Abstract
BackgroundEcological 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.Methodology/Principal FindingsWe 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.Conclusions/SignificanceOur 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.
Highlights
Gathering sample data that reliably reflects the underlying spatial and temporal variability in target measurements is a fundamental requisite for ecological studies
Our sample of seven active research areas (SD ranging from 4.8 to 40.2, Table S1) represents the meso-scale altitudinal heterogeneity found across an area of approximately 5 million km2
The Kolmogorov-Smirnov (KS) tests showed that a regular arrangement of samples could represent the distribution of altitude values within our study areas (Figure 2, Figure S7, Figure S8)
Summary
Gathering sample data that reliably reflects the underlying spatial and temporal variability in target measurements is a fundamental requisite for ecological studies. Bias (sampling bias, estimation bias, etc.) occurs when the sampling design used induces errors and artificial differences in the values among samples [1,2] These first principles form the basis of many an undergraduate statistics course but beyond lecture halls obtaining reliable and representative sample data is one of the primary challenges for any biodiversity monitoring program [3,4], where the necessary requirement is that it is sensible and meaningful to compare and contrast sample values [1,3,4]. A particular challenge to establishing payments for biodiversity services is that reliable and robust long-term monitoring of biodiversity indicators is required to ensure that services are and will continue to be provided Such long-term monitoring of biodiversity requires decisions today about the sampling distribution to recover information about changes in response to future, and often unknown, threats. 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
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
