Phenological Variation in the Neotropical Understory Shrub Piper Arielanum: Causes and Consequences

This study examines the flowering, fruiting, and leafing phenology of a Costa Rican understory shrub, Piper arieianum, with an emphasis on the complexity and consequences of incomplete synchrony of reproductive events. During 1981 and 1982, flowering in P. arieianum was a mixture of (1) highly synchronized peaks in which most plants flowered and most of their inflorescences flowered and (2) more extended flowering during which few plants flowered and few of their inflorescences flowered. The probability of fruit set was greater for large plants and for plants that flowered in synchrony with other plants than for small plants and plants that did not flower in synchrony with the rest of the population. The species is mostly self—incompatible, a possible explanation for the observed decrease in fruit set when fewer plants were flowering. However, synchronization per se did not necessarily lead to high fruit set, suggesting that certain times of the year are better for flowering than others. In general, flowering was more prevalent in February and from September to October, when rainfall is predictably the lowest and pollinator services are more reliable. The primary flowering peak (February) accounted for 64% of total inflorescence and 62% of total infructescence production in 1981, and 50% of total inflorescence and 76% of total infructescence production in 1982. Fruit and seed dispersal were much less synchronous than flowering; fruits of individual plants from a single flowering period were dispersed over a 2—4 mo period. The time from flowering to final infructescence maturation was 5—9 mo. Large fruit crops and fruit set more than once per year resulted in a longer time period over which infructescences were presented to dispersers than did small fruit crops and fruit set only once per year. Flowering time was delayed by experimental leaf area removal, but the type and strength of the effect depended on the time of year. Before the small September peak of flowering (1979), 10% removal of leaf area delayed peak flowering 1 mo compared with controls and 30% removal of leaf area delayed flowering for 1.5—2 mo. In contrast, when leaf area was removed before the large February flowering peak (1981), plants with 10, 30, or 50% leaf area removal flowered simultaneously with control plants, although complete defoliation delayed flowering up to 3 mo. Similarly, although plants in light gaps flowered more frequently and more successfully during nonpeak periods than did understory plants, the level of flowering and fruit set was equally high in plants of both locations during the February flowering peak. Thus, the reproductive events occurring during the primary flowering peak (February) were less sensitive to environmental variables than at other times of the year. Leaf production was continuous throughout the study period. Previous leaf production did not affect subsequent flowering and fruiting intensity of individual plants. Predispersal seed predators were abundant following the major flowering peaks but not the minor flowering peaks, suggesting that off—peak fruit production can lead to escape from seed predators.