Phenological periodicity in a tropical peat-swamp rainforest on Borneo, Indonesia

Griëtte van der Heide and Daniel J. Naumenko, Department of Anthropology, University of Colorado Boulder, Boulder, CO, Daniel J. Naumenko, Institute of Behavioral Science, University of Colorado Boulder, Boulder, CO, Sri Suci Utami Atmoko, Fakultas Biologi, Universitas Nasional, Jakarta, Indonesia, Erin R. Vogel, Department of Anthropology, Rutgers, The State University of New Jersey, New Brunswick, NJ, Erin R. Vogel, Center for Human Evolutionary Studies, Rutgers, The State University of New Jersey, New Brunswick, NJ

Background/Question/Methods
Peat swamp forests are endangered habitats in Southeast Asia that harbor diverse primate communities. We investigated tree species’ phenological traits to increase understanding of tropical peat-swamp forest phenology. This is important as many primates exhibit behavioral and physiological responses to fluctuations in phenology, and it is often (erroneously) assumed that tree species have regular phenological cycles. We asked whether species (1) exhibit cycles, and (2) whether these are sub-annual (<11 mo), annual (11-13 mo), or supra-annual (>13 mo).
Our long-term phenological data (April 2003 - June 2019) originate from a tropical, semi-alluvial peat-swamp forest at the Tuanan Research Station Area, Central Kalimantan, Indonesia. We monitored trees (dbh ≥ 10 cm, n = 3833) along six transects (~1 km x 10 m) monthly for fruits, flowers, and young leaves. For preliminary analyses, we used 10 tree species with the highest species’ basal areas. We included the longest observation span (≥5 years) of trees which experienced at least one phenological event. To assess phenological periodicity (cycles), we applied a modified Fourier transform analysis that identifies the dominant peak (‘peak’; largest power) and associated cycle type. Significant dominant cycles (‘sigDC’) are those for which a data-generated null model falls below the peak’s 95% confidence intervals.
Results/Conclusions
For all species, the vast majority of trees had non-significant peaks (leaves: 84-98%, flowers: 82-96%, fruits: 93-98%). Species with ≥5 sigDCs showed sub-annual leafing (6/7) and flowering (3/6), but supra-annual fruiting (2/3) with median lengths of 3.4-7.8, 3.9-7.1, and 14.6-29.6 months, respectively. Only Acronychia pedunculata trees had a large number of annual sigDCs (leafing: 16/27, flowering: 46/60). These results parallel those from all peaks. Leafing (55-88%) and flowering (19-67%) peaks were mostly sub-annual, while supra-annual fruiting peaks (20-43%) were more common than sub-annual peaks (10-29%). Non-cyclicity for fruiting was very high (31-55%). Annual peaks were rare (fruiting: 2-7%, flowering: 5-23%, leafing: 0-9%, A. pedunculata: 34%).
Preliminary results indicate that tree species of the peat-swamp forest at Tuanan experience few regular phenological cycles, and may be remarkably acyclic. Prolonged data collection is unlikely to reveal annual cycles, since annual sigDCs and peaks were rare. The question remains whether the overall low number of sigDCs for leafing and flowering is an indication of continuity, punctuated irregularity, or low detectability. Although some species may fruit supra-annually, a typical dipterocarp trait, the prevalence of non-cyclicity begs the question whether trees simply fruit sporadically.