Phenological field observations

Phenological field observations are time-consuming. To advance understanding of the controls of tree phenology, they should be well planned and executed. Here are some suggestions:

1. Clearly identify the goal of the study. Access to a stand of tropical trees alone does not justify the time investment in a phenological study. Observed trees should be representative of an ecosystem or have other unique characteristics (Fig. 1).
2. Phenology deals with plant development. Analysis of the proximate causes of tree phenology requires knowledge of organ development of the observed species and of basic plant eco-physiology.
3. Before starting field observations, examine herbarium collections of the species to be monitored (see section 3. below). They provide useful information on the timing of shoot growth and flowering (7, 21).
4. Monitor both vegetative and reproductive phenology. Flowering phenology is a function of vegetative phenology (Fig. 1). Observations of flowering and fruiting alone are therefore of little value.
5. Observe at least 5-8 conspecific trees biweekly for two growing seasons, more frequently during periods of rapid developmental change (bud break, leaf flushing, flowering; 3, 19, 21, 22).
6. In landscapes with distinct topographic variation, observe trees along the gradient from moist to dry microsites (4).
7. Measure rainfall at least weekly. Because of the patchiness of rainfall, meteorological data from a distant weather station or long-term averages of rainfall in the research area are of little value (12).

Identifying the controls of tropical tree phenology.

1. Establish correlations between phenology and seasonal variation in rainfall and photoperiod (Fig. 1). Statistical analyses of correlations between phenology and general meteorological data do not reveal the proximate causes of the observed phenology (23).
2. Analyze natural experiments revealing the effect of gradual changes in environmental controls. For example, phenology may differ widely between trees at moist lowland sites and dry upland sites (4, 11), it changes gradually along altitudinal gradients in seasonal water stress (3, 13), and responses to photoperiod vary predictably with latitude (11, 21).
3. The Missouri Botanical Garden Herbarium (St. Louis, Mo.) has millions of flowering herbarium specimens of neotropical tree species. Analyses of the collection dates revealed the flowering periodicity of many neotropical tree species (7, 21). The herbarium's new "Tropicos" data base is accessible via the internet and may permit on-line identification of flowering periodicity based on the seasonal variation in collection dates.