RESTORATION
WORKSHOP:
Linton
Park Mountain Site Development
Windsor Research Centre
&
Linton Park Mountain
Trelawny
21st Ð23rd June 2005
Workshop Sponsors:
BirdLife International UNEP-GEF -- Important Bird Areas (IBA) Programme
UN-FAO National Forest Programme Facility
Facilitator: John J. Pipoly, III (Fairchild Tropical Botanic Garden [FTBG])
Forest Management Specialist: James Gordon (FTBG)
Participants:
Susan Koenig & Michael Schwartz (Windsor Research Centre [WRC])
Lawrence Nelson, Michael Barrett, Patrick Barrett, Noel Bennett (Forestry Department [FD])
George Proctor (Institute of Jamaica [IoJ])
Kurt McLaren (University of the West Indies [UWI])
Regards sent by: Andreas Oberli (National Arboretum)
Goal:
To establish the Linton Park Mountain Forest Restoration Research & Demonstration Area: Stage I Ð Towards cost-effective techniques for improving / restoring plant and wildlife habitat in a degraded cockpit valley pastureland
1.1. Introductions & Background
Susan Koenig opened the workshop with a brief overview of Linton Park Mountain (LPM), which is located on the Cockpit Country plateau north of the community of Duanvale, Trelawny. Surveys by FD in 2003-04 of Cockpit Country Forest Reserve boundaries added approximately 50 hectares of land to the previously recognized boundary. This additional hectarage includes the valley bottomland of pasture, small farm plots of yam, corn and seasonal vegetables, and slopes of mid-succession closed-canopy forest. With funding from a UNEP-GEF grant administered by BirdLife International and a grant from UN-FAO National Forest Programme Facility, WRC proposed to FD a pilot project to evaluate potential techniques for cost-effective restoration of the pastureland, with particular reference to improving the habitat for native bird populations and with potential for application elsewhere in Jamaica. This workshop was convened to review WRCÕs proposal to evaluate Assisted Natural Regeneration (ANR) and to develop the methodologies for field trials.
1.2. Concepts of Restoration Ecology
John Pipoly facilitated the first topic of the workshop, one of the fundamental questions which must be asked before any planning or field work commences on restoration:
Why
are trees not growing back?
In the most general of terms, participants recognized that natural regeneration fails to occur when conditions move beyond the natural range of variation which a species requires for survival. The factors which may limit regeneration are typically a reflection of human activities:
|
Stress to the system |
Source of stress |
|
|
An example of the importance of assessing limiting factors was reviewed with the example of "site dominance by pasture grasses," where characteristics of grass growth and propagation (e.g., discrete clumps, root mass formed from rhizomes) will affect the type of required management interventions. George Proctor noted examples of native grasses (e.g,. Stenotaphrum secundatum, Eragrostis spp. Eriochloa spp.). Proctor also noted that, in addition to grasses competitively excluding other vegetation from becoming established, grasses provide little attraction in terms of food resources for many seed-dispersing birds, with a consequent break-down in the dispersal of seeds in many pastures. [NOTE: In a subsequent discussion, Kurt McLaren supported the observation of Andreas Oberli, cited in the workshop by Koenig, that many of Jamaica's native plants show a very limited capacity to disperse their seeds without assistance by wildlife i.e., most fruits are dropped under the mother tree.] Proctor noted that one management strategy for restoration might include erecting bird perches in pastures.
Workshop participants also briefly discussed the use of fire as a general tool for clearing by small farmers. With respect to restoration efforts, the frequency and intensity of fires may need to be considered in relation soil fertility and availability of seed banks in the soil. With respect to Jamaican farmers, the consensus of the workshop was that fire used in association for shifting agriculture is occasional while burning to remove weedy, herbaceous vegetation within a plot is annual.
To conclude this portion of the discussion, James Gordon queried whether any of the factors limiting natural regeneration where exclusive to LPM and Cockpit Country. That is, is Cockpit Country genuinely unique in how it functions or, because it is part of a continuum of the island's central limestone plateau, will the research results have relevance elsewhere in Jamaica, the Caribbean, and on mainlands? There certainly is interest in how this research might aid FD efforts to restore mined-out bauxite pits.
The second half of the discussion focused on measures of success:
How
do we know when we've done enough?
Lawrence Nelson reviewed the standards FD uses to certify restoration of bauxite pits:
a. short-term indicators over 1-2 years
b. medium-term indicators over 5-10 years
c. long-term indicators > 10 year
Examples of suitable medium-term indicators relevant for LPM could include: (1) changes in the proportional area recovered from non-desirable vegetation or bare ground; (2) percent canopy coverage; (3) increase in diversity beyond what was planted and representing species other than those which are pioneering; (4) increase in faunal diversity.
Long-term indicators also would include measures structure, function, and diversity: (1) crown closure; (2) number, size and diversity of canopy species; (3) epiphyte communities; (4) fauna associated with closed-canopy forest
1.3. Natural and Assisted Regeneration - Case Studies
James Gordon
presented examples where natural regeneration proved to be a technically simple
and cost effective method for restoring native forest in Costa Rica. At one site, Guanacaste, fire and
cattle grazing were identified as two important factors limiting
regeneration. Through land
acquisition, financial support to the Forestry Department for fire control, and
a fortuitous decline of the cattle industry, pioneering woody species of Cecropia and Guazuma became established and
facilitated successional process of recovery. A second example was of a
pasture abandoned 40 years ago in the wet forest zone of the San Carlos Region
in NE Costa Rica, which now has light-demanding hardwood species, such as Vocysia and Ochroma, which are
large enough to be harvested for timber.
At the same time, managers recognize that older, decaying snags should
be left for cavity-nesting birds and for nutrient recycling. Gordon also gave an example from
Oaxaca, Mexico, where there is conservation interest in endemic trees of dry
forest-types. Foresters are
finding that some species which are rare in the mature forest may be found
sprouting commonly in pastures.
The ejido system of communal lands, with rotation and fallow periods up to
25 years, appears beneficial for species conservation, but new development and
alternate agrarian practices are affecting these historic, beneficial dynamics.
Gordon reviewed the aforementioned case studies to highlight the value of multiple approaches for habitat restoration: in addition to exploring the specific factors which might be arresting succession, there is a value also to monitoring areas with no management interventions to see which species regenerate naturally. This latter can provide insights into how natural regeneration might be enhanced, particularly if there is a need to "recapture" land quickly. Gordon concluded the morning's discussion with a brief overview of the role enrichment plant may play in natural regeneration. While it may be relatively easy to develop a list of "preferred species," Gordon reminded the workshop of the practical realities:
Shortly after IoJ founded its field station at Mason River in 1963, efforts were directed to reclaim an area of approximately 1 ha, which was dominated by "Sour Grass" (Andropogon sp.). Recognizing the need to create shade quickly, eucalyptus was planted because (a) it grows quickly in acidic soils and (b) it was easily available. As the sour grass was shaded and died, native vegetation began to colonize from peripheral forest. Although management was planned to remove the eucalyptus, the trees have already begun dying and, of equal importance, the trees did not reproduce and spread beyond their point of introduction. This provides a good example of how a plant species can be used as a functional intermediate to restore processes of regeneration -- a tool for a purpose. Proctor concluded that areas where no eucalyptus was planted remain, at present, dominated by sour grass.
1.4. Towards a List of Suitable Species for Planting
With the recognition that some type of enrichment planting to assist natural regeneration may prove beneficial (both from an ecological stand-point and to generate support from local community members who may feel displaced by the restoration activities), workshop participants generated a list of potential species and evaluated them based on desirable characteristics:
See Annex 1 for the list of species. Michael Barrett will review which species currently are available from FD. For those not under propagation, Species Elimination Trials in the nursery (Stage I) are needed to determine which species can be grown easily and inexpensively -- important criteria for LPM enhancement planting. Stage II would involve field trials at LPM.
1.5. Towards Experimental Design at LPM
In advance of a site visit to LPM on Day 2 of the workshop, participants drafted basic questions and field methodologies designed to assess: (1) why LPM pasture is not regenerating, and (2) what cost-effective management can be developed to establish or accelerate regeneration.
Four factors were identified as the most likely sources suppressing regeneration:
Although statisticians will balk if, to address factor #3 - the role of cows - we exclude all cows but do not have an experimental control area where cows are, in deed, allowed to continue grazing, the consensus of the group was that cows should be excluded from the study plots. For all other experimental treatments, control plots will be included in a randomized block design so that they results of LPM will have the rigour to be considered for application elsewhere (i.e., workshop participants are interested in looking for more than solely restoring LPM pasture).
#1: To begin assessing whether regeneration is limited by the dispersal of seeds -- are seeds available in the adjacent forest and are they being dispersed into the pasture -- McLaren recommended setting up seed traps at variable distances from the forest edge within the pasture. In setting seed traps, McLaren recommended multiple small (e.g., 0.5 x 0.5 m), rather than a few large traps, with a goal of evaluating approximately 1% of the study area, and noted that assessments also must be conducted to monitor predation (particularly seasonal and preferential predation) at seed traps. McLaren also cautioned about the difficulties in identifying seeds but noted temporary measures, such as enumerating without the species name or sub-dividing by dispersal classification (e.g., wind vs. animal dispersed). Seeds might next be put into field plots to evaluate germination or, as suggested by Nelson, evaluated in laboratory for quality (e.g., seed weight, purity, moisture content, viability). To support Seed Trap assessments, McLaren recommended concurrent phenology studies in the peripheral forest. This will provide baseline information on the timing of seed rain, which varies seasonally and annually.
#2: The role of grass as a limiting factor should be evaluated with experimental removal. The method(s) of controlling grass will relate directly to predictions of how the grass limits regeneration: chemical herbicide spraying and burning (#4) would target root systems (i.e. subterranean and surface competitive exclusion), while a non-chemical approach of cutting grass blades would begin to address whether seeds and seedlings are failing to establish and survive due to shading effects. While herbicide spraying may prove expensive at a large scale, workshop participants agreed it would be useful to deploy a controlling method just as a local farmer would.
Workshop participants next discussed general concepts for various field experimental designs, with the recognition that two research projects were evolving: (1) research and monitoring of Natural Regeneration, to see what happens when possible limiting factors are controlled; and (2) Species Elimination Trials for Enrichment Planting, to specifically create structure, shade, attract seed dispersers, add species of economic interest to the community, etc.
With respect to Natural Regeneration research, McLaren suggested planting seeds from seed traps in the grass elimination research plots to see what can / cannot compete in the presence / absence of grass.
Nelson provided a descriptive configuration for Species Elimination Trials with a randomized block design: e.g., for 10 target species, one block is divided into 10 units. Each unit is planted with e.g, 24 individuals of a single species. The entire unit is evaluated for within-species mortality, while individuals in the core of a unit are evaluated for growth rates. Growth rates of individuals surrounding the core (i.e, which are adjacent to the neighboring unit) are not evaluated to minimize potential confounding effects of edge interactions. The block design is replicated, with species plantings randomized e.g,. Species A is next to Species B in Block 1, now Species A is next to Species C to eliminate potential for species interactions. For Species Elimination Trials, mortality data are of key interest.
A final note of the session was given by Gordon and McLaren on the collecting and use of seeds vs. collecting of wildlings for planting experiments. A "problem" in using wildlings (aside from the fact that you are disrupting the regeneration at the source of collection) is that ages of individuals are unknown. This introduces a confounding variable to the research design. Further, these individuals have already experienced pre-selection on mortality in the field. As Gordon noted, however, wildlings can be used so long as it is well-documented (e.g,. including descriptives of size at harvesting, size at planting).
1.6 Community Meeting in Duanvale
See separate report included as annex 2
2.1. Site Visit to LPM
LPM was visited during the morning session for a qualitative evaluation of conditions of the pasture and peripheral forest. Participants concluded that: (1) exclusion of cattle from the end of the valley is feasible; (2) the dominant pasture grass ( prob. Chloris sp.) form a very dense root mass, which will require greater effort to eliminate in comparison to a species which remains bunched in clumps; (3) there is adequate space for Natural Regeneration research; (4) there is probably not enough space at the end of the valley for concurrent Species Elimination trials, but other areas within the Forest Reserve boundaries of LPM would be suitable; (5) the source vegetation on the hillside shows excellent diversity in secondary succession species. Deployment of seed traps can commence immediately. Hillside vegetation surveys are needed to identify species which will be the sources for natural dispersion of seeds and to provide the baseline for comparison of restoration in the pasture.
2.2. Overview of Fairchild Tropical Botanic Garden
In the afternoon, Pipoly gave a brief history of FTBG, which was established by Daniel Fairchild and Robert Montgomery in the mid 1930s (1936-38) in Miami (see: http://www.fairchildgarden.org/resource/n_history.html). From a founding interest in cultivating tropical plants of commercial value, FTBG has evolved into a major botanic research, conservation, and education center. A start-up endowment covers approximately 45% of current operations, with the balance supported by grants, selling materials, and consulting contracts. For example, the city of Miami hires FTBG's mobile plant teaching center for educational outreach to primary schools. This proves more cost-effective than having to hire a qualified teacher for each school in the districts.
Workshop participants then discussed potential collaborative projects, such as finding "lost" tree species on Jamaica which had historic timber value but are currently rare in the wild, the rise in wider commercial interests of non-timber forest products such as "roots" drinks and the possibilities for getting over-exploited species into cultivation.
2.3. Community Meeting in Duanvale
See separate report included as annex 2
3.1. The Next Steps
Gordon provided a synthesis of the two phases of research developed by the workshop participants: (1) Enrichment Planting and (2) Natural Regeneration assisted by control of potential limited factors.
To summarize the experts' proposed block design:
(1) a randomized block design will be used to evaluate the treatments of (a) cow exclusion; (b) grass present; (c) grass blades removed manually; and (d) grass root system eliminated -- in relation to the survival and recruitment of pioneering and mid-succession plant species. Plot size within each block will be approximately 8 x 8 m, to accommodate 36 individuals planted at 1 x 1 m intervals. This is closer than FD's usual spacing of 2 x 1 m, but the workshop recognized that our interest is to maximize canopy cover to suppress grass rather than for timber production, where spacing is designed to maximize for straightness of boles with minimal canopy. Risks associated with planting too close / at high densities include the spread of pathogens and allelopathy.
Following site preparation and "beating up" - the first-month check for replanting and first weeding -- we will attempt to minimize interventions for weed and vine control -- the objective being to monitor how nature responds to the primary treatments. However, it is recognized that interventions may be required; timing and efforts will be documented as part of the "cost" assessment of the field trials.
Dependent variables of survival and recruitment, which will serve as indictors of short-term (i.e., 2-year) success, will include:
Recruitment of new species other than those planted
To meet WRC's objective for initiating field trials by October 2005, participants devised the following workplan:
|
ACTION |
DATE |
WHO |
|
1. Workshop |
1. Completed: 23rd June 05 |
1. WRC, FD, IoJ, UWI, FTBG |
|
2. Assessment of seeds / seedlings currently available |
2. 27th June |
2. M. Barrett - FD |
|
3. Seeds sown in nurseries |
3. Mid-July |
3. FD, WRC |
|
4. Seed traps deployed |
4. Mid-July |
4. WRC, consult with UWI |
|
5. Cow exclusion |
5. End of Aug |
5.WRC |
|
6. Land preparation, demarcate plots, planting |
6. Mid-October |
6. Nelson, Boswell - FD |
|
7. Social surveys |
7. End of July |
7. Bennett, P Barrett - FD, WRC, Duanvale |
|
8. Fruit tree selection workshop |
8. Third week of July |
8. WRC, RADA |
|
9. Community meeting |
9. Mid-Aug |
9. WRC |
|
10. Fruit tree planting |
10. Mid-October |
10. WRC, FD |
**************
Workshop summary prepared by:
Susan Koenig
Windsor Research Centre
Date prepared: x July 2005