Water sources used by Didymopanax pittieri at different life stages in a tropical cloud forest

Ecology, June, 1998 by Taylor S. Feild, Todd E. Dawson

INTRODUCTION

Insight into the sources of water transpired by plants has come from analyzing the stable hydrogen isotopic composition ([Delta]D) of xylem water in conjunction with the [Delta]D of available water sources (Ehleringer and Dawson 1992). The [Delta]D technique provides a powerful water "tracing" method because water sources (rain, soil water, etc.) can often differ in their hydrogen isotope composition and water passing through root tissues and into plants is not fractionated during uptake (Wershaw et al. 1966, Dawson and Ehleringer 1991, Thornburn et al. 1993, but see Lin and Sternberg 1993). Studies that have employed the use of [Delta]D as water-source tracer have focused on plant communitywide variation in water-source utilization between species in relation to competition and seasonal or diurnal changes in [Delta]D within the soil surface water due to either natural gradients, seasonal changes, or hydraulic lift (Ehleringer et al. 1991, Dawson 1993a, Thornburn and Walker 1993, Jackson et al. 1995). Few data exist on the variation in source-water utilization by a single species, particularly a species that has several unique life stages, each with the potential to utilize different sources of water.

Primary hemiepiphytes commonly undergo a tri-phasic development in growth form, from an epiphyte to a terrestrial tree when reproductive maturity is eventually achieved (Williams-Linera and Lawton 1995, Holbrook and Putz 1996). The water sources used for growth and development potentially change as these hemiepiphytes make the transition from a purely epiphytic life-style to an arborescent one when the root connection to the ground is first established and access to terrestrial water and nutrient supplies first occurs. In this study, we utilized stable hydrogen isotopic analysis of plant xylem water and water sources to determine if there were changes in water source among the three growth phases of the hemiepiphyte Didymopanax pittieri Marchal (Araliaceae) in a tropical montane cloud forest of Costa Rica. We predicted that epiphytes would exclusively use only occult precipitation (fog, cloud, and mist) and recent rainfall and thus possess isotopically distinct xylem water relative to hemiepiphytes and terrestrial trees that have both recent and older (deep) soil water sources available to them. In addition, we predicted that hemiepiphytic individuals should exhibit xylem [Delta]D "signatures" intermediate between cloud and terrestrial soil water signatures because hemiepiphytes allocate root biomass to both canopy and terrestrial soil microsites.

STUDY SITE AND METHODS

The study site was located in the Monteverde Cloud Forest Reserve (10 [degrees] 12[prime] N, 84 [degrees] 42[prime] W) situated at 1500 m in a wind-exposed "elfin" montane rain forest (Lawton and Dryer 1980). Average annual precipitation is [approximately] 3000 mm/yr; cloud moisture contributes a large but unknown amount to the water balance of this forest. Didymopanax pittieri is a shade-intolerant canopy tree from lower montane rain forest life zones of Costa Rica and western Panama (Holridge et al. 1971). D. pittieri begins life as an epiphyte in either tree canopies or on fallen nurse logs in well-lit gaps (Williams-Linera and Lawton 1995). Epiphytes frequently grow to sapling stature ([greater than] 1 m) within the canopies of their hosts. Eventual connection with the ground in D. pittieri commonly occurs when roots grow down from the canopy, though some individuals make a terrestrial connection only after the collapse of their host tree. In large gaps, hemiepiphytic saplings may grow to large-sized trees (50 cm in diameter at breast height and 10 m tall; Williams-Linera and Lawton 1995).

To determine the sources of water used by D. pittieri, all of the available water sources (cloud water, canopy soil water, rainfall, and soil water at several depths) were each sampled once (N = 1 for each water source) and their stable isotopic composition (signature) determined. Water sources were collected over a 2-d period in July 1994 in air-tight plastic scintillation vials and sealed with Par film to prevent isotopic fractionation due to evaporation. The samples were frozen until analyzed. Soil water samples were obtained by digging pits in the forest floor. Soil obtained from depths of 15 cm, 35 cm, and 55 cm was collected, and from these samples water was extracted for isotopic analysis (see below). Cloud moisture was collected using a single cotton trap placed in the direction of prevailing wind carrying mist for 1 min and then also sealed in a plastic vial. The canopy soil water was sampled by removal of debris, by hand, that collected behind the tangle of D. pittieri roots in the canopy host tree and placed into a vial and sealed quickly. Finally, rainwater was collected in a vial and sealed during a rainstorm one day previous to collection of the other water sources.

Stem water samples were taken from 10 epiphytic, 9 hemiepiphytic, and 9 trees of D. pittieri. Hemiepiphytes were distinguished from epiphytes by removing canopy mats on canopy tree trunks by hand and searching for roots from a particular plant that were connected to both the forest floor and canopy mat. Stem portions of all plants were excised and placed in air-tight plastic vials as before. Care was taken to sample only lignified and mature tissues (e.g., green tissue portions were always avoided because they are isotopically enriched relative to the source water) (Dawson and Ehleringer 1993). Each individual plant was sampled once.