Framework Species Method

Wherever there are too few natural regenerants to bring about rapid canopy closure through ANR (i.e. <3,100/ha), tree planting is necessary to rapidly restore a tropical forest ecisystem. In any tropical forest type, there may be hundreds of tree species to chose from. It is impractical to grow and plant them all, so the question becomes: which tree species to select?

The framework species method is the least intensive of the tree planting options, since it exploits natural (and free) seed dispersal mechanisms to bring about biodiversity recovery. It involves planting the fewest trees necessary to shade out weeds (i.e. site “re-capture”) and attract seed-dispersing animals.

In order for the method to work, remnants of the target forest type must survive within a few kilometres of the restoration site (as a seed source). Animals (mostly birds and bats), capable of dispersing seeds from remnant forest patches or isolated trees to restoration sites, must remain fairly common. The framework species method enhances this natural seed-dispersal service to achieve rapid tree species recruitment in restoration plots. Consequently, recovery of the biodiversity levels typical of mature forest ecosystems is attained without having to plant all the tree species that comprise the target forest type. In addition, the planted trees rapidly re-establish forest structure and functioning and create conditions on the forest floor that are conducive to germination of tree seeds and seedling establishment. The method was first conceived in Australia to restore degraded sites within Queensland’s Wet Tropics World Heritage Area and has since being adapted for use in several SE Asian countries.    What are framework tree species?
The framework species method involves planting mixtures of 20-30 indigenous forest tree species, which are typical of the target forest ecosystem, but which also share the following ecological characteristics:-
   •   high survival when planted out in deforested sites,   
   •   rapid growth,
   •   dense, spreading crowns that shade out herbaceous weeds and
   •   flowering, fruiting, or the provision of other resources, at a young age, which attract seed-dispersing wildlife.
In the seasonally dry tropics, where wild fires in the dry season are an annual hazard, an additional desired characteristic of framework species is resilience to burning. When fire prevention measures fail, the success of forest restoration plantings can depend on the ability of the planted trees to re-sprout from their rootstock after fire has burnt their above-ground parts (i.e. coppicing).
A practical consideration is that framework species should be easy to propagate and, ideally, their seeds should germinate rapidly and synchronously, with subsequent growth of vigorous saplings to a plantable size (30-50 cm tall) in less than 1 year. Furthermore, where forest restoration must yield benefits to local communities, economic criteria such as the productivity and value of products and ecological services, rendered by each species, may be taken into account.
Are framework trees pioneer or climax species?
Mixtures of framework tree species planted should include both pioneer and climax species. Goosem and Tucker (1995) recommend that at least 30% of trees planted should be pioneers. By planting both pioneer and climax trees in a single step, forest succession can be “short-circuited”.
Many climax forest tree species perform well in the open, sunny conditions of deforested areas, but they fail to colonize such areas due to lack of seed dispersal. Climax tree species often have large, animal-dispersed seeds. The decline of large mammals, over wide areas now prevents dispersal of large-seeded, climax trees into deforested sites. By including some of them amongst the trees that are planted, it is possible to overcome this limitation and accelerate recovery of climax forest.
The planted pioneer trees contribute most to early canopy closure and shading out of herbaceous weeds. The point at which the tree crowns dominate over the herbaceous sward is called “site re-capture”. Pioneer tree species mature early. Flowering and fruiting can begin as soon as 2-3 years after planting. Nectar from flowers, fleshy fruits as well as the perching, nesting and roosting sites, created within the tree crowns, attract wildlife from nearby forest. Faunal diversity increases dramatically but most importantly, many of the animals that visit the restoration sites carry with them tree seeds from climax forest. Furthermore, the cool, shady, moist, humus-rich and weed-free forest floor, created beneath the canopy of the planted trees, provides ideal conditions for seed germination.
Pioneer species begin to die back after 15-20 years, creating light gaps. This allows the saplings of in-coming tree species to grow up and replace the planted pioneers in the forest canopy. If only short-lived, pioneer tree species were planted, they might die back before sufficient numbers of incoming tree species had established, leading to the possibility of re-invasion of the site by herbaceous weeds (Lamb, 2011). Planted climax tree species form an under-storey which prevents this. They add diversity and some of the structural features and niches of climax forest right from the start of the restoration project.

What about rare or endangered tree species?
Rare or endangered tree species are unlikely to recruit into restoration sites on their own. The very fact that they are rare probably means that the seed source is limited and they may have lost their primary seed dispersal mechanisms. Including such species in forest restoration plantings, can help prevent their extinction, even if they may lack some framework characteristics. Information on the world’s endangered tree species is collated by the World Conservation Monitoring Centre of the United Nations Environment Program.
How to select framework tree species?
There are two stages to the selection of framework species i) preliminary screening, based on current knowledge, to identify “candidate” framework species for testing and ii) nursery and field experiments to confirm framework traits. At the beginning of a project, detailed information about each species is likely to be sparse. Preliminary screening must be based on existing information sources and the target forest survey. However, as the results of nursery experiments and field trials gradually accumulate, the list of acceptable framework tree species can be gradually refined. Framework species plantings gradually improve, as poor performing species are dropped from subsequent plantings and new species are tried.
Sources of information for preliminary screening include: i) floras, ii) results of the target forest survey, iii) indigenous local knowledge and iv) scientific papers and/or project reports, describing any previous work in the area.
Floras can provide basic taxonomic data on species under consideration as well as their suitability to the target forest type being restored, elevation range etc. They also indicate if the species produce fleshy fruits or nectar-rich flowers likely to attract wildlife. 
The target forest survey provides a great deal of original information that is useful for selection of candidate framework tree species including, a list of indigenous tree species from which species can be selected, as well as which species have nectar-rich flowers, fleshy fruits and dense spreading crowns capable of shading out weeds. Phenology studies yield information on which trees attract seed-dispersing wildlife.
The botanical knowledge of local people (ethnobotany) can also provide an insight into the potential of trees to act as framework species. When carrying out such studies, it is important to work with communities that have a long history of living close to the forest, especially those that practice swidden (slash and burn) agriculture. Farmers from such communities usually know which tree species readily colonize fallow fields and grow fast and which tree species attract wildlife. However, the results of such studies must be critically scrutinized. Local people sometimes provide information, which they think will please the researcher, rather than that based on actual experience. Superstition and traditional beliefs can also distort objective assessment of a tree species’ capabilities. Consequently, ethnobotanical information is reliable only if it is provided independently, by members of several different communities, with different cultural backgrounds.
Local people also know if other researchers have been active in the area and which organizations or institutions they come from. Forestry departments and protected areas authorities often carry out biodiversity surveys, although the results may remain in unpublished reports. Contact such organizations and ask for access to such reports. The local or national herbarium may also have tree specimens from your project site. Browsing through herbarium labels can reveal a lot of useful information. If any development projects have been carried out near your project site, it is likely that an environmental impact assessment (EIA), including a vegetation survey, was carried out. So it is worth contacting the agency which carried out the EIA. Universities may also be a source of more detailed information, if research students have been active in the area. Finally, there’s always the internet. Simply typing the name of your project site into a search engine may reveal major additional sources of information.
Lists of tested framework tree species currently exist only for Australia (Goosem and Tucker, 1995) and Thailand (FORRU, 2006). Trees species in the same genera as those listed for Australia and Thailand may also perform well in other countries, so including some of them in initial framework species trials is well worth a try. Two pan-tropical tree taxa deserve special mention, namely fig trees (Ficus spp) and legumes (Leguminosae). Indigenous species within these two taxa nearly always perform well as framework species. Fig trees have dense and robust root systems, which enable them to survive even the harshest of site conditions. The figs they produce are an irresistible food source for a wide range of seed-dispersing animal species. Leguminous trees often grow rapidly and have the capacity to fix atmospheric nitrogen in root nodules via symbiotic bacteria, resulting in rapid improvement of soil conditions.
What management is required?
Firstly, prevent both fire and the hunting of seed-dispersing animals. Secondly, protect and nurture any existing natural regenerants, using ANR techniques. Thirdly, plant enough framework tree species to bring the total species on-site (including natural regenerants) up to around 30, spaced about 1.8 m apart or the same distance from natural regenerants – this brings the total density of trees on site up to around 3,100/ha.
Frequent weeding and fertilizer application, of both planted trees and naturally regenerating saplings, is recommended during the first two rainy seasons. Weeding prevents herbs and grasses, particularly vines, from smothering the planted trees, enabling the tree crowns to grow above the weed canopy. Fertilizer application accelerates tree growth, resulting in rapid canopy closure. Finally, monitor the survival and growth of the planted trees and biodiversity recovery in restoration sites, so that the choice of framework tree species for future plantings can be continuously improved.