While the need to monitor biological diversity in general, and genetic diversity in particular, is well accepted, and is incorporated into all the forest criteria and indicator processes, the techniques for doing so are not yet well developed. A considerable amount of effort has been directed at this issue in the more mature criteria and indicator processes: Pan European, Montreal, ITTO and also by CIFOR, but the impetus seems to be waning. Hopefully, further indicator development work will take place, if other processes receive adequate support. The concept has been supported in a recent report of the Ad Hoc Technical Expert Group on Forest Biological Diversity, established under the aegis of the Convention of Biological Diversity (Anon, 2001).
In most criteria and indicator processes, the biological diversity indicators were formulated during the original negotiations and early discussions over the agreement, and attention has since been concentrated on finding ways to measure, assess, monitor and report on them. Limited attention seems to have been given in most processes to revision of the indicators so far.
The first issue to be resolved in any consideration of forest genetic diversity is just what aspect of biological diversity to cover. Forests are very complex biologically, the degree of complexity generally increasing from colder to warmer climates. It is obviously not feasible to evaluate biological diversity in its totality, so those processes that specify quantitative indicators have sought ways of encapsulating a trend in genetic diversity by monitoring a given, or a few (often rather arbitrarily determined) key species. The assumption has been that the key species will reflect the ecosystem as a whole and that they will thus reflect the status and trends in regard to other species. This is a rather sweeping assumption that may be more relevant in some forest types than others.
The next question is - how do we assess genetic diversity? There are number of biochemical techniques that have been used to do this at molecular level- see Box 1. A very complete explanation of the procedures involved is given by Glaubitz and Moran (2000). While biochemical techniques have been widely used for research purposes, they have not yet been adopted for routine use in a forest genetic diversity indicator. They produce a great deal of detailed information about the species under study, but making a link between changes in them to sustainability of forest management remains a complex task, not yet achieved.
Box 1: Biochemical techniques for assessment of genetic biodiversity
|
Several techniques are available for forest geneticists: chemical analysis of isozymes, terpenes, restriction fragment length polymorphisms (RFLPs), microsatellites, random amplified polymorphic DNAs (RAPDs), amplification fragment length polymorphisms (AFLPs) and polymerase chain reaction (PCR). Terpenes are not much used unless the terpene chemicals themselves are of interest. Isozyme analysis is relatively cheap to perform, but the other techniques are all expensive to perform and require a high level of technical skill, with good laboratory facilities. |
The desire for quantitative indicators of genetic diversity led to a considerable amount of work at CIFOR, as befits the role of a research organization concerned with sustainable forest management. In a preliminary account of this work Namkoong et al (1996) examined the scientific background to the evaluation of genetic diversity. They pointed out that “the evaluation of genetic diversity is complex’ and “genetic variation is difficult to measure and often cryptic in its effects on population and ecosystem dynamics.”
Making the connection between some biochemical measure and field performance is a critical issue, probably resolvable only by long term field trials.
Namkoong et al (loc.cit.) outlined the four critical processes of forest genetics that might be affected by forest management practices: random genetic drift, selection, migration and mating. Random drift refers to the changes in genotypic frequencies among generations due to random chance in small populations; selection describes the relative differences among genotypes in viability or reproductive success; migration affects the exchange of genes between populations that differ in genotypic frequencies, and mating is the process governing the recombination and assortment of genes between generations. “If these processes remain intact, then the biological capabilities are expected to be in at least as good a condition in the immediate future as they are now, and would imply a level of sustainability at least as good as now present” (Namkoong et al, 1996).
The above paper also addressed the serious issue of what species to select for study, since it is obviously not possible to apply indicators to all species in a forest. The authors suggested that there should be three criteria for selection of which species to study:
utility value or known ecological significance as a keystone species,
existence value, for species under threat of local extinction,
indicator value for species that are known to be paradigms of a large class of species.
However, the authors did not address an equally serious issue, that of sampling methods and intensity to arrive at statistically valid estimates of an attribute. There is also the fundamental issue that although a species may be forest dependent, it may well be affected, for better or for worse, by factors totally outside the forest environment.
The 1996 paper went on to propose indicators, together with several verifiers for each that, in theory, should provide information on the four critical processes listed above. Without exception the verifiers were difficult and expensive to measure and analyse. This exercise demonstrated a fundamental problem with any attempt to collect quantitative data on genetic diversity. It requires a very high level of scientific skill, and good laboratory facilities. It is inevitably expensive and resource-intensive.
Through biochemical techniques we are certainly capable of monitoring some of these critical processes. Glaubitz and Moran (loc. cit.) recommend guidelines for use of the various techniques, see Box 2.
Box 2: Guidelines for using biochemical techniques.
|
1. Applications that require large numbers of loci.
Use AFPLs or RFLPs 2. Applications that require high discrimination power
Use microsatellites. 3. Applications that require DNA sequence information
Use PCR and sequencing. |
In a later study by CIFOR, Stork et al (1997) reported that previous biological diversity indicators proposed by CIFOR were found to be deficient. Although their paper refers to forest biological diversity in general, the comments apply equally well to genetic diversity. They emphasized the complexity of the concept of biological diversity and found that it was impossible to make a direct assessment of it in anything other than a superficial manner. Consequently the authors proposed a move towards the evaluation of the processes that generate or maintain biological diversity. Stork et al also drew attention to the fragility of some basic concepts, in particular the idea that indicator groups or keystone species could provide a measure of biological diversity. Concern about the validity of this paradigm appears to be increasing.
In a still later study at CIFOR, Namkoong et al (FAO, 2002) set out a series of indicators and verifiers designed to assess the degree to which the key processes were sustained. The indicators and verifiers were an extension of those from the CIFOR Criteria and Indicators Generic Template (CIFOR, 1999). They were similarly very detailed, required extensive background information on the species under study, and concentrated on the tree component of the forest.
The Montreal Process has two national level indicators of genetic diversity:
the number of forest dependent species that occupy a small portion of their former range, and,
population levels of representative forest dependent species from diverse habitats monitored across their range.
A significant amount of research has taken place in the United States Forest Service exploring the practicability of these two indicators. Flather et al (2002), working at the Rocky Mountain Research Station, studied long-term records of bird species breeding in forest, and forest inventory data available from across the USA. They found that, although it was possible to derive broad estimates of range reduction for some species, it was not possible to quantify the relationship between range size and genetic variation between species. Mere decline in range size was also not necessarily as important as the nature of the decline, ie, whether it was a collapse into a smaller core region or a more dispersed and fragmented reduction. They also expressed concern at the definition of what constituted, “a significant decline”. How much decline in range results in too much genetic diversity decline and over what time period should the assessment be made? They concluded that much more research was required to understand better the processes affecting range occupancy and the implications of range decline.
In a related study, Sieg et al (2002) reviewed similar national datasets on population sizes of certain bird species that breed in forests. They pointed out that there is little empirical support for the basic assumptions that (a) genetic diversity can be tracked by monitoring population levels of a species, and (b) that the monitoring data will allow an understanding the factors affecting the species. There are few estimates of the rate at which genetic diversity is lost or gained as a function of changing population levels. The practice of using representative species to indicate the status of a functional group of species was commonly recommended but the general applicability of the concept is highly debateable. They also stated there is limited information on whether depression of genetic diversity does necessarily lead to a decrease in survival fitness.
This view was supported by the work of Mosseler and Rajora (1998), who studied genetic diversity in three Canadian tree species. In small, disjunct, populations of these three northern tree species, they compared a number of attributes: embryo development, seed yields, mating system parameters, and genetic diversity (measured by isozyme analyses). They concluded that the biochemical estimates of genetic diversity did not, in this case, reflect significant deterioration in reproductive capacity evident as reduced seed yield and seed quality, and suggested that direct measures of reproductive success may provide better indicators of genetic diversity.
This North American work therefore casts doubt on our ability to derive information from direct measures of genetic diversity that would provide meaningful input to an assessment of sustainable forest management. Much more research seems necessary before we can say, with any confidence, that a certain change in an indicator of genetic diversity has adverse implications for long-term forest sustainability.
The practicability of using an indicator based on biochemical data was also studied, in work done by Glaubitz and Moran for the Forest and Wood Products Research Corporation in Australia (2001, see Box 3), for two species of eucalypt that occur in a mixed forest type. Their work was designed to test the practicability of the Australian sub-national indicator, “Amount of genetic variation within and between populations of representative forest dwelling species.”(Anon, 1998b). It is especially interesting because it tackled the sampling issue in a systematic fashion and attempted to place genetic diversity in a direct management context.
This study confirmed the need for careful research before implementing an indicator based on biochemical data. It was noted that there was a distinct possibility that in a mixed species forest such as this, genetic diversity data for one species would indicate a positive situation, but in another species indicate a negative situation - and that was just for the tree component of the forest. We also need to know, for example, whether a reduction in allelic richness of 10% is something we should be concerned about. Is the critical level 10%, or is it 5% or 1%?
Boyle (2000) examined the broad issues surrounding the assessment of genetic diversity in forests. He pointed out the problems attached to the assessment of genetic diversity and drew attention to the need for further research to establish:
Selection of an appropriate conceptual framework for genetic criteria and indicators.
Choice of species to assess.
Methods to characterize or visualize genetic variation.
Methods to detect temporal changes and trends in genetic parameters.
Choice of thresholds or critical values to apply.
Combining information from multiple indicators.
Box 3: Case Study - Use of DNA markers to assess genetic diversity.
|
“This project is a case study in which DNA markers (RFLPs and microsatellites) were employed to examine the impacts of regeneration practices on genetic diversity in two contrasting tree species (Eucalyptus sieberi and E. consideniana) in the lowland mixed species eucalypt forest of East Gippsland. Genetic diversity parameters in samples of sapling regeneration from harvested coupes that were re-sown either aerially or via residual seed trees were compared to those in adjacent or nearby unharvested stands. In E. sieberi, the most common tree species on the study site, no significant differences were found for any of the diversity measures between two unharvested controls and regeneration after three different silvicultural treatments (clear-felling with aerial resowing, the seed tree system with site preparation by burning, and the seed tree system with site preparation by mechanical disturbance - two replicates each). The large data set from E. sieberi (823 trees in total) was used for empirical power analysis to determine the number of individuals that need to be sampled from pre- and post-harvest populations in order to provide adequate sensitivity to changes in genetic diversity levels. It was found that samples of 50 individuals per population would provide a greater than 80% probability of detecting a reduction in the number of alleles (allelic richness) of 10% or more (with a = 0.05). The E. consideniana study was then carried out according to these sampling guidelines. Here, diversity measures for saplings on two clear falls and two seed tree coupes were directly compared with those from bordering controls. Although E. consideniana was far less common on the study site than E. sieberi, no statistically significant reductions in diversity measures were found for this species either. However, though not quite statistically significant, there was some evidence that this species fared more poorly on seed tree coupes than on clear falls.” |
In another broad study of the possible use of genetic diversity indicators for national level State of the Environment reporting, Brown et al (1997) came to similar conclusions to Boyle. They put forward seven overall indicators that would apply equally to forests as any other ecosystem:
The number of sub-specific taxa
Population size, numbers and spatial isolation
Environmental amplitude of populations
Genetic diversity at marker loci within individuals and populations
Quantitative genetic variation
Interpopulation genetic structure
Mating.
Clearly, data collection for indicators such as these would be a massive task, even at the forest management unit level, so their practicality is highly questionable. The authors drew attention to the need for interpretation of the biological significance of changes in values between reporting periods and emphasized the need for a soundly based baseline value. They outlined some of the difficulties attached to interpretation, and pointed out the need for in-depth understanding of the processes involved.
They also acknowledged the need for further research with the comment that “without such information, the monitoring of the state of Australia’s genetic environment is open to misleading conclusions.”
There is thus a significant risk that hasty adoption of genetic indicators, without adequate understanding of how they can be interpreted, might lead to erroneous management decisions.
The Brown et al study raised another contentious issue affecting genetic diversity indicators. Just what is a “soundly based baseline value”? The issue was also discussed by Ghazoul and Hellier (2000). There is much discussion of this point taking place under the aegis of the CBD, but the sorts of proposals that are being put forward (e.g., pre-1993 condition) are essentially arbitrary. They ignore the fact that nothing is static in nature, with or without man’s intervention. Any arbitrary baseline runs the risk of generating misleading conclusions. We could also run the risk of “restoring or preserving a world that never existed” (Namkoong, 2001).
The issue of baseline values was explored in the report on the American LUCID project (Wright et al, 2002) described earlier in Section 2.2. The authors concluded, inter alia, that interpretation of data relating to indicators would be very difficult without baseline values (in their terminology, reference values), but they recognized the difficulty of setting such values without a great deal of background research. One of the benefits of a comprehensive exercise of this nature is that it does assist in focusing and prioritizing research programs.
A more indirect approach to indicators of genetic diversity was taken by the EU-Life Project, as reported by Spellman et al (2001). The genetic diversity aspect was part of a larger project, and used two indicators:
“Changes in proportion of stands managed for the conservation and utilization of forest genetic resources”
“Types and areas of regeneration”
In this case, the first indicator is phrased so that it covers the whole range of forest situations, from natural forests to plantations and focuses on the management of genetic diversity - a very pragmatic approach. The second indicator recognizes that a crucial issue in the maintenance of forest biological diversity is the success in forest regeneration, and the character of the regeneration, in terms of the species balance. Whether this is truly representative of the ecosystem as a whole needs to be established by research.
The data required in this case is available from normal forest inventory and forest management plans. The authors also stated the indicators would require considerable further explanation and interpretation in order to derive any conclusion about the sustainability of forest management. However, this is the case with almost any indicator.
Additional research on biological diversity indicators is taking place under the EU cooperative BEAR project. A report by Hansson (2000), concentrated on biological diversity at the ecosystem and species levels, but suggested that possible genetic diversity indicators might include allelic diversity, heterozygosity and effective population size. A more recent report on BEAR (Hansson 2002), however, did not elaborate on this.
The difficulty of implementing quantitative indicators of genetic diversity has led some workers to advocate a more qualitative approach. Jennings et al (2001) considered this issue in a paper directed at tropical forest management, although their arguments apply equally well to forests in other climatic zones. The authors argued that it is impractical for forest managers to measure genetic variation so it is not appropriate to use direct measurements of genetic traits as indicators of sustainable forest management.
The authors were critical of much of the current attempts to develop genetic indicators. They stated “The function of criteria and indicators is to provide a practical means by which changes in forest condition which occur as a consequence of management can be monitored. They should, in principle, be defined so that they are clear, practical and easy to monitor, and be based, as far as possible, on available research knowledge and statistics (ITTO, 1998). All of these important requirements seem to have been forgotten in the case of genetic criteria and indicators.”
They also drew attention to the critical, and neglected, issue of sampling genetic parameters.
They took a pragmatic, silvicultural approach to assessing the impact of a disturbance (in their case, timber harvesting) on genetic diversity. Essentially, they advocated that if the disturbance resulted in adequate subsequent regeneration of the affected species, it was likely that their population genetic structure and diversity will not have been severely and permanently affected. The risk of adverse effects on genetic variation within a species are greater for rare species, for those where selection harvesting removes a particular segment of the population (such as all the straight trees), and for those species where reproducing individuals are widely scattered.
Jennings et al took the view that the key to maintaining genetic diversity was designing the silvicultural system to take account of the nature of the species affected by disturbance. They noted that different strategies were required, for example, for species that reproduced only when large and those that reproduced at an early age. They proposed that forest managers use a list of simple ecological criteria to assess which species may be at risk from disturbance and then target management so as to alleviate any possible adverse genetic impacts.
In other words, Jennings et al argued that conservation of genetic diversity is better addressed if it is given due weight in a comprehensive strategy for forest management. It is the development of such a strategy that integrates considerations of silvicultural knowledge, research on impacts of factors such as fire, timber and NTFP harvesting, etc, that is the key issue.
ITTO has, in fact, moved in that direction. It has taken a different line from some other criteria and indicator processes in not requiring quantitative indicators for genetic diversity. Instead, it has an overall indicator (ITTO 1998) that requires: “Existence and implementation of a strategy for in situ and /or ex situ conservation of the genetic variation within commercial, endangered, rare and threatened species of forest flora and fauna.”
The indicator is supported by management guidelines and a requirement for “the existence and implementation of monitoring and evaluation procedures for assessing biodiversity changes of the production forests, compared with areas in the same forest type kept free from human intervention.”
The latter requirement is a weakness as it is very difficult, if not impossible, to find and maintain any areas of forest free from human intervention. Despite this weakness, the ITTO approach is a pragmatic and practical one. It leaves the way open for an individual country or FMU to implement a strategy, at the designed to fit its own unique situation. The guideline is likely to be very difficult for any country or FMU to implement, given the current state of development of monitoring and evaluation procedures.
