eprintid: 11238
rev_number: 9
eprint_status: archive
userid: 1305
dir: disk0/00/01/12/38
datestamp: 2019-08-13 06:30:50
lastmod: 2019-08-13 06:31:07
status_changed: 2019-08-13 06:30:50
type: article
metadata_visibility: show
contact_email: Library-ICRISAT@CGIAR.ORG
creators_name: Chitale, V S
creators_name: Behera, M D
creators_name: Roy, P S
icrisatcreators_name: Roy, P S
affiliation: Centre for Oceans, Rivers, Atmosphere and Land Sciences, Indian Institute of Technology (Kharagpur)
affiliation: International Centre for Integrated Mountain Development (ICIMOD) (Kathmandu)
affiliation: School of Water Resources, Indian Institute of Technology (Kharagpur)
affiliation: ICRISAT (Patancheru)
country: India
country: Nepal
title: Deciphering plant richness using satellite remote sensing: a study from three biodiversity hotspots
ispublished: pub
subjects: Bio1
subjects: RS1
divisions: CRPS4
full_text_status: restricted
keywords: Satellite data, Generalized linear models, India, Tropical forests, Biodiversity, hotspots, satellite remote sensing
note: The authors are thankful to Dr. PS Roy, Project Director, Biodiversity Characterization
project for providing the field sampling data for this study.
abstract: The ‘spectral variation hypothesis (SVH)’ assumes spectral variability as a result of variation in species richness. In the present study, we explore the potential of satellite datasets in identifying the patterns in species richness in part of three global biodiversity hotspots falling in India viz., Himalaya, Indo-Burma, and Western Ghats. We used generalized linear models to correlate remote sensing based vegetation indices (VIs) and physiographic indices (PIs) with plant richness calculated using 1264, 1114, and 1004 field plots across 21 different forest vegetation classes in Himalaya, Indo-Burma, and Western Ghats respectively. Three different vegetation indices ranked highest in explaining the variance in plant richness in the three hotspots. The variance in species richness explained by models based on only VIs was highest (69%, P < 0.01) for Bamboo vegetation in Indo Burma hotspot with Normalized Difference Vegetation Index, followed by that for dry deciduous forest in Western Ghats (57%, P < 0.001) with Normalized Difference Water Index, and for grasslands (54%, P < 0.05) in Himalaya by Modified Soil Adjusted Vegetation Index. The explained variance increased with combined models that are based on PIs and VIs to up to 85% (P < 0.05). Overall, we observed very high correlation between VIs and plant richness in open canopy vegetation classes with low species richness such as grasslands, scrubs, and dry deciduous forests, followed by vegetation classes with moderately dense canopy. Our study provides crucial insights on utility of satellite datasets as a proxy for estimating plant richness for better conservation of diverse ecosystems.
date: 2019-07
date_type: published
publication: Biodiversity and Conservation (TSI)
volume: 28
number: 8-9
publisher: Springer
pagerange: 2183-2196
id_number: 10.1007/s10531-019-01761-4
refereed: TRUE
issn: 0960-3115
official_url: https://doi.org/10.1007/s10531-019-01761-4
related_url_url: https://scholar.google.co.in/scholar?hl=en&as_sdt=0%2C5&q=Deciphering+plant+richness+using+satellite+remote+sensing%3A+a+study+from+three+biodiversity+hotspots&btnG=
related_url_type: pub
citation:   Chitale, V S and Behera, M D and Roy, P S  (2019) Deciphering plant richness using satellite remote sensing: a study from three biodiversity hotspots.  Biodiversity and Conservation (TSI), 28 (8-9).  pp. 2183-2196.  ISSN 0960-3115     
document_url: http://oar.icrisat.org/11238/1/Deciphering%20plant%20richness%20using%20satellite%20remote%20sensing.pdf