Forest fragmentation can alter herbivory on tree recruits with possible consequences for regeneration. We assessed effects of forest-fragment quality (tree diversity, vegetation complexity, relative abundance of pioneer trees) and matrix habitat on arthropods and herbivory in KwaZulu-Natal, South Africa. We compared arthropod abundances and herbivory on woody seedlings and saplings among four forest-fragment types differing in size and matrix (large fragments and small fragments surrounded by natural grassland, eucalypt and sugarcane plantations; nplots = 24) using analyses of covariance. We recorded 3385 arthropods and inspected 897 seedlings (71 species) and 876 saplings (91 species). Relative abundance of predators increased with fragment quality; that of herbivores decreased. Herbivory responses to fragment quality varied: seedling herbivory decreased with relative abundance of pioneers and sapling herbivory increased with vegetation complexity. Matrix effects were low with little variation in relative abundance of predators (0.39–0.53) and herbivores (0.22–0.32), proportion of seedling (8.3–11.0%) and sapling herbivory (12.4–14.3%) among the forest-fragment types. These findings indicate that herbivory on tree recruits is mediated by forest-fragment quality rather than matrix habitat. Future studies should evaluate whether contrasting effects of fragment quality on arthropods and herbivory are caused by weak trophic interactions and variable herbivore compositions.

Levels of leaf damage due to insect folivory have been investigated in forests of different latitudes all over the world, but most research has concentrated on a few common forest types. Most studies of insect herbivory were conducted in (sub)tropical rain forests (Barone 1998, Basset 1996, Coley 1983, Lowman 1985), or in temperate forests (Landsberg & Ohmart 1989, Lowman & Heatwole 1992). In contrast, little is known about insect folivory of woody plants in tropical savannas (Fowler & Duarte 1991, Marquis et al. 2001, Ribeiro 2003, Stanton 1975), and no such data are available for the West African savanna ecosystem (Andersen & Lonsdale 1990). Savannas cover about 40% of the land surface of Africa and 20% of the world (Scholes & Walker 1993), and savanna trees may host considerable numbers of insects, including many herbivores (Grant & Moran 1986, Mody et al. 2003). Therefore, insect herbivory can be considered a potentially important aspect of plant–animal interactions for vast areas of tropical ecosystems, where it has been studied remarkably rarely so far.

Effects of litter cover on insect herbivory, leaf-fungal damage, and seedling performance (survival, growth in height and leaf number), were assessed on established Nectandra ambigens seedlings in three sites of a Mexican tropical rain forest. Naturally occurring seedlings were measured for height, leaf number and standing levels of leaf damage by herbivores and pathogens. Thereafter, seedlings were exposed for 1 y to three litter treatments: (1) litter addition, (2) control and (3) continuous litter removal. There was a significant effect of site on the proportion of surviving seedlings and on the relative growth rates in terms of leaf number. Overall, seedlings with higher initial leaf number showed higher survival and produced more leaves after 12 mo than seedlings with lower initial leaf number. Litter addition resulted in seedlings with lower relative growth rates in height and higher levels of damage by herbivores and pathogens. On average, leaf area damaged by pathogens was < 6% and herbivory was < 27%. Seedlings with higher levels of herbivory showed higher levels of pathogen attack. Herbivory and pathogen attack had no effect on seedling performance. This study reveals the importance of litter-cover on the interactions between seedlings, herbivores and pathogens, as well as on the performance of tropical seedlings.

Leaf defences, leaf nutritional quality and leaf expansion rates may vary with resource availabilities to plants. Such variation could affect rates of leaf loss to herbivores, particularly along the steep resource gradients in disturbed forests. Intraspecific and interspecific variation in leaf damage and leaf expansion rates were measured on dipterocarp seedlings planted into secondary forests 1, 5 and 15 y after logging, and in adjacent primary forest of Sabah, Malaysia. Herbivory rates or amounts of leaf damage were compared across habitats and species for expanding, recently expanded, and mature leaves of Shorea leprosula and Dryobalanops lanceolata (Dipterocarpaceae). In all four habitats, leaves of the faster growing S. leprosula sustained higher rates and amounts of leaf-area loss than did the tougher leaves of slower growing D. lanceolata. Expanding leaves accumulated more leaf-area loss per week than did mature leaves. In all habitats and in both species, more than 25% of expanding leaves disappeared entirely. Rates of leaf-area loss per week differed among habitats for expanding leaves but not for mature leaves. In a relatively open, 1-y-old logged forest, faster leaf expansion reduced the time leaves spent in the most vulnerable stage; however, in S. leprosula a greater rate of leaf area loss countered the shorter expansion time. Thus, leaves accumulated similar total damages across habitats, and herbivory did not produce differences among habitats in seedling growth or mortality. High levels of resources may increase both leaf palatability and leaf expansion rates, with counteracting effects on herbivory.

Variation in herbivory is often associated with plant density and light environment. The effect of these variables was studied on leaf production and herbivory of Manilkara bidentata in Puerto Rico. The major herbivore of M. bidentata is the leaf miner Acrocercops sp. To determine the effect of plant density, twenty-four 20-m × 20-m plots were established and the densities of saplings, juveniles and adults were determined. Leaf production, herbivory and growth were measured on saplings. Plant density was determined in eight 20-m × 20-m plots surrounding the focal plots. The effect of light was determined by comparing leaf phenology, leaf quality and herbivory in vertical and horizontal profiles. Sapling density in 60-m × 60-m plots was associated with herbivory. In the vertical profile, leaf production was continuous only for the canopy: herbivory increased from the canopy (1.3%) towards the understorey (35.6%). In the horizontal profile leaf production was related to light. Saplings in low light environments had narrow peaks in leaf production compared with saplings in high light environments. Differences in leaf phenology did not result in differences in herbivory possibly because of variation in herbivory among leaves. Although many saplings lost more than 80% of new leaf area, there was no detectable effect on growth. High levels of intra-specific variation in herbivory suggest that to better understand plant/herbivore interactions it is necessary to consider all size classes in a population. It is also important to document these pattern across different spatial scales.