BCNet     Community Monitoring of Ngali Nuts in the Solomon Islands

A Report on the Field Implementation of a Biological Survey


by John Parks, Victor Kohaia, and Francis Tarihao

August 1996


Related BCNet Links

  • Ngali Nut Oil Soap
  • Nut Oil Processing and Ecotourism, Makira Island, Solomon Islands

    • TABLE OF CONTENTS

    1. INTRODUCTION
    1.1 Background to the Biological Survey
    1.2 Survey Goal and Objectives
    1.3 Survey Participants
    2. SURVEY DESIGN AND METHODS
    2.1 Pilot Investigation
    2.2 Methodology and design of the Survey
    2.2.1 Fundamentals in the Design
    2.2.2 Identifying Control Groups for the Survey
    2.2.3 Finalizing the Design
    2.3 Survey Locations and Dates
    2.4 Survey Methods
    2.4.1 How did we get it set up?
    2.4.2 How was it done?
    2.4.3 What data did we collect?
    2.5 Survey Constraints
    2.5.1 Climate
    2.5.2 Lack of Control Sites
    2.5.3 Sampling Was Largely Limited to Footpaths
    2.5.4 Replication and Confidence of Data
    3. RESULTS
    3.1 What have the collected data told us?
    3.1.1 Results from the East Bauro (Warihito) District
    3.1.2 Results from the Wainoni District
    3.1.3 Presentation and Community Interpretation of the Results
    3.2 What human impacts did the team observe?
    3.3 What observations and consequences did the participatory community members communicate?
    4. PROCESS CONCLUSIONS AND FOLLOW-UP RECOMMENDATIONS
    4.1 Conclusions related to the Survey Goal & Objectives
    4.2 Conclusion related to Sustainability Issues
    4.3 Follow-up Needs and Corresponding Recommendations
    4.3.1 Need for Continued On-site Technical Support
    4.3.2 Need for Analysis of the Baseline Data
    4.3.3 Need for Community Interpretation of Results and Analysis
    4.3.4 Recommended Next Steps for Follow-up
    APPENDIX ONE

    1. INTRODUCTION

    1.1 Background to the Biological Survey
    A core element of all BCN-funded projects is the monitoring of project impacts. Therein, BCN strives to support its grantee partners in the collection of relevant biological and social information which can be utilized by participating communities for appropriate and effective decision-making in the adaptive management of local resources.

    Through communications with the Solomon Islands Development Trust (SIDT), Maruia Society, and Conservation International (CI) during April 1996, the need for technical assistance in the facilitation of a community-based biological monitoring strategy within the SIDT Makira Conservation in Development (CID) Program became apparent. As BCN's May 1996 site visit to the CID Project presented an ideal opportunity to directly offer technical assistance while in the field, discussions regarding the relevance and potential for such assistance occurred. As a result of these discussions, support and approval for BCN's offer was given by all three grantee partners (SIDT, CI, and Maruia Society).

    In addition to the site visit team of Hank Cauley (BCN), Robin Conner (Maruia), and John Parks (BCN), Julian Ash, Associate Professor of the Australian National University, Canberra, was contracted by CI/Maruia to accompany the team and take the lead in developing an appropriate and feasible monitoring strategy. This strategy was aimed at collecting relevant biological information on Canarium indicumtrees within the forest areas of participating communities in the CID Project. Upon arrival into Makira on May 13th, the visiting team was joined by CID Project Personnel Francis Tarihao, Victor Kohaia, and Alfred Gari, as well as Roger James, who is contracted by CI to do research on frugivorous pigeons.

    1.2 Survey Goal and Objectives
    The overall goal of the biological survey process was to identify, develop, facilitate, and establish a community-based protocol for the periodic collection of key biological information on the ngali nut tree (Canarium indicum) in relation to the CID ngali nut oil extraction enterprise.

    Three objectives of the survey process were to:

    1. Compile the necessary biological baseline data within the CID project area to monitor the sustainability of ngali nut harvest over time;

    2. Identify the feasibility and viability of community-based biological monitoring activities within the CID project area;

    3. Provide on-site technical assistance to the national NGO grantee partner within the biological component of the project and provide capacity building at the local level to participating project communities;

    Additional objectives of the visit to Makira were to:

    1. Assess the need for further technical assistance within any of the three project components (i.e. biological, socioeconomic, and enterprise);

    2. Strengthen the working relationships and increase communication between BCN and its grantee partners; and

    3. Act upon recommendations outlined within the BCN Mid-Term Evaluation Report.

    1.2.1 What do we want to know and why?
    Previous studies completed by CID Project staff and participating communities were focused on: 1) estimating the local population of fruiting C. indicum harvested within the forest area held by landowners in the CID Project Area; and 2) estimating nut production values for local trees. The results of a census conducted by the project estimated that the number of known, harvestable trees within the project area exceeded 3,000, representing a potential yield of over 45 tons of nuts. Conservative estimates generated from these studies suggested that an average yield of 15 kg of dry kernels could be harvested each year per ngali tree. Additional yield information is currently being generated through semi-structured interviewing techniques conducted by the press manager and assisting village residents, who record information regarding the number and size of trees harvested by collectors at the time when the collectors bring their nuts to the nut press for sale.

    Although the CID Programme has already begun to generate information regarding the number of fruiting trees and potential yield, the issue of how the operation of the ngali nut press (i.e., the enterprise component of the project) effects the biology of local ngali nut trees had not been adequately explored prior to the May 1996 site visit. Thus, the visiting team defined their survey intentions through a single question:

    In an attempt to identify those data which required collection to answer this question, the process inevitably boiled down to a focus on how to most efficiently and appropriately gather regeneration information, particularly in the form of ngali nut recruitment and growth rates within the forest canopy. More specifically, size class structure/distribution, growth, and seedling recruitment information needed to be compiled in order to begin developing a strong baseline from which a more concrete and precise understanding of the harvesting impacts on C. indicum biology could be assessed over time. Consequently, the site visit team developed an appropriate community-based methodology for a survey designed to gather such information during the latter half of the week of May 13th (see below, § 2.1).

    It is important to keep in mind that the ultimate purpose for completing the biological survey was to begin gathering information which would eventually enable the participating communities and CID project staff to more effectively gauge the impact of ngali nut collection on their surrounding forest communities. Through community examination of this information it is hoped that appropriate decisions can be made in regards to ensuring the sustainability of nut harvesting from local forests. Moreover, it is assumed that through the continued monitoring of natural resources on the community-level, the recognition of anthropogenic impacts on forest biodiversity will be reinforced.

    1.3 Survey Participants
    The day following the design of the survey methods by the visiting team (see below,§ 2.1), Ash, Cauley, and Conner returned to Honiara (May 18th). Thus, Tarihao, Kohaia, Gari, James, and Parks took the responsibility of implementing the survey in three geographic areas within the CID Project Area between the 19th of May and the 5th of June. Most importantly, SIDT Mobile Team Members (MTMs) and local villagers actively participated with the survey team in completing the biological studies within each geographic area.

    Due to time constraints, it was decided that the survey team would divide into two separate groups and split the work of completing the survey in the three geographic locations. Kohaia and Wagner proceeded with surveying ngali trees within the Districts of East Bauro (Warihito) and Wainoni, while James returned to his research post in the Bauro Highlands to survey trees within that geographic area. Despite their prior commitments to community development activities within the highlands and along the weather coast during the survey time-frame, Tarihao and Gari managed to actively participate in the Warihito area surveys for a few days prior to their departure from the Warihito survey area, and thus engaged in survey activities long enough to adequately familiarize themselves with the methods. In addition to the participating SIDT staff, a major driver in the successful implementation of the survey efforts was the consistent enthusiasm displayed by participating community residents.

    2. SURVEY DESIGN AND METHODS

    2.1 Pilot Investigation
    An initial "pilot investigation" of five days (May 13-17) was completed in the East Bauro (Warihito and Bauro Highlands Areas) and Wainoni Districts by the visiting team and several participating community members. This resulted in the generation of preliminary information regarding the growth, size class distribution, and recruitment of C. indicum.This information was obtained through visits to abandoned village and garden sites where knowledge on the ages of specific individuals and/or cohorts of ngali trees was held by the local community. Circumference breast height (CBH) measurements (later converted to diameter, or DBH) and observations on seedling abundance within a specific area (relative density), paired with information on the known ages of previously planted or naturally recruited trees, assisted in shaping the initial hypotheses and assumptions generated by the team while working in the field during the first week. Circumference data were chosen for collection because of: 1) the relative ease, speed, and guaranteed precision inherent in a circumference measurement by local village participants, who lacked experience in accurately estimating diameter measurements; and 2) the fact that the survey team did not posses a DBH tape. Differentiations between forest habitat types and the scale of human impact within specific geographic areas of the CID Project Conservation Area were also observed and discussed by the visiting team and local participants during the pilot investigation period.

    2.2 Methodology and Design of the Survey
    During a meeting held on the afternoon and evening of May 17, discussion of the results and conclusions of the visiting team's findings for the week occurred. During this meeting, Ash led the group in the conceptualization of a community-based survey strategy, introducing a few ideas on how to design an implementable and practical sampling process for obtaining the necessary biological data to begin to determine harvesting impact effects on ngali regeneration.

    2.2.1 Fundamentals in the Design
    Given the team's observations during the week of pilot investigations, coupled with Ash's knowledge and experience with Canarium spp. in the Pacific, it was clear that the design of the survey would have to incorporate the following fundamentals:

    1. The proposed survey would have to be completed in three geographic locations within the CID Project Area.Biological information on C. indicum trees would be collected by the survey team and participating community members within three, separate geographic survey areas. The rationale for breaking the survey efforts down over these three areas was two-fold:

      • to measure impact disparities between habitats -- Each of these three geographic areas can be characterized under a different habitat type. Ngali nut harvesting occurs within each area, and each area actively contributed to the 1995 nut take by the nut press; thus the team deemed it relevant to sample within each habitat area in an attempt to examine impact disparities due to habitat differentiation.

      • to measure impact disparities between harvesting methods -- Additionally, differences in the harvesting methods used by local residents in the three geographic areas has subsequently resulted in discrepancies in the degree of impact between these areas.

      Once adequate sampling had occurred over these three geographic areas, impact variability as a function of the incongruent habitat types and harvesting methods over the project area could thereby be extrapolated, analyzed, and interpreted. An index of these differentiations over the three geographic survey areas is listed in Table 1, below.

    2. Paired surveys would need to be completed at each geographic area.Within each of the three geographic areas, ngali trees would be surveyed in the following two locations (i.e., paired surveys):

      • In places where harvesting of ngali nuts occurs (i.e., a treatment site)

      • In places where no harvesting occurs (i.e., a control site)

      The group recognized that in order to answer the proposed "impact of nut harvesting" question, the design of an appropriate survey strategy would focus on data comparison between harvested (anthropogenic + natural/autogenic impacts) and non-harvested (only natural/autogenic impacts) sites of ngali nut trees within each respective geographic area.


    2.2.2 Identifying Control Groups for the Survey
    As the identification of treatment locations would be a relatively simple task given the fact that the area within which the sampling would occur is essentially a large agroforestry system, the identification of paired control (i.e., non-harvested area) sites within the same area became the major focus within the group's discussion. Following the team's discussion on the two survey fundamentals detailed above, a methodological conversation occurred on how to identify control areas against which harvesting impacts on ngali nut biology could be measured. This discussion culminated in two conclusions:

    1. There are two general 'habitats' of concern over which the survey should focus, one of which would be used as the survey control. Local distributions of ngali nut trees can generally be classified under two geographic areas, or 'habitats', of differential harvest pressure:

      • "garden" habitat -- incorporating all areas under which ngali nuts are subjected to harvesting by village residents. The "garden" habitat would be used as the survey's treatment.

      • "bush" habitat -- forest areas in which the harvesting of ngali nuts does not occur. This definition of "bush" therefore includes both undisturbed primary forest, as well as previously disturbed (secondary) forest which has long since been abandoned by humans and is now barely, if at all, distinguishable from undisturbed primary forest. The "bush" habitat would be used as the survey's control.


      Table 1: A breakdown of the different habitat types and local harvesting methods within each geographic area of the ngali nut surveys.
      Geographic area of survey Habitat type Harvesting method Survey dates
      East Bauro District:
      Warihito Area       		lowland hills/riverine forest
      		gathering of fallen nuts
      (in situon forest floor)       		May 21-25 &
      June 1; July ÷
      East Bauro District:
      Bauro Highlands Area
      		sub-montane forest (up to ~600 m
      elevation)
      		extraction of fruiting
      branch from tree
      		May 27-31
      Wainoni District:
      Naharahau & Nagatare Area
      		coastal ridge forest
      		gathering of fallen nuts
      (in situon forest floor)
      		May 27-30
      ÷ CID staff and community residents completed surveys within the Warihito Area during July 1996 in continuing efforts to gather the necessary baseline data.

      Thus, a survey within each of the three geographic areas would require the sampling of two paired sites, one "bush" (i.e., a control, or non-harvested site) and the other "garden" (i.e., a treatment, or harvested site).

    2. Traditionally 'tambu' areas would have to serve as control sites in place of pristine bush areas. In consideration of completing paired surveys, it was clear that due to the frequency and extent of nut harvesting activities within the project area, locating a control, or "bush", site which was geographically adjacent to the harvest area surveyed would be difficult. CID Staff members pointed out that the sphere of influence of any given village on the surrounding environment typically extended as far as a two or three day travel from the village through the forest. This is particularly true along designated forest footpaths, where these "corridors of harvest" are frequented by groups while hunting for wild pigs or tending gardens in remote family plots. In the opinion of the CID Staff, one would have to travel three full days before reaching truly "bush" areas containing nut trees which are completely free of harvest activity. Even if it were possible for the survey team to travel three days into the bush to sample a truly pristine, non-harvested site of ngali trees, the simple fact that this survey would no longer be "paired" in proximity to the harvest area surveys alerted the team of the potential error which could enter into the data collected. Thus, the issue was how to locate and conduct a survey within a non-harvested site which was "paired" to a harvest survey site when there were reportedly no ngali trees in the area which were not subjected to local harvesting.

      Resolution of this issue came through the decision to use traditionally "tambu" (i.e., forbidden or sacred) sites in lieu of pristine "bush" areas. Such tambu sites represented both a historical moratorium on the harvest of ngali nuts by local residents, as well as a guarantee that this local custom of obeying non-harvest practices in the area would likely be maintained into the future. However, two confounding factors were recognized within the use of tambu areas:

      • Not all tambu sites forbid the harvesting of ngali nuts from trees within their area.

      • Within any given geographic area, the frequency of tambu sites whose total area contains an adequate density and/or population of non-harvested ngali nuts to allow for the necessary sample size is very low.

      Thus, even when using tambu areas a substitutes for pristine bush, the survey team was faced with the realization that the utilization of this solution was not without constraints of its own.


    2.2.3 Finalizing the Design
    Eventually, the discussions regarding the design of the survey enabled the group to agree upon the proposed design of a survey strategy, built principally upon the pilot investigation observations and conclusions discussed above. According to the proposed design, survey participants would run either belt transects (of varying widths) or 40 m2 quadrats (which the team called "plots") along fixed compass bearings within any particular survey site (i.e., "garden" or control site). Any C. indicum trees or seedlings encountered within a given transect or plot were measured, recorded, and marked. A full description of the survey methods utilized follows (see below, § 2.4).

    The meeting ended following a brief conversation on how the survey team would have to adequately "fine tune and de-bug" the proposed survey procedures during the first few trials in the field.

    Table 2: A descriptive listing of the Warihito and Wainoni Area ngali nut surveys completed.
    Dates Geographic Area Impact Status Survey Location Type # trees/area m2
    Measured
    May 21-23
    		Warihito (East Bauro District)
    		 harvested
    ("garden")
  • Wanauhiuhi (old village)
  • Puruwapare (ridge trail)
  • Birobiro (ridge trail)
  • plot a
  • plot a
  • transect
    		• 57/4200 m2
    May 24; mid-July B Warihito (East Bauro District) non-harvested
    ("bush")
  • Tararihu (tambu site)
  • 3 transects
    		• 50/7100 m2
    May 25 & June 1 Warihito (East Bauro District) harvested
    ("garden")
  • Warohino (current village gardens)
  • 3 transects
    		• 33/14380 m2
    May 27 & 28 Nagatare (Wainoni District) harvested
    ("garden")
  • Wanaparoro (ridge trail)
  • transect
    		• 13/1470 m2
    May 28-29 Naharahau (Wainoni District) harvested
    ("garden")
  • Nabomeraha/Namuria (current village gardens)
  • 2 transects
    		• 58/10460 m2
    May 30 Robogena (Wainoni District) harvested
    ("garden")
  • Roto (ridge trail) to Mahuganae (old village)
  • transect
    		• 18/5150 m2
    a Plot surveys were initially conducted in an effort to test their utility; the survey team decided to drop the use of plots due to their inherently time-consuming nature and because they require a relatively constant slope to perform with precision, thus being an impractical design when sampling along mountain ridges.

    B CID staff and community residents completed the third transect at the tambu site during July in an effort to continue the collection of baseline data from the Warihito non-harvested (i.e., "bush") area.

    2.3 Survey Locations and Dates
    Table 2, above, details the locations and dates of the survey team's efforts within the Warihito and Wainoni Areas. Within the table, note that only one of the ten total survey days during the May/June site visit reflects sampling under control conditions (June 3 was planned as a control survey day within the Warihito area, but torrential rains and subsequent high floods prevented the survey team from obtaining any additional control data). It is also important to point out that supplemental control data from the Warihito tambu site has been collected by community members and CID staff since the conclusion of the May/June site visit (surveys were completed during the middle of July). Additional plans to complete the remainder of the control (tambu) survey in Warihito, as well as to begin a "bush" survey in Wainoni, have been tentatively planned for August/September 1996 by CID Programme Staff. Survey efforts within the Bauro Highlands was coordinated through frugivorous pigeon researcher Roger James (representing CI) between May 27 and 31. Due to other research obligations, only a "garden" survey was possible in the highlands during this period of time. Locations and specifications of the completed "garden" survey have not as yet been made available, and thus are not discussed within this report. Plans to complete paired surveys (i.e., "garden" and "bush") within the highlands during 1996 have reportedly been discussed.

    2.4 Survey Methods
    2.4.1 How did we get it set up?

    Communication with the Local Communities
    Prior to engaging in sampling activities within each geographic area, the survey team met with the respective communities and discussed with the local residents the purpose and utility behind performing such biological studies of C. indicum. These community meetings were well attended, and took place within Warohino and Naharahau Villages on the 20th and 26th of May, respectively. Most importantly during these meetings, the survey team requested authorization from the participating communities to conduct such surveys on their land. After receiving consent from the local chief(s) and elders on the proposed survey, the team briefly summarized the methods which they were to perform during sampling activities.

    Standardization of the Survey Procedures
    Within each of the three geographic areas of the study, the corresponding survey team members set up the sampling procedures in each survey site identically. Because sampling procedures deployed within control sites were the same techniques deployed within treatment sites, the team assumed that bias was minimized within the sampling regime.

    Parameters on the Paired Survey Sites
    During the May 11th meeting in which the visiting team finalized the survey design, comments were made in regard to how actually to go about setting up surveys within the paired sites. As the paired sites represent two different harvesting impact levels (i.e., impact and no impact), the conclusions reached during the design meeting led to the use of the following parameters during the set up of survey:


    --->


    WHY BIODIVERSITY   FIELD STORIES   PARTICIPATE   RESULTS   MARKETSPACE

    ABOUT THE NETWORK   WHAT'S NEW   LEARNING MATERIALS   SEARCH   LINKS   SITE MAP   HOME