Towards monitoring biodiversity in Amazonian forests: how regular samples capture meso-scale altitudinal variation in 25 km2 plots.

<h4>Background</h4>Ecological monitoring and sampling optima are context and location specific. Novel applications (e.g. biodiversity monitoring for environmental service payments) call for renewed efforts to establish reliable and robust monitoring in biodiversity rich areas. As there i...

Full description

Saved in:
Bibliographic Details
Main Authors: Darren Norris, Marie-Josée Fortin, William E Magnusson
Format: Article
Language:English
Published: Public Library of Science (PLoS) 2014-01-01
Series:PLoS ONE
Online Access:https://doi.org/10.1371/journal.pone.0106150
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850127102589796352
author Darren Norris
Marie-Josée Fortin
William E Magnusson
author_facet Darren Norris
Marie-Josée Fortin
William E Magnusson
author_sort Darren Norris
collection DOAJ
description <h4>Background</h4>Ecological monitoring and sampling optima are context and location specific. Novel applications (e.g. biodiversity monitoring for environmental service payments) call for renewed efforts to establish reliable and robust monitoring in biodiversity rich areas. As there is little information on the distribution of biodiversity across the Amazon basin, we used altitude as a proxy for biological variables to test whether meso-scale variation can be adequately represented by different sample sizes in a standardized, regular-coverage sampling arrangement.<h4>Methodology/principal findings</h4>We used Shuttle-Radar-Topography-Mission digital elevation values to evaluate if the regular sampling arrangement in standard RAPELD (rapid assessments ("RAP") over the long-term (LTER ["PELD" in Portuguese])) grids captured patters in meso-scale spatial variation. The adequacy of different sample sizes (n = 4 to 120) were examined within 32,325 km2/3,232,500 ha (1293×25 km2 sample areas) distributed across the legal Brazilian Amazon. Kolmogorov-Smirnov-tests, correlation and root-mean-square-error were used to measure sample representativeness, similarity and accuracy respectively. Trends and thresholds of these responses in relation to sample size and standard-deviation were modeled using Generalized-Additive-Models and conditional-inference-trees respectively. We found that a regular arrangement of 30 samples captured the distribution of altitude values within these areas. Sample size was more important than sample standard deviation for representativeness and similarity. In contrast, accuracy was more strongly influenced by sample standard deviation. Additionally, analysis of spatially interpolated data showed that spatial patterns in altitude were also recovered within areas using a regular arrangement of 30 samples.<h4>Conclusions/significance</h4>Our findings show that the logistically feasible sample used in the RAPELD system successfully recovers meso-scale altitudinal patterns. This suggests that the sample size and regular arrangement may also be generally appropriate for quantifying spatial patterns in biodiversity at similar scales across at least 90% (≈5 million km2) of the Brazilian Amazon.
format Article
id doaj-art-6217ad16aa514b618459db4f238d1afe
institution OA Journals
issn 1932-6203
language English
publishDate 2014-01-01
publisher Public Library of Science (PLoS)
record_format Article
series PLoS ONE
spelling doaj-art-6217ad16aa514b618459db4f238d1afe2025-08-20T02:33:44ZengPublic Library of Science (PLoS)PLoS ONE1932-62032014-01-0198e10615010.1371/journal.pone.0106150Towards monitoring biodiversity in Amazonian forests: how regular samples capture meso-scale altitudinal variation in 25 km2 plots.Darren NorrisMarie-Josée FortinWilliam E Magnusson<h4>Background</h4>Ecological monitoring and sampling optima are context and location specific. Novel applications (e.g. biodiversity monitoring for environmental service payments) call for renewed efforts to establish reliable and robust monitoring in biodiversity rich areas. As there is little information on the distribution of biodiversity across the Amazon basin, we used altitude as a proxy for biological variables to test whether meso-scale variation can be adequately represented by different sample sizes in a standardized, regular-coverage sampling arrangement.<h4>Methodology/principal findings</h4>We used Shuttle-Radar-Topography-Mission digital elevation values to evaluate if the regular sampling arrangement in standard RAPELD (rapid assessments ("RAP") over the long-term (LTER ["PELD" in Portuguese])) grids captured patters in meso-scale spatial variation. The adequacy of different sample sizes (n = 4 to 120) were examined within 32,325 km2/3,232,500 ha (1293×25 km2 sample areas) distributed across the legal Brazilian Amazon. Kolmogorov-Smirnov-tests, correlation and root-mean-square-error were used to measure sample representativeness, similarity and accuracy respectively. Trends and thresholds of these responses in relation to sample size and standard-deviation were modeled using Generalized-Additive-Models and conditional-inference-trees respectively. We found that a regular arrangement of 30 samples captured the distribution of altitude values within these areas. Sample size was more important than sample standard deviation for representativeness and similarity. In contrast, accuracy was more strongly influenced by sample standard deviation. Additionally, analysis of spatially interpolated data showed that spatial patterns in altitude were also recovered within areas using a regular arrangement of 30 samples.<h4>Conclusions/significance</h4>Our findings show that the logistically feasible sample used in the RAPELD system successfully recovers meso-scale altitudinal patterns. This suggests that the sample size and regular arrangement may also be generally appropriate for quantifying spatial patterns in biodiversity at similar scales across at least 90% (≈5 million km2) of the Brazilian Amazon.https://doi.org/10.1371/journal.pone.0106150
spellingShingle Darren Norris
Marie-Josée Fortin
William E Magnusson
Towards monitoring biodiversity in Amazonian forests: how regular samples capture meso-scale altitudinal variation in 25 km2 plots.
PLoS ONE
title Towards monitoring biodiversity in Amazonian forests: how regular samples capture meso-scale altitudinal variation in 25 km2 plots.
title_full Towards monitoring biodiversity in Amazonian forests: how regular samples capture meso-scale altitudinal variation in 25 km2 plots.
title_fullStr Towards monitoring biodiversity in Amazonian forests: how regular samples capture meso-scale altitudinal variation in 25 km2 plots.
title_full_unstemmed Towards monitoring biodiversity in Amazonian forests: how regular samples capture meso-scale altitudinal variation in 25 km2 plots.
title_short Towards monitoring biodiversity in Amazonian forests: how regular samples capture meso-scale altitudinal variation in 25 km2 plots.
title_sort towards monitoring biodiversity in amazonian forests how regular samples capture meso scale altitudinal variation in 25 km2 plots
url https://doi.org/10.1371/journal.pone.0106150
work_keys_str_mv AT darrennorris towardsmonitoringbiodiversityinamazonianforestshowregularsamplescapturemesoscalealtitudinalvariationin25km2plots
AT mariejoseefortin towardsmonitoringbiodiversityinamazonianforestshowregularsamplescapturemesoscalealtitudinalvariationin25km2plots
AT williamemagnusson towardsmonitoringbiodiversityinamazonianforestshowregularsamplescapturemesoscalealtitudinalvariationin25km2plots