Integrating rapid assessment, variable probability sampling, and machine learning to improve accuracy and consistency in mapping local spatial distribution of plant species richness

Bo Hao Perng; Tzeng Yih Lam; Sheng Hsin Su; Mohamad Danial Bin Md Sabri; David Burslem; Dairon Cardenas; Álvaro Duque; Sisira Ediriweera; Nimal Gunatilleke; Vojtech Novotny; Michael J. O’Brien; Glen Reynolds

doi:10.1093/forestry/cpad041

Integrating rapid assessment, variable probability sampling, and machine learning to improve accuracy and consistency in mapping local spatial distribution of plant species richness

Bo Hao Perng, Tzeng Yih Lam^* (Corresponding Author), Sheng Hsin Su, Mohamad Danial Bin Md Sabri, David Burslem, Dairon Cardenas, Álvaro Duque, Sisira Ediriweera, Nimal Gunatilleke, Vojtech Novotny, Michael J. O’Brien, Glen Reynolds

^*Corresponding author for this work

Aberdeen Centre For Environmental Sustainability

Research output: Contribution to journal › Article › peer-review

Abstract

Conserving plant diversity is integral to sustainable forest management. This study aims at diversifying tools to map spatial distribution of species richness. We develop a sampling strategy of using rapid assessments by local communities to gather prior information on species richness distribution to drive census cell selection by sampling with covariate designs. An artificial neural network model is built to predict the spatial patterns. Accuracy and consistency of rapid assessment factors, sample selection methods, and sampling intensity of census cells were tested in a simulation study with seven 25–50-ha census plots in the tropics and subtropics. Results showed that identifying more plant individuals in a rapid assessment improved accuracy and consistency, while transect was comparable to or slightly better than nearest-neighbor assessment, but knowing more species had little effects. Results of sampling with covariate designs depended on covariates. The covariate I_freq, inverse of the frequency of the rapidly assessed species richness strata, was the best choice. List sampling and local pivotal method with I_freq increased accuracy by 0.7%–1.6% and consistency by 7.6%–12.0% for 5% to 20% sampling intensity. This study recommends a rapid assessment method of selecting 20 individuals at every 20-m interval along a transect. Knowing at least half of the species in a forest that are abundant is sufficient. Local pivotal method is recommended at 5% sampling intensity or less. This study presents a methodology to directly involve local communities in probability-based forest resource assessment to support decision-making in forest management.

Original language	English
Pages (from-to)	282-294
Number of pages	13
Journal	Forestry
Volume	97
Issue number	2
Early online date	14 Aug 2023
DOIs	https://doi.org/10.1093/forestry/cpad041
Publication status	Published - Apr 2024

Bibliographical note

We would like to thank three anonymous reviewers and the editors for their valuable comments that have significantly improved the manuscript. We also like to thank Dr George Weiblen from the University of Minnesota for assisting with the Wanang dataset. The 25-ha Long-Term Ecological Research Project of Amacayacu is a collaborative project of the Instituto Amazónico de Investigaciones Científicas Sinchi and the Universidad Nacional de Colombia Sede Medellín, in partnership with the Unidad de Manejo Especial de Parques Naturales Nacionales and the Forest Global Earth Observatory of the Smithsonian Tropical Research Institute (ForestGEO). The Amacayacu Forest Dynamics Plot is part of ForestGEO a global network of large-scale demographic tree plots. We acknowledge the Director and staff of the Amacayacu National Park for supporting and maintaining the project in this National Park. BCI: The BCI forest dynamics research project was made possible by National Science Foundation grants to Stephen P. Hubbell: DEB-0640386, DEB-0425651, DEB-0346488, DEB-0129874, DEB-00753102, DEB-9909347, DEB-9615226, DEB-9615226, DEB-9405933, DEB-9221033, DEB-9100058, DEB-8906869, DEB-8605042, DEB-8206992, DEB-7922197, support from the Forest Global Earth Observatory, the Smithsonian Tropical Research Institute, the John D. and Catherine T. MacArthur Foundation, the Mellon Foundation, the Small World Institute Fund, and numerous private individuals, and through the hard work of over 100 people from 10 countries over the past three decades. The plot project is part the Forest Global Earth Observatory (ForestGEO), a global network of large-scale demographic tree plots. Danum: The Danum plot is a core project of the Southeast Asia Rain Forest Research Partnership (SEARRP). We thank SEARRP partners, especially Yayasan Sabah for their support, and HSBC Malaysia and the University of Zurich for funding. We are grateful to the research assistants who are conducting the census, in particular the team leader Alex Karolus, and to Mike Bernados and Bill McDonald for species identifications. We thank Stuart Davies and Shameema Esufali for advice and training. Fushan: The establishment and first census of the Fushan 25-ha plot is a collaborative project by the Taiwan Forestry Research Institute (data provider), the Taiwan Forestry Bureau, the National Taiwan University (Institute of Ecology and Evolutionary Biology) and the ForestGEO (formerly the CTFS). We thank the agencies for providing the datasets and all field staff. S.H.S. also thanks Dr. I-Fang Sun for his long-term support for the Fushan project. Pasoh: Data from the Pasoh Research Forest was provided by the Forest Research Institute Malaysia-Forest Global Earth Observatory, Smithsonian Tropical Research Institute collaborative research project. Negeri Sembilan Forestry Department is the custodian of Pasoh Research Forest and I/we acknowledge the department for preserving the research forest. Sinharaja: The 25-ha Long-Term Ecological Research Project at Sinharaja World Heritage Site is a collaborative project of the Uva Wellassa University, University of Peradeniya, the Forest Global Earth Observatory (ForestGEO) of the Smithsonian Tropical Research Institute, with supplementary funding received from the John D. and Catherine T. Macarthur Foundation, the National Institute for Environmental Science, Japan, and the Helmholtz Centre for Environmental Research-UFZ, Germany, for past censuses. The PIs gratefully acknowledge the Forest Department, Uva Wellassa University, and the Post-Graduate Institute of Science at the University of Peradeniya, Sri Lanka for supporting this project, and the local field and laboratory staff who tirelessly contributed in the repeated censuses of this plot. Wanang: The 50-ha Wanang Forest Dynamics Plot is a collaborative project of the New Guinea Binatang Research Center, the Forest Global Earth Observatory (ForestGEO) of the Smithsonian Tropical Research Institute, the Forest Research Institute of Papua New Guinea, the Czech Academy of Sciences (19-28126X), and the University of Minnesota supported by NSF DEB-1027297 and NIH ICBG 5UO1TW006671. We acknowledge the government of Papua New Guinea and the customary landowners of Wanang for supporting and maintaining the plot.

Funding Information:
The funding of this project is provided by the National Science and Technology Council Taiwan (Grant no. MOST 111-2628-B-002-042 and MOST 111-2326-B-002-005-MY3).

Publisher Copyright:
© The Author(s) 2023. Published by Oxford University Press on behalf of Institute of Chartered Foresters. All rights reserved.

Data Availability Statement

The datasets analyzed for this study are available upon request from The Forest Global Earth Observatory (ForestGEO) database (forestgeo.si.edu).

Keywords

biodiversity conservation
design-based sampling
forest inventory
rapid biodiversity assessment
species diversity
variable probability sampling

Access to Document

10.1093/forestry/cpad041Licence: Unspecified

Cite this

Perng, B. H., Lam, T. Y., Su, S. H., Bin Md Sabri, M. D., Burslem, D., Cardenas, D., Duque, Á., Ediriweera, S., Gunatilleke, N., Novotny, V., O’Brien, M. J., & Reynolds, G. (2024). Integrating rapid assessment, variable probability sampling, and machine learning to improve accuracy and consistency in mapping local spatial distribution of plant species richness. Forestry, 97(2), 282-294. https://doi.org/10.1093/forestry/cpad041

Integrating rapid assessment, variable probability sampling, and machine learning to improve accuracy and consistency in mapping local spatial distribution of plant species richness. / Perng, Bo Hao; Lam, Tzeng Yih (Corresponding Author); Su, Sheng Hsin et al.
In: Forestry, Vol. 97, No. 2, 04.2024, p. 282-294.

Research output: Contribution to journal › Article › peer-review

Perng, BH, Lam, TY, Su, SH, Bin Md Sabri, MD, Burslem, D, Cardenas, D, Duque, Á, Ediriweera, S, Gunatilleke, N, Novotny, V, O’Brien, MJ & Reynolds, G 2024, 'Integrating rapid assessment, variable probability sampling, and machine learning to improve accuracy and consistency in mapping local spatial distribution of plant species richness', Forestry, vol. 97, no. 2, pp. 282-294. https://doi.org/10.1093/forestry/cpad041

@article{a6357ec437f64f6e81211b6d5066822e,

title = "Integrating rapid assessment, variable probability sampling, and machine learning to improve accuracy and consistency in mapping local spatial distribution of plant species richness",

abstract = "Conserving plant diversity is integral to sustainable forest management. This study aims at diversifying tools to map spatial distribution of species richness. We develop a sampling strategy of using rapid assessments by local communities to gather prior information on species richness distribution to drive census cell selection by sampling with covariate designs. An artificial neural network model is built to predict the spatial patterns. Accuracy and consistency of rapid assessment factors, sample selection methods, and sampling intensity of census cells were tested in a simulation study with seven 25–50-ha census plots in the tropics and subtropics. Results showed that identifying more plant individuals in a rapid assessment improved accuracy and consistency, while transect was comparable to or slightly better than nearest-neighbor assessment, but knowing more species had little effects. Results of sampling with covariate designs depended on covariates. The covariate Ifreq, inverse of the frequency of the rapidly assessed species richness strata, was the best choice. List sampling and local pivotal method with Ifreq increased accuracy by 0.7%–1.6% and consistency by 7.6%–12.0% for 5% to 20% sampling intensity. This study recommends a rapid assessment method of selecting 20 individuals at every 20-m interval along a transect. Knowing at least half of the species in a forest that are abundant is sufficient. Local pivotal method is recommended at 5% sampling intensity or less. This study presents a methodology to directly involve local communities in probability-based forest resource assessment to support decision-making in forest management.",

keywords = "biodiversity conservation, design-based sampling, forest inventory, rapid biodiversity assessment, species diversity, variable probability sampling",

author = "Perng, {Bo Hao} and Lam, {Tzeng Yih} and Su, {Sheng Hsin} and {Bin Md Sabri}, {Mohamad Danial} and David Burslem and Dairon Cardenas and {\'A}lvaro Duque and Sisira Ediriweera and Nimal Gunatilleke and Vojtech Novotny and O{\textquoteright}Brien, {Michael J.} and Glen Reynolds",

note = "We would like to thank three anonymous reviewers and the editors for their valuable comments that have significantly improved the manuscript. We also like to thank Dr George Weiblen from the University of Minnesota for assisting with the Wanang dataset. The 25-ha Long-Term Ecological Research Project of Amacayacu is a collaborative project of the Instituto Amaz{\'o}nico de Investigaciones Cient{\'i}ficas Sinchi and the Universidad Nacional de Colombia Sede Medell{\'i}n, in partnership with the Unidad de Manejo Especial de Parques Naturales Nacionales and the Forest Global Earth Observatory of the Smithsonian Tropical Research Institute (ForestGEO). The Amacayacu Forest Dynamics Plot is part of ForestGEO a global network of large-scale demographic tree plots. We acknowledge the Director and staff of the Amacayacu National Park for supporting and maintaining the project in this National Park. BCI: The BCI forest dynamics research project was made possible by National Science Foundation grants to Stephen P. Hubbell: DEB-0640386, DEB-0425651, DEB-0346488, DEB-0129874, DEB-00753102, DEB-9909347, DEB-9615226, DEB-9615226, DEB-9405933, DEB-9221033, DEB-9100058, DEB-8906869, DEB-8605042, DEB-8206992, DEB-7922197, support from the Forest Global Earth Observatory, the Smithsonian Tropical Research Institute, the John D. and Catherine T. MacArthur Foundation, the Mellon Foundation, the Small World Institute Fund, and numerous private individuals, and through the hard work of over 100 people from 10 countries over the past three decades. The plot project is part the Forest Global Earth Observatory (ForestGEO), a global network of large-scale demographic tree plots. Danum: The Danum plot is a core project of the Southeast Asia Rain Forest Research Partnership (SEARRP). We thank SEARRP partners, especially Yayasan Sabah for their support, and HSBC Malaysia and the University of Zurich for funding. We are grateful to the research assistants who are conducting the census, in particular the team leader Alex Karolus, and to Mike Bernados and Bill McDonald for species identifications. We thank Stuart Davies and Shameema Esufali for advice and training. Fushan: The establishment and first census of the Fushan 25-ha plot is a collaborative project by the Taiwan Forestry Research Institute (data provider), the Taiwan Forestry Bureau, the National Taiwan University (Institute of Ecology and Evolutionary Biology) and the ForestGEO (formerly the CTFS). We thank the agencies for providing the datasets and all field staff. S.H.S. also thanks Dr. I-Fang Sun for his long-term support for the Fushan project. Pasoh: Data from the Pasoh Research Forest was provided by the Forest Research Institute Malaysia-Forest Global Earth Observatory, Smithsonian Tropical Research Institute collaborative research project. Negeri Sembilan Forestry Department is the custodian of Pasoh Research Forest and I/we acknowledge the department for preserving the research forest. Sinharaja: The 25-ha Long-Term Ecological Research Project at Sinharaja World Heritage Site is a collaborative project of the Uva Wellassa University, University of Peradeniya, the Forest Global Earth Observatory (ForestGEO) of the Smithsonian Tropical Research Institute, with supplementary funding received from the John D. and Catherine T. Macarthur Foundation, the National Institute for Environmental Science, Japan, and the Helmholtz Centre for Environmental Research-UFZ, Germany, for past censuses. The PIs gratefully acknowledge the Forest Department, Uva Wellassa University, and the Post-Graduate Institute of Science at the University of Peradeniya, Sri Lanka for supporting this project, and the local field and laboratory staff who tirelessly contributed in the repeated censuses of this plot. Wanang: The 50-ha Wanang Forest Dynamics Plot is a collaborative project of the New Guinea Binatang Research Center, the Forest Global Earth Observatory (ForestGEO) of the Smithsonian Tropical Research Institute, the Forest Research Institute of Papua New Guinea, the Czech Academy of Sciences (19-28126X), and the University of Minnesota supported by NSF DEB-1027297 and NIH ICBG 5UO1TW006671. We acknowledge the government of Papua New Guinea and the customary landowners of Wanang for supporting and maintaining the plot. Funding Information: The funding of this project is provided by the National Science and Technology Council Taiwan (Grant no. MOST 111-2628-B-002-042 and MOST 111-2326-B-002-005-MY3). Publisher Copyright: {\textcopyright} The Author(s) 2023. Published by Oxford University Press on behalf of Institute of Chartered Foresters. All rights reserved.",

year = "2024",

month = apr,

doi = "10.1093/forestry/cpad041",

language = "English",

volume = "97",

pages = "282--294",

journal = "Forestry",

issn = "0015-752X",

publisher = "Oxford University Press",

number = "2",

}

TY - JOUR

T1 - Integrating rapid assessment, variable probability sampling, and machine learning to improve accuracy and consistency in mapping local spatial distribution of plant species richness

AU - Perng, Bo Hao

AU - Lam, Tzeng Yih

AU - Su, Sheng Hsin

AU - Bin Md Sabri, Mohamad Danial

AU - Burslem, David

AU - Cardenas, Dairon

AU - Duque, Álvaro

AU - Ediriweera, Sisira

AU - Gunatilleke, Nimal

AU - Novotny, Vojtech

AU - O’Brien, Michael J.

AU - Reynolds, Glen

N1 - We would like to thank three anonymous reviewers and the editors for their valuable comments that have significantly improved the manuscript. We also like to thank Dr George Weiblen from the University of Minnesota for assisting with the Wanang dataset. The 25-ha Long-Term Ecological Research Project of Amacayacu is a collaborative project of the Instituto Amazónico de Investigaciones Científicas Sinchi and the Universidad Nacional de Colombia Sede Medellín, in partnership with the Unidad de Manejo Especial de Parques Naturales Nacionales and the Forest Global Earth Observatory of the Smithsonian Tropical Research Institute (ForestGEO). The Amacayacu Forest Dynamics Plot is part of ForestGEO a global network of large-scale demographic tree plots. We acknowledge the Director and staff of the Amacayacu National Park for supporting and maintaining the project in this National Park. BCI: The BCI forest dynamics research project was made possible by National Science Foundation grants to Stephen P. Hubbell: DEB-0640386, DEB-0425651, DEB-0346488, DEB-0129874, DEB-00753102, DEB-9909347, DEB-9615226, DEB-9615226, DEB-9405933, DEB-9221033, DEB-9100058, DEB-8906869, DEB-8605042, DEB-8206992, DEB-7922197, support from the Forest Global Earth Observatory, the Smithsonian Tropical Research Institute, the John D. and Catherine T. MacArthur Foundation, the Mellon Foundation, the Small World Institute Fund, and numerous private individuals, and through the hard work of over 100 people from 10 countries over the past three decades. The plot project is part the Forest Global Earth Observatory (ForestGEO), a global network of large-scale demographic tree plots. Danum: The Danum plot is a core project of the Southeast Asia Rain Forest Research Partnership (SEARRP). We thank SEARRP partners, especially Yayasan Sabah for their support, and HSBC Malaysia and the University of Zurich for funding. We are grateful to the research assistants who are conducting the census, in particular the team leader Alex Karolus, and to Mike Bernados and Bill McDonald for species identifications. We thank Stuart Davies and Shameema Esufali for advice and training. Fushan: The establishment and first census of the Fushan 25-ha plot is a collaborative project by the Taiwan Forestry Research Institute (data provider), the Taiwan Forestry Bureau, the National Taiwan University (Institute of Ecology and Evolutionary Biology) and the ForestGEO (formerly the CTFS). We thank the agencies for providing the datasets and all field staff. S.H.S. also thanks Dr. I-Fang Sun for his long-term support for the Fushan project. Pasoh: Data from the Pasoh Research Forest was provided by the Forest Research Institute Malaysia-Forest Global Earth Observatory, Smithsonian Tropical Research Institute collaborative research project. Negeri Sembilan Forestry Department is the custodian of Pasoh Research Forest and I/we acknowledge the department for preserving the research forest. Sinharaja: The 25-ha Long-Term Ecological Research Project at Sinharaja World Heritage Site is a collaborative project of the Uva Wellassa University, University of Peradeniya, the Forest Global Earth Observatory (ForestGEO) of the Smithsonian Tropical Research Institute, with supplementary funding received from the John D. and Catherine T. Macarthur Foundation, the National Institute for Environmental Science, Japan, and the Helmholtz Centre for Environmental Research-UFZ, Germany, for past censuses. The PIs gratefully acknowledge the Forest Department, Uva Wellassa University, and the Post-Graduate Institute of Science at the University of Peradeniya, Sri Lanka for supporting this project, and the local field and laboratory staff who tirelessly contributed in the repeated censuses of this plot. Wanang: The 50-ha Wanang Forest Dynamics Plot is a collaborative project of the New Guinea Binatang Research Center, the Forest Global Earth Observatory (ForestGEO) of the Smithsonian Tropical Research Institute, the Forest Research Institute of Papua New Guinea, the Czech Academy of Sciences (19-28126X), and the University of Minnesota supported by NSF DEB-1027297 and NIH ICBG 5UO1TW006671. We acknowledge the government of Papua New Guinea and the customary landowners of Wanang for supporting and maintaining the plot. Funding Information: The funding of this project is provided by the National Science and Technology Council Taiwan (Grant no. MOST 111-2628-B-002-042 and MOST 111-2326-B-002-005-MY3). Publisher Copyright: © The Author(s) 2023. Published by Oxford University Press on behalf of Institute of Chartered Foresters. All rights reserved.

PY - 2024/4

Y1 - 2024/4

N2 - Conserving plant diversity is integral to sustainable forest management. This study aims at diversifying tools to map spatial distribution of species richness. We develop a sampling strategy of using rapid assessments by local communities to gather prior information on species richness distribution to drive census cell selection by sampling with covariate designs. An artificial neural network model is built to predict the spatial patterns. Accuracy and consistency of rapid assessment factors, sample selection methods, and sampling intensity of census cells were tested in a simulation study with seven 25–50-ha census plots in the tropics and subtropics. Results showed that identifying more plant individuals in a rapid assessment improved accuracy and consistency, while transect was comparable to or slightly better than nearest-neighbor assessment, but knowing more species had little effects. Results of sampling with covariate designs depended on covariates. The covariate Ifreq, inverse of the frequency of the rapidly assessed species richness strata, was the best choice. List sampling and local pivotal method with Ifreq increased accuracy by 0.7%–1.6% and consistency by 7.6%–12.0% for 5% to 20% sampling intensity. This study recommends a rapid assessment method of selecting 20 individuals at every 20-m interval along a transect. Knowing at least half of the species in a forest that are abundant is sufficient. Local pivotal method is recommended at 5% sampling intensity or less. This study presents a methodology to directly involve local communities in probability-based forest resource assessment to support decision-making in forest management.

AB - Conserving plant diversity is integral to sustainable forest management. This study aims at diversifying tools to map spatial distribution of species richness. We develop a sampling strategy of using rapid assessments by local communities to gather prior information on species richness distribution to drive census cell selection by sampling with covariate designs. An artificial neural network model is built to predict the spatial patterns. Accuracy and consistency of rapid assessment factors, sample selection methods, and sampling intensity of census cells were tested in a simulation study with seven 25–50-ha census plots in the tropics and subtropics. Results showed that identifying more plant individuals in a rapid assessment improved accuracy and consistency, while transect was comparable to or slightly better than nearest-neighbor assessment, but knowing more species had little effects. Results of sampling with covariate designs depended on covariates. The covariate Ifreq, inverse of the frequency of the rapidly assessed species richness strata, was the best choice. List sampling and local pivotal method with Ifreq increased accuracy by 0.7%–1.6% and consistency by 7.6%–12.0% for 5% to 20% sampling intensity. This study recommends a rapid assessment method of selecting 20 individuals at every 20-m interval along a transect. Knowing at least half of the species in a forest that are abundant is sufficient. Local pivotal method is recommended at 5% sampling intensity or less. This study presents a methodology to directly involve local communities in probability-based forest resource assessment to support decision-making in forest management.

KW - biodiversity conservation

KW - design-based sampling

KW - forest inventory

KW - rapid biodiversity assessment

KW - species diversity

KW - variable probability sampling

UR - http://www.scopus.com/inward/record.url?scp=85187508463&partnerID=8YFLogxK

U2 - 10.1093/forestry/cpad041

DO - 10.1093/forestry/cpad041

M3 - Article

AN - SCOPUS:85187508463

SN - 0015-752X

VL - 97

SP - 282

EP - 294

JO - Forestry

JF - Forestry

IS - 2

ER -

Integrating rapid assessment, variable probability sampling, and machine learning to improve accuracy and consistency in mapping local spatial distribution of plant species richness

Abstract

Bibliographical note

Data Availability Statement

Keywords

Access to Document

Other files and links

Fingerprint

Cite this