Volunteer biomass production between multipurpose tree hedgerows after two years of fallow in southern Cameroon

The ability of multipurpose hedgerow tree species to out-compete undesired regrowth during fallow phases was examined. Biomass and spatial distribution of grass and broad leaf volunteers was measured after two years of fallow, in two alley cropping systems planted at six m interrow distance, at the Humid Forest Ecoregional Centre Research Station, Mbalmayo, southern Cameroon. The two experiments had been continuously cropped for five and six years previously. In the experiment cropped for six years, the presence of Senna spectabilis [(DC.) Irwin and Barnaby] hedgerows reduced the biomass of the volunteer regrowth from 9.2 to 4.3 Mg ha⁻¹. Tillage during the previous cropping phase increased the broad leaf biomass from 3.0 to 4.4 Mg ha⁻¹, reduced the biomass of grasses from 3.4 to 2.7 Mg ha⁻¹ but had no effect on the total volunteer biomass. Volunteer biomass was significantly lower within 1.5 m of the S. spectabilis hedgerows than at positions further away. In the experiment cropped for five years, S. spectabilis reduced the volunteer regrowth biomass significantly. Two other hedgerow species, Dactyladenia barteri [(Hook f ex Oliv.) Engl.] and Flemingia macrophylla [(Willd.) Merrill] had no effect on the total amount of volunteer regrowth but did reduce volunteer biomass within 0.5 m of the hedgerows. S. spectabilis caused a stronger reduction of volunteer biomass than D. barteri and F. macrophylla at almost all distances from the hedgerows. The competitiveness of D. barteri and F. macrophylla is insufficient and their growth habit is unsuitable to out-compete undesired species in this alley cropping system. This revised version was published online in June 2006 with corrections to the Cover Date.     

Weed seedbank characteristics of arable fields under different fallow management systems in the humid tropical zone of southeastern Nigeria

The composition and pattern of weed flora in arable fields are determined by their seedbank structure; but the influence of fallow management practices on weed seedbank structure is presently unknown. The objective of this study was to investigate weed seedbank characteristics and weed population dynamics in arable fields in natural and planted-fallow systems. The study plots were at Mbaise, a densely populated area of southeastern Nigeria, where farmers regenerate their exhausted soils by maintaining planted fallows of Dactyladenia barteri (Hook. F. ex Oliv.) Prance & F. White, and at Umuahia, a less-densely populated area in the same region, where farmers depend on natural bush fallow for soil regeneration. The effect of three years of fallow on the weed flora of arable fields in the two fallow management systems differed remarkably. The first flush of weeds on fields that were cultivated after three years of planted D. barteri fallow (Mbaise) consisted of 80% broadleaf weeds, 7% grass weeds and 13% sedges. On the other hand, the first flush of weeds on the natural bush fallow fields (Umuahia) of the same fallow duration as the D. barteri fallow system consisted of 17% broadleaf weeds, 70% grasses and 13% sedges. Three years of planted fallow caused 36% decrease in weed seedbank at Mbaise relative to the cropped field while the same duration of natural bush fallow caused a 31% increase in weed seedbank at Umuahia. These results show that the planted D. barteri fallow system has a higher potential to reduce weed pressure in smallholder agriculture than the natural bush fallow system and may explain in part why farmers in this humid forest zone have adopted the practice. This revised version was published online in June 2006 with corrections to the Cover Date.     

Traditional fallow systems of the Americas

Six neotropical traditional fallow systems are described: 1) enriched fallows of the Amazon, 2) Babassu palm (Orbignya phalerata Mart.) forests of central and northern Brazil, 3) bracatinga (Mimosa scabrella Benth) improved fallow of southeastern Brazil, 4) carbon negro (Mimosa tenuiflora Willd.) fallow of the wet-dry zone of Mesoamerica, 5) frijolillo (Senna guatemalensis Donn. Smith) of high-elevation zones in southern Honduras, and 6) caragra (Lippia torresii) fallow of humid zones of Costa Rica.These systems include both biologically and economically enriched fallows. Some (e.g., the enriched fallows of the Amazon) require a considerable degree of human intervention, while others (e.g., the bracatinga and carbon negro fallows) seem to form with little human encouragement, following the burn in shifting cultivation systems. Some of the systems (e.g., frijolillo and caragra) are quite site-specific and have a limited distribution; others (e.g., babassu, bracatinga, and carbon negro) occur over large areas and could be adapted to considerable areas in the Americas. As economic analyses are generally lacking, it is unclear what benefits would be achieved from a wider use of these systems. Most of the biologically enriched fallows seem able to maintain low yields of food crops at low input levels. Some of the economically enriched fallows seem able to produce higher economic returns.     

Modelling species and spacing effects on root zone water dynamics using Hydrus-2D in an Amazonian agroforestry system

Modelling the root zone water dynamics in a tree crop agroforestry system is a useful approach to understanding small-scale effects in tree crop systems and may be helpful for optimizing tree spacing in agroforestry system planning. The agroforestry system in this study consists of the species Theobroma grandiflorum (Willd ex Spreng) Schum (Cupuaçu), Bactris gasipaes H.B.K. (peach palm) and the cover crop Pueraria phaseoloides (Roxb.) Benth (Pueraria). The soiltype is an oxisol. Calibration was conducted for each of the three species separately. Calibration results show good conformity between simulated and measured data. Simulated scenarios examine the influence of different spacing between trees on root water uptake, evaporation and drainage. Mean interception and crop factors of the whole flow region vary with spacing or are held constant to examine below-ground effects only. Also a fictitious scenario of an older agroforestry system with deeper roots is calculated. In order to overcome restrictions of the computer program Hydrus-2D, correction factors in the root zone were introduced and a calculation scheme for root water uptake of a flow subregion was developped. Below-ground effects of spacing between trees are not or almost not present, but the depth of the tree roots has a significant influence on root water uptake, evaporation and drainage. When mean interception and crop factor vary, drainage increases and root water uptake decreases slightly with spacing. The modelling approach has been found promising for optimizing agroforestry systems although it can only be seen as a first beginning. In an agroforestry systems under drier conditions differences in results will probably be larger.     

Soil water regime under rotational fallow and alternating hedgerows on an Ultisol in southern Cameroon

Soil moisture depletion during dry seasons by planted hedgerows to lower levels than under natural fallow, would reduce drainage and nutrient losses in the following rainy season when food crops are grown. The volumetric water content of the 0–150 cm soil profile was measured under planted hedgerows (alternating Leucaena leucocephala and Gliricidia sepium) and natural fallow, both either annually cropped to sole maize or in a two-year crop/two-year fallow rotation, in the humid forest zone (annual rainfall 1700 mm) of southern Cameroon during the 1995–1996 and 1996–1997 dry seasons. Hedgerows were cut to 0.05 m height, largely eliminating trees’ water consumption during cropping phases. Differences in total soil water content at 0–150 cm depth, between systems, occurred only in the early phases of the 1996–1997 dry season. In both dry seasons, differences between systems in water content were found in some soil layers, all within 0–60 cm depth, yet, without consistent advantage of any system in exploiting the topsoil water resources. Soil water content was lower under L. leucocephala than G. sepium at 20–40 cm depth only. Below 60 cm depth, no differences in water regimes between systems were found. Under southern Cameroonian conditions it is unlikely that any of the systems has an advantage in accessing or recovering water and thus, if available, nutrients from the sub-soil. None of the systems examined was capable of delaying drainage and thus it appears unlikely that downward displacement of nutrients is delayed after the start of the rains.     

From alley cropping to rotational fallow: Farmers' involvement in the development of fallow management techniques in the humid forest zone of Cameroon

Alley cropping was introduced in the humid forest zone of Cameroon to increase soil fertility in 1987, but until 1992 the adoption rate had remained low. To better understand the reasons behind this, three types of on farm trials were established from fully researcher controlled to fully farmer controlled. During the evaluation of the technology with farmers a number of modifications were registered (1) Pruning height and frequency: Because of the difficulties to consistently cut back at 30 cm, farmers decided to slash at ground level as they normally do while slashing the natural fallow vegetation. This equally allowed for more flexibility in time of pruning. (2) Cropping intensity and pattern: Initially, alleys were cropped each year. This however had several shortcomings. Having observed the positive impact of incidental fallow period in a farmer's field, it was decided to introduce a fallow phase of at least one year. During the fallow period the plot can be used for fodder production, bee farming and production of stakes. (3) Residue management: Fire went incidentally in a farmer's tree plot after slashing, and the trees were not affected. This gave farmers an alternative way to manage the residue, by controlling the fire, before bringing in other crops such as groundnut and cassava. (4) Agroforestry species: Because Leucaena leucocephala, in spite of it soil fertility restoration potential, rapidly became a noxious weed, farmers have asked for a less invasive species. Calliandra calothyrsus was introduced for this purpose and became a good bee forage. With these modifications, the original alley cropping system has evolved into a rotational tree fallow with higher adoption potential. From about 15 farmers who were testing the technology in 1992, the number increased to 52 in 1996, 120 in 1997 and 236 in 1998. This revised version was published online in June 2006 with corrections to the Cover Date.     

Yields of maize/cassava intercrops grown with hedgerows of three multipurpose trees on an acid Ultisol of Cameroon

Maize/cassava were intercropped between hedgerows of Senna spectabilis [(DC.) Irwin and Barneby], Flemingia macrophylla [(Willd.) Merrill] and Dactyladenia barteri [(Hook f ex Oliv.) Engl.] for five consecutive years on an Ultisol in southern Cameroon. Crop yields and hedgerow biomass production in the third to fifth year of cropping are reported. S. spectabilis produced more biomass than F. macrophylla and D. barteri in all years. Cumulative maize grain and cassava tuber yields were highest in F. macrophylla alley cropping, outyielding the no-tree control consistently by 42 to 67% (average 56%). Between hedgerows of D. barteri and S. spectabilis, crops yielded 17% and 16% more than the no-tree control, respectively. However, between S. spectabilis hedgerows, yields were highly variable between years (–15% to +35% compared to the no-tree control) and thus the system is at risk of failure. F. macrophylla is recommended for continuous alley cropping of maize/cassava intercrop. The use of D. barteri may require fallow phases for biomass accumulation followed by cropping phases with rigorous pruning. Although this may lead to lower cumulative yields, the products of the fallow phase, such as stakes and firewood, may provide some compensation.This revised version was published online in November 2005 with corrections to the Cover Date.     

Soil fertility enhancement by planted tree-fallow species in the humid lowlands of Cameroon

Four shrub/tree species, Alchornea cordifolia, Pennisetum purpureum, Chromolaena odorata, and Calliandra calothyrsus were evaluated for their potential contribution to soil fertility restoration after two years fallow. Standing biomass, root distribution, nutrient content in the biomass, decomposition and nutrient release patterns, and association with mycorrhizae were the evaluation parameters. Alchornea and Pennisetum produced thehighest above-ground biomass, 66 t and 54 t/ha respectively. Pennisetum had more than 19 t/ha of root, 92% of which was in the 20 cm top soil. Alchornea had 74% of it roots in this soil layer, mostly as coarse roots while Calliandra had a deeper root system. Alchornea fallow accumulated more N and Ca, and Pennisetum fallow, more K than others, and mycorrhizae were mostly associated with Alchornea roots. The ranking of the different species for the decomposition rate was: Chromolaena > Pennisetum > Calliandra = Alchornea. Also release of nutrients during decomposition followed the order K > N > Ca. Alchornea and Pennisetum could be recommended as green manure species especially when high quantities of material are needed for weed or erosion control. Calliandra and Chromolaena, because of the flush of nutrient during early mass, loss can be used as mulch when the crop demand of nutrient is high. Alchornea decomposed slowly and therefore could be used to improve Chromolaena mulch, thus contributing to the build up of soil organic N and providing both short- and long-term nutrient release. This revised version was published online in June 2006 with corrections to the Cover Date.     

Nutrient composition of the leaf biomass of three selected woody leguminous species

In order to assess the value of leaf biomass the foliar nutrient content of Leucaena leucocephala, Gliricidia sepium and Flemingia macrophylla was studied under field conditions. In total 5 series of leaf samples were collected and analysed for nitrogen, phosphorus, potassium, calcium and magnesium content over a period of 42 months of growth observation. Results are compared with mean values derived from a literature study. Factors contributing to the variation observed are discussed and suggestions towards variation control are made.     

Effect of planted fallow systems on <Emphasis Type="Italic">Zonocerus variegatus</Emphasis> (Orthoptera: Pyrgomorphidae) abundance in the humid forest zone of Southern Cameroon

Zonocerus variegatus (L.) is known as an agricultural pest in Central Africa. Since legumes are introduced in the Chromolaena odorata fallows in the southern Cameroon as short fallow improvement strategies, we evaluated the abundance of Z. variegatus in short fallows (natural versus planted leguminous fallows) and adjacent fields. Existing Randomized Completed Block Design were used, with three treatments (C. odorata, Calliandra callothyrsus and Pueraria phaseoloides) and three replicates, split in two sites (land cleared from forest and land cleared from chromolaena fallow). Research was carried out from February 2000 to February 2002 by captures (every 14 days) in each fallow and adjacent cropped fields. Within 30 min, 24 ± 75 (SD) individuals were captured in the fallows, while only 10 ± 20 (SD) were captured in adjacent cropped fields. The abundance was greater in fields after forest clearing than in fields after chromolaena clearing. More individuals were captured in C. odorata fallows in both sites. Z. variegatus abundance in planted fallows depended on the site. After forest clearing, Z. variegatus abundance was highest in the P. phaseoloides and lowest in C. callothyrsus fallow, while after chromolaena clearing the abundance in P. phaseoloides fallow was lowest. Z. variegatus abundance between cropped fields adjacent to C. odorata, C. callothyrsus and P. phaseoloides fallow was not significantly different, but significant linear relationships were found between Z. variegatus abundance in 2 years old fallows and the adjacent cropped fields. These results indicate that in areas where chomolaena dominates the fallow, planted P. phaseoloides could be useful in managing Z. variegatus.     

Dynamics of Particulate Organic Matter following biomass addition from fallow-improvement species in southern Mali

Knowledge about the nature and turnover rates of soil organic matter (SOM) is a key to soil fertility maintenance, especially in traditional farming systems. Particulate organic matter (POM), rather than total SOM content, is now considered a better measure of the soil's fertility. Fractionation of SOM, based on particle-size distribution after dispersion, was carried out on soil samples collected from a field experiment at N'Tarla, Mali (12.25° N latitude, 5.42° W longitude, Typic Plinthustalf). Biomass of off-site grown Gliricidia sepium (Jacq.) Walp. (gliricidia), Pterocarpus erinaceus Poir.(pterocarpus), and Stylosanthes hamata (L.) Taub.(stylosanthes) was applied to a maize (Zea mays L.) crop for two consecutive years. A significant decrease was noted in the soil's POM and POM-N contents after harvest of maize in both years as compared with the initial value measured at the beginning of the experiment. Small POM fraction (0.053 to 0.250 mm, POM₅₃) contributed more to total POM than large POM fraction (0.250 to 2.000 mm, POM₂₅₀). The POM N content was better correlated with maize grain yield than POM total weight, and the N content of POM₅₃ was better related to maize grain yield (R² = 0.55) than that of POM₂₅₀ (R² = 0.20). Total SOM was poorly correlated with maize grain yield (R² = 0.14). The results indicate that the quality of the applied biomass is of greater value than its quantity in sustaining maize grain yield, and POM₅₃ might be a better indicator of soil fertility status than POM₂₅₀ and total POM. This revised version was published online in June 2006 with corrections to the Cover Date.     

Structure and composition of cocoa agroforests in the humid forest zone of Southern Cameroon

The distribution and composition of the tree component inside cocoa agroforests plays an important role in the economic and ecological services offered by these plantations. The presence of these plant components appears to be influenced by several factors controlling the introduction and management of associated plants inside cocoa agroforests. To date, few studies have tried to evaluate the horizontal and vertical distribution of plants inside cocoa plantations in Cameroon. This study determines the structure of cocoa plantations in Southern Cameroon. Field data were collected in 60 cocoa plantations belonging to 12 villages located along a contiguous gradient of market access, population density and resource use intensity in the humid forest zone of southern Cameroon. This study area comprises (i) the sub-region of Yaoundé, (ii) the sub-region of Mbalmayo, and (iii) the sub-region of Ebolowa. Market access, population density and resource use intensity all decreased from the first to the third sub-region. For cocoa and associated plants, we quantified (1) the density (2) the individual number, the species composition and the group uses of plants (edible, timber, medicinal, etc…) distribution across strata, and (3) the basal area in the 60 cocoa plantations located in the three main sub-regions. Results are presented for each sub-region and the whole study area. The paper develops cocoa agroforest typologies and discusses possible implications of cocoa agroforest structure diversity in the achievement of economic and ecological services.     

Plant growth, biomass production and nutrient accumulation by slash/mulch agroforestry systems in tropical hillsides of Colombia

Planted fallow systems under ‘slash and mulch’ management were compared with natural fallow systems at two farms (BM1 and BM2) in the Colombian Andes. The BM1 site was relatively more fertile than the BM2 site. Planted fallow systems evaluated included Calliandra calothyrsus CIAT 20400 (CAL), Indigofera constricta (IND) or Tithonia diversifola (TTH). During each pruning event slashed biomass was weighed, surface-applied to the soil on the same plot and sub-samples taken for chemical analyses. While Indigofera trees consistently showed significantly greater (p < 0.05) plant height and collar diameter than Calliandra trees at both study sites, only collar diameter in Indigofera was significantly affected at all sampling times by differences between BM1 and BM2. After 27 months, TTH presented the greatest cumulative dry weight biomass (37 t ha^–1) and nutrient accumulation in biomass (417.5 kg N ha^–1, 85.3 kg P ha^–1, 928 kg K ha^–1, 299 kg Ca ha^–1 and 127.6 kg Mg ha^–1) among planted fallow systems studied at BM1. Leaf biomass was significantly greater (P < 0.05) for CAL than IND irrespective of site. However, CAL and IND biomass from other plant parts studied and nutrient accumulation were generally similar at BM1 and BM2. At both sites, NAT consistently presented the lowest biomass production and nutrient accumulation among fallow systems. Planted fallows using Calliandra and Indigofera trees had the additional benefit of producing considerable quantities of firewood for household use. This revised version was published online in June 2006 with corrections to the Cover Date.     

Competition for light between hedgerows and maize in an alley cropping system in Hawaii, USA

Successful agroforestry systems depend on minimizing tree-cropcompetition. In this study, field experiments and a simulation model were usedto distinguish between tree-crop competition for light and belowgroundcompetition in an alley cropping system. Maize (Zea maysL.) was harvested periodically in three treatments: between vertical barriers ofshade cloth, hedgerows of Flemingia macrophylla (Willd.)Merr., and sole maize. Radiation intercepted by the maize was calculated using asimulation model based on measured values for direct and diffuse light, hedgerowdimensions and leaf area, and solar trajectory. Radiation use efficiency wascalculated as biomass production per unit of intercepted radiation. Maizebiomass and yield in both the alley crop and the shade cloth treatment weregreatest in the center of the alleys. Grain yield between hedgerows was 3.5Mg ha⁻¹ (averaged across the alley), significantlyless than in the shade cloth (7.4 Mg ha⁻¹) or thesole maize (7.7 Mg ha⁻¹) treatments. Lightintercepted by the maize in the alley crop was about half that intercepted bythe maize in the sole crop. The shade cloth intercepted less light than thehedgerows because it did not have an appreciable width. Radiation use efficiencyin the three treatments was 0.75 g mol⁻¹ PAR anddid not differ significantly among treatments. Tree-crop competition wasoverwhelmingly for light. This revised version was published online in June 2006 with corrections to the Cover Date.     

Tree-root systems and herbaceous species-characteristics under tree species introduced into grazing lands in subhumid Cameroon

The effect of tree species on the characteristics of the herbaceous stratum, during the first five years of a fallow, was evaluated in the North of Cameroon (average annual temperature 28.2 °C, total annual rainfall 1050 mm). Treatments included a natural grazed herbaceous fallow, a natural ungrazed herbaceous fallow and three planted tree fallows (Acacia polyacantha Willd. ssp. campylacantha (Hochst. ex A. Rich.), Senna siamea Lam. and Eucalyptus camaldulensis Dehnh.), which were protected against grazing. Because tree species influenced light interception in different ways, as well as having different root patterns, they had different effects on the herbaceous stratum in terms of species composition and biomass. The grazed herbaceous fallow maintained the greatest species richness. Protection against grazing or the introduction of tree species associated with the absence of grazing induced both a progressive evolution to a particular species composition. The ungrazed herbaceous fallow consisted mainly of Andropogon gayanus Kunth, which provided the greatest biomass (8 t dry matter ha^–1 at the end of the fallow period). E. camaldulensis provided little shade and the lowest fine root mass in the top layer allowing the growth of A. gayanus and thus a greater herbaceous biomass (3.5 t DM ha^–1) than that found under the other tree species. Under the heavy shade of A. polyacantha, the herbaceous stratum consisted mainly of annual Pennisetum spp. (2.2 t DM ha^–1) and showed the greatest N concentration (1.3%), probably due to N₂ fixation by the tree species. After the fourth year, despite the relatively open tree canopy, S. siamea, which showed the highest fine root mass, had a strong depressive effect on the herbaceous stratum. This revised version was published online in June 2006 with corrections to the Cover Date.     

Initial agronomic evaluation of Parkia biglobosa in the humid zone of Nigeria

Parkia biglobosa is an important traditional economic tree legume of considerable multipurpose potentials that has not been well researched. It is used for fodder, human food, fuel wood, timber, green manure, medicine, provides shade for forage grasses and livestock and protects soil from heat and it is important in soil nutrient cycling.The objective of this study was to determine the effect of management regimes (cutting frequency and height) on the coppicing ability, fodder production, and nutritive value potentials of cultivated Parkia trees. This initial study has shown that the tree has a potential as fodder for livestock as well as human food and should be further studiedto fully understand its biology, agronomy and feeding value under various agroforestry or silvopastoral systems of humid tropics in West Africa.     

Establishment of Inga edulis and Calliandra calothyrsus in improved fallow systems in southern Cameroon

The adoption of planted fallow largely depends on the cost and feasibility of using the technology; easy, inexpensive and simple fallow establishment methods are known to greatly enhance adoption. It was the objective of this study to assess the effects of weeding regime on the establishment of Calliandra calothyrsus and Inga edulis on degraded acid soils in southern Cameroon. A combination of the two fallow species and two weeding regimes, weeding or not weeding, were compared to a natural fallow. The trial was conducted in two sites of different base saturation levels with four replications. The results indicate that differences between the two species and the two weeding regimes were statistically significant (p 005) on both sites for all measured tree growth parameters, as well as the residual effects on subsequent maize grain yield. Presence of weeds reduced stem diameter and height of C. calothyrsus and I. edulis at the early stage of their establishment. Weeding doubled the leaf biomass of both species. The highest woodmass was produced by Inga in plots with weeding treatment, with 48 t/ha of dry material. Tree fallow improved the yield of succeeding crops by twofold over the natural grass fallow. Weeding treatment improved maize yield, from 1.9t/ha to 2.8 t/ha after Calliandra fallow, and from 2.22 t/ha to 3.0 t/ha after Inga fallow. The significant effects of weeding treatments implies that fallow-improving tree species should be planted in relay intercropping for trees to benefit from the weeding of crops, thus reducing the labour spent on fallow establishment. This revised version was published online in June 2006 with corrections to the Cover Date.     

Woody legumes as live support systems in yam cultivation

This paper reports on the use of three leguminous perennials (Leucaena leucocephala, Flemingia macrophylla and Gliricidia sepium) as live support systems in yam cultivation. In a planting arrangement in which yam rows alternated with rows of the woody species the tuber yields per ha were 3.4 (leucaena), 5.3 (flemingia) and 10.1 (gliricidia) ton fresh weight. TUKEY's L.S.D. value equalled 2.9 ton. Leucaena leucocephala is unsuitable as live support since the species shows a strong competitive power expressed in terms of leaf productivity and relative density of the root mass in the upper soil stratum, the zone explored by the yam crop. Flemingia macrophylla is unsuitable as support species mainly because of its structural weakness: Its branches do not sufficiently lignify to carry the yam leaf mass. The significantly higher tuber yield of the yam crop grown with Gliricidia sepium is a function of specific properties of the tree species: Low leaf productivity, a relatively weakly developed root system and an open architecture, which leaves sufficient space for a yam crop grown in association.     

Crown development and biomass production of pollardedErythrina berteroana,E. fusca andGliricidia sepium in the humid tropical lowlands of Costa Rica

Leguminous trees are widely used to support climbers such as black pepper (Piper nigrum L.) and vanilla (Vanilla planifolia Andr.), to provide shade to crops and to maintain soil fertility. Pruning or pollarding provides the means to maximize benefits from the trees, particularly through the production of biomass as a soil amendment. At the same time, excessive shading is reduced. In order to quantify the degree of shading of black pepper by the support trees during a six-month pollarding cycle, this study monitored crown development (part I) and light transmission (part II) of three widely used species,Erythrina berteroana Urban,E. fusca Loureiro andGliricidia sepium (Jacq.) Steud.The two sites were in the humid Atlantic Lowlands of Talamanca, southern Costa Rica (mean annual rainfall 2460 mm, no distinct dry season), on alluvial soils (typic Tropofluvents) with low levels of K, P, Mn and Zn. Two-year-old trees, that had been established from cuttings as live supports for black pepper, were used for the study. They were pollarded twice per year. The variables measured/estimated monthly were: stem diameter at breast height (dbh); height, foliated height, depth, diameter, leaf area and leaf biomass of crowns; length, diameter, number and inclination of branches. Leaf nutrient contents were also determined.Following pollarding,G. sepium was the first to resprout, followed byE. berteroana andE. fusca. G. sepium with its few but erect and long branches had slender, columnar crowns. while those ofE. berteroana andE. fusca were more spherical. Four months after pollarding,G. sepium started shedding leaves at the base of its branches. Average crown diameter after six months were 2.2 m forE. berteroana, 1.9 m forE. fusca and 1.5 m forG. sepium; average crown depths after six months were 2.8 m, 2.1 m, and 2.7 m, respectively, for the three species. For 1600 trees ha^–1 and two prunings per year, foliar biomass production from prunings alone (i.e., without litter fall), calculated from regressions with length and basal diameter of branches as independent variables, was 3.8 t, 3.4 t and 2.3 t dry matter ha^–1 a^–1 forE. berteroana, E. fusca andG. sepium, respectively; these estimates agreed well with measured values. The corresponding N contents were 146 kg, 124 kg and 90 kg, respectively. While N contributions from the prunings exceeded 50% of the fertilizer recommendations for black pepper, the contributions were <10% for P and <40% for K. Linear regressions between leaf area and branch dimensions, and quadratic regressions between foliar biomass and crown diameter showed high coefficients of determination (0.83>R ²>0.99). Correlations between foliar biomass, dbh, and dbh increments were generally weak. Conclusions from the study appear to be valid also for other agroforestry systems where the same species are planted under similar ecological conditions for reasons other than as live supports.

---

Resumen Los árboles leguminosos se usan frecuentamente como soportes vivos para plantas trepadoras como pimienta negra (Piper nigrum L.) y vainilla (Vanilla planifolia Andr.), para dar sombra y para reducir el desgaste de los suelos. La poda de los árboles reduce la cantidad mantener la fertilidad de sombra y produce biomasa que sirve como enmiendas. Con el objectivo de cuantificar el grado del sombreado de pimienta negra por los árboles durante un cíclo de poda de seis meses, se monitoreó en este estudio el desarrollo de la copa (parte I) y la transmisión de luz (parte II) de tres especies usados como soportes vivos:Erythrina berteroana Urban,E. fusca Loureiro yGliricidia sepium (Jacq) Steud.Los dos sitios estan en la zona tropical húmeda de Talamanca Baja, en el sureste de Costa Rica (precipitación promedia 2460 mm, sin estación seca) en suelos aluviales (Typic Tropofluvents) y pobres en K, P, Mn y Zn. Los árboles fueron establecidos de estacas. Cuando este estudio empezó, los árboles tenían dos años de edad y fueron podados dos veces por año. Los parámetros medidos ó estimados mensualmente fueron: diámetro del tronco a la altura del pecho (DAP); altura, altura foliada, profundidad, diámetro, área foliar y biomasa foliar de las copas; largo, diámetro, número y inclinación de las ramas. Además se determinó el contenido de nutrientes de las hojas.Despues de la poda,G. sepium fue el primero en rebrotar, seguido porE. berteroana yE. fusca. G. sepium con sus ramas escasas pero largas y erectas, formó una copa delgada y columnar, mientras que las copas deE. berteroana yE. fusca fueron más esféricas. Cuatro meses después de la poda,G. sepium empezó a perder las hojas en la base de sus ramas. El promedio del diámetro de la copa después de seis meses fue de 2.2 m paraE. berteroana, 1.9 m paraE. fusca y 1.5 m paraG. sepium; el promedio de la profundidad de la copa después de seis meses fú de 2.8 m, 2.1 m y 2.7 m, respectivamente para las tres especies. Para 1600 árboles ha^–1 y dos podas por año, la producción estimada de biomasa foliar de la poda (sin tomar en cuenta la hojarasca caída antes), calculada de regresiónes con el largo y el diámetro basal de las ramas como variables independientes, fue de 3.8 t, 3.4 t y 2.3 t de matéria seca ha^–1 a^–1 paraE. berteroana, E. fusca yG. sepium, respectivamente; estos valores calculados concordaron bien con valores medidos. El contenido de N correspondiente fué de 146 kg, 124 kg, y 90 kg, respectivamente. Mientras que la contribución de N procedente de las podas excedió 50% de una recomendación de fertilización común para pimienta negra, la contribución de P fue inferior al 10%, y menos del 40% para K. Regresiones lineales entre área foliar y las dimensiones de la ramas, y regresiones cuadraticas entre biomasa foliar y el diámetro de la copa mostraron altos coeficientes de determinación (0.83>R ²>0,99). Las correlaciones entre biomasa foliar, el DAP y incrementos del DAP fueron generalmente débiles. Las conclusiones de éste estudio parecen válidas tambien para sistemas agroforestales donde las especies estudiadas son plantadas bajo condiciones ecológicas similares para propósitos diferentes de como soportes vivos.