Water relations, phenology and drought adaptation of understorey trees in tropical lowland rain forest.

Abstract

1. The ecology of common understorey species associated with ridges was compared with species found on lower-slopes and those species occurring ubiquitously in two 4-ha plots in lowland rain forest at Danum, Sabah, East Malaysia (4" 58’ N, 117’ 46’ E) over 3 years. 2. During the study period one dry period occurred (psychrometer-measured ridge soil water potential, 20 cm depth -0.67 MPa), but other, more severe, dry periods have occurred since records began in 1985 (estimated ridge water potential -1.21 MPa, March 1992). Lower soil water potentials occurred on ridges which had up to 0.22 MPa lower water potentials than lower-slopes (estimated difference March 1992, 0.40 MPa). 3. At dry times, Dimorphocatyx muricatus (ridge species) had higher pre-dawn (-0.21 v. -0.57 MPa; all quoted differences are significant at p<0.05) and mid-day (-0.59 v. -1.77 MPa) leaf w'ater potentials than Mallotus wrayi (ubiquitous). Leaf osmotic potentials of D. muricatus were higher (-1.11 v. - 1.58 MPa), and both species osmotically adjusted between wet and dry times. D. muricatus trees were more deeply rooted (mean root depth, 97.4 cm) than M. wrayi trees (69.8 cm). M. wrayi seedlings on ridges had lower assimilation rates than on lower-slopes (1.9 v. 3.4 pmol CO2 m " s '). 4. Leaf production of Arciisia co/orata (ubiquitous), Cleistanthus glaber (ridge), D. muricatus and A/. wrayi trees varied from year to year. C. glaber, D. muricatus, and A/, wrayi all had peaks in leaf and flower production associated with sunnier drier spells, A. colorata did not. A. colorata leaves were longer lived (5.7 years, estimated from leaf turn-over) than D. muricatus (2.7 years) and M. wrayi (2.2 years). There were no size or site differences for leaf life-span or leaf production. More large trees flowered than small trees. More trees flowered on ridges than lower-slopes but more trees set fruit on lower-slopes. 5. D. muricatus seedlings grew equally well on ridges and lower-slopes under different water regimes. A/, wrayi and D. muricatus cutting mortality was higher than D. muricatus seedling mortality for all treatments. M. wrayi cutting mortality was highest in lower-slope drought plots. Cutting mortality was higher in logged forest than primary forest. 6. A/, wrayi seed germination rates (1 %) were much lower than Baccaurea stipulata (lower-slope) (65 %) and D. muricatus (53 %). Germination rates were higher on ridge than lower-slope sites. D. muricatus nursery germination rates (32 %) were lower than in the field. B. stipulata seed took longer to germinate (35 days) than D. muricatus (21 days). 7. Under nursery conditions A/, wrayi wildings grown under 2.7 % daylight grew taller and increased leaf area, specific leaf area (SLA) and fine and coarse root weight than under 1.2 % daylight. Addition of phosphate had no effect on growth 8. Un-watered D. muricatus seedlings took significantly longer to wilt (36 v. 16 days) and die (46 v. 29 days) than B. stipulata seedlings. 9. The results suggest that droughts may have an important effect on understorey ridge community species composition at Danum.