Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Can Community Members Identify Tropical Tree Species for REDD+ Carbon and Biodiversity Measurements?

  • Mingxu Zhao,

    Affiliations Key Laboratory for Plant Diversity and Biogeography of East Asia, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming, Yunnan, China, University of Chinese Academy of Sciences, Beijing, China, World Agroforestry Centre, East and Central Asia, Kunming, Yunnan, China

  • Søren Brofeldt,

    Affiliations Department of Food and Resource Economics, Faculty of Science, University of Copenhagen, Copenhagen, Denmark, Nordic Foundation for Development and Ecology (NORDECO), Copenhagen, Denmark

  • Qiaohong Li,

    Affiliations Key Laboratory for Plant Diversity and Biogeography of East Asia, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming, Yunnan, China, World Agroforestry Centre, East and Central Asia, Kunming, Yunnan, China

  • Jianchu Xu,

    Affiliations Key Laboratory for Plant Diversity and Biogeography of East Asia, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming, Yunnan, China, World Agroforestry Centre, East and Central Asia, Kunming, Yunnan, China

  • Finn Danielsen,

    Affiliation Nordic Foundation for Development and Ecology (NORDECO), Copenhagen, Denmark

  • Simon Bjarke Lægaard Læssøe,

    Affiliation Department of Food and Resource Economics, Faculty of Science, University of Copenhagen, Copenhagen, Denmark

  • Michael Køie Poulsen,

    Affiliation Nordic Foundation for Development and Ecology (NORDECO), Copenhagen, Denmark

  • Anna Gottlieb,

    Affiliation Nordic Foundation for Development and Ecology (NORDECO), Copenhagen, Denmark

  • James Franklin Maxwell †,

    † Deceased.

    Affiliation Biology Department, Faculty of Science, Chiang Mai University, Chiang Mai, Thailand

  • Ida Theilade

    Affiliation Department of Food and Resource Economics, Faculty of Science, University of Copenhagen, Copenhagen, Denmark


Biodiversity conservation is a required co-benefit of REDD+. Biodiversity monitoring is therefore needed, yet in most areas it will be constrained by limitations in the available human professional and financial resources. REDD+ programs that use forest plots for biomass monitoring may be able to take advantage of the same data for detecting changes in the tree diversity, using the richness and abundance of canopy trees as a proxy for biodiversity. If local community members are already assessing the above-ground biomass in a representative network of forest vegetation plots, it may require minimal further effort to collect data on the diversity of trees. We compare community members and trained scientists’ data on tree diversity in permanent vegetation plots in montane forest in Yunnan, China. We show that local community members here can collect tree diversity data of comparable quality to trained botanists, at one third the cost. Without access to herbaria, identification guides or the Internet, community members could provide the ethno-taxonomical names for 95% of 1071 trees in 60 vegetation plots. Moreover, we show that the community-led survey spent 89% of the expenses at village level as opposed to 23% of funds in the monitoring by botanists. In participatory REDD+ programs in areas where community members demonstrate great knowledge of forest trees, community-based collection of tree diversity data can be a cost-effective approach for obtaining tree diversity information.


Biologists working in the tropics and elsewhere have always relied on local people for guidance. Indigenous and local communities possess knowledge about the landscape they inhabit [1]. In tropical forests, indigenous cultures sometimes have meticulous classification systems to distinguish between vegetation types on the landscape [28].

While individual human societies may differ considerably in their conceptualization of plants and animals, folk biological classification systems share a number of strikingly similar structural principles [9,10]. All languages seem to have linguistically recognised groupings of organisms (or taxa) of varying degree of inclusiveness, and all languages seem to group the taxa into hierarchical categories. Moreover, taxa assigned to each rank are usually mutually exclusive, and folk biology taxa are about as inclusive as the scientific genera [11].

Few studies have investigated the salience of folk classification of vegetation and plants. From our review, field data on plant species collected by community members and those collected by botanists were compared in only five studies (summary in Table 1). While scientific plant names are designed to prevent the same name from being used for different species, such rules do not apply to vernacular names [12]. Local people often split taxa of great cultural significance into many ethnoforms while species that are less important or less distinctive are often lumped into one single ethnoform with a common name [13,14]. Local names are often based on different criteria from those of scientific taxonomy, such as use or spiritual status [12]. Vernacular names cannot thus be equated consistently to particular scientific names [12]. Nevertheless, some vernacular names do show a one-to-one correspondence with scientific taxa. For example, Jinxiu et al. [15] found a high correlation between folk and scientific plant species among the Dai people of Xishuangbanna, China (Table 1). Likewise, Cardoso et al. [16] found the classification criteria used for fungi by several Brazilian indigenous groups to be similar to those used in classical, morphology-based scientific studies.

Table 1. Previous scientific studies comparing community members’ classification of vegetation and identification of plant species to those of scientists.

Deforestation and forest degradation in the tropics are responsible for approximately 20% of anthropogenic carbon emissions [17] and compromise both livelihoods and biodiversity. In response, the United Nations Framework Convention on Climate Change (UNFCCC) has agreed to establish an international framework that will provide developing countries with financial incentives to reduce emissions from deforestation and forest degradation. While the primary purpose of this framework, known as REDD+ (Reducing Emissions from Deforestation and Forest Degradation), is climate change mitigation, it may have enormous co-benefits for biodiversity because tropical forests are exceptionally rich in exclusive biodiversity reservoirs [18]. However, REDD+ may have a negative influence on biodiversity when low-carbon, high-biodiversity forests are replaced with high-carbon, low-biodiversity forests (e.g. tree plantations), or when the protection of high-carbon forest in one area leads to the displacement of other high-biodiversity forests [19, 20]. Biodiversity monitoring is therefore needed [2123].

A key obstacle in accounting for biodiversity is a lack of consensus as to what to monitor [24] because there is, so far, no single, agreed metric of biodiversity, unlike carbon (Mg ha-1 of carbon). While biomass estimates are often based on number of trees per hectare and their diameter at breast height (DBH), monitoring of biodiversity requires understanding parameters such as species richness, composition, abundance and the distribution of many taxa, which is a tall order given the widespread lack of human and financial resources [25]. One approach to minimizing the risk of overburdening REDD+ programs is to deploy a limited set of ‘indicator’ taxa [26]. The use of indicator taxa in biodiversity monitoring for REDD+ needs to satisfy the following four requirements: low monitoring costs, ease of identification, surrogates of ecosystem integrity, and cross-taxon congruency, see below [26].

Tree assemblage, i.e. a community of canopy-tree species including species richness and abundance as attributes, has been suggested as a suitable indicator. Firstly, the sampling of trees is relatively easy and inexpensive [27]. REDD+ remote sensing relies on tree density and DBH to ground-truth satellite images. Tree density data will thus be collected irrespective of how biodiversity safeguards are monitored. Secondly, unlike most other organisms, tree taxonomy is relatively well described. Thirdly, tree assemblages have a high cross-taxon congruency, in which tree species richness and composition are correlated with those of other taxa [28, 29], probably because trees provide other taxa with resources and habitats.

Imai et al. [25] state that, for trees, the availability of local experts is relatively adequate compared with other taxa. While this may be true for some parts of the world, the enormous challenge of flora projects in lower-income countries, where the most diverse terrestrial ecosystems are found [30], is exacerbated by the short supply of taxonomic experts available [3133]. Researchers have pointed to the vast number of indigenous and local botanical experts, representing a potentially valuable, yet largely unrecognised and untapped, resource [3336]. The participation of ‘parataxonomists’, defined as resident, field-based, biodiversity inventory specialists with no formal training [37], has been shown to enhance biodiversity inventories for both arthropods [35,3739], fungi [16,40], and plants [15,41,42].

UNFCCC texts and guidance documents on the technical aspects of REDD+ outline explicit roles for indigenous people and local communities in implementing REDD+ [[4346]. Yet little has been published on how community-based REDD+ should be implemented in practice, including community-level monitoring of carbon, livelihoods or biodiversity [47,48]. The degree of local participation may vary from virtually no local involvement to an entirely local effort, with data collection, interpretation and reporting undertaken by local people [49]. Studies suggest that locally-based monitoring may be advantageous in terms of lower costs [50], enhanced local ownership, greater cultural relevance and improved institutional strength at the community level [5153]. Moreover, local people’s participation in monitoring can potentially enhance decision-making at the operational level of forest management [54,55]. Community monitoring of forest carbon and tree biodiversity may therefore contribute to a fair and equitable REDD+ [47].

One of the functions of the newly-established Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) is to bring different knowledge systems, including indigenous and local knowledge systems, to the science–policy interface [56]. One key challenge lies in how to use information generated by different knowledge systems within synthetic assessments at the science-policy interface [57]. It is therefore important to understand how folk biological classification systems connect or otherwise with scientific classification systems.

Here, we explore one aspect of folk biological classification systems of particular relevance to REDD+: we compare local community-collected data on canopy trees in forest vegetation plots with that collected by trained botanists, using the botanists’ findings as a benchmark. In addition, we compare the costs of trained botanists’ and community monitors’ identification of tree species.


Ethics Statement

This research did not involve human or other animal subjects. For plant collections, we collected the minimum number of specimens required to appropriately voucher field identifications. The field studies did not involve endangered or protected species.

Permission to conduct research in Manlin was obtained through a bilateral agreement between Kunming Institute of Botany (Chinese Academy of Sciences) and the Forestry Bureau of Xishuangbanna Autonomous Prefecture at the regular meeting between these parties. Free, prior and informed consent was obtained for all community monitors participating in the study.

Literature Search

We reviewed previous studies that have compared scientist and local knowledge on vegetation or species level by searching the databases Web of Science, PubMed, CABI, AGRICOLA and AGRIS using the following keywords: participatory monitoring, local monitoring, community monitoring, parataxonomist, local ecological knowledge, traditional ecological knowledge, and indigenous knowledge.

Study Site and Data Collectors

The study site was chosen opportunistically. The criteria were that the site was appointed by the government as a potential REDD+ site, and that local communities used the forest area. The study area was the forest near Man Lin village, located at 1180 m.a.s.l. (above sea level) in Xiangming township of Xishuangbanna autonomous prefecture, Yunnan province. The climate is monsoonal with an average annual temperature of 25° C and an average annual precipitation of 1700 mm. Slope inclinations range between 30° and 70°, and in some areas attain up to 90°. The vegetation is tropical mountain rainforest at around 900–1400 m.a.s.l. The forest area surveyed covered 761 ha and is characterised by Castanopsis mekongensis and Schima wallichii. The canopy can be divided into 3 layers: the overstory reaches 35 m in height and is dominated by emergent trees such as Pometia pinnata; the midstory reaches 25 m and is dominated by Castanopsis spp., and Schima wallichii while the understory contains a multitude of species, such as Cratoxylon cochinchinensis, Phoebe puwenensis, Machilus spp., Lithocarpus spp. Elaeocarpus spp. Mallotus spp. Shrub and herbaceous layers at the edges and inside forest areas are rich in species. Community members and scientists both measured two forest strata (homogenous forest areas in terms of structure and composition). The stratum closer to the village (291 ha) is classified as collective forest. It is moderately disturbed forest and consists of abandoned shifting cultivation fields and ancient tea trees with an overstory of natural forest. The second stratum (470 ha) is classified as state forest and consists mainly of natural old-growth forest on steep to very steep slopes. Few trees were being extracted 40–60 years ago. Shifting cultivation was practised in small plots on more gradual slopes from the 1950s to the1990s and then gradually abandoned. The forest recovered and is in a good condition with a profusion of lianas and epiphytes. Tea seedlings are planted in the collective forest and dominate the understory in smaller areas. Firewood to dry the tea leaves as well as timber for house construction and furniture are harvested from nearby private forests. The rural Yi community is connected by road and most villagers are employed in newly-established rubber plantations.

Data were collected from permanent vegetation plots. Plots were surveyed by botanists in July 2012 and community members in March 2013. Representatives of the local Yi community selected three community participants for tree species identification, based on their interest in and experience of forest resources. These community monitors are thus probably more skilled than the average villager. All community monitors were male, reflecting the fact that men visit the forest more frequently than women, often for hunting and collecting of non-timber forest products (NTFPs). Women do not venture into steeply-sloped forest areas distant from the village. All community monitors had attended primary school, which is the usual length of schooling in the village. The community monitors received 1–2 days’ training from an intermediate organisation (research institution) on how to establish plots and measure tree girth, as required for assessing the above-ground biomass. The community identification of trees relied solely on existing local ecological knowledge. The botanical team consisted of the late J.F. Maxwell, botanist and curator at Chiang Mai Herbarium (CMU), who had more than 40 years’ experience of floristic work in Indochina, and PhD fellow Mingxu Zhao from Kunming Institute of Botany (KIB).

Methods for Measuring Forest Tree Diversity and Costs

The community monitors and the staff of the intermediate organisation divided the forest into two homogenous strata in terms of tree species composition and level of degradation, using the available knowledge of the forest and its history (i.e. previous logging or shifting cultivation). Based on this pre-analysis, staff of the intermediate organisation randomly placed 30 circular plots in each stratum making a total of 60 plots. The community monitors and professional botanists then independently carried out forest inventories in each plot. All trees with a girth of ≥ 30 cm (as a proxy for DBH ≥ 10 cm) were identified within a radius of 9 m from the plot centre and all trees with a girth of ≥ 100 cm (proxy for DBH ≥ 30 cm) were identified within a radius of 15 m from the plot centre. Local names and scientific names were recorded by pencil on pre-printed paper forms. The community monitors worked as a team and discussed identifications internally but not with the botanists. Botanists were allowed to ask local guides about flower and fruit characteristics and phenology as some trees were not in flower/fruit at the time of the survey. Specimens were collected for herbaria work by the botanists. Botanists used Flora of Thailand [65], Flora of Yunnan [66], and Flora of China [67] plus herbarium material at Xishuangbanna Botanic Garden (XTBG), Kunming Institute of Botany (KIB) and Chiang Mai (CMU). Voucher specimens were deposited at both CMU and KIB.

We estimated the costs of community-based and professionally-executed identifications on the basis of the actual expenses incurred for local transport and during the training and fieldwork [47,50]. The cost of tree species identification was accounted for separately from other research activities. The chief botanist’s airfare from Thailand to China was not included. We calculated the number of genera and species identified by both community monitors and botanists.


Tree Identification by Botanists and Community Monitors

In total, 1071 trees were recorded by both the botanists and the community monitors (S1 File). We first examined how many taxa and morphospecies (species distinguished from others only by their morphology) the botanists could identify. We found that, of the 1071 trees, the botanists were able to identify 1052 trees belonging to 50 families, 104 genera, and 142 species. In addition, the botanists recognised 19 morphospecies (1 identified to family level, 7 to genus level and 11 unidentified). Of the 161 recognised taxa, the botanists named 149 to genus level and 142 to species level.

We found no significant difference between the number of trees in the plot network identified to at least genus level by botanists (99.3%; n = 1071 trees) and community monitors (94.7%; n = 1071 trees). The community monitors were able to name 1013 trees belonging to 42 families, 90 genera, and 111 species that showed a one-to-one correspondence with the botanists’ named genera and species. Of the 161 taxa recognised by the botanists, the community monitors named 128 to genus level and 111 to species level (Table 2).

Table 2. Comparison of the number of trees identified to genus or species level by botanists and by community monitors, and the number of genera and species that the identified trees belonged to, in montane forest in Yunnan, China (n = 1071 trees).

Numbers for community monitors are calculated using only those with a one-to-one correspondence to scientific taxa.

Community monitors grouped 27 species (262 trees), mainly of the genera Castanopsis, Engelhardtia, and Schima, into 11 ethnotaxa. The 11 ethnotaxa referred one-to-one to 11 scientific genera. The lumping of species that morphologically appear very similar makes up half (52%, n = 31) the difference between the number of species identified by botanists and community monitors. Community monitors split 2 species (7 trees) into four ethnospecies. In addition, the community monitors did not have a name for 58 trees (5%, n = 1071 trees) belonging to 32 species. We found that 3 trees (0.3%, n = 1071 trees) seemed to be misidentified by the community monitors based on the observation that the community monitors consistently identified other trees of the same species as being of a different ethnospecies.

We investigated whether the trees that the community monitors did not have a name for shared any common characteristics that might make them difficult or irrelevant for them to identify. We examined the composition of the unnamed trees against six criteria: family, wood density, size, habitat (primary and secondary forest), abundance, and usefulness for the community members as source of timber, fruits and other products.

The 58 unidentified trees belonged to what the botanists identified as 25 genera of 18 families. Five families represented 57% of the unidentified trees (Magnoliaceae, Meliaceae, Myristicaceae, Rubiaceae, and Rutaceae). Forty-five (78%, n = 58) of the trees were rare, i.e. only 1–3 individuals were encountered in the plot network. Thirty-four trees (59%, n = 58) were small (DBH<20 cm). Thirty-six trees (62%, n = 58) were classified as light wood by community monitors using a scale from one to three. Forty-one (71%, n = 58) of the unidentified trees were found in the primary forest on steep slopes (>45 degrees), 22% in the disturbed forest closer to the village, and 7% in the ancient tea plantations in the vicinity of the village. Twenty-two trees (37%, n = 58) were useful for timber. None of the unidentified trees (0%, n = 58) had edible fruits. Thirteen (22%, n = 58) did not have any known uses (Table 3 and S2 File).

Table 3. Characteristics of trees unidentified by community monitors compared to total number of trees in montane forests in Yunnan, China.

* Light wood was classified by community monitors on a scale from 1 to 3 (low to high wood density).

Community monitors had between 2 and 4 vernacular names for 29 species (18%, n = 161 taxa) recognised by the botanists (S1 File). For example, Colona floribunda is named pao huo sheng when using its fibres to make fires and fan peng shu when using its hardy and non-perishable timber to construct outdoor kitchens. Some species also had one local name for their use and another for a conspicuous characteristic. For example, Ficus auriculata is named both after its edible fruit and its conspicuous ear-shaped leaves and Engelhardtia spicata is named both for its soft and easily cut wood and for its bark, which resembles the flab around the waist of a fat woman. Toona ciliata, a luxury wood, has three local names; a traditional local name, a local name adopted from the Flora of China and a local name referring to the quality of the timber. If local synonyms, as in the examples above, were used consistently (n > 2 times) for the same scientific species, we considered the synonyms useful in identification.

Costs of Tree Identification

Finally, we estimated the costs of tree identifications (Table 4). The costs of the plot survey were 4.5 USD/ha for botanists and 1.3 USD/ha for community monitors. Salary and domestic travel were the main expenses in the botanist-executed plot survey. The community monitors all lived in an adjacent village. Salaries paid to community monitors were higher than the daily rates in the rubber plantations due to the strenuous work on the slopes. Botanists and community monitors both spent 7 days visiting all plots. The cost of equipment was minimal and constituted spray paint for marking the plots, a pair of clippers to collect specimens and cardboard and newspapers for the plant press. The relative amount of expenses disbursed at village level was 23% for the botanist-led survey and 89% for the community-led survey (S3 File).

Table 4. Costs of tree species identification in the plot network by professional botanists and community monitors in montane forest in Yunnan, China (in USD).

* Two villagers acted as guides and assistants during botanists’ tree identification.


Our results suggest that local experts from among the Yi people can reliably identify tree species in Yunnan’s forests without having access to identification guides and herbaria. Moreover, these local experts are able to collect large volumes of tree diversity data at relatively low cost. The fact that the local experts’ and trained botanists’ results matched cannot be a result of the pre-study training exercise because this focused only on plot establishment and tree girth measurements.

How representative are our findings? We looked at the ability of experienced Yi community members to match botanists’ identifications of tree species in one area at one time. Our results are similar to field investigations in Xishuangbanna, China, and Ecuador and Brazil [14,15,63], but contrast with results from Indonesia where plant names provided by local informants, who had their experience from timber companies, could not be equated to particular taxa [12,64]. Not all local monitors are thus able to handle this task [69]. Unfortunately, few publications on the topic describe how the community monitors were chosen (Table 1). Local knowledge may vary with age, proximity to the resource and experience [70]. In the present study, the community monitors were carefully selected by representatives of the local community, based on their interest in and experience of forest resources. Moreover, the taxonomic identifications in our study were made by the late J.F. Maxwell, who was acknowledged as the most skilled field botanist in Indochina [71]. Further studies are needed to examine the replicability of our findings among carefully selected community member experts and botanists in other areas.

The tree taxa that were not identified by the community monitors differed from the majority of trees in that they were mainly light-wood, low-density taxa of primary forest, of limited use-value to the local communities. These findings concur with previous studies which demonstrate that community member assessments may be suitable for monitoring organisms or phenomena that are meaningful to community members e.g. as source of food or income or with cultural or spiritual value [12,16,34]. If the aim is to monitor attributes that are not relevant from a local perspective then local community members’ assessments may not be suitable [72].

While studies on the local uses of plants are numerous, in-depth and well-documented studies on the principles underlying folk biological taxonomy and nomenclature in non-Western societies are still lacking. This lack of understanding of the conceptual foundations of ‘ethnoscience’ (sensu UNESCO, [73]) as practised by non-Western people may partly explain the few attempts to bring indigenous and local knowledge systems into the science-policy interface [56, 74]. Linguistic and cultural barriers continue to hamper such efforts [34,40].

Attempts have been made to elaborate standards for monitoring biodiversity in relation to REDD+ [21,23,75]. Yet few REDD+ programs include community-level monitoring of biodiversity [47].

Accuracy is essential when forest carbon stocks are measured in order to track emissions from deforestation. Most forest carbon stock monitoring in REDD+ programs is undertaken by national consultants and foreign experts using remote sensing [76]. There may, however, be markedly divergent estimates of forest carbon density when measured from ground plots and satellites [77]. Variations in biodiversity can matter greatly when determining carbon stocks but neither wood density nor species assemblages can be mapped from space. Our findings suggest that community monitors may be important partners not only in terms of measuring stem diameters but also identifying tree species (at least to genus level), thus enabling stems to be matched to wood density information. Community monitors’ tree identifications can thus complement, and add value to, remote sensing.

In REDD+ pilot schemes where community members have been involved in monitoring forest biomass, their role has been largely limited to data collection [47,50,52,53,69,76]. In every case, an intermediate organisation has helped establish the plot network and interpret the biomass data. Since we found a high degree of one-to-one correspondence between the vernacular names and the botanists’ named taxa, an intermediate organisation could also translate the local experts’ tree data into Latin names, thereby connecting the vernacular names to the current scientific knowledge of each taxon.

A key debate surrounding the development of REDD+ relates to costs [51,76,78]. Although opportunity costs are generally considered the largest cost component of REDD+, monitoring may also form a significant component of the total project costs [78]. Our results suggest that sustaining a network of field plots is 70% cheaper (1.3 USD/ha/year) when tree diversity data are collected by community experts instead of botanists (4.5 USD/ha/year). Moreover, our findings suggest that community monitoring of canopy trees in vegetation plots resulted in 89% of all expenditure being disbursed at community level as opposed to 23% when the monitoring was led by professional scientists.

Monitoring costs at six Peruvian REDD+ sites ranged from 0.2–4.0 USD/ha/year [78]. The most important factors determining costs per hectare are: i) the size of area to be monitored [50,69,76], ii) the desired level of accuracy [78], and iii) the salary and time taken ([50], this study). The monitoring cost per ha decreases as the size of the forest area increases. The start-up costs of community monitoring may be high but, with time, the community monitors’ skills improve and community monitoring becomes a cost-effective alternative to professional forest monitoring [50]. In addition, past works suggest that community involvement in monitoring enhances feelings of ownership and improves governance while building local capacity [7980].

There are challenges to using tree data from permanent vegetation plots to provide input to forest management. Permanent plots may be treated differently from the rest of the forest. Over time, the tree composition in the plots may therefore no longer be representative of the forest area. Moreover, if pressures on the forest are high, the frequency of data collection may not match the speed with which the forest is undergoing transformation. Monitoring of tree diversity in REDD+ programs therefore cannot stand alone. Monitoring the status of threatened species, potential threats and changes in the use of the forest and its resources, may also be necessary [81]. Participatory REDD+ programs will require complementary participatory biodiversity monitoring tools that can quickly provide reliable information with which to guide action, at a low cost. One such approach is that of focus group discussions with knowledgeable local community members on the status of particular natural resources and species of significance due to their role, value or conservation status [82].

There was no conflict over the forest and its resources in this study. If community-based biodiversity monitoring is to become a key element in the monitoring of participatory REDD+ programs, periodic triangulation of the monitoring results will be required, although this is no different from any well-designed natural resource management initiative, whether the monitoring is implemented by communities, the government or the private sector [69]. To help practitioners choose suitable approaches for biomass and biodiversity monitoring in REDD+ programs, we have developed a decision tree (Fig 1).

Fig 1. Decision tree to guide practitioners in choosing methods for biomass and biodiversity monitoring in REDD+ programs.

The arrows indicate the flow of the decisions. REDD+ programs using permanent forest vegetation plots as part of their monitoring of the above-ground biomass of a forest area can take advantage of data from the same plots for monitoring the richness and abundance of canopy trees.

There are major international efforts underway by botanists to inventory tropical forest trees [32,83,84]. Local involvement is also relevant in this context. Initial reliance on local nomenclature for documenting tree richness should be augmented with the positive identification of voucher specimens [12,34,64]. It is an arduous task to collect specimens due to the infrequent flowering of tropical plants. A long-term collection of fertile material, when available (month by month), would benefit hugely from the involvement of local experts [33,36,85].

In conclusion, we have shown that if community members with significant knowledge of forest trees are already assessing the biomass in a network of vegetation plots as part of Participatory Measurement, Reporting and Verification for REDD+ programs [48,76], then minimal further effort is required for them to collect data on the diversity of trees in the same plots. Such an approach could generate large volumes of high-quality tree diversity data at a relatively low cost.

Supporting Information

S1 File. List of species, author, local names, number of trees in plot network, wood density, and usefulness as timber, fruit, or other uses.

Wood density was classified by communities on a scale from 1 to 3 (low to high wood density). Information on uses is based on Flora of Yunnan (1977–2006) and Flora of China (2014).


S2 File. Characteristics of the 58 trees that remained unidentified by community monitors.

Information on uses is based on Flora of Yunnan (1977–2006) and Flora of China (2014). * Wood density was classified by communities on a scale from 1 to 3 (low to high wood density).


S3 File. Details of the costs of botanist and community-collected tree data.



We dedicate this paper to the memory of J.F. Maxwell, a naturalist with a compulsion for collecting and describing, who will be deeply missed for his honest character and unique knowledge of the Thai and Indochina flora.

Author Contributions

  1. Conceived and designed the experiments: IT MZ SB JX FD MKP JFM.
  2. Performed the experiments: MZ JFM SB IT SBLL MKP.
  3. Analyzed the data: MZ SB QL IT.
  4. Contributed reagents/materials/analysis tools: MZ SB IT AG.
  5. Wrote the paper: IT MZ SB FD MKP.


  1. 1. Pitman N, Cecilio MP, Pudicho MP, Graham JG, Núñez MPV, Valenzuela M, et al. Indigenous perceptions of tree species abundance across an upper Amazonian landscape. J Ethnobiol. 2011;31: 233–43.
  2. 2. Abraão MB, Nelson BW, Baniwa JC, Yu DW, Shepard GH Jr. Ethnobotanical ground-truthing: Indigenous knowledge, floristic inventories and satellite Imagery in the upper Rio Negro, Brazil. J Biogeogr. 2008;35: 2237–48.
  3. 3. Fleck DJ, Harder JD. Matses Indian rainforest habitat classification and mammalian diversity in Amazonian Peru. J Ethnobiol. 2000;20: 1–36.
  4. 4. Gilmore MP, Young JC. The Maijuna participatory mapping project: Mapping the past and the present and the future. In Gilmore MP, Vriesendorp C, Alverson VS, del Campo A, von May R, López Wong C, Río Ochoa S. editors. Peru: Maijuna. The Field Museum, Chicago; 2010. P. 233–242.
  5. 5. Hernandez-Stefanoni JL, Pineda JB, Valdes-Valdez G. Comparing the use of indigenous knowledge with classification and ordination techniques for assessing the species composition and structure of vegetation in tropical forests. Environ Manage. 2006;37: 686–702. pmid:16508801
  6. 6. Shepard GH Jr, Yu DW, Lizzaralde M, Italiano M. Rain forest habitat classification among the Matsigenka of the Peruvian Amazon. J Ethnobiol. 2001;21: 1–38.
  7. 7. Torre-Cuadros MA, Ross N. Secondary biodiversity: local perceptions of forest habitats, the case of Solferino, Quintana Roo, Mexico. J Ethnobiol. 2003;23: 287–308.
  8. 8. Shepard GH Jr, Yu DW, Nelson BW. Ethnobotanical Ground-Truthing and Forest Diversity in the Western Amazon. Advances in Economic Botany. 2004;15: 133–171.
  9. 9. Berlin B, Breedlove DE, Raven PH. General principles of classification and nomenclature in folk biology. Am Anthropol. 1973;75: 214–42.
  10. 10. Berlin B. Ethnobiological classification: Principles of categorisation of plants and animals in traditional societies. Princeton: Princeton University Press; 1992.
  11. 11. Holman EW. The relationship between folk and scientific classifications of plants and animals. J Classif. 2002;19: 131–59.
  12. 12. Wilkie P, Saridan A. The limitations of vernacular names in an inventory study, Central Kalimantan, Indonesia. Biodivers Conserv. 1999;8: 1457–67.
  13. 13. Martin GJ. Ethnobotany. A methods manual. 4th ed. People & Plants. London: Earthscan, 2004.
  14. 14. de Lacerda AEB, Nimmo ER. Can we really manage tropical forests without knowing the species within? Getting back to the basics of forest management through taxonomy. For Ecol Manage. 2010;259: 995–1002.
  15. 15. Jinxiu W, Hongmao L, Huabin H, and Lei G. Participatory approach for rapid assessment of plant diversity through a folk classification system in tropical rainforest: Case study in Xishuangbanna, China. Conserv Biol. 1992;18: 1139–42.
  16. 16. Cardoso DBOS, de Queiroz LP, Bandeira FP, Góes-Neto A. Correlations between indigenous Brazilian folk classifications of fungi and their systematics. J Ethnobiol. 2010;30: 252–264.
  17. 17. van der Werf GR, Morton DC, DeFries RS, Olivier JGJ, Kasibhatla PS, Jackson RB, Collatz GJ, Randerson JT. CO2 emissions from forest loss. Nat Geosci. 2009;2: 737–8.
  18. 18. CBD. REDD+ and biodiversity; Technical series No 59; Convention on Biological Diversity. Montreal, Canada; 2011.
  19. 19. Miles L, Kapos V. Reducing greenhouse gas emissions from deforestation and forest degradation: Global land-use implications. Science. 2008;320: 1454–5. pmid:18556549
  20. 20. Paoli GD, Wells PL, Meijaard E, Struebig MJ, Marshall AJ, Obidzinski K, et al. Biodiversity Conservation in the REDD. Carbon Balance Manag. 2010;5: 7. pmid:21092321
  21. 21. The Climate, Community & Biodiversity Alliance (CBBA). Climate, Community & Biodiversity Project Design Standards. 2nd ed. Arlington, VA; 2008.
  22. 22. Dickson B, Kapos V. Biodiversity monitoring for REDD+. Curr Opin Environ Sustain. 2012;4: 717–25.
  23. 23. Gardner TA, Burgess ND, Aguilar-Amuchastegui N, Barlow J, Berenguer E, Clements T, et al. A framework for integrating biodiversity concerns into national REDD+ programmes. Biol Conserv. 2012;154: 61–71.
  24. 24. Struebig MJ, Harrison ME, Boonman A, Cheyne SM, Husson S, Marchant NC. Biodiversity monitoring protocols for REDD+: can a one-size-fits-all approach really work? Trop Conserv Sci. 2012;5(1): 1–11.
  25. 25. Imai N, Tanaka A, Samejima H, Sugau JB, Pereira JT, Titin J, et al. Tree community composition as an indicator in biodiversity monitoring of REDD+. For Ecol Manage. 2014;313: 169–79.
  26. 26. Su J, Debinski D, Jakubauskas ME, Kindsher K. Beyond Species Richness: Community Similarity as a Measure of Cross-Taxon Congruence for Coarse-Filter Conservation. Cons Biol. 2009;18: 167–73.
  27. 27. Gardner TA, Barlow J, Araujo IS, Ávila-Pires TC, Bonaldo AB, Costa JE, et al. The cost-effectiveness of biodiversity surveys in tropical forests. Ecol Lett. 2008;11: 139–50. pmid:18031554
  28. 28. Kati V, Devillers P, Dufrene M, Legakis A, Vokou D, Lebrun P. Testing the value of six taxonomic groups as biodiversity indicators at a local scale. Conserv Biol. 2004;18: 667–75.
  29. 29. Barlow J, Gardner TA, Araujo IS, Ávila-Pires TC, Bonaldo AB, Costa JE, et al. Quantifying the biodiversity value of tropical primary, secondary, and plantation forests. Proc Natl Acad Sci U S A. 2007;104: 18555–60. pmid:18003934
  30. 30. Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J. Biodiversity hotspots for conservation priorities. Nature. 2000;403: 853–8. pmid:10706275
  31. 31. Prance GT, Beentje H, Dransfield J, Johns R. The tropical flora remains undercollected. Ann Mo Bot Gard. 2000;87: 67–71.
  32. 32. Phillips OL, Vasquez Martinez R, Nunez Vargas P, Monteagudo AL, Chuspe Zans M., Galiano Sanchez W, et al. Efficient plot-based floristic assessment of tropical forests. Journal of Tropical Ecology. 2003;19: 629–45.
  33. 33. Webb CO, Slik JWF, Triono T. Biodiversity inventory and informatics in Southeast Asia. Biodivers Conserv. 2010;19: 955–72.
  34. 34. Pfeiffer J, Uril Y. The role of indigenous parataxonomists in botanical inventory: from Herbarium Amboinense to Herbarium Floresense. Telopea. 2003;10: 61–72.
  35. 35. Basset Y, Novotny V, Miller SE, Weiblin GD, Missa O, Stewart AJ. Conservation and biological monitoring of tropical forests: the role of parataxonomists. J Appl Ecol. 2004;41: 163–74.
  36. 36. Sheil D, Lawrence A. Tropical biologists, local people and conservation: new opportunities for collaboration. Trends Ecol Evol. 2004;19(12): 634–638. pmid:16701325
  37. 37. Janzen DH. Setting up tropical biodiversity for conservation through non-damaging use: participation by parataxonomists. J Appl Ecol. 2004;41: 181–7.
  38. 38. Basset Y, Novotny V, Miller SE, Pyle R. Quantifying biodiversity: Experience with parataxonomists and digital photography in Papua New Guinea and Guyana. Bio Sci. 2000;50: 899–908.
  39. 39. Janzen DH, Hallwachs W. Joining inventory by parataxonomists with DNA barcoding of a large complex tropical conserved wildland in Northwestern Costa Rica. PLoS ONE. 2011;6(8). e18123. pmid:21857894
  40. 40. Ellen R. Ethnomycology among the Nuaulu of the Moluccas: Putting Berlin’s “General principles” of ethnobiological classification to the test. Econ Bot. 2008;62: 483–96.
  41. 41. Martin GJ, Agama, AL, Beaman JH, Nais, J. Projek Etnobotani Kinabalu: The making of a Dusun Ethnoflora (Sabah, Malaysia). People and Plants Working Paper. UNESCO; 2002.
  42. 42. Dalle SP, Potvin C. Conservation of useful plants: an evaluation of local priorities from two indigenous communities in Eastern Panama. Econ Bot. 2004;58: 38–57.
  43. 43. GOFC-GOLD. A sourcebook of methods and procedures for monitoring and reporting anthropogenic greenhouse gas emissions and removals caused by deforestation, gains and losses of carbon stocks in forests remaining forests, and forestation. GOFC-GOLD Report version COP 16–1, Alberta, Canada; 2010.
  44. 44. Epple C, Dunning E, Dickson B, Harvey C. Making biodiversity safeguards for REDD+ work in practice. Developing operational guidelines and identifying capacity requirements. Cambridge, UK: United Nations Environment Program–World Conservation Monitoring Centre; 2011.
  45. 45. UNFCCC. Outcome of the work of the Ad Hoc Working Group on Long-term Cooperative Action under the Convention. Draft decision [-/CP.17]. Bonn, Germany: UNFCCC; 2011.
  46. 46. UNFCCC. UNFCCC/SBSTA/2011/L.25/Add.1. Framework Convention on Climate Change, Subsidiary Body for Scientific and Technological Advice (SBSTA), Methodological guidance for activities relating to reducing emissions from deforestation and forest degradation and the role of conservation, sustainable management of forests and enhancement of forest carbon stocks in developing countries. Draft conclusions proposed by the Chair, Thirty-fifth session Durban, 28 November to 3 December 2011. Bonn, Germany: UNFCCC. 2011.
  47. 47. Danielsen F, Adrian T, Brofeldt S, van Noordwijk M, Poulsen MK, Rahayu S, et al. Community monitoring for REDD+: international promises and field realities. Ecol Soc. 2013;18(3): 41.
  48. 48. Boissière M, Beaudoin G, Hofstee C, Rafanoharana S. Participating in REDD+ Measurement, Reporting, and Verification (PMRV): Opportunities for Local People? Forests 2014; 5(8): 1855–78.
  49. 49. Danielsen F, Burgess ND, Balmford A, Donald PF, Funder M, Jones JPG, et al. Local Participation in Natural Resource Monitoring: a Characterization of Approaches. Conserv Biol. 2009; 23: 31–42. pmid:18798859
  50. 50. Brofeldt S, Theilade I, Burgess ND, Danielsen F, Poulsen MK, Adrian T, et al. Community monitoring of carbon stocks for REDD+: Does accuracy and cost change over time? Forests. 2014;5: 1834–54.
  51. 51. Fry BP. Community forest monitoring in REDD+: the ‘M’ in MRV? Environ Sci Policy. 2011;14:181–7.
  52. 52. Larrazábal A, McCall MK, Mwampamba TH, Skutsch M. The role of community carbon monitoring for REDD+: a review of experiences. Curr Opin Environ Sustain. 2012;4: 707–16.
  53. 53. Pratihast AK, DeVries B, Avitabile V, de Bruin S, Kooistra L, Tekle M, et al. Combining Satellite Data and Community-Based Observations for Forest Monitoring. Forests. 2014;5(10): 2464–89.
  54. 54. Danielsen F, Mendoza MM, Tagtag A, Alviola PA, Balete DS, Jensen AE, et al. Increasing conservation action by involving local people in natural resource monitoring. Ambio. 2007;36: 566–70. pmid:18074893
  55. 55. Danielsen F, Burgess ND, Jensen PM, Pirhofer-Walzl P. Environmental monitoring: the scale and speed of implementation varies according to the degree of peoples involvement. J Appl Ecol. 2010;47: 1166–8.
  56. 56. [UNEP] United Nations Environment Programme. Report of the Second Session of the Plenary Meeting to Determine Modalities and Institutional Arrangements for an Intergovernmental Science-Policy Interface on Biodiversity and Ecosystem Services. Report no. UNEP/IPBES.MI/2/9. UNEP. 2012. Available: Accessed 4 February 2014.
  57. 57. Sutherland WJ, Gardner TA, Haider LJ, Dicks LV. How can local and traditional knowledge be effectively incorporated into international assessments? Oryx 2014;48: 1–2.
  58. 58. Hellier A, Newton A, Gaona SO. Use of indigenous knowledge for rapidly assessing trends in biodiversity: a case study from Chiapas, Mexico. Biodivers Conserv. 1999;8: 869–89.
  59. 59. Naidoo R, Hill K. Emergence of indigenous vegetation classifications through integration of traditional ecological knowledge and remote sensing analyses. Environ Manage. 2006;38: 377–87. pmid:16832592
  60. 60. Chalmers N, Fabricius C. Expert and generalist local knowledge about land-cover change on South Africa’s Wild Coast: can local ecological knowledge add value to science? Ecol Soc. 2007;12(1): 10. URL:
  61. 61. Halme KJ, Bodmer RE. Correspondence between scientific and traditional ecological knowledge: Rain forest classification by the non-indigenous riberenos in Peruvian Amazonia. Biodivers Conserv. 2007;16: 1785–1801.
  62. 62. Vergara-Asenjo G, Sharma D, Potvin C. Engaging stakeholders: assessing accuracy of participatory mapping of land cover in Panama. Conserv Lett. 2015;
  63. 63. Oldekop JA, Bebbington AJ, Berdel F, Truelove NK, Wiersberg T, et al. Testing the accuracy of non-experts in biodiversity monitoring exercises using fern species richness in the Ecuadorian Amazon. Biodivers Conserv. 2011;20: 2615–26.
  64. 64. Theilade I, Rutishauser E, Poulsen MK. Community assessment of tropical tree biomass: challenges and opportunities for REDD+. Carbon Balance Management 2015;10: 17. Available: pmid:26229548
  65. 65. Flora of Thailand (Vol. 1–12). Forest Herbarium. Royal Forest Department, Bangkok, Thailand. 1970–2011.
  66. 66. Zy W. Flora of Yunnan. Flora Yunnanica (Vol. 1–16). Beijing: Science Press (In Chinese). 1977–2006.
  67. 67. Zy W, Raven PH, Hong DY. Flora of China (Vol. 1–25). Beijing: Science Press, St. Louis: Missouri Botanical Garden Press. Available: Accessed 10 February 2014.
  68. 68. Corlett RT. What is secondary forest? J Trop Ecol. 1994;10: 445–447.
  69. 69. Danielsen F, Skutsch M, Burgess ND, Jensen PM, Andrianandrasana H, Karky B, et al. At the heart of REDD+: a role for local people in monitoring forests? Conserv Lett. 2011;4: 158–167.
  70. 70. Garcia NT, Meilby H, Sørensen M, and Theilade I. 2015. Wild edible plant knowledge, distribution and transmission: a case study of the Achí Mayans of Guatemala. J Ethnobiol Ethnomed 2015;11:52. pmid:26077151
  71. 71. Webb EL. Wijedasa LS, Theilade I, Merklinger F, van de Bult M, Steinmetz R, et al. 2016. James F. Maxwell: Classic Field Botanist, Inimitable Character. Biotropica 2016;48: 132–133.
  72. 72. Danielsen F, Jensen PM, Burgess ND, Altamirano R, Alviola PA, Andrianandrasana H, et al. A multi-country assessment of tropical resource monitoring by local communities. Bioscience 2014;64: 236–251.
  73. 73. UNESCO Local and Indigenous Knowledge Systems (LINKS). UNESCO portal Available: Accessed 22 January 2016.
  74. 74. Padmanaba M, Sheil D, Basuki I, Liswanti N. Accessing Local Knowledge to Identify Where Species of Conservation Concern Occur in a Tropical Forest Landscape. Environ Manage. 2013; 52(2):348–359. pmid:23633002
  75. 75. Pitman N. Social and Biodiversity Impact Assessment Manual. Climate, Community and Biodiversity Alliance. 2011.
  76. 76. Skutsch M, Zahabu E, Karky B, Danielsen F. The costs and reliability of community forestry monitoring. In M. Skutsch ed. Community forest monitoring for the carbon market. London: Earthscan. 2011 p. 73–81.
  77. 77. Mitchard ETA, Feldpausch TR, Brienen RJ, Lopez-Gonzalez G, Monteagudo A, Baker TR, et al. Markedly divergent estimates of Amazon forest carbon density from ground plots and satellites. Global Ecology and Biogeography. 2014. pmid:26430387
  78. 78. Thompson OR, Paavola J, Healey JR, Jones JPG, Baker TR, and Torres J. Reducing emissions from deforestation and forest degradation (REDD+): Transaction costs of six Peruvian projects. Ecology and Society 2010; 18: 17. URL:
  79. 79. Danielsen F, Burgess ND, Balmford A, editors. Special issue: Monitoring matters: examining the potential of locally-based approaches. Biodiv and Cons. 2005;14: 2507–820.
  80. 80. Gibson CC, Williams JT, Ostrom E. Local enforcement and better forests. World Development. 2005;33: 273–284.
  81. 81. Danielsen F, Balete DS, Poulsen MK, Enghoff M, Nozawa CM, Jensen AE. A simple system for monitoring biodiversity in protected areas of a developing country. Biodiv and Cons 2005; 9:1671–1705.
  82. 82. Danielsen F, Jensen PM, Burgess ND, Coronado I, Holt S, Poulsen MK, et al. Testing focus groups as a tool for connecting indigenous and local knowledge on abundance of natural resources with science-based land management systems. Conserv Lett. 2014;7: 380–389.
  83. 83. Higgens MA, Ruokolainen K. Rapid Tropical Forest Inventory: A comparison of techniques based on Inventory Data from Western Amazonia. Cons Biol. 2004;18: 799–811.
  84. 84. Gordon JE, Newton AC. 2006. Efficient floristic inventory for the assessment of tropical tree diversity: A comparative test of four alternative approaches. For Ecol Manage. 2006;237: 564–573.
  85. 85. Beehler BM. Using village naturalists for tree plot biodiversity studies. Tropical Biodiversity. 1994;2: 333–8.