OPERATIONAL USE OF THE PROPOSED INDICATORS AND VERIFIERS

To use the above indicators effectively in forest management, we recommend using:

1. A species selection process to provide a preliminary screening of vulnerable species, followed by,

2. Risk assessment based on the indicators.

Based on the information generated, and assessment of the probability that management actions will reduce risks to sustainability (Indicator 5), it is recommended to apply a decision making system to determine if the targeted species can be deemed to be managed sustainable or, at least, cannot be deemed to be threatened under the prevailing or proposed forest management system. As the species selection process is intended to identify those species which are highly susceptible to loss of genetic variation, it can be concluded that, if these species are found to be sustainable in the management system applied, the other species (which by definition are less vulnerable and less susceptible), are also in the same condition.

It is expected that the indicator species will be examined first using at least one of the demographic verifiers and, subsequently, if added genetic precision is desired, using at least one of the genetic verifiers. To provide information for comparison, since no common standard exist the verifiers must be assessed both in a reference population and in the affected population.

The value of using the above indicators stems from their regular assessment over time, to reveal trends. Without the benefit of baseline data, the value of initial assessments of verifiers will be limited. This drawback can be overcome by sampling populations which are thought to represent different levels of sustainability. By taking such an approach, the initial values of verifiers can be used to provide a preliminary understanding of the state of forest management in relation to sustainability, without awaiting true time series data derived form repeated measurements. However, the use of such a “pseudo-chronosequence” confounds variation in verifiers due to environmental heterogeneity, and comparisons must therefore be treated with caution. It is only through second and subsequent assessments, of the same site, that direct estimates of temporal dynamics are possible.

If the forest management unit is large, the distribution of samples at the appropriate scales must ensure that micro- and macro-scale influences are captured and measured. The time scale becomes very important when changes and rates of change are to be assessed. However, it is rarely possible to directly assess changes that occur in evolutionary time scales and apply them to forest management decisions which have to be made in time scales of months or a few years. Therefore, important differences will often be estimable only by assessing, at regular intervals, differences among contemporary reference populations with differences in age or environmental factors, and not by following cohorts over several years or generations. While applying such compromise solutions, research should be carried out in parallel to validate the assumptions on which the substitutions of variables are made.

Species selection

The use of criteria and indicators to assess sustainability will usually be undertaken when an intervention is planned or, alternatively, following a management intervention (harvesting of wood and NTFP, etc.). Naturally occurring, spatial and temporal variation in genetic variation will make it impossible to assess historical trends in genetic variation, and thus to establish baseline data prior to forest management interventions. Only through a series of subsequent assessments will such information become available. To establish baseline information from other comparable locations, or “reference populations”, must generally be used. Such data can provide information used to assess the changes due to the management interventions.

The indicators of genetic processes cannot be applied to all species in the forest and hence a system for selecting genetic “indicator species” is needed. As noted by Brown et al. (1997), ideally taxa monitored should be selected from all major groups of organisms. However, the amount of information that can be derived from verifiers partly depends on the amount of prior knowledge of the species and its genetics. Especially in tropical forests, little is known about forest related species in general, however, trees are often the best studied organisms, and are also critical in providing habitats for other organisms. Standard forest inventories provide information from which many verifiers can be derived, however, usually only information on forest tree species is included in them.

We present below an example of a tree species selection process, noting however that, ideally, other species of plants, vertebrates and invertebrates should also be included in the exercise, resources permitting.

Since the assessment of any target species requires a certain amount of prior research and documentation of their status in reference populations as well as in the test sites, there are obvious opportunities for reductions in costs if the same species are selected in all FMU’s within a given region in a country. The accumulation of data from reference populations over time will then make the process of assessing and monitoring genetic indicators progressively cheaper.

Species selected should be those which are thought to be most susceptible to adverse affects from the forest management interventions under consideration. The species will vary according to the location and the type of intervention. If several different types of management intervention are practiced and different types of product harvested, a number of different target species will likely need to be identified. As Brown et al. (1997) noted, indicator species, or indicator taxa, should be biologically, ecologically and taxonomically representative of the types of species found in the area. Thus, although trees may be the major initial species monitored, opportunities to incorporate information on other organisms, such as birds, mammals, and invertebrates, should be sought when possible. Criteria for selection of indicator tree species for different types of interventions are proposed below.

A. Selective harvesting of wood, selective logging

In the context of this paper, “logging” refers to the planned and regulated harvesting of wood by means of selective removal of individual large trees. The intensity of harvesting can vary from less than one tree per hectare in parts of central Africa to more than 10 trees per hectare in parts of southeast Asia. As noted above, “logging” is not used here in reference to the process of converting forests to other lands uses, as this is obviously a non-sustainable activity as concerns forestry, and is designed as such. Similarly, the unregulated cutting of any and all mature individuals of some species due to their high value is usually undertaken with little consideration for sustainability, and is therefore not considered in this context. An example of this latter situation is mahogany (Swietenia mahagoni), which over the past decades has reportedly undergone extreme genetic erosion due to the harvesting of superior phenotypes in its natural area of occurrence, in the Caribbean and Central America (Rodan et al., 1992).

The proposed characteristics of indicator species used for assessing the impact of logging can be sub-divided into two categories. The first category includes those characteristics for which information will normally be easily available, probably without the need for additional field visits. The second category includes those characteristics in which little information is available, and which may require additional field work and other investigations. The proposed characteristics are generally listed in order of importance, but it is not essential that a selected species meet all characteristics, and indeed it is desirable that a mixture of species is included. For example, while the first characteristic refers to the value of the species for logging, some non-commercial species should also be included.

a) Characteristics for which information is readily available:

• The species is harvested, and has high value (influences drift, selection)

• The species is rare (drift, mating system)

• The species has a spatial distribution which makes it susceptible to adverse impacts of logging - see note below

• The reproductive niche of the species requires shade (mating system)

• The species is dioecious (drift, mating system).

b) Characteristics for which information is generally less readily available:

• The species is ecologically valuable - it is a “keystone” species^[3]

• The species flowers only after attaining a large size or advanced age (influences mating system, drift)

• The pollinators/seed dispersers of the species are highly specific (mating system).

Most of the above characteristics are exclusive: species either meet them or do not meet them. This allows species to be screened out, leaving relatively few candidate indicator species. The remaining candidate species can then be prioritized based on those characteristics which are scaleable, such as largest maximum size, degree of aggregation and, possibly, economic value. In this way, a number of indicator species, perhaps up to 10, can be prioritized in the final list.

As mentioned above, the characteristics can be applied in any area to select those species which are the most appropriate ones to use as indicator species in a given ecological region. It is envisaged that this would be done in consultation with an ecologist with detailed local knowledge. As examples of this process, species which may be selected as indicator species in southern Cameroon, and in Kalimantan, are shown in Table 4.

Table 4. Species potentially susceptible to loss of genetic variation from logging in Southern Cameroon and Kalimantan.

B. NTFP Harvesting

Selecting indicator species for effects of the harvesting of non-timber forest products is complicated by the fact that so many different products are harvested, involving a large range of different harvesting procedures. The likely susceptibility of species can be ranked according to the impact of the harvesting method on the individuals or populations from which such products are collected. For example, collection of reproductive structures (flowers, fruits or seeds) can have a potentially large impact on genetic variation. Harvesting of complete individuals to derive products from them, which includes reproductive structures, would be expected to have an even larger impact, as it not only removes the possibility of reproduction in the current year, but also future reproduction. Also the harvesting of immature individuals will have a large potential impact on genetic variation. The proposed characteristics for indicators species for NTFP collection are:

a) Characteristics for which information is readily available:

• Number of high value individuals being harvested in whole or in part (influences drift, selection)

• Level of species rareness (drift, mating system)

• Distribution of the species: grows in clumps and has a predictable distribution (e.g. always on ridge top) making it easy to find and to harvest (drift, migration)

• Complete individuals are harvested, often when immature, or (if no species in this category), reproductive structures are harvested (drift, selection)

• The species is not amenable to cultivation (drift)

• The species is dioecious (drift, mating system).

b) Characteristics for which information is generally less readily available:

• The species is ecologically valuable - it is a “keystone” species (influence is direct and indirect)

• The species flowers only after attaining a large size (drift).

As examples of this process, species which could be selected as indicator species in southern Cameroon, and in Kalimantan are shown in Table 5.

Table 5. Species potentially susceptible to loss of genetic variation from collection of non-timber forest products in Southern Cameroon and Kalimantan.

C. Fire

In natural circumstances, fire is an unusual event in humid forests. However, in drier forest ecosystems species are frequently ecologically adapted to fire. Human activities often greatly increase the incidence of fire, and wildfires may result in considerable economic and ecological damage. Proposed characteristics of indicator species used for assessing effects of fire are:

• The species does not have fire-adapted characteristics such as thick bark or underground dormant structures (influences drift)

• The reproductive niche of the species requires organic matter and/or shade (drift, migration)

• Fruit/seed of the species is mature at the beginning of, or during, the dry season (drift)

D. Additional considerations

In addition to identifying indicator species for different types of intervention or effects, additional considerations may help identify species of special interest. Such considerations may include:

• known state of threat or endangerment

• degree of endemism (influences drift, migration)

USE AND APPLICATION OF INDICATOR SPECIES

The complexities and interactions among indicators makes it difficult to ensure sustainability over time, even if these are regularly monitored. It is much easier to identify conditions of non-sustainability, though still with some considerable margins of error. If management interventions do not have effects on given species which render the management clearly unsustainable from a genetic point of view, it can be assumed that the levels of risk are acceptable. However, there can never be a full guarantee that given species do not go extinct, or are not impoverished genetically due to natural or manmade influences. It is necessary that the standards of acceptability are sufficiently rigorous to ensure that even moderate threats to sustainability will flag a warning or lead to recommendations for changes in forest management practices. Thus, in a simplified manner, one could say that forest practices can either:

• lead to non-sustainability, or

• do not pose unacceptable risk of causing species extinction or severe genetic degradation.

A third scenario is that, although likely non-sustainable under present conditions, if modified forest practices can help alleviate negative impacts in the future - “conditionally acceptable” (see discussion under Indicator 5, above). If the forest practices planned or in effect, pose no unacceptable risk to the most sensitive processes acting on the most sensitive species, we then infer that there is also likely to be no unacceptable risks for less susceptible species. We propose the following process to determine the sustainability of management in terms of genetic conservation:

Step 1.

The types of interventions within the FMU are recorded and mapped. Using Table 3 as a guide, the specific indicators that should be assessed are determined.

Step 2.

With advice from forest ecologists knowledgeable of local conditions, those species which are likely to be most susceptible to negative impacts of the forest management interventions planned, or taking place, in the FMU are identified. These species should be identified on a regional basis within each country, to promote ease of comparison among FMUs.

Step 3.

The most susceptible species for each intervention is assessed, initially using demographic verifiers, subsequently supplemented by genetic verifiers if greater precision is required.

Step 4.

Information derived from verifiers of Indicators (1) through (4) is combined to obtain an overall assessment of sustainability. The most meaningful way in which this information can be used should be further researched. However, general principles can be established for a first evaluation based on combining information from multiple indicators. For example, if management philosophy calls for a rigorous assessment of sustainability, it follows if verifiers of any one of the four indicators fall below the critical level, this indicates overall non-sustainability. A more conservative approach may recognize that, for example, if the critical values for indicator 2 (directional change) are exceeded, this may not indicate non-sustainability if the values for indicators 3 (migration) and 4 (reproductive system) are very good. Such an approach recognizes the possibility of compromises.

Irrespective of the rigour of evaluation adopted, two possible conclusions can be reached:

• IF the most susceptible species are found not to face unacceptable risks, it is concluded that the management interventions undertaken can be considered to support sustainability.

• IF the effects of management interventions on the most susceptible species are found to lead to non-sustainability of the resource, then a system is used to establish the critical points in the species lists. This may be a bracketing strategy, in which the second species tested may be the 5th most susceptible, and if it fails the 15th, etc. Then, if resources permit, a species halfway between the positions of failure and lack of failure can be tested until the point of non-sustainability is established and action taken to gradually improve the situation (see below).

Step 5.

Any species for which the conclusion in Step 4 was “non-sustainability” is then assessed in order to establish what measures must be taken to improve the situation. If such action is feasible, and implementable in the immediate future, the status of “unsustainable” is modified to “conditionally acceptable”.

PREVENTIVE OR RESTORATIVE ACTIONS

The potential of the genetic system to recover would indicate that sustainable forestry is in fact biologically possible. For the targeted species, various management options may be available, and some of them would be feasible depending on the state of the system as determined by Indicators (1) through (4).

For each of these four indicators, critical levels of the verifiers can be determined for developing a genetic “triage” system, as described above (Koshy et al. 2001). For example, if levels of genetic variation are so low that demographic recovery is unlikely, if prospects for adaptation are so poor that future adaptation will not be possible, if migration is so scarce that colonization is unlikely, or if the mating system cannot generate acceptable levels of genetic variation, then little can be done to conserve the system, and the methods used can be considered to have led to non-sustainable management. If assessment of the indicators shows acceptable levels of risk, the methods of management can be considered to promote sustainability. Finally, if some of the indicators are below critical levels, a modification of the management practices may improve overall sustainability. If such modifications in management interventions help improve the status of those indicators which have been below critical levels, then conditional acceptability of management methodologies could be declared.

The form that corrective measures take will depend on the degree to which improvement in sustainability is required. Some examples of interventions in a FMU in which harvesting of wood is the prime activity, in a rough order, from the most drastic to the more benign, include:

• planting a genetically variable population for species that have become genetically impoverished (“artificial recolonization”),

• changing species to be harvested to more “robust” ones,

• reducing harvest levels or modifying cutting regimes to leave sufficient residual populations for regeneration,

• shifting the wood harvesting operations to environmentally less sensitive areas,

• releasing potential parents from competitive suppression through silvicultural interventions,

• carrying out site amelioration to favor regeneration of target species,

• increasing the rotation age to allow more time for recovery, raising the cutting size limits, or implementing other measures to moderate the indirect effects of logging.

It may then be possible to array the forest management practices recommended for sets of species occurring in a given area, and that set would constitute a recommended forest management system from the point of view of genetic conservation as an integral part of management. It is expected that a finite set of practices will be found to be satisfactory for the threats facing many or all species in a given area and hence that a single management prescription for an area can be derived. Such management prescriptions must be continuously monitored, and must be flexible enough to allow the introduction of adjustments in accordance with new know-how and experiences.

The verifiers for the success of these actions are the same as those for Indicators (1) through (4). Added to these is the existence of ex situ resources, and the operational feasibility of restoration techniques.

Further development of decision support systems are needed to assist forest managers in assessing risks and in judging when the political value of additional documentation and justifications is worth its cost.