Weed suppression by canopy closure and mulching
Many studies have shown alley cropping to be very effective in reducing weed incidence, which -- as a result of decreased weed competition -- indirectly leads to an increase in crop yield. For example, in a maize alley-cropping study in Costa Rica, Rippin (1991) found that weed biomass was over 50% lower in the alleys of E. poeppigiana and G. sepium than in the sole crop control; release from weed competition by complete removal of weed only marginally increased alley-cropped maize yield but led to over 50% higher yield in the sole crop, suggesting that approximately half of the higher maize yield in the alley crop could be a result of reduced weed competition.

Weed suppression in alley cropping is largely accomplished through shading of the undergrowth by hedgerow canopy closure and surface application of prunings as mulch. Most weeds in their natural forest environment are pioneer species that quickly colonize a site after the canopy is opened and light provided (Kang 1993). Since these weeds are usually intolerant of shade, limiting available light resource can significantly depress their germination and growth. In alley cropping systems in Nigeria, Yamoah et al. (1986) found that weed incidence was positively correlated with available radiation. Similarly, Jama et al. (1991) concluded that shading was responsible for the weed reduction under closely spaced Leucaena alleys in Kenya. Since hedgerow trees are pruned during cropping cycles to provide mulch or green manure and to prevent shading of the crop, the shading effect of canopy closure is significant mainly between cropping cycles when hedgerows are allowed to grow freely. This is important in limiting temporally segregated weed competition where weeds take up nutrients that would otherwise be available for the next crop. Limiting weed growth between cropping cycles also reduces the labor needed for weeding before planting the new crop. This is necessary because crop seedlings do not compete well with well established weed population. Mulch covering the soil throughout the growth of the crop not only decreases light infiltration but can also physically smother weed germination and growth.

Factors influencing effectiveness of weed control
The extent to which weed is suppressed depends partly on the tree species, i.e., on their canopy development and shading characteristics, as well as the rate of decomposition of pruned tree residues (Szott et al. 1991, Rippin et al. 1994). Studying maize/weed competition in an alley cropping trial with Erythrina poeppigiana and Gliricidia sepium in Costa Rica, Rippin et al. (1994) found the mulch materials of the two tree species showed considerable differences in their potential weed-suppressing ability. E. poeppigiana was more effective in checking weed growth, most like because the rate of decomposition of E. poeppigiana mulch material was lower than that of G. sepium mulch. E. poeppigiana mulch covered the ground for a longer time and very effectively hindered the development of the weed seedlings. On the contrary, a comparison between the alley crop and sole-crop treatment of G. sepium mulch shows that the fast decay of G. sepium mulch material limited its ability to suppress weed. Szott et al. (1991) also reported that weed suppression by prunings in alley cropping was related to mulch quality: slowly decomposing mulches such as Inga suppressed weeds more effectively than mulches that decomposed more rapidly.

Interhedgerow spacing can also affect the weed-suppressing capacity of an alley cropping system. Rosecrance et al. (1992) studied the effects of Calliandra calothyrsus and Gliricidia sepium hedgerows, established 4, 5 and 6 m apart, on weed populations in Western Samoa. Weed growth was lowest in the 4-m alleys, due to rapid canopy closure between cropping cycles and the large amounts of mulch applied (due to higher biomass produced by hedgerows) throughout the growth of the taro crop. The 5-m and 6-m alleys were too wide for canopy closure during the fallow and substantial weed populations were growing in these plots just prior to new crop planting. The 6-m alleys supported the highest weed populations (even greater than the no-tree no-mulch control), which Rosecrance et al. (1992) attributed to minimal shading from the hedgerows and improved soil characteristics from mulch application. Only spot weeding was required in the 4-m alleys throughout the cropping period in this study, which is noteworthy since weed control is a major expense -- in terms of labor requirement and high cost of herbicides -- for farmers in most developing countries.