Fig trees (Ficus) are often ecologically significant keystone species because they sustain populations of the many seed-dispersing animals that feed on their fruits. They are prominent components of riparian zones where they may also contribute to bank stability as well as supporting associated animals. The diversity and distributions of riparian fig trees in deciduous and evergreen forests in Chiang Mai Province, Northern Thailand were investigated in 2010–2012. To record the diversity and abundance of riparian fig trees, we (1) calculated stem density, species richness, and diversity indices in 20×50 m randomly selected quadrats along four streams and (2) measured the distances of individual trees from four streams to determine if species exhibit distinct distribution patterns within riparian zones. A total of 1169 individuals (from c. 4 ha) were recorded in the quadrats, representing 33 Ficus species (13 monoecious and 20 dioecious) from six sub-genera and about 70% of all the species recorded from northern Thailand. All 33 species had at least some stems in close proximity to the streams, but they varied in their typical proximity, with F. squamosa Roxb. and F. ischnopoda Miq the most strictly stream-side species. The riparian forests in Northern Thailand support a rich diversity and high density of Ficus species and our results emphasise the importance of fig tree within the broader priorities of riparian area conservation. Plans to maintain or restore properly functioning riparian forests need to take into account their significance.

The phenology of plants reflects selection generated by seasonal climatic factors and interactions with other plants and animals, within constraints imposed by their phylogenetic history. Fig trees (Ficus) need to produce figs year-round to support their short-lived fig wasp pollinators, but this requirement is partially de-coupled in dioecious species, where female trees only develop seeds, not pollinator offspring. This allows female trees to concentrate seed production at more favorable times of the year. Ficus squamosa is a riparian species whose dispersal is mainly by water, rather than animals. Seeds can float and travel in long distances. We recorded the leaf and reproductive phenology of 174 individuals for three years in Chiang Mai, Northern Thailand. New leaves were produced throughout the year. Fig production occurred year-round, but with large seasonal variations that correlated with temperature and rainfall. Female and male trees initiated maximal fig crops at different times, with production in female trees confined mainly to the rainy season and male figs concentrating fig production in the preceding months, but also often bearing figs continually. Ficus squamosa concentrates seed production by female plants at times when water levels are high, favouring dispersal by water, and asynchronous flowering within male trees allow fig wasps to cycle there, providing them with potential benefits by maintaining pollinators for times when female figs become available to pollinate.

Seed size is a key trait for understanding and predicting ecological processes in a plant community. In a tropical forest, trees and lianas are major components driving ecosystem function and biogeochemical processes. However, seed ecological research on both components remains limited, particularly phylogenetic patterns and relationships with other traits. Here, we compiled a unique dataset of seed size (seed mass and geometrical size metrics) based on collections of more than 5,200 seeds of 196 woody plant species, covering >98 and 70% of tree and liana stems, respectively, located on a 30-ha plot in a tropical evergreen forest in central Thailand. We aimed to (1) develop allometric equations among seed size metrics to predict seed mass; (2) examine phylogenetic influence on seed size variation; and (3) examine relationships among seed traits and several other functional plant traits. Our allometric equations relating seed mass, seed volume, and width were well-fitted with data (R2= 0.94, 0.87 respectively). A phylogenetic signal test found that seed size was randomly distributed across the phylogeny. To study the functional trait relationships, we separately tested seed size data of the tree and liana communities (146 and 50 species, respectively), against mean body size of frugivores, successional niches, leaf, and structural traits. For the tree community, seed size was significantly related to mean body size of frugivores, which we believe is a basic driver of seed size because it is related to the gape width affecting dispersal effectiveness. Nearly all leaf traits were significantly positively correlated with seed size (p < 0.03). The significant positive correlation of leaf area and greenness suggested the high-energy demand of large-seeded species. We found a strong positive correlation between seed size and leaf toughness, suggesting a coordination between seed size and leaf defense. However, all these patterns disappeared in the same analysis applied to the liana community. Liana seed size variation was lower than that of trees, perhaps because lianas grow in relatively more uniform conditions in the forest canopy. Frugivore size was the strongest driver of seed size variation. Our study shows a surprising contrast between trees and lianas that is worth further investigation.