Postgraduados, Montecillo, Mexico. Address: Km 36.5, Carretera México-Texcoco, Km 36.5. México 56230 e-mail: alejvela@colpos.mx

Jorge Etchevers Barra. Agronomist Engineer, Universidad de Concepción, Chile. Ph.D., North Dakota State University, USA. Professor Researcher, Colegio de Postgraduados, Montecillo, Mexico. e-mail: jetchev@colpos.mx

Resumen

Se estudió el potencial de producción de materia orgánica y la incorporación total de nitrógeno al suelo de plantas de Cocoite (Gliricidia sepium) (Jacq.) Walp. en diferentes regímenes de poda. El estudio se realizó en una plantación de 6 meses de edad (6670 plantas/ha) bajo un arreglo factorial 4×2 en un diseño de parcelas divididas con cuatro repeticiones. La frecuencia de las podas (4, 8, 12 y 24 semanas) se asignó a las parcelas principales, mientras que la altura de poda (0,5 and 1,0m) se asignó a las subparcelas. La altura de la poda no afectó la producción de materia orgánica ni su calidad; sin embargo, las variables bajo estudio fueron significativamente modificadas por la frecuencia de la poda. La producción de materia seca aumentó desde 0,50 hasta 10,52ton·ha^-1 para los regímenes de 4 y 24 semanas, respectivamente. La proporción de la biomasa de los tallos aumentó desde 20 hasta 53% en el régimen de 4 y 24 semanas. La concentración total de nitrógeno en la biomasa aérea disminuyó de 3,19 a 2,64% al reducirse la frecuencia de podas de 4 a 24 semanas, pero el nitrógeno total acumulado fue superior en el régimen de 24 semanas (272kg·ha^-1) que en otras frecuencias de poda debido a la mayor producción de biomasa en aquel. Sin embargo, en las muestras de suelo colectadas dos meses después de la última poda en ningún régimen se observó incremento en la materia orgánica o nitrógeno del suelo. Para detectar cambios en la materia orgánica y nitrógeno del suelo, se requiere de mayores períodos de tiempo que permitan la descomposición de la biomasa.

Summary

The aboveground dry matter (ADM) production and nitrogen accumulation in Gliricidia sepium (Jacq.) Walp. under several pruning regimes was studied to evaluate the potential for organic matter production and total nitrogen incorporation into the soil. The study was carried out in a 6-months-old plantation (6670 plants/ha) established under a 4×2 factorial design in a split-plot design with four replications. The pruning frequency (4, 8, 12 and 24 weeks) was assigned to main plots, whereas the pruning height (0.5 and 1.0m) was assigned to subplots. Pruning height did not affect ADM production or its quality; however, all variables were significantly modified by the pruning frequency. ADM increased from 0.50 up to 10.52ton·ha^-1 in the 4- and the 24-weeks regime, respectively. The proportion of stem biomass increased from 20 up to 53% in the 4- and the 24-weeks regime. Overall nitrogen concentration (NC) in ADM diminished from 3.19 to 2.64% upon reducing the pruning frequency from 4 to 24 weeks, but total nitrogen accumulated was much higher in the 24-weeks regime (272kg·ha^-1) than in the other pruning frequencies, due to the larger biomass production. However, in soil samples taken two months after the last pruning, no significant increase in soil organic matter (SOM) or soil nitrogen (SN) was observed in any of the pruning regimes. In order to detect changes in SOM and SN, a longer evaluation period might be required to allow for biomass decomposition.

Resumo

Estudou-se o potencial de produção de matéria orgânica e a incorporação total de nitrogênio ao solo de plantas de Gliricidia (Gliricidia sepium) (Jacq.) Walp. em diferentes regimes de poda. O estudo se realizou em uma plantação de 6 meses de idade (6.670 plantas/ha) sob um arranjo fatorial 4×2 em um desenho de parcelas divididas com quatro repetições. A freqüência des podas (4, 8, 12 e 24 semanas) foi designada para as parcelas principais, enquanto que a altura de poda (0,5 e 1,0m) foi designada para as sub parcelas. A altura da poda não afetou a produção de matéria orgânica nem sua qualidade; no entanto, as variáveis sob estudo foram significativamente modificadas pela freqüência da poda. A produção de matéria seca aumentou de 0,50 até 10,52 ton·ha^-1 para os regimes de 4 e 24 semanas, respectivamente. A proporção da biomassa dos caules aumentou de 20 até 53% no regime de 4 e 24 semanas. A concentração total de nitrogênio na biomassa aérea diminuiu de 3,19 a 2,64% ao reduzir-se a freqüência de podas de 4 a 24 semanas, mas o nitrogênio total acumulado foi superior no regime de 24 semanas (272kg·ha^-1) que em outras freqüências de poda devido à maior produção de biomassa naquele. No entanto, nas amostras de solo recolhidas dois meses depois da última poda, não foram observados, em nenhum regime, incremento na matéria orgânica ou nitrogênio do solo. Para detectar mudanças na matéria orgânica e nitrogênio do solo, se requer de maiores períodos de tempo que permitam a decomposição da biomassa.

Keywords / Pruning Frequency / Pruning Height / Biomass Distribution / Soil Organic Matter /

Received: 04/15/2004. Modified: 02/09/2005. Accepted: 02/15/2005

Introduction

Gliricidia sepium (Jacq.) Walp. is a leguminous tree native from Mexico to Central America, commonly used in several tropical regions of the world. G. sepium is a fast-growing species, easy to propagate, with broad plasticity for adapting to diverse environmental conditions, so it is highly appreciated by farmers and widely used in living fences, as shade tree for other crops, or for forage and fuelwood production. Moreover, its coppicing ability, high rate of aboveground biomass production, and biological nitrogen-fixing capability, make this multi-purpose tree species a potential source to improve soil fertility in alley cropping systems (Kang et al., 1984; Sanginga et al., 1994).

In Mexico, research on G. sepium has focused on the productivity and survival of living fences (Llera et al., 1998), on the genetic and geographic variation of seed sources growing under adverse soil moisture conditions (Marín, 1996; Román et al., 1996; García y Vargas, 2000) and on the selection of native Rhizobium strains for high effectiveness of seedling inoculation (Melchor et al., 1999). In a previous study carried out in an alley cropping system in Veracruz, México, Melchor (1992) found that by pruning G. sepium twice a year increased maize production 50% in relation to the average yield obtained by the traditional cropping system (600kg·ha^-1). This increase was attributed to improved soil fertility and water retention by the aboveground biomass incorporated into the soil. Even though this fact highlights the potential of G. sepium to increase yield of agricultural crops under agroforestry systems in this tropical region, no attempt has been made to evaluate the effects of pruning on biomass production and potential incorporation into the soil in this ecological region.

In Niger, Sanginga et al. (1991) found that with an adequate pruning regime G. sepium produces 10-20ton dry matter·ha^-1·year^-1 and up to 130kg nitrogen·ha^-1·year^-1. Incorporation of the pruned biomass to the soil has been shown to improve its physical and chemical properties as well as its biological activity and productivity (Kang, 1994). In an alley cropping system with G. sepium, Lal (1989) found that both carbon and nitrogen levels in the soil were higher than in plots without this species. Similarly, Akonde et al. (1986) observed a 35 to 50% increase in maize yield in Cameroon, attributing this to both organic matter and N incorporated into the soil.

However, several studies have shown that biomass production in N₂-fixing leguminous trees might be influenced by the frequency and height of pruning (Duguma et al., 1988; Sanginga et al., 1994). Duguma et al. (1988), in a G. sepium trial established in Niger, found that harvested aboveground biomass increased from 7.6 to 9.2ton dry matter·ha^-1 as the height of pruning increased from 0.5 to 1.0m, and from 0.5 to 5.4ton dry matter·ha^-1 when the pruning frequency decreased from 4 to 24 weeks. Besides, the effect of pruning regimes on biomass production is expected to vary from site to site. On the other hand, most studies have focused primarily on the effect of pruning regimes on the total biomass production, but there is not enough information about the impact of this management practice on quality of biomass harvested (proportion of foliage to stem tissues, and N concentration).

In other fast-growing tree species it is common to observe a strong age-trend in the aboveground biomass allocation pattern, favoring stems as trees get older or pruning frequency decreases (Dickson, 1989; Haddad et al., 1995; Tschaplinski and Blake, 1995). If this trend exists in G. sepium, reducing the frequency of pruning to increase biomass production might also increase the proportion of mature foliage, stems, or woody biomass, and reduce the N concentration of the foliage fraction harvested. These effects, in turn, could affect negatively the total amount of N and its rate of incorporation into the soil, since mature foliage and woody biomass has lower decomposition rates than young, less-lignified, foliage (Handayanto et al., 1994; Lehman et al., 1995). The object of this study was to evaluate the production and quality (stem proportion and N concentration) of aboveground biomass in G. sepium under several pruning regimes, as well as the potential effect of incorporating this biomass on soil fertility. This information will help to develop more efficient agroforestry systems and to design appropriate pruning regimes for alley cropping systems in Mexico involving G. sepium.

Materials and Methods

Establishment of field trial

The study was carried out at the Ixtacuaco Agroforestry Station (20º04'N and 97º04'W, 152masl) located in Martínez de la Torre County, in northern Veracruz, Mexico. The climate is tropical humid with an annual mean temperature and precipitation of 24.5ºC and 1778mm, respectively. About 45.8% of rainfall is distributed during the summer, from June to October. The experimental site has a slope of 10%; the soil is a 40-50cm depth vertisol with a pH between 6.5 and 7.0. Average organic matter and total nitrogen content are 2.26 and 0.11%, respectively. The soil commonly cracks during the dry season because of heavy clay content. Seeds of G. sepium from a local natural population were sown in polyethylene bags (20×25cm) filled with local soil and sand (3:1 vol) previously disinfected with Captan (5g·l^-1). Before planting, seedlings were maintained in the nursery during 4 months to ensure its quality. The plantation was established on Feb 5, 1996 at a spacing of 1.5 by 1.0m (6670plants/ha). Plants were left free to grow for six months and then pruned to the specified height before starting the various pruning regimes. During this period weeds were manually controlled to avoid competition.

Pruning treatments

A 4×2 factorial design in split plots with four replications was used. Pruning frequency, with four levels (4, 8, 12 and 24 weeks), was assigned to the main plots (48m² useful harvested area) and pruning height, with two levels (0.5 and 1.0m stem height from the ground), was assigned to the subplots (24m² useful area). Each subplot was completely surrounded by one row of border trees, which received the same pruning treatment as the experimental unit (i.e., there were two border rows between adjacent plots). After six months of free growth a homogenization pruning at the established heights (0.5 and 1.0m) was carried out to remove all foliage and branches, leaving only one main stem. Starting from that date, the pruning regimes were applied systematically in each of the experimental units during a 24-weeks cycle (i.e. six prunings every 4-weeks, three prunings every 8-weeks, two prunings every 12-weeks, and one pruning at 24-weeks). The 24-weeks regime was considered as the control treatment.

Aboveground dry matter and nitrogen concentration

In each experimental unit the total fresh weight of foliage and stems was quantified separately at each harvest and one sub-sample (100g, approximately) of each tissue was taken for further analysis. At the laboratory, samples were oven-dried at 72ºC until constant weight. Using the dry/fresh weight ratio of the sample, total dry matter (ton·ha^-1) of foliage and stems was estimated for each treatment plot. Nitrogen concentration (NC, %) in each tissue sample was determined by using the microkjeldahl method (Kalra and Maynard, 1991) and the total N content in the pruned biomass (kg·ha^-1) was then estimated. At each pruning, all the aboveground biomass harvested was incorporated on the top of the soil within its corresponding experimental plot.

Soil organic matter and total soil nitrogen content

Thirty days before applying the first pruning in the 4-weeks regime, and sixty days after the whole pruning cycle ended, a compound soil sample (three randomly selected sampling points, 0-20cm depth) was taken from each experimental plot. At the laboratory, these samples were processed to determine soil organic matter (SOM) by the Walkley and Black method and total soil nitrogen (SN) content by the microkjeldahl method (Kalra and Maynard, 1991). The interest was to determine whether significant changes in SOM and SN can be detected shortly after the pruning, so as to influence productivity of a potential companion crop within the same growing cycle. There are evidences that young and succulent shoots can loose over 50% of their initial mass and release a similar fraction of N within 8 weeks after being applied to the soil (Henriksen et al., 2002).

Statistical analysis

Data was analyzed using the GLM procedure of the Statistical Analysis System (SAS, 1990), according to the statistical model of the experimental design used, to determine the effect of the pruning regime on the variables measured. ADM, NC, and total nitrogen content were analyzed at the end of the pruning cycle as well as for each pruning event separately. Before the analysis, NC data were transformed using the arc sine square-root function (Sabin and Stafford, 1990) and mean values for pruning regimes were back-transformed after the analysis. Variance analysis for SOM and SN was done using the multivariate approach of the repeated-measures data model (Moser et al., 1990; Gumpertz and Brownie, 1993), with sampling date as the repeated factor, to test for differences between initial and final values, as well as among pruning regimes and the interaction of both factors:

Y_ijkl = µ+B_i +P_j +(BP)_ij +H_k +(PH)_jk + e_ijk + T_l +(BT)_il +(PT)_jl *f_ijl +(HT)_kl +(PHT)_jkl +g_ijkl

where µ+B_i +P_j +(BP)_ij +H_k +(PH)_jk + e_ijk: the between-plot part of the model (also used for analysis of ADM, NC and total nitrogen content) with an overall mean (µ), a random block effect (B_i), a fixed pruning frequency effect (P_j), a random main plot error ((BP)_ij), a fixed pruning height effect (H_k), a pruning frequency ´ height interaction effect ((PH)_jk), and a random plot to plot variation (e_ijk), while T_l +(BT)_il +(PT)_jl +f_ijl +(HT)_kl +(PHT)_jkl +g_ijkl: represents the within-plot part of the model, with a fixed time effect (T_l), a random block ´ time effect ((BT)_il), a pruning frequency ´ time interaction effect ((PT)_jl), a random effect for observations on the same main plot (f_ijl), a pruning height x time interaction effect ((HT)_kl), a pruning frequency ´ height ´ time interaction effect ((PHT)_jkl), and a random effect for observations on the same subplot (g_ijkl). Main and subplot factors, as well as time, are considered fixed effects, but block effects are random. Additional assumptions in this model are that within-plot effects are correlated over time, measurements over time within the same block are also correlated, and all random effects are normally distributed.

For those variables where significant differences between treatments were found, the Tukey test (p£0.05), which is slightly more conservative in detecting significant differences than other methods, was used for multiple-means comparisons.

Results and Discussion

Aboveground dry matter production

No significant differences (p>0.20) in total aboveground dry matter (ADM) production or NC were found between the two pruning heights used (data not shown). Previous studies have shown contrasting results in this regard; Duguma et al. (1988) found a significant effect of pruning height on biomass production of several woody perennials, including G. sepium, grown in alley cropping systems, but Erdmann et al. (1993) obtained similar biomass production with pruning heights of 25 and 100cm above the ground. The length of the study might have influenced these results; the study by Duguma et al. (1988) was much longer than the others, so differences could have accumulated over time in that study. Other factors, such as initial age and size of plants, plantation density, pruning season, etc., could also play an important role on the effects of pruning height over biomass production, since all of them would affect the amount of carbohydrate reserves along the stem (Erdmann et al., 1993). However, all variables were affected by the pruning frequency, so only this factor will be discussed further. Total ADM drastically increased upon reducing the pruning frequency. The 24-weeks pruning regime was the most productive (10.52ton·ha^-1), followed by the 12-, 8- and 4-weeks pruning frequencies with 5.05, 3.53 and 0.50ton·ha^-1, respectively (Table I). This trend is not unexpected, since earlier studies in other regions of the world have found similar results (Duguma et al., 1988; 1994), and relative growth rate of plants commonly increases with time after pruning. However, biomass production obtained in our study was generally higher than in other published reports. For instance, ADM produced by the 24-weeks regime was about 30% higher than that obtained by Sanginga et al. (1994) in Niger for a similar pruning frequency but with a much higher plant density (20000 plants·ha^-1). In another study, Erdmann et al. (1993) estimated an average ADM production of 1.9ton·ha^-1 for a 12-weeks regime with a density of 5000 plants·ha^-1. Other studies have shown that ADM production in G. sepium, commonly varies between 6 and 8 ton·ha^-1·year^-1 in plants pruned 2-4 times per year using densities from 15 to 20000 plants·ha^-1 (Atta-Krah and Sumberg, 1987; Ella et al., 1989). These comparisons indicate that in the northern region of Veracruz, Mexico, under similar plantation densities, G. sepium could have a high biomass productivity using the appropriate pruning regime, partly due to a favorable climate, deep soils and the fact that the species is native to this region.

Even though both foliage and stems followed the same trend of increasing biomass as the pruning frequency was reduced, the proportion of these components varied among the pruning regimes. In the 4-weeks regime the stem proportion was only 20% whereas in the 24-weeks regime it increased to 53%. This trend is attributed to the shift in biomass allocation associated with age or time after pruning, favoring accumulation in stems over foliage, as has been found in other woody species (Haddad et al., 1995; Tschaplinski and Blake, 1995).

For all the pruning frequencies, ADM harvested increased with each consecutive cut; thus, in the 4-weeks regime ADM increased about eight times from the first (0.017ton·ha^-1) to the last (0.13ton·ha^-1) pruning (Figure 1a). Likewise, in the 8- and 12-weeks regimes ADM increased seven-fold from the first (0.34 and 0.52ton·ha^-1 respectively), to the last (2.49 and 4.51ton·ha^-1, respectively) pruning (Figures 1b and 1c). Erdmann et al. (1993) observed a similar dynamics in biomass production of G. sepium in a field trial established in Niger when a 3-weeks regime was used; in that study ADM increased between three to seven times during three consecutive prunings. However, when they used a 6-week pruning regime, ADM production in the first pruning was about seven times higher than in the last one; they attributed this reverse trend to the drought that occurred during the second half of the evaluation period. Granados (1998) in Tabasco, Mexico, also found that stem and foliage biomass production during the rainy season was much higher than during the dry season. These results confirm that the dynamics of ADM production does not necessarily follow a definite pattern, but rather depends on the environmental conditions along the evaluation period, as well as on age, plant size and growing rhythm of the ecotype used (Wiersum and Dirdjosoemarto, 1987; Kwesiga, 1994; Scroth and Zech, 1995).

Nitrogen concentration

Quality of foliage and stem biomass (nitrogen concentration, NC%) decreased with the gradual reduction in pruning frequency, but was only significantly different for the 4-weeks regime (Table II). In the last pruning, overall average NC in the 4-weeks regime was 3.19%; in the same harvest, overall mean N concentration in the 8- 12- and 24-weeks regimes was 2.70, 2.68 and 2.64%, respectively. Despite the differences in NC between foliage and stems, reduction in NC associated with a larger pruning interval was similar in both biomass components, from 3.97 to 3.30% in foliage and from 2.41 to 1.97% in stems, even though it was only significant for the foliage (Table II).

The higher NC found in young plant tissues (as in the 4-weeks pruning regime) can be explained by the high metabolic and expansion rates that developing leaves commonly have, representing strong sinks for N compounds. Growing organs demand a higher translocation rate of assimilates from mature leaves or other storage zones, as well as mineral nutrients absorbed and metabolized by the root system of the plant (Dickson, 1989; Parsons et al., 1993; Shantharam and Mattoo, 1997). As the pruning frequency decreased, most leaves in the plant reached maturity or even senility. Since metabolic rates normally decrease with age of leaves after maturity is reached, less mineral nutrients, including N, are required. In addition, as leaves senesce, N-derived and other soluble compounds are quickly exported to other actively growing or storage tissues, causing a decrease in N concentration in the foliage (Dalling, 1987; Shantharam and Mattoo, 1997).

No defined trend in NC was observed along the pruning cycles in the 4-weeks regime, but in the 8- and 12-weeks regimes a gradual reduction in NC was found for both foliage and stem as the successive cuts were applied. In these regimes, NC in the last pruning were between 15 and 40% lower than those obtained in the first cut; in the 24-weeks regime, NC was similar to that obtained in the last pruning of the 12-weeks regime (Figure 2). It is not known whether the reduction in NC observed for both components in successive cuttings is an effect of inhibition in N₂ fixation as a result of the stress caused by the sequential prunings, as discussed by Erdmann et al. (1993), or it is just a dilution effect associated with a higher growth rate in the final cycles as compared to the initial pruning. Either way, these results show that the relative distribution of nitrogen in the plant varies with the pruning regime and the pruning cycle.

Total nitrogen

In all the pruning regimes, total nitrogen (TOTN) accumulation was higher in foliage than in stems, due mainly to the higher NC found in foliage (Table II); thus, even though in the 24-weeks regime stem biomass was almost 10% higher than foliage biomass (Table I), TOTN was on average 60% higher in foliage (Table III). In addition, TOTN increased as the pruning frequency was reduced, which can be attributed to the large increase in aboveground biomass obtained, despite the reduction in N concentration associated to these lower-frequency pruning regimes. The highest average TOTN was obtained with the 24-weeks regime (272.1kg·ha^-1), whereas the 4-weeks regime accumulated only 7.2% of that amount. It seems that from the point of view of potential TOTN incorporated into the soil, it is not a wise idea to use high-frequency prunings in this species. However, using an 8- or 12-weeks regime would incorporate above 150kgN·ha^-1 per year under the conditions of this study (Table III). These estimates are within the range of values reported earlier for the same species during a similar evaluation period. Atta-Krah and Sumberg (1987) obtained between 85 and 95kgN·ha^-1, whereas Juo and Kang (1989) estimated an average of 84.5kgN·ha^-1. Duguma et al. (1988) obtained an average accumulation of 131.5kgN·ha^-1 during a six-month period, which is similar to our results for a 12-weeks regime.

In general, TOTN increased with each consecutive pruning. Since TOTN accumulation is associated to the increase in ADM, the trend was more notorious in both the 8- and the 12-weeks regimes than in the 4-weeks regime (Figure 3). In the 8- and 12-weeks regimes, TOTN increased 4-5 times from the first to the last pruning cycle whereas in the 4-weeks regime there was not much difference after the second cycle of pruning, reflecting the fluctuations in biomass production observed in this regime (Figure 1). These results highlight the potential of G. sepium for enhancing crop productivity in alley cropping systems where the observed dynamics in TOTN can be synchronized to the N requirements of agricultural crops along its growing cycle (Lehman et al., 1995), assuming that N is readily available for the companion crop, or allowing for some time advantage in the pruning of G. sepium before sowing the agricultural crop.

Soil organic matter and soil nitrogen concentration

Using the repeated-measures data model, no statistical differences were observed (Table IV) between sampling dates (i.e., final vs. initial) or among pruning regimes in both soil organic matter (SOM) and soil nitrogen (SN). Before the pruning, average SOM and SN varied from 2.11 to 2.54% and from 0.11 to 0.12%, respectively, with no differences between plots assigned to different pruning regimes. Two months after the final pruning, these values varied from 2.02 to 2.70% and from 0.12 to 0.14%, with no significant change (p>0.20) from the initial values, and no significant differences between pruning regimes.

Even though the 24-weeks regime accumulated much higher amounts of TOTN in the biomass than the more frequent pruning regimes, the results found in SOM and SN are not surprising, given that the soil samples were taken only two months after the final pruning cycle. Whereas in the 24-weeks regime all biomass was harvested and incorporated to the ground only two months before the sampling, in all the other regimes some biomass was harvested and incorporated at least five months before the samples were taken. In addition, as the pruning frequency was reduced, the proportion of mature foliage and woody material (stems) increased, requiring a longer period to fully decompose.

In Hawaii, Oglesby and Fownes (1992) found that three months after incorporation of young leaves and twigs of G. sepium to the soil, about 65% of its TOTN had been mineralized. In fact, the rates of decomposition of plant materials depend on both tissue type and chemistry (Oglesby and Fownes, 1992); a fast decomposition rate has normally been linked to low polyphenol content (Palm and Sánchez 1990), low polyphenol/N ratio (Oglesby and Fownes, 1992), low lignin/N ratio (Kachaka et al., 1993), or low (lignin+ polyphenol)/N ratio (Fox et al., 1990), as in young foliage. The slow decomposition rate of mature foliage and stem tissue helps to explain the lack of increase in SOM and SN in all pruning regimes, despite the large differences among them in biomass and TOTN harvested. The differences among these treatments might only become apparent in a longer evaluation period, once the biomass incorporated to the ground has been fully decomposed and mineralized.

Using a 12-weeks pruning regime, Lal (1989) found an increment of 0.60% in SOM only after a year of biomass incorporation. Gavina (1989) found an increase of 0.36% in SOM using a similar pruning regime and evaluation period in Philippines, whereas Duguma and Tonye (1994) found an increase of 0.44% (from 3.5 to 3.94%) in SOM after a 3-years period of yearly prunings of G. sepium in Cameroon. Published information of SN changes attributed to biomass incorporation to the soil are much more variable; for instance, Gavina (1989) found an increase of 0.52% during a one-year evaluation period, which is about 17 times higher than the results of this study. On the other hand, Duguma and Tonye (1994) reported a decrease of 0.03% in SN after a three-year evaluation period of G. sepium in an alley cropping system. While the discrepancies among studies might be mostly attributable to differences in methodology, length of the evaluation period, and particular conditions of the studies, it seems obvious that a more careful interpretation is required to fully evaluate and understand the effects of biomass incorporation on soil fertility in alley cropping systems.

Conclusions

Some benefits from using agroforestry systems are the maintenance of soil organic matter and soil fertility levels (particularly N) when the aboveground biomass of the trees pruned periodically is incorporated to the soil. In our study, this situation was confirmed even though no significant increase in both SOM and SN was found during a 6-month evaluation period. Even though the study did not include an agricultural crop, it was shown that with an 8- or 12-weeks pruning regime of the tree species would incorporate above 150kgN·ha^-1 per year, enough to supply the needs of a maize crop in the region. These results highlight the potential of using G. sepium in alley cropping systems under the particular environmental conditions of the tropical humid regions in Mexico. The pruning regime of G. sepium has to be considered in the management of these agroforestry systems, because the pruning frequency affects the total aboveground biomass production, its composition, and its quality (N concentration). According to the results of this study in terms of the potential nitrogen incorporation to the soil, 8- to 12-weeks pruning regimes might be appropriate in the G. sepium-maize association under the alley cropping technique: However, a longer evaluation period is required to determine whether aboveground biomass production and nitrogen contribution using this pruning regime could be synchronized to the nutrient requirements of the agricultural crop along its phenological stages. In addition, it is required to validate these results under different field conditions, in collaboration with local farmers. In developing optimal pruning regimes it is also required to consider the interaction between trees and crops; the pruning regime not only affects the crop by enhancing soil properties, but also by changing microclimate, particularly light interception, or increasing root competition.

REFERENCES

1. Akonde TPN, Lame N, Kummerer E (1986) Adoption of alley cropping in the Province of Atlantique, Benin. In Kang BT, Reynolds L (Eds.) Alley farming in the humid and subhumid tropics. IDRC. Ottawa, Canada. pp. 141-142.        [ Links ]

2. Atta-Krah AN, Sumberg JE (1987) Studies with G. sepium for crop/livestock production systems in West Africa. In Whithington D, Glover N, Brewbaker JL (Eds.) Gliricidia sepium (Jacq.) Walp.: management and improvement. Special publication 87-01. NFTA. Turrialba, Costa Rica. pp. 31-43.        [ Links ]

3. Dalling MJ (1987) Proteolyc enzymes and leaf senescence. In Thomson WW, Nothangele EA, Huffaker RC (Eds.) Plant senescence: its biochemistry and physiology. American Society of Plant Physiologists. Maryland, USA. pp. 54-69.        [ Links ]

4. Dickson RE (1989) Carbon and nitrogen allocation in trees. Ann. Sci. For. 46(suppl.): 631-647.        [ Links ]

5. Duguma B, Kang BT, Okali DU (1988) Effect of pruning intensities of three woody leguminous species grown in alley cropping with maize and cowpea on an alfisol. Agrofor. Syst. 6: 19-35.        [ Links ]

6. Duguma B, Tonye J (1994) Screening of multipurpose tree and shrub species for agroforestry in the humid lowlands of Cameroon. For. Ecol. Manag. 64: 135-143.        [ Links ]

7. Duguma B, Tonye J, Kanmegne J, Manga T, Enoch T (1994) Growth of ten multipurpose tree species on acid soils in Sangmelina, Cameroon. Agrofor. Syst. 27: 107-119.        [ Links ]

8. Ella A, Jacobsen C, Stur W, Blair G (1989) Effect of plant density and cutting frequency on the productivity of four tree legumes. Tropical Grasslands 1: 28-34.        [ Links ]

9. Erdmann TK, Nair PK, Kang BT (1993) Effects of cutting frequency and cutting height on reserve carbohydrates in G. sepium. For. Ecol. Manag. 57: 45-60.        [ Links ]

10. Fox RH, Myers RJ, Vallis L (1990) The nitrogen mineralization rate of legume residues in soil as influenced by their polyphenol, lignin and nitrogen contents. Plant Soil 129: 251-259.        [ Links ]

11. García FM, Vargas HJ (2000) Growth and biomass allocation of Gliricidia sepium seed sources under drought conditions. J. Sust. For. 10: 45-50.        [ Links ]

12. Gavina LD (1989) Preliminary assessment of nutrient content of soils under Gliricidia sepium. Nitrogen Fixing Tree Research Reports 7: 94-95.        [ Links ]

13. Granados ZL (1998) Crecimiento y valor nutritivo del cocoite (Gliricidia sepium) como criterio para emplearlo como forraje. Tesis. Colegio de Postgraduados. México. 86 pp.        [ Links ]

14. Gumpertz ML, Brownie C (1993) Repeated measures in randomized block and split-plot experiments. Can. J. For. Res. 23: 625-639.        [ Links ]

15. Haddad Y, Clair-Maczulajtys D, Bory G (1995) Effects of curtain-like pruning on distribution and seasonal patterns of carbohydrates reserves in plane (Platanus acerifolia Wild) trees. Tree Physiol. 15: 135-140.        [ Links ]

16. Handayanto E, Cadish G, Giller KE (1994) Nitrogen release from prunings of legume hedgerow trees in relation to quality of the prunings and incubation method. Plant Soil 160: 237-248.        [ Links ]

17. Henriksen I, Michelsen A, Schlönvoigt A (2002) Tree species selection and soil tillage in alley cropping systems with Phaseolus vulgaris L. in a humid premontane climate: biomass production, nutrient cycling and crop responses. Plant Soil 240: 145-159.        [ Links ]

18. Juo ASR, Kang BT (1989) Nutrient effects of modification of shifting cultivation in West Africa. In Proctor J (Ed.) Mineral nutrients in tropical forest and savanna ecosystems. Blackwell. London, England. pp. 289-300.        [ Links ]

19. Kachaka S, Vanlauwe B, Merckx R (1993) Decomposition and nitrogen mineralization of prunings of different quality. In Mulongoy K, Merckx R (Eds.) Soil organic matter dynamics and sustainability of tropical agriculture. Wiley. Chichester, UK. pp. 199-208.        [ Links ]

20. Kalra YP, Maynard DG (1991) Methods manual for forest soil and plant analysis. For. Can. Northwest Reg., North. For. Cent., Inf. Rep. NOR-X319. XXX pp.        [ Links ]

21. Kang BT (1994) Cultivos en callejón: logros y perspectivas. In Krishnamurthy L, Leos JA (Eds.) Agroforestería en desarrollo: educación, investigación y extensión. Universidad Autónoma de Chapingo. Mexico. pp. 61-72.        [ Links ]

22. Kang BT, Wilson GF, Lawson TL (1984) Alley cropping: an alternative to shifting cultivation. IITA. Ibadan, Nigeria. 22 pp.        [ Links ]

23. Kwesiga F (1994) Performance of fifteen provenances of G. sepium in eastern Zambia. For. Ecol. Manag. 64: 161-170.        [ Links ]

24. Lal R (1989) Agroforestry systems and soil surface management of a tropical alfisol: III. Changes in soil chemical properties. Agrofor. Syst. 8: 113-132.        [ Links ]

25. Lehman J, Schroth G, Zech W (1995) Decomposition and nutrient release from leaves, twigs and roots of three alley-cropped tree legumes in central Togo. Agrofor. Syst. 29: 31-36.        [ Links ]

26. Llera ZM, Vargas HJ, Velázquez MA, Jasso MJ (1998) Análisis de crecimiento de propágulos vegetativos y sexuales de cocoite (G. sepium). Agrociencia 32: 365-371.        [ Links ]

27. Marín QM (1996) Estimación de parámetros genéticos en dos ciclos de crecimiento de plantas de cocoite blanco (Gliricidia sepium). Tesis. Colegio de Postgraduados. México. 87 pp.        [ Links ]

28. Melchor MJI (1992) El cultivo en callejones: una estrategia para mejorar la agricultura tradicional en suelos de ladera en Papantla, Ver. In Ruíz JF (Ed.) I Reunión Nacional de Manejo y Conservación del Suelo y Agua. Colegio de Postgraduados. México. pp. 323-328.        [ Links ]

29. Melchor MJI, Vargas HJ, Ferrera CR, Krishnamurthy L (1999) Screening Rhizobium spp. Strains associated with Gliricidia sepium along an altitudinal transect in Veracruz, México. Agrofor. Syst. 46: 25-38.        [ Links ]

30. Moser EB, Saxton AM, Pezeshki R (1990) Repeated measures analysis of variance: Applications to tree research. Can. J. For. Res. 20: 524-535.        [ Links ]

31. Oglesby KA, Fownes JH (1992) Effects of chemical composition on nitrogen mineralization from green manures of seven tropical leguminous trees. Plant Soil 143: 127-132.        [ Links ]

32. Palm CA, Sánchez PA (1990) Decomposition and nutrient release patterns of the leaves of three legumes. Biotropica 22: 330-338.        [ Links ]

33. Parsons R, Stanforth A, Raven JA, Sprent JI (1993) Nodule growth and activity may be regulated by a feedback mechanism involving phloem nitrogen. Plant Cell Envir. 16:125-136.        [ Links ]

34. Román JA, Vargas HJ, Velázquez MA, Jasso MJ, Ortega E (1996) Variación geográfica y genética en el crecimiento y producción de biomasa del cocoite (Gliricidia sepium) en dos condiciones de humedad del suelo. Agrociencia 30: 549-558.        [ Links ]

35. Sabin TE, Stafford SG (1990) Assessing the need for transformation of response variables. Special Publication 20. Forest Research Lab, College of Forestry, Oregon State University. Corvallis OR, USA. 31 pp.        [ Links ]

36. Sanginga N, Manrique K, Hardarson G (1991) Variation in nodulation and N2 fixation by G. sepium/Rhizobium sp. symbiosis in a calcareous soil. Biol. Fertil. Soils 11: 273-278.        [ Links ]

37. Sanginga N, Danso SKA, Zapata F, Bowen GD (1994) Influence of pruning management on P and N distribution and use efficiency by N2 fixing and non-N2 fixing trees used in alley cropping systems. Plant and Soil 167: 219-226.        [ Links ]

38. SAS (1990) SAS/STAT User’s Guide. Version 6. 4th ed. SAS Institute Inc. Cary, NC, USA.        [ Links ]

39. Scroth G, Zech W (1995) Above and below-ground biomass dynamics in a sole cropping and alley system with G. sepium in the semi-deciduous rainforest zone of West Africa. Agrofor. Syst. 31: 181-198.        [ Links ]

40. Shantharam S, Mattoo A (1997) Enhancing biological nitrogen fixation: an appraisal of current alternative technologies for N input into plants. Plant Soil 194: 205-216.        [ Links ]

41. Tschaplinski TJ, Blake TJ (1995) Growth and carbohydrate status of coppice shoots of hybrid poplar following shoot pruning. Tree Physiol. 15: 333-338.        [ Links ]

42. Wiersum F, Dirdjosoemarto S (1987) Past and current research with G. sepium. In Withington D, Glover N, Brewbaker JL (Eds.) Gliricidia sepium (Jacq.) Walp.: management and improvement. Special publication 87-01. NAFTA. Turrialba, Costa Rica. pp. 20-29.        [ Links ]