Landscape approach to forest conservation
With reference to the article ‘Does natural forest size really matter’, published in the February 2013 issue of SA forestry magazine, written by Dr. Coert J. Geldenhuys, Dr Geldenhuys makes both direct, and indirect reference to my PhD thesis (Berliner 2009): a systematic plan for the forest biome of South Africa. The discussion below refers both directly and indirectly to these references.
by Dr Derek Berliner
|Natural forest loss in the Wild Coast.|
The issues raised are not trivial, nor are they merely academic debates, rather they have important implications for the management and conservation of South Africa’s forest biodiversity. Because our indigenous forest make up such an extremely small area of South Africa (less than 0,4%), as well as the fact that they are inherently highly fragmented, makes them particularly vulnerable to surrounding land use changes, non-sustainable use, and increasingly, global climate change.
The issues raised are not necessarily new; some have been the subject of ongoing and often heated debates, while others stem from differences in perception and terminology, or failure to consider the full complexity of the issues at hand. I will not go into any great detail regarding these debates, but would like to pull out a few points pertinent to three key issues. The information and views presented here are based on my experience, studies, observations, as well as an understanding of my readings of both national and international literature on forest landscape ecology.
Three issues raised by Dr Geldenhuys include:
- The emphasis placed on large forest patches as being of higher conservation value as compared to smaller forest patches, is often not founded, and that small forest patches can or should have high conservation values (due to their often unique suites of species assemblages).
- South Africa’s natural forest are inherently fragmented, but this fragmentation has decreased with changes in land use surrounding forests.
- The total natural indigenous forest area in South Africa has increased, rather than decreased.
Small verses large patches
The issue of large reserves-verses-small reserves has been an ongoing debate in conservation biology for many years, and there is no simple answer to this, mostly because it depends on context. If the context of systematic conservation planning is used, then the creation of protected areas are seen as an investment in the limited resource of available land. The question then is, would it be better to conserve one large protected area or many small, covering the same total area?
There are arguments for and against each approach but in general, larger reserves will mostly be favoured and considered to have higher conservation value that many small reserves, simply because species richness increases with habitat area (a well established law in ecology). This means that a larger block of habitat would support more species than any of the smaller blocks. However, from a process point of view, larger areas can maintain landscape processes that many small reserves would be unable to do. In addition, many species, in particular territorial apex predators (for example leopards and crowned eagles), require a minimum size territory for suitable habitation.
Despite the above arguments, the debate is still not that clear cut. There is an implicit assumption hidden in this that may not always hold – this is that the one larger reserve will have all of the species that occur in all of the smaller reserves. If the smaller reserves had unique/unshared species, then it is possible that two smaller reserves could have more species than a single large reserve. As Dr Geldenhuys correctly points out, the relic nature of South African forests are the by-product of successive expansions and contractions over geological time, and this has, in cases, given rise to unique species within small patches (these are often highly immobile species such as anthropoids). For example, the Dlinza Forest Pinwheel Snail, occurring only in the Dlinza Forest, or the Pondoland Cannibal Snail, occurring only in the forests of the Wild Coast (Herbert, et. al, 2004), and the critically endangered velvet worm species Opisthopatus roseus, endemic only to the Ngele forest in the KwaZulu-Natal province of South Africa (see figure 1 below).
|Figure 1. The critically endangered Pondoland cannibal snail occurring within the limited coastal forests of the Transkei region (from Herbert et. al 2004).|
These species, being highly immobile, are poorly dispersed and highly vulnerable to fragmentation effects.
Although unique species assemblages in small patches may be exceptions, in most cases, small patches consist of a nested subset of species found within larger patches, at least within the same bio-geographic zone. Perhaps the most important reason why small patches may have an unusually high conservation value, is their importance as part of ecological corridors, i.e. small patches may act as ‘stepping stones’ connecting larger forest patches. The importance of this becomes particularly evident within the context of fragmentation and matrix transformation, discussed below.
Because the science of systematic conservation planning aims to find optimal solutions to reserve design to ensure the persistence of as many species as possible, within the constraints of limited resources and ongoing threats (Margules & Pressey, 2000) one large reserve would mostly be preferable to many small. Patchy habitats such as forests typically occur in a matrix of surrounding land that may not be conducive to forest species conservation, or may even be a direct threat to the forest (such as slash and burn agriculture, alien invasive plants etc). The ratio of forest edge to total forest patch area will almost always be lower in a single large reserve as opposed to many small reserves, but the degree of this will vary depending on the shape of the patch (see figure 2 below).
|Figure 2. The shape of a forest patch will influence its vulnerability to outside influences such as fires, and matrix transformation. This can be measured by the ratio of edge to area (the shape index). Because of this, a single larger reserve will always be less vulnerable to outside disturbances and have lower management costs than many small reserves, making it in the long term a more viable choice for a protected area, all else being equal.|
Natural forest fragmentation
Globally, habitat fragmentation is considered one of the greatest threats to forest biodiversity, as it significantly increases vulnerability to biodiversity loss. Fragmentation can be measured in various ways, basically by the structural configuration of patches (patch size and distance from each other), and by its functional effect (how well connected patches are, and edge to area ratio).
Dr Geldenhuys states that “fragmentation has decreased due to changing land use, (corridors developed via timber plantations, invader plant stands, woody re-growth in protected areas)”, and that “plantation forestry, with plantation stands planted in the fire-prone environments within the fire pathways, have became corridors or nurse stands for the connection between former isolated natural forest patches”.
South African forests are naturally highly fragmented, but without a doubt human impacts in the form of fires, clearing, and agricultural land use change have exacerbated this fragmentation. This is particularly evident by the complete loss of many smaller forest patches, shrinkage of larger patches and the transformation of the matrix surrounding forest (Lawes et al., 2004; Berliner et al., 2006). Coastal forests, historically once an almost contiguous band along the coast, have been particularly affected by anthropogenic fragmentation over the last 100 to 150 years, mostly due to expansion of plantation forestry and sugar cane farming, and more recently the invasion of alien plants.
Natural forest/plantation interface
Could it be that when the matrix surrounding small isolated forest patches, once natural grasslands, when converted to plantations, will result in improved connectivity between the patches (and hence reduce the fragmentation effect)? Again the answer depends on context and detail. Clearly, certain forest occurring species are benefited by exotic plantations, as these can provide an expanded habitat (for example, certain accipiter raptors), and many forest tree species, in particular, pioneer trees will spread into plantations if given half a chance. In addition, plantation forestry management will provide fire protection, often allowing forests to expand into the new fire protected areas.
Unfortunately, forest biodiversity is complex, and the argument is just not that simple. Forest biodiversity cannot be represented by the expansion of a few trees species, or transient goshawks. Recent research into forest dependent and forest associated bird species show significant declines in the recorded occurrences of these birds within certain areas of South Africa. Preliminary work by Cooper et al. (in prep), and Professor Mike Cherry of Stellenbosch (personal communication), using South African bird atlas data covering more than 10 years of monitoring, have noticed that for the 57 forest associated and forest dependent bird species, 39 (or about 68%) have declining ranges, five are stable, and 13 are increasing (based on preliminary assessment of the SABAP 1 and 2 data). Although these results will require further investigation, in particular as there are differences in methodology between the two bird atlas data sets (SABAP 1 was based on quarter degree grid squares, and SABAP 2 on pentads), there is cause for concern, in particular as similar declines have been found using completely different data sets by Trimble, et al. (2011) for KwaZulu-Natal dune forests. They found 76% of forest associated or forest dependent bird species had declined, with 57% significantly so, at an average rate of 13.9% per year. They drew their conclusions from an independent 13-year, long-term monitoring programme designed to assess the success of coastal dune forest.
When the distribution maps of declining forest birds were studied, two re-occurring geographical areas stood out: the Mpumalanga mist belt, and the Ugie-Maclear-Kokstad region. For these two regions, as well as the KwaZulu-Natal dune forest area, all have one thing in common: significant habitat transformation has occurred within the last 10 to 15 years. While natural forest loss may have stabilised in these areas, matrix transformation has not, and these areas have undergone considerable expansion in agriculture, mining and importantly, plantation forestry. Although research into the probable causes of these declines is still in progress, my guess is that we are witnessing the effects of habitat fragmentation as predicted by landscape ecological theory. In particular, it is the associated disruption or complete loss of natural forest ecotones that concerns me most. Ecotones are particularly important habitat areas for birds as they usually contain the highest species diversity and they also form natural buffers against fire and invasion of alien plants into forests. Natural ecotones between forest-grasslands or forest-fynbos are maintained by natural, landscape-scale ecosystem process such as fire, wind, light, moisture changes etc. These are typically disrupted when the matrix is converted into plantations or agriculture, and particularly when fire breaks are established along them, as is common practise.
It is of interest that for these areas, it is not the forest area that has changed, but the forest matrix (mostly grasslands converted into plantations). Despite this, it is still forest associated and forest dependent birds that are being impacted. To me, this implies that the problem lies at a larger, landscape scale, involving loss of ecosystem connectivity and the degradation of valuable ecotone habitat, and not due to localised forest habitat loss.
It is well documented that birds are good indicators of biodiversity, and that bird declines can be a warning of emerging environmental problems, so these changes require further investigation.
The important point here, at least as far as the forestry industry is concerned, is that while there is not much that can be done to reverse past matrix transformation, there is much that can be done to reduce the impacts on natural forests. This can be done by improving the management of the plantation-natural forest ecotone interface, as well as by allowing for the recovery or reforestation of natural strips of vegetation that can act as ecological corridors between remnant forest patches. The first step is for plantation managers to understand the importance of maintaining natural forest ecotones and the processes that are responsible for maintaining them.
Have natural indigenous forests in SA increased or decreased?
Dr Geldenhuys’s article he states that “it’s mostly a perception that we have lost forest and that in reality, we have gained forest through changed land use practices (with changed fire regimes, including exclusion of fire)”.
What is often not realised, is that most forest destruction took place at the hands of European settlers in the period 1860-1940, (King, 1938). The extent of more recent forest loss is not clear.
Limited quantitative data is available to accurately determine forest loss across the whole country, and early estimates of forest loss (for example given by King, 1938) cannot be accurately tested. However, evidence suggests that in certain areas, indigenous forests were significantly larger than at present, but at least for some areas these losses have been exaggerated, and more likely to be in the order of 10 to 15% loss, mainly through boundary contraction (Prof. Mike Lawes, personal communication).
Certainly, forests may have expanded in some areas, particularly where fires have been kept out (by roads or habitat transformation), but in cases, this expansion is not true forest, but bush encroachment or woodlands. In my experience, particularly in the communal areas of the eastern Cape, significant numbers of forest have been cleared for slash and burn agriculture: many of these lands now lie fallow, overrun by alien plants. Overall forest loss is still difficult to determine, but certainly for some areas, this has been extensive. For example, McKenzie (1989) looked at habitat change in a number of plots in the former Transkei between the years of 1937 and 1982. He found that forests had declined by 3% and 7% for two different plots. When I re-analysed these same plots to determine loss in forest cover since 1982, up to 2008, I found that there had been a significant acceleration in forest loss, particularly after 1994. For two of these plots, there was a 25% and a 60% loss in total forest area since 1937.
The total forest loss in South Africa is unlikely ever to be accurately determined, however recent advancements in remote sensing and land cover mapping will be able to track, at least for more recent times, changes in forest cover with far greater accuracy.
Berliner, D.D., van der Merwe, I., Benn, G., Rouget, M. 2006. Systematic Conservation Planning for the Forest Biome of South Africa: approach, methods and results used for the selection of priority forests. Updated September 2006. Funded by UK, DFID for the Department of Water Affairs and Forestry.
Berliner, D.D. 2009. Systematic Conservation Planning for South Africa’s Forest Biome: An assessment of the conservation status of South Africa’s forests and recommendations for their conservation. PhD dissertation, University of Cape Town.
Berliner, D.D., 2011. The Conservation Status of Forest on the Eastern Cape’s. Wild Coast, South Africa. The Wild Coast Project. Presented at the Fifth Natural Forest and Woodlands Symposia, Richards Bay, April 2011.
Cooper, T., Wannenburgh, A, Cherry, M. (in prep). Why are forest birds declining? (working title). PhD. dissertation. University of Stellenbosch.
Geldenhuys, C.J. 2013. Does natural forest size really matter? February 2013 issue of SA Forestry magazine.
Herbert, D.G. & Kilburn, R.N. (2004) Field guide to the land snails and slugs of eastern South Africa. 340pp. Natal Museum, Pietermaritzburg.
King, N.L. 1938. Historical Sketch of the Development of Forestry in South Africa. Journal of the South African Forestry Association 1: 4-16.
Lawes, M.J., Macfarlane, D.M., Eeley H.A.C. 2004b. Forest landscape pattern in the KwaZulu-Natal midlands, South Africa: 50 years of change or stasis? Austral Ecology 29, 613-623.
Margules, C.R. &.Pressey, R. L. Systematic conservation planning. Nature 405, 243-253.
Mckensie, B.1989. Medium-Term Changes of Vegetation Pattern in Transkei. South African Forestry Journal. No.150 September.
Trimble, M.J, van Aarde RJ. 2011. Decline of Birds in a Human Modified Coastal Dune Forest Landscape in South Africa. PLoS ONE 6(1): e16176. doi:10.1371/journal.pone.0016176.
*Published in June 2014