Improving aquaculture in post-tsunami Aceh, Indonesia: experiences and lessons in better management and farmer organizations
Abstract
Coastal aquaculture in Aceh was severely affected by the Asian Tsunami in December 2004. Capacity building among stakeholders was one of the key activities implemented by various agencies during the post-tsunami aquaculture rehabilitation and subsequent development phase. The main objective was improving production efficiencies and farmer incomes. This article describes the process of implementation of the approach and crop outcomes until the end of 2009. In 2009, 1135 farmers and their 1296 ponds were ‘organized’ in 27 clusters across 84 villages of three districts to implement better management practices (BMP). Interventions reduced the prevalence of shrimp disease outbreaks significantly in participating ponds (22.45%) compared with non-participating ponds (62.64%). Among the normal harvested ponds, though there was no significant improvement in shrimp yield but costs of shrimp production were significantly reduced by 12.1% and benefit cost ratio was significantly increased by 0.523 points in normal harvested participating ponds compared with normal harvested non-participating ponds. Multivariable logistic regression analysis with 27 independent variables including a set of BMP showed that many factors were significantly associated with the normal harvests, improved yield and profits. The study shows that simple management improvements can reduce risks and improve economic returns.
Introduction
Aquaculture is an important source of food security, employment and income for many people in the coastal regions of the Aceh Province in Indonesia. The sector was severely damaged by the earthquake and tsunami that struck the region on 26th December 2004 (Phillips & Budhiman 2005).
The main commodities produced by aquaculture in Aceh are milkfish, Chanos chanos Forsskal 1775 and black tiger shrimp, Penaeus monodon Fabricius 1798, from coastal ponds (tambak) and a range of freshwater fish species from inland ponds and rice fields. A diversity of minor fish, shrimp and crab species are also farmed in diverse farming systems across the province, including ponds, rice fields, fish pens and cages, hatcheries and nurseries located in marine, brackish and fresh waters. Total production as provided in Provincial Government (Dinas Kelautan dan Perikanan – DKP) statistics was around 32 000 tonnes in 2006, although this is almost certainly an overestimate. The economic value of aquaculture is highly significant, estimated at IDR 692 billion (US$76.9 million considering the currency exchange rate of US$1 = IDR 9000) in Provincial Government (DKP) statistics for 2006, of which brackishwater pond farming, mostly shrimp, contributes 75% of the value. Provincial Government (DKP) statistics, estimate 42 000ha of brackishwater ponds, 2900 ha of freshwater ponds and 2500 ha of rice field, plus cages and pens used to produce various species of fish and crustaceans. An estimated 20 000 ha of brackishwater ponds were damaged by the tsunami, and according to the Bureau of Rehabilitation and Reconstruction of Aceh and Nias (BRR) only 13 500 ha (or 67%) were rehabilitated by the end of 2007. Overall, productivity of ponds is low, reflecting traditional farming practices and technical/managerial constraints, with yields well below the Southeast Asian average for traditional extensive ponds. Aquaculture in Aceh is widespread, but the economically most important aquaculture districts are along the northeast coast, districts that were heavily affected by both the tsunami and preceding political conflict.
Aquaculture in Aceh was and still is a highly significant livelihood activity. Provincial Government statistics in 2006 indicate there are around 40 065 households involved in brackishwater ponds (23 347), freshwater ponds (10 572), rice field culture (5776) and cage/pen culture (370). These figures appear to include only farmer owners or operators and therefore exclude many other stakeholders, significantly underestimating the number of people involved as employees and labourers, suppliers of inputs, hatchery operators and workers, and in trading, marketing and service provision. There may be 100 000 people and their families involved with brackishwater pond farming, and over 20 000 in freshwater aquaculture. With new estimates of brackishwater pond area, the sector may be important in the livelihoods of considerably more. The sector is clearly a critical one for livelihoods and food security in the Province.
Aquaculture in Aceh, like much of Indonesia, is dominated by small-scale farmers, in both inland and coastal areas, who individually produce small amounts, but collectively contribute significant amounts to total output volumes and values. There are some more ‘commercially’ oriented aquaculture ventures, but the majority of stakeholders are small-scale rural farmers and aquaculture development is important for poverty reduction. Brackishwater pond farming in Aceh is dominated by traditional, low input – low output, farms producing shrimp and milkfish in polyculture and monoculture, with most farm areas less than 2 ha. The small-scale nature of the sector, as in many countries in Asia, poses special challenges with servicing and market access in a competitive and rapidly changing market place. Final markets for aquaculture products from Aceh range from household consumption, sale to local markets for consumption within the province, export to other provinces in Indonesia, and export to other countries, within Asia and other regions. Economically important export commodities are shrimp, crab and groupers. Major products for domestic consumption are milkfish and freshwater fish. Acehnese black tiger shrimp is a highly valued export commodity previously with a good reputation among buyers in Medan in North Sumatra Province, but concerns over poor quality, use of chemicals, erratic supplies and poor traceability has reduced interest in this product in recent years, and particularly since the tsunami.
Assistance to rehabilitation of aquaculture facilities and the livelihoods of people involved in Aceh started soon after the December 2004 earthquake and tsunami, in early 2005. A number of national and international agencies were involved in various ways, including Australian Center for International Agricultural Research (ACIAR), Asian Development Bank Earthquake and Tsunami Emergency Support Programme (ADB ETESP), AusAid, Aquaculture Without Frontiers (AwF), Bureau of Rehabilitation and Reconstruction of Aceh and Nias (BRR), Food and Agriculture Organization of the United Nations (FAO), French Red Cross (FRC), German Technical Cooperation (GTZ), International Finance Corporation (IFC), Japan International Cooperation System (JICS), Ministry of Marine Affairs and Fisheries, Government of Indonesia (MMAF), Network of Aquaculture Centers in Asia-Pacific (NACA), WorldFish Center, World Wildlife Fund (WWF), United Nations Development Programme (UNDP), and others. The initial work of the agencies was oriented towards physical restoration of ponds and canals, and providing some assistance with inputs such as feed, seed and fertilizers for farmers to start farming again.
Following an initial period of physical restoration, it became apparent that the coastal farming sector suffered from a number of significant constraints, similar to other small-scale farming areas in other parts of the region, but perhaps made more challenging in Aceh because of the poor institutional support caused by the tsunami, and more generally by the years of neglect related to the preceding period of political conflict. These include poor farming practices, shrimp disease risks, and limited extension capacity, amongst others. Subsequent initiatives of several agencies then became more oriented towards supporting the rebuilding of services and institutions, and testing and adopting of improved management practices (eventually referred to as better management practices, or so-called BMP) among coastal farms to reduce risks and secure improved and more consistent yields.
Historically, the need for development and implementation of responsible coastal shrimp aquaculture has been felt both nationally and internationally following questions arose on its impact on environmental and socio-economic sustainability starting in the early 1990s. As a result, a global consortium led by FAO, NACA, UNEP, WWF and World Bank reviewed various farming practices in various aquaculturally important countries (WorldBank, NACA, WWF, FAO 2001, 2002). This exercise culminated in publication of a manual on ‘International Principles for Responsible Shrimp Farming’ in 2006 (FAO, NACA, UNEP, WB & WWF 2006). In parallel, a collaborative project between MPEDA and NACA in India on disease and environmental management in coastal shrimp farming starting 2000 has scientifically studied shrimp disease using epidemiology and developed set of risk management practices that were referred to as BMP which were field tested (Padiyar, Phillips, Primphon, Mohan, Bhat, Rao, Ravi Babu, Mohan, Murthy & Chanratchakool 2003; Padiyar 2008; Umesh, Mohan, Phillips, Bhat, Ravi Babu, Chandra Mohan & Padiyar 2008; Umesh, Chandra Mohan, Ravibabu, Padiyar, Phillips, Mohan & Bhat 2010). The MPEDA/NACA project published a shrimp health extension manual during 2003 (MPEDA/NACA 2003) and the success of shrimp farming by implementation of BMP and more organized clusters by farmer groups (AquaClubs) among small-scale holders in India has significantly influenced other shrimp farming nations in the region to adopt similar ways for sustainability of the small-scale shrimp farming sector (Padiyar et al. 2003; Padiyar 2005a, 2005b; Umesh et al. 2010). Furthermore, FAO has widely promoted ‘cluster management in small-scale shrimp aquaculture’ (Subasinghe 2005). Since the shrimp farming sector in Aceh resembles small-scale shrimp farming sector of India, it was considered appropriate to explore a similar approach with traditional low input farmers in Aceh.
This article is part of a series on aquaculture rehabilitation in Aceh province but focusing on explaining the results from the BMP and cluster management work, which we believe will be of wider relevance to initiatives to support improved farming practices among small-scale farming systems in brackishwater areas in Asia.
Materials and methods
Approach to BMP development
A participatory description of the coastal brackishwater ponds systems and management practices was conducted in 2005, leading to formulation of a locally appropriate set of BMP for pond production in 2005 (subsequently published in 2007 – ADB, ACIAR, AwF, BRR, DKP, FAO, GTZ, MMAF, NACA & WWF 2007). The BMP are based on the international principles for responsible shrimp farming (FAO, NACA, UNEP, WB & WWF Bangkok, Thailand), international experiences on the better management of shrimp aquaculture in coastal areas (WorldBank, NACA, WWF, FAO 2001; MPEDA/NACA 2003; Padiyar et al. 2003; Padiyar 2005a, 2005b, 2008; Umesh et al. 2008; Umesh et al. 2010), local aquaculture stakeholder consultations in Aceh (farmers, hatchery operators, government extension service units, farm input suppliers etc.) and field experiences of the authors from other regional countries. These BMP were ‘tested’ on 47 farmer volunteers (22 ha of ponds) in 11 villages during 2007 and 260 farmers (184 ha) in 34 villages during 2008. After completion of every crop cycle the BMP were modified/fine-tuned through stakeholder consultations to discuss the crop results and next crop planning at district level. The list of BMP tested in this study is given in Box 1. Of these BMP, screening shrimp PL by two-step PCR test, no use of seed from commercial nurseries, water depth of >30 cm at shallowest part of the pond, no use of pesticides and harmful chemicals and no use of live and fresh feed were considered as mandatory for farmers to follow under the programme. The rest of the BMP were optional (voluntary) to implement based on farmer willingness and affordability to implement them.
Approach to BMP promotion
Various types of BMP extension materials were prepared and disseminated to the stakeholders in Aceh province including the farmers in the target sub-districts (Table 1). The BMP leaflet and manuals distributed to farmers in target villages, BMP posters posted at farming community operated Aceh Aquaculture Livelihoods Service Center (ALSC) (Ravikumar & Yamamoto 2009), village community centres, government offices, farm huts; BMP video road-shows at villages; BMP radio talk-shows and jingles (twice a week for entire crop period broadcasted from local radio stations in target area). The BMP promotion approaches used were modified with time using a monitoring and evaluation system as described elsewhere in this article.
| Extension method | Extension outreacha | Number of beneficiariesb |
|---|---|---|
| BMP manual | 5642 | 5642 |
| BMP leaflet | 8920 | 8920 |
| BMP poster | 650 | >650 |
| BMP video road-shows | 96 | 19 907 |
| BMP radio talk-shows and advertisement spots | >250 | All the farmers in 3 target districts. Exact number of radio listener farmers not available |
| Hands-on training to farmer leaders (Kontak Petambaks) on BMP, farmer group management etc. | 2 | 49 |
| Village level hands-on training to farmers | 54 | 887 |
| Hands-on training to extension staff (project and local government) | 1 | 24 |
| Local field days (farmer study tour) | 5 | 152 |
| International study tour to India (visit to NaCSA-MPEDA promoted farmer societies in Andhra Pradesh) | 1 | 15 |
| On-farm direct technical assistance during 2009 | – | 1135 |
- a Number of print materials distributed/audio-visual shows conducted/trainings conducted.
- b Number of beneficiary farmers that received/influenced by the method.
Technical organization
The existing systems of extension in Aceh were extremely weak, in some cases as a result of loss of personnel in the earthquake and tsunami disaster, but often on the north-east coast because of years of conflict and lack of effective skills. To support extension and conduct BMP testing with farmers, a team of Acehnese field facilitators (21) were recruited during January 2007–July 2008 with minimum selection criteria of a degree from university or fisheries and aquaculture vocational school and ability to speak the Acehnese language. Some of them had 1–5 years of experience working with pond farmers in Aceh and Indonesia. They were trained in extension, tambak BMP, monitoring and data collection. They were housed at target sub-district centres for community rapport building and for extension services to farmers on a full-time basis from the period July 2008 to December 2009.
Enrolment of participating farmers
Following positive outcomes and the lessons learned in 2007 and 2008, a wider recruitment drive for volunteer farmers was made for the 2009, which is the main subject of this article. The recruitment drive was conducted during November 2008–March 2009. All the villages in the target sub-districts were informed on the study objective, which is organizing the farmers in groups for collective implementation of BMP to reduce the impacts of shrimp disease outbreak. Also, they were informed on the BMP (mandatory and voluntary) (Box 1), criteria for selection of ponds (Table 2), the on-farm extension services provided by the study and shrimp seed quality ‘incentive’ that they would receive from the project for participation. The quality incentive was intended to encourage farmers to use quality, virus tested post-larvae. A maximum of 50% of the shrimp PL price (ca. 8000 IDR, or US$0.9 per 1000 seed) was provided by the study as incentive for participation to compensate for the additional cost that hatchery operators have to bear for improving the seed production practices and for PCR screening of post-larvae. This amounted to about US$18 ha−1 at a stocking density of 20 000 shrimp ha−1. The canvassing for the BMP intervention study was through information posters, community meetings, meeting community leaders, one-to-one meetings with farmers. Based on their willingness to cooperate and acceptance of the conditions of the cooperation and BMP, and based on the fulfilment of pond selection criteria, farmers were recruited for the study. The study team monitored and recorded the activities of farmers, in particular, the BMP by visiting the farms at least once a week and extended technical support to farmers for BMP implementation.
| Pond condition | Pond condition evaluation grade | |||
|---|---|---|---|---|
| A | B | C | D | |
| Condition of embankment (perception of technical team) | Strong and good | Medium | Bad | Very bad |
| Water depth at shallowest part of the pond (mid bed, not side ditches) just prior to stocking | >60 cm | 45–60 cm | 30–45 cm | <30 cm |
| Recommended number of seed and for which the seed subsidy is provided | 30 000 seed ha−1 | 20 000 seed ha−1 | 10 000 seed ha−1 | Do not stock shrimp |
In total, 1135 farmers with 1296 ponds situated in 84 villages in the target sub-districts were enrolled in the study. Monoculture of shrimp and polyculture of shrimp, milkfish and/or tilapia were followed in 1033 (79.7%) and 263 (20.3%) ponds respectively.
Farmer organization
According to some senior brackishwater farmers in Aceh, Petuah Neuhun (meaning water canal manager in Acehnese language) system existed as a traditional farmer organization in Aceh for management of brackishwater water canals and dependent aquaculture ponds for the previous three generations. However, the organization became weak starting in the early 1980s as a result of leasing out of some farms to non-Acehnese investor farmers who were less concerned about local customs and tradition. During the 1980s brackishwater farmer organization named BMPT (Badan Musyawarah Petani Tambak) and during 1990s UPP (Unit Pelayanan Pengembangan) were established and recognized by the local government for channelling government development funds and activities to the farmer community. These organizations were established at district and sub-district level and not at village level. These two organizations focused only on distribution of government aid to farmers. As a result, they became notorious among the farming community because of financial irregularities and poor or non-delivery of any technical services to farmers. During 2005–2007, various post-tsunami aquaculture programmes in Aceh were attempted with BMPT and UPP but eventually attention shifted towards revival of the Petuah Neuhun system. The study reported here worked towards assisting farming communities with revival of the Petuah Neuhun system from 2008.
Farmers were assisted with organization by facilitating farmer groups at village and cluster level. At village level, the farmers were helped to organize into informal self-help groups (Kelompok) with 10–20 farmer members per Kelompok which was managed by a Kontak Petambak (Kelompok leader). There were 113 Kontak Petambaks in 84 target villages. Tambaks sited in villages situated along the same water canal that are inter-dependent for sourcing the water and community management of the shrimp disease were grouped in to clusters (Figure 1). Farmers in each cluster selected a Petuah Neuhun (Cluster leader) to manage the cluster. There were 27 clusters and Petuah Neuhun in the study.
Investments were also made in improving some extension services. ADB ETESP in cooperation with NACA, FAO and WorldFish Center supported establishment of farmer owned, operated and legalized Aquaculture Livelihoods Service Centers (ALSC) under the leadership of one of the local Petuah Neuhun at four places in the target districts. An Aceh Aquaculture Communication Center (AACC) was also established at BBAP Ujung Batee during early 2009. ALSCs provided services (technical, marketing and other services) to farmers situated in surrounding clusters with each ALSC connected to the AACC by internet for technical assistance (Ravikumar & Yamamoto 2009).
PCR screening of shrimp PL for WSSV and purchase of seed direct from a reliable hatchery were considered mandatory and the study team supported participating farmers to purchase shrimp seed from a local hatchery in Bireuen following contract hatchery system (Padiyar 2005a, 2005b). PCR tests were conducted at BBAP Ujung Batee on 37 batches of shrimp PL selected from 37 different larval rearing tanks from the hatchery using IQ2000™ WSSV Detection and Prevention System kit (GeneReach Biotechnology Corp., Taichung, Taiwan). All the batches tested two-step PCR negative for WSSV. In total, 20.651 million two-step PCR negative for WSSV shrimp PL were procured by participating farmers in 2009. Stocking of shrimp seed started on 28 March 2009 and ended on 20 June 2009.
Analytical design
An intervention design was used for the study of the 2009 crop. In this design, ponds were regularly monitored throughout the cropping cycle for various pond parameters and the implementation of BMP (interventions) by farmers. The follow-up period was the summer brackishwater pond crop of 2009 (from 1 January 2009 to 20 October 2009). All the ponds situated in 15 coastal sub-districts in three adjacent districts (Bireuen, Aceh Utara and Kota Lhokseumawe) (Figure 1), were the target population of the study. The basic unit of study was pond. Farmed species were black tiger shrimp (P. monodon) in monoculture system or in polyculture with milkfish (C. chanos) and/or tilapia, Oreochromis niloticus Linnaeus 1758 and/or Oreochromis mossambicus Peters1852). The farming system was chosen by farmers.
During the follow-up period, non-participating ponds were identified through interactions with participating farmers, through group discussions with farming community at village level and by the field facilitators during the regular field visits. In total, 870 ponds in the study villages did not participate in the study and stocked shrimp. Monoculture of shrimp and polyculture of shrimp and milkfish were followed in 440 (50.6%) and 430 (49.4%) ponds respectively. And the remaining ponds in the target sub-districts were either farmed fish in monoculture system or were maintained fallow.
Monitoring and evaluation
Data collection
Once the farmer ponds were enrolled in the study, each pond was identified with unique pond ID and their GPS (geographical positioning system) coordinates. Field facilitators made visits to all the participating ponds once in a week by following a fixed time-table. They provided on-farm technical support to farmers to implement the BMP and helped farmers to record the farm data in pond books. The pond books (daily recording sheets) were distributed to farmers by the study. The farm data included ownership status, farm size, pond preparation practices (sludge removal, water filling, use of piscicide, fertilizer and lime), stocking details (seed source, seed quantity), daily farm management practices (water depth, feed and other farm inputs), daily farm observations (water quality parameters, shrimp health status), crop results (fish and shrimp yield, disease outbreak) and financial details (expenditures and revenue). Soil pH, water pH and salinity were measured using soil pH meter (Takemura Electric Works, Tokyo, Japan), pH meter (electronic) and refractometer (Atago, Tokyo, Japan).
Pond books were regularly and randomly checked at the pond site by the project manager and monitoring and evaluation (M&E) assistants for consistency of data recording. Upon completion of harvest in each pond, the pond books were returned to the study field office located in Bireuen town for data checking by two full-time M&E assistants.
For 870 non-participating ponds, a cross-sectional census was conducted using a questionnaire between 21 and 31 October 2009. The information collected included stocking details (fish and shrimp seed quality testing and seed source), crop outcomes (fish and shrimp yield, disease outbreak, disease outbreak and harvest dates), and financial details (expenditures and revenue). Completed questionnaires were cross-checked by M&E assistants.
Case definitions for crop outcomes
The case definition for a shrimp disease outbreak was developed prior to the cropping cycle in consultation with farmers in target sub-districts and considering the definitions given in an epidemiological study in small scale modified traditional shrimp farms in India (Padiyar 2008). For this study, a disease outbreak was defined as a case if there was a sudden and rapid increase in number of dead or moribund shrimp (up to 100% mortality within 3–5 days from the day of first observation of moribund or dead shrimp) in the pond leading to a farmer-initiated emergency harvest or abandoning of the pond by the farmer. However, during the study period in 93 ponds (7.17%) there were no live shrimp at harvest and no disease outbreak. Such ponds were neither considered as normal harvests nor disease outbreak and are excluded from the analysis. To confirm the disease affected stock for WSSV infection, the diseased shrimp were tested by Shrimple® dotblot kit (Fujikura Kasei, Tokyo, Japan). The dotblot test was conducted to confirm the presence of WSSV in sick shrimp.
The case definition for poor shrimp yield was determined upon completion of the crop. Both the mean (112.52 kg ha−1) and median (83 kg ha−1) values of the shrimp yield were considered for determining the cut off value of 100 kg ha−1, and poor yield (≤100 kg ha−1) and good yield (≥100 kg ha−1) was finalized upon consultation with the Petuah Neuhun.
The economic loss was defined as the negative return on investment. It was calculated by subtracting the actual crop expenditures (adjusted to the financial incentive provided by the study for seed procurement and for PCR testing) from the total revenue from sale of fish and shrimp. The crop expenditures included those for pond preparation activities (sludge removal, embankment repair, water gate repair, water intake, fertilization), shrimp and fish seed procurement and transportation, feed, lime, chemicals water exchange and labour wages.
Data processing and statistical analyses
Separate spreadsheets were created for participating ponds and non-participating ponds in Microsoft Excel 2007. Statistical analyses were performed in SPSS 15.0 (Statistical Package for the Social Sciences, SPSS; SPSS Inc., Chicago, IL, USA). All the descriptive statistics were produced in SPSS 15.0. Crop results (shrimp and fish yield, shrimp size, crop age at harvest, crop age at disease outbreak, shrimp survival) between the groups (participating ponds and non-participating ponds) were compared by anova test. Statistical significance was defined as P-value <0.05.
Potential factors (BMP, pond details and pond observations) associated with the three major crop outcomes with the case definitions: disease outbreak = yes/no, poor yield = yes/no, and economic loss = yes/no, were first screened individually through univariable logistic regression in SPSS 15.0. All the factors identified as having a P-value ≤0.20 in univariable analysis were then incorporated in multivariable logistic regression models using a backward step-wise approach with probability for step-wise entry and removal of P = 0.05 and P = 0.10 respectively. Factors included in the models were examined for multicollinearity through simple correlation in SPSS 15.0. Factors with R-values >0.7 were not included in the same model. The overall model was considered significant, if the −2 log likelihood test resulted in P ≤ 0.05. The overall model was considered valid (i.e. the model is a good fit to the data), if the Hosmer-Lemeshow goodness-of-fit test resulted in P ≥ 0.05. This analysis identified factors that had the strongest statistical association with the outcome, which would form the basis of further research.
Results
Descriptive statistics for continuous and categorical independent variables (BMP, pond details and observations) are given in Table 3 and Table 4. This analysis concentrates on the results from 2009.
| Variable | N | Mean | Median | SD | Min | Max |
|---|---|---|---|---|---|---|
| 1. Pond size (ha) | 1296 | 0.795 | 0.7 | 0.483 | 0.08 | 4.00 |
| 2. Shrimp stocking density (shrimp ha−1) | 1296 | 23 292 | 20 000 | 18 464 | 2000 | 300 000 |
| 3. Soil pH | 1270 | 6.15 | 6.2 | 0.714 | 3.0 | 8.2 |
| 4. Average salinity | 1278 | 17.37 | 16.5 | 7.01 | 1.0 | 33.5 |
| 5. Average water pH | 1281 | 8.00 | 8.1 | 0.53 | 5.25 | 9.05 |
| 6. Average water depth (cm) | 1294 | 51.72 | 50 | 15.75 | 15 | 105 |
| Variable | N | n | % |
|---|---|---|---|
| 1. Ownership | 1296 | ||
| 1 (Owned and operated) | 965 | 74.45 | |
| 0 (Leased) | 331 | 25.55 | |
| 2. Pond evaluation grade | 1272 | ||
| 1 (A) | 245 | 19.26 | |
| 2 (B) | 732 | 57.55 | |
| 3 (C) | 295 | 23.19 | |
| 3. Sludge removal | 1296 | ||
| 1 (Yes) | 893 | 68.90 | |
| 0 (No) | 403 | 31.10 | |
| 4. Source of shrimp seed | 1296 | ||
| 1 (Hatchery) | 1188 | 91.67 | |
| 0 (on-farm nursery) | 108 | 8.33 | |
| 5. Polyculture | 1296 | ||
| 1 (Yes) | 263 | 20.29 | |
| 0 (No) | 1033 | 79.71 | |
| 6. Feed usage | 1296 | ||
| 1 (Yes) | 303 | 23.38 | |
| 0 (No) | 993 | 76.62 | |
| 7. Lime usage | 1296 | ||
| 1 (Yes) | 617 | 47.61 | |
| 0 (No) | 679 | 52.39 | |
| 8. Fertilizer application | 1296 | ||
| 1 (Yes) | 1270 | 97.99 | |
| 0 (No) | 26 | 2.01 | |
| 9. Saponin usage | 1296 | ||
| 1 (Yes) | 1251 | 96.53 | |
| 0 (No) | 45 | 3.47 | |
| 10. Use of water pump | 1296 | ||
| 1 (Yes) | 763 | 58.87 | |
| 0 (No) | 533 | 41.13 |
Crop results
Crop results in participating and non-participating ponds are given in Table 5. Crop results among the pond condition evaluation grades in participating ponds are given in Table 6.
| Outcome | Participating ponds | Non-participating ponds | P-value | ||||||
|---|---|---|---|---|---|---|---|---|---|
| n | Mean or % | SD | Min–Max | n | Mean or% | SD | Min–Max | ||
| 1. Ponds with disease outbreak | 291 | 22.45% | 545 | 62.64% | <0.001 | ||||
| 2. Ponds without disease outbreak but without any shrimp survival | 93 | 7.17% | 0 | 0% | – | ||||
| 3. Ponds without any disease outbreak (normal) | 912 | 70.37% | 325 | 37.36% | <0.001 | ||||
| 4. Shrimp yield (kg ha−1) | 1296 | 112.52 | 126.88 | 0–1640 | 870 | 71.48 | 173.65 | 0–3000 | <0.001 |
| 5. Fish yield in polyculture ponds(kg ha−1) | 263 | 224.2 | 321.38 | 6–4000 | 430 | 208.76 | 281.99 | 10–4500 | 0.508 |
| 6. Mean shrimp size (g) | 1081 | 21.25 | 6.57 | 5–50 | 597 | 23.84 | 6.07 | 6–50 | <0.001 |
| 7. Shrimp survival (%) | 1296 | 25.88 | 22.14 | 0–116 | 870 | 14.83 | 15.96 | 0–144 | <0.001 |
| 8. Crop age at harvest (days) | 1296 | 97.17 | 34.93 | 1–203 | 870 | 80.24 | 38.16 | 1–169 | <0.001 |
| 9. Crop age at disease outbreak | 291 | 51.9 | 26.67 | 1–140 | 545 | 42.36 | 24.02 | 1–140 | <0.001 |
| 10. Profitable ponds | 945 | 73% | 553 | 63.6% | <0.001 | ||||
| 11. Shrimp stocking density (shrimp ha−1) | 1296 | 23 290 | 18 464 | 2000–300 000 | 870 | 20 686 | 18 176 | 1200–250 000 | 0.001 |
| Outcome | Total observations | Pond condition evaluation grades | P-value | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| A | B | C | ||||||||||||
| n | Mean or% | SD | Min–Max | n | Mean or% | SD | Min–Max | n | Mean or% | SD | Min–Max | |||
| Total ponds | 1272 | 245 | 732 | 295 | ||||||||||
| 1. Ponds with disease outbreak | 289 | 59 | 24% | 169 | 23% | 61 | 20.7% | |||||||
| 2. Ponds without disease outbreak but without any shrimp survival | 12 | 5% | 55 | 7.5% | 22 | 7.4% | – | |||||||
| 3. Ponds without any disease outbreak (normal) | 894 | 174 | 71% | 508 | 69.5% | 212 | 71.9% | |||||||
| 4. Shrimp yield (kg ha−1) | 1272 | 245 | 159.99 | 179.10 | 0–1640 | 732 | 101.39 | 111.52 | 0–1000 | 295 | 96.57 | 95.62 | 0–526.67 | <0.001 |
| 5. Fish yield in polyculture ponds(kg ha−1) | 260 | 81 | 224.86 | 248.95 | 6–1176 | 125 | 229.39 | 397.13 | 12.5–4000 | 54 | 183.89 | 135.39 | 30–600 | 0.659 |
| 6. Mean shrimp size (g) | 1063 | 211 | 20.65 | 6.80 | 5–50 | 604 | 21.35 | 6.63 | 5–50 | 248 | 21.38 | 6.36 | 6.67–40 | 0.376 |
| 7. Shrimp survival (%) | 1272 | 245 | 24.29 | 17.10 | 0–87.2 | 732 | 23.72 | 21.40 | 0–116.67 | 295 | 31.11 | 24.76 | 0–114.75 | <0.001 |
| 8. Crop age at harvest (days) | 1272 | 245 | 100.29 | 34.04 | 14–179 | 732 | 96.50 | 33.88 | 1–174 | 295 | 96.32 | 37.82 | 1–203 | 0.301 |
| 9. Crop age at disease outbreak | 289 | 59 | 52.42 | 27.05 | 1–125 | 169 | 53.82 | 28.24 | 1–140 | 61 | 46.41 | 21.25 | 17–106 | 0.177 |
| 10. Profitable ponds | 928 | 196 | 80% | 496 | 67.7% | 236 | 80% | <0.001 | ||||||
| 11.Shrimp stocking density (shrimp ha−1) | 1272 | 245 | 33 430 | 27 567 | 7100–300 000 | 732 | 22 706 | 15 850 | 2200–200 000 | 295 | 16 645 | 9901 | 2000–80 000 | <0.001 |
Prevalence of shrimp disease outbreak was significantly lower (P < 0.001) in participating ponds 22.45% and than in non-participating ponds 62.64%. In participating ponds, sick shrimp from 181 of the 291 disease outbreak ponds were tested by Shrimple® dotblot test kit to confirm the presence of WSSV and samples from 174 ponds confirmed positive in this test. This test could not be performed in all the disease outbreak ponds due to delays in communication between the farmers and field facilitators. There was no significant difference in prevalence of disease outbreak among the participating ponds with different pond condition evaluation grades (refer to the Table 2 for definition of grades). In 93 (7.17%) participating ponds there was no disease outbreak and no shrimp yield (0% shrimp survival). This situation did not occur in non-participating ponds.
The averages of shrimp yield, shrimp survival, crop age at harvest and crop age at disease outbreak were significantly higher (P < 0.001) in participating ponds when compared with those in non-participating ponds. Correspondingly, they were 112.52 kg ha−1, 25.88%, 97.17 days and 51.9 days in participating ponds, and 71.48 kg ha−1, 14.83%, 80.24 days and 42.36 days in non-participating ponds. But the mean shrimp size was significantly smaller (P < 0.001) in participating ponds (21.25 g) when compared with that in non-participating ponds (23.84 g). Average shrimp stocking density was significantly higher by 2604 shrimp ha−1 (P = 0.001) in participating ponds (23 290 shrimp ha−1) than that in non-participating ponds (20 686 shrimp ha−1). In polyculture ponds, average fish yield in participating ponds (224.2 kg ha−1) and non-participating ponds (208.76 kg ha−1) was not significantly different (P = 0.508).
Not surprisingly, shrimp yield in disease outbreak ponds was significantly lower (P < 0.001) than the normal harvest ponds in both participating (48.98 and 144.26 kg ha−1 respectively) and non-participating ponds (23.5 and 151.95 kg ha−1 respectively) (Table 7). Between the participating and the non-participating ponds, though the shrimp yield was not significantly different (P = 0.498) in normal harvest ponds (144.26 and 151.95 kg ha−1 respectively), it was significantly higher (P < 0.001) in participating disease outbreak ponds compared with non-participating disease outbreak ponds (48.98 and 23.5 kg ha−1 respectively). In the disease outbreak ponds, shrimp stocking density was significantly higher (P < 0.001) in participating ponds (24 435 shrimp ha−1) than that in non-participating ponds (18 999 shrimp ha−1). However, in the normal harvest ponds in participating and non-participating ponds there was no difference (P = 0.736) in shrimp stocking density.
| Participating ponds | Non-participating ponds | P-value (participating vs. non-participating) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| N | Mean | SD | Min–Max | n | Mean | SD | Min–Max | |||
| Shrimp yield (kg ha−1) | Normal harvest | 912 | 144.26 | 132.82 | 1.88–1640 | 325 | 151.95 | 261.04 | 10–3000 | 0.498 |
| Disease outbreak | 291 | 48.98 | 72.04 | 0–500 | 545 | 23.5 | 37.52 | 0–320 | <0.001 | |
| P-value (disease outbreak vs. normal) | 0 | 0 | ||||||||
| Stocking density (shrimp ha−1) | Normal harvest | 912 | 23 059 | 19 650 | 2000–300 000 | 325 | 23 516 | 24 269 | 1667–250 000 | 0.736 |
| Disease outbreak | 291 | 24 435 | 15 500 | 5000–120 000 | 545 | 18 999 | 13 011 | 1200–100 000 | <0.001 | |
| P-value (disease outbreak vs. normal) | 0.276 | 0 | ||||||||
These results indicated that the BMP programme (the study) was effective in significantly minimizing the risk of disease outbreaks in participating ponds when compared with non-participating ponds. However, BMP did not significantly improve the yield at normal harvests in participating ponds, compared with normal harvests in the surrounding non-participating ponds.
Among the participating ponds with different pond condition evaluation grades, shrimp yield was significantly higher (P < 0.001) in A grade ponds (159.99 kg ha−1) compared with B and C grade ponds (101.39 and 96.62 kg ha−1 respectively). However, the shrimp yield was not significantly different (P = 0.515) between B and C grade ponds. Although the shrimp survival was not significantly different between A and B grade ponds (24.29% and 23.72%), it was significantly higher (P < 0.001) in C grade ponds (31.11%). There was no significant difference (P = 0.376; 0.301; 0.177) in mean shrimp size, crop age at disease outbreak and crop age at harvest among A, B and C grade ponds. In polyculture ponds, there was no significant difference (P = 0.659) in fish yield among A, B and C grade ponds (224.86, 229.39 and 183.89 kg ha−1 respectively).
Crop economics
Table 8 gives the results of crop economic analysis. Proportion of profitable ponds was significantly higher in participating ponds (73%) than in non-participating ponds (63.6%). Although the average cost of production was not significantly different between participating (US$296.5 ha−1) and non-participating ponds (US$315.22 ha−1), the profit margin and benefit cost ratio (BCR) was significantly higher (P < 0.001) in participating ponds. Among the profitable ponds, the cost of production was lower by 12.1% and the BCR was higher by 0.523 points in participating ponds than those in non-participating ponds. However, among unprofitable ponds, the quantum of loss was significantly higher (P < 0.001) in participating ponds (US$−142.61 ha−1) than that in non-participating ponds (US$−95.91 ha−1).
| Participating ponds | Non-participating ponds | P-value | |||||||
|---|---|---|---|---|---|---|---|---|---|
| N | Mean | SD | Min–Max | n | Mean | SD | Min–Max | ||
| Cost of production (US$ ha−1) | |||||||||
| All ponds | 1296 | 296.50 | 304.23 | 18.81–4321.09 | 870 | 315.22 | 532.80 | 61–10 242 | 0.299 |
| Ponds with profit | 945 | 324.21 | 332.87 | 27.47–4321.09 | 553 | 368.89 | 642.31 | 103–10 242 | 0.077 |
| Ponds with loss | 351 | 221.90 | 189.63 | 18.81–1499.16 | 317 | 221.58 | 214.78 | 61–2793 | 0.984 |
| Profit margin (US$ ha−1) | |||||||||
| All ponds | 1296 | 271.08 | 481.61 | −1499.16–4097.26 | 870 | 158.23 | 508.32 | −1372.37–6773.03 | <0.001 |
| Ponds with profit | 945 | 424.74 | 472.19 | 0.21–4097.26 | 553 | 303.91 | 584.17 | 0.70–6773.03 | <0.001 |
| Ponds with loss | 351 | −142.61 | 146.47 | −1499.16–1.89 | 317 | −95.91 | 111.84 | −1372.37–0.00 | <0.001 |
| Benefit cost ratio | |||||||||
| All ponds | 1296 | 1.884 | 1.602 | 0–14.44 | 870 | 1.361 | 1.164 | 0.00–21.39 | <0.001 |
| Ponds with profit | 945 | 2.480 | 1.469 | 1–14.44 | 553 | 1.863 | 1.165 | 1.0–21.39 | <0.001 |
| Ponds with loss | 351 | 0.279 | 0.365 | 0–0.99 | 317 | 0.485 | 0.378 | 0.00–1.00 | <0.001 |
- Currency conversion rate: US$1 = 9500 IDR.
In both the polyculture (249/263) and shrimp monoculture (696/1033) farming systems, participating ponds had a significantly higher proportion of profitable ponds when compared with non-participating ponds (polyculture 300/419 and shrimp monoculture 253/441 respectively) (Table 9).
| Farming system | Participating ponds | Non-participating ponds | OR (95% CI) | P-value | ||
|---|---|---|---|---|---|---|
| Profit | Loss | Profit | Loss | |||
| Polyculture | 249 | 14 | 300 | 129 | 7.65 (4.18–14.24) | <0.001 |
| Monoculture | 696 | 337 | 253 | 188 | 1.53 (1.21–1.94) | <0.001 |
| OR (95% CI) | 8.61 (4.83–15.63) | 1.73 (1.29–2.31) | ||||
| P-value | <0.001 | <0.001 | ||||
Multivariate logistic regression analysis demonstrated that (Table 11) shrimp disease, poor shrimp yield and high cost of production all contributed to ponds being unprofitable. Pond location (district) was significantly associated with the unprofitable ponds. Polyculture, use of lime and fertilizer, higher soil pH (alkaline) and lower water pH reduced the risk of economic loss.
Association of factors with crop outcomes
Results of univariable and multivariable logistic regression analyses are given in Table 10 and Table 11. In total 27 independent variables were considered. In the univariable logistic regression analysis, disease outbreak was significantly associated (P ≤ 0.20) with 16 variables, and poor yield was associated with 21 variables. There was no multicollinearity among the continuous variables. These significantly associated variables were included in the multivariable backward step-wise conditional logistic regression analysis. Multivariable analysis showed that the disease outbreak was strongly associated with location of ponds (cluster and villages). Risk of disease outbreak was significantly higher (P ≤ 0.10) in clusters with higher prevalence of disease outbreaks in non-participating ponds, larger ponds, rental ponds, stocking direct hatchery sourced shrimp seed (over on-farm nurseries), shallower water depth, not using feed, lime and water pump, lower (acidic) soil pH and higher average salinity. Poor yield was associated with pond location (district). Risk of poor yield was significantly higher in ponds with disease outbreaks, clusters with higher prevalence of disease outbreaks in non-participating ponds, larger ponds, poorer grades of pond condition evaluation, non-removal of sludge, lower shrimp stocking density, not using feed, lime and water pump, lower (acidic) soil pH and lower average salinity. Both the final models were significant and valid with −2 log likelihood test results of P ≤ 0.05 and the Hosmer-Lemeshow goodness-of-fit test result of P ≥ 0.05 respectively.
| Variables considered under analyses | Shrimp disease | Poor shrimp yield (<100 kg ha−1) | Economic loss | ||||
|---|---|---|---|---|---|---|---|
| Crude odds ratio (95% CI) | P-value | Crude odds ratio (95% CI) | P-value | Crude odds ratio (95% CI) | P-value | ||
| 1 | Shrimp disease outbreak | – | 6.105 (4.392–8.486) | <0.001 | 12.226 (8.913–16.771) | <0.001 | |
| 2 | Poor shrimp yield | – | – | 24.015 (14.860–38.812) | <0.001 | ||
| 3 | Cost of production (US$ ha−1) | – | – | 0.998 (0.997–0.998) | <0.001 | ||
| Pond location | |||||||
| 4 | Districts | 1.545 (1.170–2.040) | 0.002 | 2.972 (2.211–3.955) | <0.001 | 2.421 (1.866–3.140) | <0.001 |
| 5 | Clusters | 1.021 (1.005–1.038) | 0.010 | 1.021 (1.007–1.034) | 0.002 | ||
| 6 | Villages | 1.006 (1.000–1.012) | 0.037 | 1.006 (1.002–1.011) | 0.009 | ||
| Participation | |||||||
| 7 | Cluster-level | 0.971 (0.965–0.978) | <0.001 | 0.979 (0.973–0.986) | <0.001 | ||
| 8 | Village-level | 1.008 (1.001–1.014) | 0.016 | 0.986 (0.981–0.991) | <0.001 | 0.994 (0.989–1.000) | 0.037 |
| Prevalence of disease outbreak in non-participating ponds | |||||||
| 9 | Cluster level | 1.009 (1.004–1.014) | <0.001 | 1.011 (1.007–1.015) | <0.001 | 1.005 (1.001–1.010) | 0.017 |
| 10 | Village-level | 1.010 (1.007–1.013) | <0.001 | ||||
| Pond details | |||||||
| 11 | Pond size | 1.533 (1.183–1.986) | 0.001 | 4.025 (2.957–5.478) | <0.001 | 1.435 (1.127–1.827) | 0.003 |
| 12 | Owned and operated ponds | 1.245 (0.912–1.700) | 0.167 | 1.759 (1.368–2.262) | <0.001 | ||
| 13 | Pond evaluation grade | 1.595 (1.339–1.900) | <0.001 | ||||
| Farming practices | |||||||
| 14 | Sludge removal | 0.507 (0.396–0.649) | <0.001 | 0.715 (0.552–0.926) | 0.011 | ||
| 15 | Average water depth | 0.968 (0.959–0.978) | <0.001 | 0.983 (0.976–0.990) | <0.001 | 0.979 (0.970–0.987) | <0.001 |
| 16 | Shrimp stocking month | 0.875 (0.721–1.062) | 0.178 | 0.875 (0.730–1.049) | 0.148 | ||
| 17 | Shrimp stocking density | 0.598 (0.538–0.665) | <0.001 | ||||
| 18 | Hatchery shrimp seed | 2.890 (1.525–5.479) | 0.001 | 1.432 (0.965–2.124) | 0.075 | 1.410 (0.874–2.273) | 0.159 |
| 19 | Polyculture | 0.116 (0.067–0.202) | <0.001 | ||||
| 20 | Feed usage | 0.432 (0.302–0.618) | <0.001 | 0.435 (0.334–0.565) | <0.001 | 0.726 (0.536–0.983) | 0.038 |
| 21 | Lime usage | 0.574 (0.438–0.751) | <0.001 | 0.716 (0.574–0.892) | 0.003 | 0.662 (0.516–0.848) | 0.001 |
| 22 | Fertilizer application | 0.363 (0.166–0.790) | 0.011 | ||||
| 23 | Saponin usage | ||||||
| 24 | Use of water pump | 0.432 (0.331–0.565) | <0.001 | 0.308 (0.243–0.391) | <0.001 | 0.431 (0.336–0.553) | <0.001 |
| Pond soil and water quality parameters | |||||||
| 25 | Soil pH | 0.724 (0.601–0.872) | 0.001 | 0.610 (0.516–0.721) | <0.001 | 0.552 (0.463–0.657) | <0.001 |
| 26 | Average salinity | 1.034 (1.014–1.054) | 0.001 | 0.968 (0.953–0.983) | <0.001 | 0.985 (0.967–1.002) | 0.089 |
| 27 | Average water pH | 0.533 (0.421–0.676) | <0.001 | 0.579 (0.463–0.724) | <0.001 | 0.681 (0.544–0.854) | 0.001 |
| Variables considered under analyses | Shrimp disease (n = 347 missing) | Poor shrimp yield (<100 kg ha−1) (n = missing 359) | Economic loss (n = 347 missing) | ||||
|---|---|---|---|---|---|---|---|
| Adjusted odds ratio (95% CI) | P-value | Adjusted odds ratio (95% CI) | P-value | Adjusted odds ratio (95% CI) | P-value | ||
| 1 | Shrimp disease outbreak | 8.098 (4.599–14.260) | <0.001 | 13.420 (8.124–22.166) | <0.001 | ||
| 2 | Poor shrimp yield | 32.426 (11.675–90.059) | <0.001 | ||||
| 3 | Cost of production (US$ ha−1) | 1.003 (1.002–1.005) | <0.001 | ||||
| Pond location | |||||||
| 4 | Districts | 5.081 (2.768–9.326) | <0.001 | 5.185 (2.726–9.861) | <0.001 | ||
| 5 | Clusters | 1.392 (1.141–1.697) | 0.001 | ||||
| 6 | Villages | 0.887 (0.828–0.951) | 0.001 | ||||
| Participation | |||||||
| 7 | Cluster-level | ||||||
| 8 | Village-level | ||||||
| Prevalence of disease outbreak in non-participating ponds | |||||||
| 9 | Cluster level | 1.011 (1.004–1.018) | 0.001 | 1.010 (1.004–1.016) | 0.002 | ||
| 10 | Village-level | ||||||
| Pond details | |||||||
| 11 | Pond size | 2.728 (1.863–3.996) | <0.001 | 6.303 (3.481–11.410) | <0.001 | ||
| 12 | Owned and operated ponds | 0.554 (0.361–0.850) | 0.007 | ||||
| 13 | Pond evaluation grade | 1.726 (1.234–2.413) | 0.001 | ||||
| Farming practices | |||||||
| 14 | Sludge removal | 0.401 (0.257–0.625) | <0.001 | ||||
| 15 | Average water depth | 0.964 (0.951–0.977) | <0.001 | ||||
| 16 | Shrimp stocking month | ||||||
| 17 | Shrimp stocking density | 0.480 (0.385–0.598) | <0.001 | ||||
| 18 | Hatchery shrimp seed | 4.192 (1.780–9.875) | 0.001 | ||||
| 19 | Polyculture | 0.083 (0.040–0.174) | <0.001 | ||||
| 20 | Feed usage | 0.435 (0.254–0.747) | 0.003 | 0.306 (0.193–0.485) | <0.001 | ||
| 21 | Lime usage | 0.609 (0.410–0.905) | 0.014 | 0.446 (0.301–0.662) | <0.001 | 0.498 (0.292–0.852) | 0.011 |
| 22 | Fertilizer application | 0.207 (0.058–0.738) | 0.015 | ||||
| 23 | Saponin usage | ||||||
| 24 | Use of water pump | 0.334 (0.226–0.493) | <0.001 | 0.309 (0.207–0.459) | <0.001 | ||
| Pond soil and water quality parameters | |||||||
| 25 | Soil pH | 0.801 (0.628–1.020) | 0.072 | 0.516 (0.384–0.695) | <0.001 | 0.724 (0.535–0.979) | 0.036 |
| 26 | Average salinity | 1.032 (1.004–1.060) | 0.023 | 0.963 (0.936–0.990) | 0.009 | ||
| 27 | Average water pH | 1.565 (0.985–2.484) | 0.058 | ||||
| P-value of model statistics | |||||||
| −2 Log likelihood test | <0.001 | <0.001 | <0.001 | ||||
| Hosmer and Lemeshow Goodness-of-fit test | 0.251 | 0.677 | 0.269 | ||||
| A. Pond preparation practices |
| 1. Completely drain out the water from the tambak. |
| 2. Remove the organic waste (sludge) from the pond bottom and dispose it away from the pond site. |
| 3. Dry the pond bottom. |
| 4. Check the soil pH. If acidic soil (pH<7), use lime. |
| 5. Fill the pond with water using double layered fine mesh (300 micron) water filter nets. |
| 6. Maintain a water depth of >30cm and ideally >80cm in the shallowest part of the pond.a |
| 7. Do not use pesticides and other harmful chemicals.a |
| 8. Stabilize the plankton bloom (green/brown water colour) before stocking. |
| B. Shrimp seed selection and stocking practices |
| 9. All the ponds in the village and farming cluster should stock the shrimp seed at the same time by following crop calendar system. Avoid different batches of seed from different hatcheries. |
| 10. Uniform sized and coloured PLs, actively swimming against the water current. |
| 11. Two-step PCR negative PLs for WSSV.a |
| 12. Pack the seed in good condition with enough oxygen and transport quickly (within 6h from hatchery to farm). |
| 13. Acclimatise the seed before releasing in to ponds. |
| 14. Avoid seed from poorly managed commercial nurseries. Stock on-farm nursery reared juveniles or PL-20 from hatchery.a |
| C. Feed management practices |
| 15. No use of live and fresh feed (crabs, shrimp, molluscs etc.)a |
| 16. Properly and judiciously use high quality commercial pellet feeds by following feed ration table and check tray method. |
| D. Water management practices |
| 17. Exchange water carefully (only when it is essential). |
| 18. Regularly (on weekly basis) check the water quality parameters. |
| 19. Remove benthic algae, filamentous algae and hydrilla. |
| 20. Maintain adequate oxygen levels (4–6ppm) in the water. |
| E. Health management practices |
| 21. Regularly (once a week) monitor the shrimp health and growth by cast-net sampling. |
| 22. Remove and safely dispose (bury) the sick and dead shrimp. |
| 23. If there is increasing mortality trend within 1–2days after first observation of moribund shrimps, decide emergency harvesting of the stock |
| 24. Do not drain or abandon the disease affected ponds, but harvest all the shrimps. |
- a Mandatory better management practices.
Discussion
The study clearly showed that farmers participating in the BMP programme had significantly reduced risk of disease occurrence and increased the proportion of profitable ponds, but the programme was not effective in improving the shrimp yield at normal harvests in participating ponds when compared with normal harvests in surrounding non-participating ponds. Similar results were reported in a MPEDA/NACA BMP programme among small-scale modified extensive farming systems in Andhra Pradesh, India (Padiyar 2008). Increasing yield will likely require further improvements in pond management, infrastructure (e.g. increasing the area of ponds with depth >0.3 m) and perhaps inputs. Further analysis of the top performing farmers is needed to understand the factors leading to higher yields in these traditional farming systems.Adoption of BMP reduced cost of production and increased BCR in profitable ponds, which again is similar to the outcomes of the Indian programme (Umesh et al. 2008; Umesh et al. 2010). Some, if not allof the recommended BMP were significantly associated with the better crop outcomes (no disease outbreaks and higher yield) and profitability. However, it is important here to note that there was no significant adverse effect (harm) on the crop by using the BMP. Since the crop outcomes are also associated with pond location (villages/clusters/districts), perhaps a detailed spatial study using GIS would help in determining the factors, including geological and social, within each location that influence the crop outcomes.
The study shows that simple management improvements can reduce risks in brackishwater ponds of Aceh. However, investments in increasing yield are required. In a post-disaster situation, it also suggests that broader review of the risks of different aquaculture practices should be conducted. Organizational systems such as farmer groups, and supporting services are also necessary for successful crops from higher risk aquaculture systems such as shrimp, which may not always be present in rural farming areas.
Conclusion
This study has investigated a farmer group approach and implementation of BMP for small-scale aquaculture farming Aceh. The concept has been well accepted by local farmers during and after the study period due to the immediate farm level benefits it offered to farmers (Box 2). However, further improvements that generate income for farmers will need additional investments in yield improvements and better connections between farmer groups and the backward and forward linkages in the value chain. Modern market trends for certification, traceability and branding of farm produce for small-scale farmers will require even further investment in services. Scaling up of the activities by extending the programme to other farming areas of Aceh is most important for inclusion of larger numbers of farmers and to bring the economies of scale in technical assistance and extension efforts required, such as from indigenous institutions like BBAP Ujung Batee and DKP. The community managed ALSCs and improved organizational system involving farmer groups and clusters provide a basis for sustaining programme investments and sustaining services needed by small-scale aquaculture farmers.
| 1. The programme gave information and improved the skills of farmers |
| 2. Successful shrimp crop and increase in income from the pond |
| 3. Reduced cost of production due to better management practices implementation and direct procurement of farm inputs at wholesale prices from the input companies through ALSC. |
| 4. Better quality of farm inputs, especially shrimp seed, due to contract hatchery system followed by the ALSC. |
| 5. Unity among farmers to solve their problems through local kelompok and ALSC. Regular meetings of kelompok and ALSC members helped to discuss the problems and find the local solutions. |
| 6. There is need to expand the programme to other neighbouring subdistricts and districts to benefit more farmers. |
Acknowledgments
The authors are grateful to the American Red Cross, which funded this activity under a broader FAO implemented project OSRO/INS/601/ARC ‘Rehabilitation and sustainable development of fisheries and aquaculture affected by tsunami in Aceh, Indonesia’. The authors wish to thank all the 21 Field Facilitators, tambak farmers, hatchery operators, farm input suppliers, staff of BBAP Ujung Batee, DKP Aceh Province, DKP Aceh Utara, DKP Bireuen, DKP Kota Lhokseumawe districts, ACIAR, AusAid, ADB ETESP, NACA, WWF and The Packard Foundation for the support and cooperation extended during the study.
