Contemporary defaunation of fragmented forests potentially alters patterns of seed predation and dispersal. Alternatively, the remaining fauna may compensate for missing animals, resulting in equivalent rates of seed dispersal and predation. In the Los Tuxtlas region of southern Mexico, populations of large terrestrial fruit-eating mammals are diminished or absent from many forest remnants. This study reports fruit removal and seed predation patterns of Poulsenia armata (Moraceae), in forest fragments and a continuous forest (LTBS). Contrary to expectation, we found no differences in seed predation (mean ± SD) between LTBS (7.2 ± 1.8 seeds per station) and forest fragments (5.6 ± 1.1). However more fruits were removed in the LTBS (11.4 ± 0.9 fruits per station) than in forest fragments (8.1 ± 0.8). Animal activity, recorded by camera traps, differed between animal guild with fewer seed dispersers in forest fragments (mean = 0.43 ± 0.02 photos wk−1) than in the LTBS (mean = 0.68 ± 0.05). Fruits and seeds attracted many species of mammal (n = 12) in both habitats, indicating substantial redundancy. Remnant forest patches in the Los Tuxtlas landscape retain great ecological value, both as refuges for mammals and habitats for trees, such as P. armata.

Little is known about how land-use changes affect interspecific interactions such as fruit–frugivore mutualisms. Forest age could affect both fruit sugar concentrations via differences in light availability or disperser abundance, and fruit removal rates via differences in bird and plant community composition. We examined how these two factors are affected by forest age in a Costa Rican rain forest. We compared seven young-secondary forest species, seven old-growth forest species, and Miconia nervosa growing in both forests. We measured sugar concentrations in fruits and manipulated the location of paired fruiting branches, measuring subsequent fruit removal. Sugar concentration means were on average 2.1 percentage points higher in young-secondary forest species than in old-growth forest species, but did not differ among Miconia nervosa fruits from the two forests. Fruit removal rates were higher in young-secondary forest for 86% of young-secondary forest species, 71% of old-growth forest species, and on average for both young-secondary and old-growth forest Miconia nervosa individuals. Higher sugar concentrations in young-secondary forest plants could reflect stronger competition for dispersers, while experimental fruit removal results suggests the opposite patterns of competition; fruits are more likely to be removed by dispersers in young-secondary forest independent of fruit nutrient concentration.

In four field and glasshouse experiments designed to alter the supply of resources through manipulation of nutrients, root tissue, leaf area and fruit number in Alliaria petiolata (Brassicaceae), more than 99% of ovules per plant showed signs of fertilization, suggesting that seed production in this plant was not pollen limited. However, in all treatments a significant proportion of fruits and seeds did not develop to maturity. Total fruit and seed production did not differ significantly from controls when plants were given nutrient supplements at flowering. Removal of 50–75% of the root tissue in 1-yr-old plants significantly reduced fruit set, but had no effect on individual seed development. Removal of cauline leaves significantly reduced most measures of fruit and seed production, suggesting that current photosynthate is critical for fruit and seed filling. Seed maturation was significantly affected by both fruit position within an infructescence and ovule position within a fruit. Basally located fruits and ovules (within fruits) developed more mature seeds than distally positioned fruits and ovules. Plants responded to removal of basal fruits by re-allocating resources to distal fruits that would normally have aborted. Our results suggest that fruits and seeds act as reproductive sinks competing for parental photosynthate. Patterns of resource allocation within infructescences and fruits were also modified by our experiments.

Maternal-environmental effects on subsequent progeny life-history traits were evaluated in squash (Cucurbita pepo L.) in terms of the amount of time available for seed development, and the timing of fruit production. Progeny arising from three kinds of fruit were compared. Plants from which fruits were removed 3 d post-pollination throughout the growing season developed only ‘late’ fruits (during 10–15 d) at the end of the growing season; on control plants both ‘early’ and ‘late’ fruits developed (both types allowed to ripen fully). Seed from each type of fruit was weighed individually and categorized into three size classes, then germinated and raised to maturity, including regular harvesting of all fruits 3 d post-pollination. Maternal effects were evident for both vegetative and reproductive traits and carried over to later stages. In contrast, effects due simply to seed size disappeared by day 30 for leaf variables and day 60 for male flower production. Within a seed-size class, progeny arising from fruit of treated plants produced significantly more leaves, with greater size, and more male flowers than those arising from fruit of control plants, while the reverse was true for fruit number and fruit mass. This result is discussed in terms of possible gibberellic acid involvement. In control plants, progeny arising from seeds in the large, fully mature ‘early’ fruits produced significantly more, and larger leaves by day 30 than did those from late fruits (suggesting differential provisioning in seeds during development). Male flower production had a highly significant positive correlation with vegetative mass and a significant negative correlation with fruit production.