Study ID 329

Twenty years tree demography in an undisturbed Dipterocarp permanent sample plot at Uppangala Western Ghats of India

Download dataset

Realm: Terrestrial
Climate: Tropical
Biome: Tropical and subtropical moist broadleaf forests
Central latitude: 12.537500
Central longitude: 75.662778
Duration: 6 years, from 1990 to 2010

2397 records

102 distinct species

Across the time series Diospyros sylvatica is the most frequently occurring species


Pélissier, R., Pascal, J.-P., Ayyappan, N., Ramesh, B. R., Aravajy, S. & Ramalingam, S. R. (2011) Tree demography in an undisturbed Dipterocarp permanent sample plot at Uppangala, Western Ghats of India. Ecology, 92, 1376.


We report a data set on demography of trees monitored over 20 years in Uppangala permanent sample plot (UPSP) in undisturbed old-growth wet evergreen Dipterocarp forest located within the Pushpagiri Wildlife Sanctuary in Indias Western Ghats biodiversity hotspot. During 1989–1990 all trees ? 30 cm girth at breast height (gbh) were sampled in five north–south transects 20 m wide and 180 to 370 m long covering a total area of 3.12 ha. In 1992–1993 additional rectangular plots were established bringing the area sampled to 5.07 ha in total. In all 3870 trees were identified tagged mapped and provided with permanent dendrometer bands. Since then the sampled area has been regularly censused at 3–5 year intervals recording tree recruitment mortality and growth. We present data from censuses conducted in 1990–1993 1994 1997–1998 2001–2002 2007 and 2010. These data have been used to study the natural forest dynamics and to calibrate spatially explicit simulation models. ____ Sampling design consists of five north–south oriented transects viz. A B C D and E each 20 m wide 180 to 370 m long and 100 m apart center to center. The 5 transects were installed in 1989–1990 and collectively represent a 3.12-ha systematic sample of the forest compartment whose main dendrometric and floristic characteristics are given in Pascal and Pélissier (1996). Subsequently additional rectangular sampling plots viz. H R and S which overlap the transects and represent an additional area of 1.95 ha were established in 1990–1993 to study the dynamics of the forest in particular topographic situations (Pélissier 1997 1998). A piece of R source code is given in section III-V-D below to extract data according to various sampling configurations. Table 1 gives a summary of plot size and location. Because of the uneven site topography (see B.1.Watersheds and hydrology) each plot was designed as a grid of 10 × 10 m contiguous elementary quadrats adjusted to account for local slope angle. The purpose of the correction was to ensure that each side of a quadrat was approximately 10 m flat so that it contains in total a 100 m2 area in planar projection. The slope angle (alpha) was measured along each quadrat sides which length parallel to the ground was adjusted to l = 10/cos(alpha). In order to accurately locate the trees in (xy) co-ordinates within each quadrat they were temporarily delineated with a plastic rope. All trees from 30 cm girth at breast height (gbh) were inventoried measured tagged mapped and identified to species. The point of girth measurement of each tree (pom) was marked with paint in the initial inventory and then fitted with a permanent dendrometer band. For mapping the trees within each quadrat the flat distance from the centre of the trunk base to the nearest quadrat boundary was measured in two perpendicular directions parallel to the rope. Each quadrat was mapped in the field at a scale of 1:100 with indication of tree locations main rocks fallen trees stream courses trails etc. These maps are stored at IFP Botany Lab. Instrumentation: Slope angles were measured with a 1-degree precision using a Suunto hand-held clinometer. Distance measurements were taken with a 1-cm precision thanks to a 20 m fiberglass tape. Initial tree girths were measured at 1.30 m from the ground or above the buttresses with a precision of 0.1 cm using a 3 or 5 m flexible steel tape. A calibrated 1.30 m stick was used to determine pom height. The dendrometer bands were prepared by technical IFP staff using stainless steel bands adjusted to tree girth and equipped with a 1:50 vernier (Hall 1944; Fig. 5) that allows girth to be measured with a theoretical precision of 0.02 cm. The dendrometers were then fitted at pom on the trees and vernier readings taken as initial gbh measurement. Recensuses. Recensuses have been conducted every 3–5 years in all plots since May 1990 (census 0) in May 1994 (census 1) November 1997–April 1998 (census 2) December 2001–May 2002 (census 3) April 2007 (census 4) and April 2010 (census 5). Time intervals between censuses are given in Julian day calendar since day 0 on 1st March 1990. However some plots (H extension R and S) installed in March 1992 and November 1993 are included with a census 0 date on day 762 and 1341 respectively. Similarly censuses 2 and 3 which were completed over several field trips refer to more than one Julian date (see Table 1). At each recensus all plots were screened for tree mortality tree growth by girth increments of alive trees reading from dendrometer bands and tree recruitment i.e. those trees attained the girth threshold of 30 cm gbh since the previous census. The recruited trees were tagged mapped identified and installed with a dendrometer band following the initial census protocol (see II.B.3. Research methods). Summary data of plots demography are given in Tables 3 (all plots pooled together) and 4 (tansects A to E pooled together). These summaries have been computed from the raw data file using a piece of R code provided in section III-V-D below. Unit of abundance = IndCountInt, Unit of biomass = NA