Get Involved
About Us
Our Members

Sustainable Harvest of Wild Plant Populations


Harvesting plant resources from wild populations affords a number of benefits. For many medicinal and aromatic species, wild material is thought to be qualitatively superior to cultivated stock because plants that grow in natural, more adverse environments typically produce increased amounts of secondary metabolites.1 Wild-harvested plants frequently provide an important source of income for local communities and can also offer a powerful incentive for conserving natural habitats.2,3 Finally, harvesting plants from the wild, as opposed to cultivating them, allows their populations to keep growing, regenerating, and evolving in response to an ever-changing array of selective pressures. Wild plant populations typically exhibit a high degree of genetic diversity; cultivated plants, as a rule, do not.

The risk of harvesting wild populations is that they can be overexploited and degraded easily. Unfortunately, this appears to be happening with increasing frequency all over the world. It is estimated that about 20% of the wild-harvested sources of medicinal plants worldwide are currently exhausted or threatened by overexploitation.4 For example, several of the most valuable wild rattan species (from various genera in the Arecaceae family) in the Greater Mekong region of Southeast Asia are seriously depleted,5 numerous native fruit trees in Amazonia are disappearing,6 and the gaharu (Aquilaria malaccensis, Thymelaeaceae) trees of Indonesia and Malaysia — whose fungal-infected heartwood is a valuable incense, perfume, and medicine — have been in Appendix II of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) since 1995.7 The interaction between wild plant populations, uncontrolled exploitation, and demanding market economies has been, to say the least, rather dismal.

There is a finite quantity of leaves, stems, roots, fruits, seeds, or exudates that can be harvested each year from a wild population of plants. Once this limit is passed, the regeneration dynamics of the population are affected and the number of individuals that compose it will decrease. If exploitation continues at the same intensity, the population will eventually disappear. To avoid the dangers of overexploitation, several assessment tools have been developed with criteria for ranking a species’ resilience to harvesting based on life history characteristics, habitat specificity, and demand.8,9 For species that receive a low score (i.e., “high-risk” species), cultivation rather than wild harvesting is recommended. While these efforts are undeniably useful, it is important to note that even the most “low-risk” species can be overexploited and depleted, and wild populations of even the most vulnerable species can be harvested sustainably. What is needed to avoid the former and achieve the latter is a clearer understanding of the productive capacity of the resource, a conscientious management effort, and a modicum of control over both the people and the plants.

Over the past 35 years, I have been involved with numerous projects in collaboration with local communities focused on the management and sustainable harvest of wild populations of valuable plant resources.10,11 This work was conducted in tropical regions around the world with different ethnic groups and different types of botanical resources. Most of the resources I studied were trees (harvested for timber, carving wood, fruit, oleoresin, or useful bark) and vines (such as rattan), but several species of woody perennials or herbaceous annual plants were included in the management initiatives of some communities. Similar management protocols were used in every case. Building on this research, the purpose of this article is to review the conceptual foundation that underlies the sustainable harvest of wild plant populations and outline the basic data requirements for developing a management plan. The benefits of harvesting botanicals from the wild are numerous, and doing so in a sustainable manner is not as complicated or difficult as it might seem.

Basic Population Management Concepts

From a management perspective, a wild resource is most usefully described in terms of two parameters. The stock of a resource is the number of stems or individual plants of the resource-producing species (whether tree, shrub, or herb) that is found in the forest or field at one point in time. The yield is the rate at which a particular resource grows, multiplies, or increases in quantity each year. The amount of new timber (cubic meters/hectare), rattan cane (m/ha), bark (kg/ha), or latex (liter/ha), or the number of fruits, leaves, or stump sprouts that a species produces each year is the yield of that resource.

There is a close relationship between the current stock and the yield of a wild resource. Abundant species with dense populations have a large stock and produce a large amount of harvestable resource each year, while sparse, low-density populations exhibit a much lower yield. As the stock of the resource increases within a given area, so does the annual yield. The converse is also true.

The relationship between stock and yield can have profound consequences for the sustainable exploitation of a wild plant resource. In order to exploit the same species year after year in the same place, it is important to harvest no more than its annual growth each year. Harvesting more than the annual growth in a single year will diminish the current stock of the resource, and the resource will be depleted over time. This functions in much the same way as the relationship between the principal and the interest of an endowment or a savings account. As long as annual withdrawals are less than or equal to the interest generated by the principal each year — and the principal is left intact — such withdrawals can continue (theoretically) for perpetuity. However, if withdrawals are larger than this each year, the principal gets smaller and the account is eventually overdrawn.

Resource Depletion Scenario

A graphic example of this process is shown in Figure 1. The initial stock of rattan in the forest is assumed to be 1,000 commercial canes (at least four meters long), with each of these canes exhibiting an annual growth rate of two meters per year. Based on the stock and the growth rate, the annual yield from this rattan population is estimated to be 500 canes. By the end of Year 1, the initial stock of 1,000 rattan canes has produced 500 new canes (i.e., the existing stock is now 1,500 canes). During the first harvest, an order for 700 canes is fulfilled, reducing the stock to 800 canes. In Year 2, the reduced stock yields less new material (i.e., 400 canes), but harvest rates are held constant at 700 canes to satisfy the demanding buyer. By Year 3, the population now exhibits a stock of 500 canes and barely grows enough to produce 250 new canes. The final harvest of 700 canes reduces the stock to 50 canes, which will produce only 25 new canes and certainly not support another commercial harvest. In this example, the rattan population is severely overexploited in only a few years. To repeat, it is of utmost importance that no more than the annual growth of a wild resource be harvested each year. Defining this critical harvest limit will inevitably require the collection of baseline data.

Data Requirements

The ability to exploit a wild plant population with minimal ecological impact improves dramatically when more is known about the species.12,13 Regardless of the species, habitat, or plant part harvested, the most important ingredient required to achieve a truly sustainable form of resource use is information,14 such as quantitative data on the stock and annual yield of the plant.

The stock of a population is assessed through a forest inventory. Foresters and ecologists have developed a variety of plot sizes and shapes and methodologies to survey wild plants, and trade-offs of time, cost, and statistical precision are inherent in each one.15,16 Inventories conducted at the village level are strongly recommended, but they require a sampling methodology that is easy to understand and implement and that does not involve the use of specialized or expensive field equipment. Based on these considerations, and after many years of experimentation with different inventory methods in collaboration with villagers, a systematic sample composed of parallel, 10-meter-wide transects appears to work best for counting and measuring forest resources.

Transects should run straight along a pre-determined compass bearing. The bearing should be chosen so that the transects run across topographical features (i.e., up and down slopes and across rivers, rather than parallel to them). Orienting the transects in this way will maximize the number of different habitats encountered in the inventory and provide a more representative sample of local habitats. The distance between transects determines the sample intensity (i.e., the percentage of the total area that is included in the inventory). The closer the transects are together, the higher the sample intensity. For example, separating each 10-meter-wide transect by 100 meters would give a sample intensity of 10%, while separating the transects by 200 meters would yield a sample intensity of 5%. Given the paucity of density data that exists for even the most economically important forest resources, a 5% sample would be sufficient to estimate the existing stock of a wild resource.

Finally, in addition to counting the plants of a particular species in the inventory, individual specimens should also be measured or visually estimated into size classes. Diameter at breast height (DBH) is the most convenient measurement for trees, and basal diameter is an appropriate size-class parameter for shrubs. Height classes can be used for vines, palms, and smaller woody perennials, and life stages (i.e., seedling, sapling, juvenile, adult) can also be used for species such as agaves (Agave spp., Asparagaceae) that are difficult to measure. The reason for assigning individuals to size classes or life stages is to divide the population into groups that are likely to be experiencing the same growth conditions.

The yield of different plant resources is quantified through a growth or yield study. Growth studies are used when the resource of interest is stem or root tissue (e.g., timber, rattan, ginseng [Panax spp., Araliaceae] roots); yield studies are used to measure fruit production, latex yield, or bark growth. The objective is the same: to quantify the size-specific annual production of the resource of interest in different habitats. How much rattan, timber, bark, latex, or floral nectar, or how many native fruits or leaves, are produced by a given species within a particular habitat? This is an important consideration, because it will ultimately determine how much of a given resource can be harvested from the wild.

In selecting the sample individuals to measure for yield studies, care should be taken to choose individuals that represent a range of different sizes (or ages), canopy covers, and habitats. Plants typically grow faster (i.e., produce more wood, leaves, or fruit) when there is more available light and/or nutrients; taller plants with a better canopy position also usually grow faster than suppressed individuals. Both fast-growing and slow-growing specimens should be included in the selection of sample plants for the yield study. If only fast-growing individuals are sampled, the annual yield of the population will likely be overestimated, and too much of the resource will be harvested from the area. Conversely, if the selection of sample plants contains too many slow-growing individuals, yield will be underestimated, and resources that could have been harvested sustainably will be left in the field. While this may seem obvious, it is often tempting to bias the selection of sample individuals toward the smaller-size classes, which are often more numerous, accessible, and easier to measure. These individuals also usually grow more slowly.

Defining a Sustainable Harvest

With data on the density and size-class distribution of a plant population and relatively precise yield estimates, it is possible to calculate the total quantity of the resource produced by the species in a single year. Multiplying the size-specific growth rate by the number of individuals in that class and then totaling the result over all classes provides an estimate of total population productivity. For resources like timber, rattan cane, bark, and roots, the harvest of which inevitably kills the plant, this estimate represents the limit of how much material can be sustainably harvested from the population in one year. If the population produces a total of 1,000 kilograms of new bark each year, then this is all that can be exploited. The exact number of trees that are harvested will depend on the size and bark volume of the trees that are felled or de-barked, but the total amount of bark extracted from the site in a single year should never exceed 1,000 kg.

For fruits and seeds, which can be harvested without killing the plant but impact the regeneration dynamics of the population, defining a sustainable harvest limit is a bit more complicated. Clearly, not all of the fruit produced by the population can be harvested year after year. It is necessary to determine how many fruits need to be left on the site to facilitate the recruitment of new seedlings into the population. This can be accomplished in two main ways: through successive approximation (i.e., harvesting a certain percentage of the fruit crop, monitoring the impact on regeneration, and adjusting subsequent harvests as warranted17) or by dividing the management area into separate harvest units of comparable size and leaving one unit untouched or fallow each year; the fallow plot should be rotated sequentially and be different each year.

Impact Monitoring

Even when harvest limits are respected and maintained, collecting commercial quantities of resources from wild plants can cause changes in the population being exploited. Shifts in biophysical parameters (e.g., rainfall, temperature, presence or absence of pollinators or predators) can cause plant populations to produce varying numbers of seedlings or to exhibit varying rates of mortality each year. These demographic changes, in either direction, should not go undetected. It is best to set up a series of permanent plots in the forest and re-inventory them every five years or so.14

In particular, changes in the size-class distribution of the harvest population that suggest adverse effects on rates of regeneration should be noted. To assess this change, the results from the initial inventory should be used to produce a size-class histogram (i.e., a graph that shows the number of individuals in each size class). A histogram serves as a baseline and represents the initial structure or “pre-harvest” condition of the population.

In spite of the variety of different reproductive and growth strategies used by plants, wild populations exhibit a limited number of size-class distributions. Three of the most common distributions are shown in Figure 2 with a representative example of each type. Size classes indicated are diameter classes (cm DBH) for the tree species (tengkawang nyamuk and uvos) and height classes (m) for the rattan species (may sap).

The Type I size-class distribution, illustrated by tengkawang nyamuk (Shorea atrinervosa, Dipterocarpaceae), a valuable seed oil-producing tree in western Borneo, displays a greater number of small individuals than large ones and an almost constant reduction in number from one size class to the next. This type of population structure is characteristic of shade-tolerant plants that maintain a relatively constant rate of recruitment. It is probable that the death of an adult tree will be supplanted by the growth of individuals from the smaller size classes.

The Type II size-class distribution, illustrated by uvos (Spondias mombin, Anacardiaceae), a popular native fruit tree from the Peruvian Amazon, is characteristic of species that show discontinuous or periodic recruitment. The actual level of seedling establishment may be sufficient to maintain the population, but its infrequent recruitment causes notable discontinuities in the structure of the population as the newly established seedlings and saplings grow into the larger size classes. This type of diameter distribution is common among tree species that depend on canopy gaps for regeneration. Many of the mast-fruiting Dipterocarpaceae in Southeast Asia also exhibit a Type II distribution (although not tengkawang nyamuk, which fruits every year).18 The large number of individuals in the first size class, as shown in the uvos histogram, suggests that gap colonization by this species was particularly successful in the past.

The final size-class distribution, Type III, is illustrated by may sap (Calamus dioicus, Arecaceae), an important commercial rattan from the Greater Mekong Region. A Type III distribution is seen in species whose regeneration is severely limited for some reason. Population density is low, seedling numbers are greatly reduced, and very few individuals are in the intermediate size classes. Type III distributions are frequently encountered among light-demanding, pioneer species that require large canopy gaps for regeneration. In the absence of factors required for regeneration, these species may disappear from the forest. The situation with may sap is aggravated by intensive commercial harvesting of adult plants for their cane.

The size-class distribution of a population is extremely sensitive to the population’s regeneration rate. A Type I distribution can easily change into a Type II if existing recruitment levels are diminished or interrupted. Further constraints on regeneration may drive the population to a Type III distribution. It is perhaps most useful to view these three distribution types as a single sequence through which a wild species passes on its way to extinction.

Size-class histograms should be constructed every time a harvest area is re-inventoried, and the distribution of individuals in the current population should be compared to the “baseline” histogram from the original inventory. Are there notable changes in the shape of the distribution? Have the number of individuals in the initial size class decreased (i.e., is the population recruiting fewer new individuals)? Or, ideally, has the general shape of the size-class distribution remained relatively unchanged over the five-year period?

Pronounced changes in the size-class distribution should be accompanied by immediate reductions in the amount of material harvested from the population each year. These harvest reductions should be maintained until the structure of the population returns to baseline conditions. If, after five years, the number of individuals in the smaller size classes continues to decline, the harvest intensity should be further decreased and the response should be monitored. By conscientiously monitoring the population response to different harvest levels, a level of resource extraction that the population can support will eventually be determined. This is where sustainability happens.

Recommendations and Next Steps

The sustainable management of wild plant populations is both a science and an art. In most cases, the actual sustainability of an enterprise hinges more on the willingness of the collectors and buyers to follow the prescribed harvest limits than on the statistical rigor of the data used to estimate them. It is important not only to count the plants and measure the growth as accurately as possible, but also to make sure that everyone understands how and why these data were collected and to continually affirm that things will get better if the harvested controls are respected.

In terms of next steps:

Start small. Identify a specific plant resource-collector-market system in which information exists about the biology of the plant species, market conditions for the resource are favorable, and stakeholders are enthusiastic about the prospects of using wild resources sustainably.

Conduct a quantitative inventory of the plant population and start a yield study, ideally involving the collectors and/or local community.

Calculate a sustainable harvest level in collaboration with all stakeholders and initiate harvesting. Promote the sustainability initiative as much as possible.

Periodically re-inventory the harvest area to assess the impact of harvesting. Adjust harvest rates as necessary and carefully explain to all stakeholders the reason(s) for the modification.

Freely share information with colleagues, local authorities, retailers, business representatives, and the general public.

Both the harvested and the harvester ultimately benefit from sustainable resource use.

Charles M. Peters, PhD, is the Kate E. Tode Curator of Botany at the Institute of Economic Botany at the New York Botanical Garden and an adjunct professor of tropical ecology at the Yale School of Forestry and Environmental Studies. His recent book, Managing the Wild: Stories of People and Plants and Tropical Forests (Yale University Press and NYBG Press, 2018) describes more than three decades of fieldwork with local communities in tropical forests to sustainably manage wild plant resources.


    1. Schippmann U, Leaman D, Cunningham AB. A comparison of cultivation and wild collection of medicinal and aromatic plants under sustainability aspects. In: Bogers RJ, Craker LE, Lange, D, editors. Medicinal and Aromatic Plants. Wageningen, the Netherlands: Springer Netherlands; 2006;75-95.
    2. Shackleton C, Shackleton S. The importance of non-timber forest products in rural livelihood security and as safety nets: A review of evidence from South Africa. South African Journal of Science. 2004;100(11-12):658-664.
    3. Townson IM. Forest Products and Household Incomes: A Review and Annotated Bibliography. Tropical Forestry Papers No. 31. Oxford, UK and Bogor, Indonesia: Oxford Forestry Institute and Center for International Forestry Research; 1995.
    4. Hamilton AC. Medicinal plants, conservation and livelihoods. Biodiversity and Conservation. 2004;13(8):1477-1517.
    5. Evans TD. The status of the rattan sectors in Lao People’s Democratic Republic, Vietnam and Cambodia-with an emphasis on cane supply. In: Dransfield J, Tesoro F, Manokaran N, eds. Rattan: Current Research Issues and Prospects for Conservation and Sustainable Development. Rome, Italy: Food and Agriculture Organization; 2002;115-144.
    6. Vazquez R, Gentry AH. Use and mis-use of forest harvested fruits in the Iquitos area. Conservation Biology. 1989;3(4):350-361.
    7. Wyn LT, Anak NA. Wood for trees: A review of the agarwood (gaharu) trade in Malaysia. Petaling Jaya, Malaysia: TRAFFIC Southeast Asia; 2010.
    8. Castle LM, Leopold S, Craft R, Kinscher K. Ranking tool created for medicinal plants at risk of being overharvested in the wild. Ethnobiology Letters. 2014;5:77-88.
    9. American Herbal Products Association. Section 4: Wild collection assessment tool. In: Good Agricultural and Collection Practices (GACP). Silver Spring, MD: AHPA. July 7, 2017. Available at: Accessed March 4, 2018.
    10. Peters CM. Community forestry and sustainability research at The New York Botanical Garden. Brittonia. 2016;68(3):290-298.
    11. Peters CM. Managing the Wild: Stories of People and Plants and Tropical Forests. New Haven, CT: Yale University Press; 2018.
    12. Foster S. Harvesting medicinals in the wild: The need for scientific data on sustainable yields. HerbalGram. 1994;24:10-16.
    13. Cunningham AB. Applied Ethnobotany: People Wild Plant Use and Conservation. London, UK: Earthscan; 2001.
    14. Peters CM. Sustainable Harvest of Non-timber Plant Resources in Tropical Moist Forest: An Ecological Primer. Washington, DC: Biodiversity Support Program; 1994.
    15. Husch B, Beers TW, Kershaw JA. Forest Mensuration. 4th ed. Hoboken, NJ: John Wiley and Sons, Inc.; 2003.
    16. Peters CM. Beyond nomenclature and use: A review of ecological methods for ethnobotanists. In: Alexiades MN, ed. Selected Guidelines for Ethnobotanical Research: A Field Manual. Bronx, NY: NYBG Press; 1996;241-276.
    17. Peters CM. The Ecology and Management of Non-Timber Forest Resources. World Bank Technical Paper No. 322. Washington, DC: The World Bank; 1996:82-85.
    18. Peters CM. Illipe nuts (Shorea spp.) in West Kalimantan: Use, ecology, and management potential of an important forest resource. In: Padoch C, Peluso NL, eds. Borneo in Transition: People, Forests, Conservation, and Development. Kuala Lumpur, Malaysia: Oxford University Press; 1996;230-244.