In Costa Rican tropical wet forest the distribution of seedlings of most tree species colonizing 51 tree-fall gaps is clumped, probably because of differences in the proximity of gaps to fruiting trees. Different gap zones (root, bole and crown) are more similar to one another, in terms of their species composition, than they are to other zones in the same gap. This nonrandom pattern is established soon after gap formation, indicating that mortality of young seedlings is both high and related to gap zones. Some tree species are strongly positively associated with one of the gap zones, whereas others are associated with the age of the gap at the time the census was taken or the species of tree whose fall caused the gap. Root zones are dominated by fewer species than are bole or crown zones. Results are consistent with the assumption that an initial random distribution of seedlings is quickly changed to a strongly non-random pattern by selective mortality of seedlings of different tree species in the different gap zones. The internal heterogeneity of gaps is probably one of the factors helping to maintain the high tree species richness characteristic of tropical wet forests.

We describe forests from three areas of Jamaica, all on White Limestone but with markedly different rainfall regimes. The areas are Hog House Hill in the north-east with lower montane rain forest at c. 450 m altitude with a rainfall of c. 4000 mm yr−1; Broom Hall in the centre of the island with evergreen seasonal forest at c. 670 m altitude and with a rainfall of c. 1600 mm yr−1 and a marked dry season; and Round Hill near the south coast with dry semi-evergreen forest at c. 300 m altitude with an irregularly distributed rainfall of c. 1000 mm yr−1. Species lists were made from c. 180 ha at Hog House Hill, c. 5 ha at Broom Hall and c. 50 ha at Round Hill, and detailed inventories made of five sample sites of c. 1000 m2, two at Hog House Hill, one at Broom Hall and two at Round Hill.

At Hog House Hill we listed 280 vascular plant species, including 118 species of trees and larger shrubs; at Broom Hall 247 and 135; at Round Hill 129 and 81. Species-area and species-individuals curves confirm that Broom Hall was richer in tree species than Hog House Hill. The wetter forests contain high proportions of species endemic to Jamaica: 40% of the total flora at Hog House Hill and 36% at Broom Hall. Canopy height decreased from c. 26–28 m at Hog House Hill to c. 13–24 m at Broom Hall to c. 8–15 m at Round Hill. Predominant leaf size decreased from mesophyll at Hog House Hill to notophyll at Broom Hall to microphyll at Round Hill.

Compared with forests on other Caribbean islands, the Jamaican forests appear to be as species-rich as any, but lower in stature than natural forest in Trinidad and Dominica. Continental Neotropical forests are both more species-rich and taller.

Fruit production, frugivore foraging activity and seed dispersal was studied at 10 Cymbopetalum baillonii (Annonaceae) trees during an entire fruiting season. Fruits dehisced during the first six months of the year offer to potential dispersal agents a package of 8–24 arillate seeds. Insects and fungi, however, killed up to 14% of the seeds potentially available for dispersal before fruit dehiscence. Visitation by 24 species of birds and one mammal to the trees closely followed the availability of arillate seeds. The foraging activity of Peromyscus and Heteromys rodents accounted for ≥ 80% mortality of the seeds deposited under the crown of the parent tree. Among those seedlings that became established under the crown mortality was ≥ 90% after 12 months. Seeds dispersed under conspecific fruiting trees experienced intense predation by vertebrates. Field experiments showed that seeds and seedlings planted ≥ 30 m away from adult trees survived significantly longer than those planted under the crown and at distances ≤ 30 m. An unequivocal advantage to dispersal away from the tree and under allospecific trees was evident from the data. A very narrow range of frugivores (N = 8 species) accounted for the quality component of dispersal services to the trees. The tree with the lowest animal visitation and fruit crop was the most efficient in the dispersal of its seeds.

Soil water potentials were measured weekly by psychrometers at 20 cm depth during the dry season in a tropical moist forest on Barro Colorado Island, Panama. There was a persistent gradient of decreasing soil moisture from gap centre to gap edge to adjacent understorey at both a large and a small gap. On both a north-south and an east-west transect, the soil was drier (water potentials were more negative) on an upland surface than on moderate slopes. This trend was reflected in the predawn, total water potentials of shallow-rooted Psychotria horizontalis and deep-rooted Trichilia tuberculata measured in the understorey during the last two months of the dry season. P. horizontalis from the wettest sites on the transects had higher osmotic potentials at full hydration and at zero turgor, indicating less drought resistance than for conspecifics from the driest sites.

Dormancy and germination ecology of two Harpagophytum species (Pedaliaceae) from an open Acacia savanna in Botswana were investigated. The maintenance of dormancy is governed partly by the seed coat but mainly by the endosperm and the embryo itself, as demonstrated by removal of the endosperm. Dissemination of the seed from the fruit can be delayed for several years without affecting the viability of the embryo, due to very low respiration rates.

Germination can be enhanced slightly by high temperatures under natural conditions, and by gibberellic acid or removal of the endosperm under laboratory conditions. Relative growth rates for both Harpagophytum species are lower than for subtropical grasses and legumes, due to a strong investment in root and tuber biomass. Germination and seedling growth is discussed in relation to the drought avoidance syndrome and the animal disperser syndrome.

Aechmea magdalenae is a terrestrial bromeliad that dominates areas of forest understorey on Barro Colorado Island, Panama. Nocturnal CO2 uptake via crassulacean acid metabolism was the primary means of carbon gain under well-watered conditions and all light regimes. The ability to maintain a positive carbon balance under conditions of very low light was demonstrated by laboratory measurements of photosynthesis and forest measurements of growth. Low-light-grown juvenile rosettes had the same daily net assimilation whether tested at photon flux densities of 15 or 300 μmol m−2 s−1 Growth rates of rosettes in treefall gaps were similar to those of plants in closed canopy forest. Growth rates of forest plants were increasingly correlated with canopy openness as the wet season progressed due to increased growth by gap plants, suggesting that water availability rather than light may limit growth during the annual dry season.

Tropical forests offer many more products than the category for which they are best known, viz. wood products. They can supply a host of raw materials, such as resins, oils, fibres and fruits, which collectively represent a commercial value far higher than is generally recognized. In addition they provide an exceptional abundance and variety of genetic resources, which make substantial contributions to modern agriculture, medicine, industry and energy. On top of these materials, tropical forests supply significant environmental benefits, such as protection of the soil and safeguards for watershed systems. While it is sometimes difficult to quantify the economic values of these diverse goods and services, they are often to be reckoned as equal to, if not higher than, the marketplace values represented by the commercial hardwood timber. Yet when the forests are exploited for their hardwood timber, the process is usually so disruptive to forest ecosystems that the other outputs are severely reduced. By contrast, one may exploit the forests for their other outputs with virtually no disruption to forest ecosystems. The paper describes the main categories of non-wood products available from tropical forests, analyses their relative worth, and offers summarized proposals for an expanded approach to forest development in order to take integrative account of the full range of forest benefits.