boblme-2012-ecology-10 · report of the hilsa fisheries assessment working group ii . 1 . 1....

BOBLME-2012-Ecology-10

The designations employed and the presentation of material in this publication do not imply the expression of any opinion whatsoever on the part of Food and Agriculture Organization of the United Nations concerning the legal and development status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. The BOBLME Project encourages the use of this report for study, research, news reporting, criticism or review. Selected passages, tables or diagrams may be reproduced for such purposes provided acknowledgment of the source is included. Major extracts or the entire document may not be reproduced by any process without the written permission of the BOBLME Project Regional Coordinator. BOBLME contract: PSA-GCP 32/2/2012 For bibliographic purposes, please reference this publication as: BOBLME (2012) Report of the hilsa fisheries assessment working group II, 24-25 April 2012, Mumbai, India BOBLME-2012-Ecology-10

ii

TABLE OF CONTENTS

1. Opening of the meeting and adoption of the agenda .................................................................... 1

2. Introduction to selected stock assessment topics of hilsa assessment in BOBLME Region ........... 1

2.1. Information needs for a defensible stock assessment – Dr Rishi Sharma .............................. 1

2.2. Hilsa fisheries assessment in Bay of Bengal (Bangladesh)– Dr M. Anisur Rahman ................ 1

2.3. Hilsa fisheries assessment in Bay of Bengal (India) – Dr Naskar ............................................. 4

2.4. Hilsa fisheries assessment in bay of bengal (Myanmar) – Mr Soe.......................................... 4

2.5. Effort standardization and survey for hilsa – Mr Gulati .......................................................... 5

3. Hilsa (Tenualosa ilisha) stock assessment workgroup recommendations and work plan ............. 5

3.1.1. Capacity building in newer stock assessment techniques .............................................. 5

3.1.2. Develop a fishery management plan for the region in hilsa ........................................... 6

3.1.3. Develop a standardized model framework for stock assessment .................................. 6

3.1.4. Develop a standardized short and long term strategic action plan (SAP) for ecosystem health and resource evaluation for the river systems that hilsa occupy in BOBLME region ............................................................................................................ 7

3.1.5. Other priority items for consideration ............................................................................ 7

3.2. Proposals for research and M&E work in 2011-2013 ............................................................. 8

3.3. New framework with technical management reference point to provide fishing advice ...... 8

3.4. Stock status advice for hilsa in the BOBLME region ............................................................. 10

4. Adoption of the report .................................................................................................................. 10

Appendix I List of participants ......................................................................................................... 11

Appendix II Agenda .......................................................................................................................... 12

Appendix III List of documents presented to the meeting ............................................................... 14

Appendix IV Draft document outlining the stock assessment and management framework .......... 16

LIST OF TABLES

Table 1 Catch by marine and inland sectors since 1984 (source M. A. Rahman, BFRI) .......................... 2 Table 2 Results from FISAT of multiple years of length based data assessments .................................. 3 Table 3 Stock Status and Management Advice for the region................................................................ 8

LIST OF FIGURES

Figure 1 Target reference points for Hilsa Management based on recommendations from Hilsa FAWG (Mumbai, April 25th, 2012) .......................................................................................... 9

Report of the hilsa fisheries assessment working group II

1

1. Opening of the meeting and adoption of the agenda

1. A Workshop developing the framework for stock assessment on Hilsa for the Bay of Bengal region was held in Mumbai, India on April 24th and 25th, 2012. The BOBLME Stock Assessment Coordinator, Dr Rishi Sharma welcomed the participants and wished them well in their work

2. Dr Sharma reminded the meeting that BOBLME Project is mandated to develop regional fishery assessments for Hilsa, and this meeting was the fourth one to that end. The focus of the meeting was to develop a common regional management plan with some key reference points that would be critical in determining common management measures for the region. Follow up on some activity from the previous meeting in Dhaka would be addressed, and an update on the research and M&E initiatives proposed by the countries that will further the understanding of Hilsa in the BOBLME Region would also be discussed.

3. The meeting was opened by Dr Vijaykumaran (NC, India). A meeting address was made by Dr Vijaykumaran, followed by Dr Sharma. The rest of the meeting was chaired by Dr Rishi Sharma.

4. The participants of the meeting are listed in Appendix I and the agenda for the Meeting was adopted as presented in Appendix II.

5. Dr Sharma informed the meeting about the scope of the project, and how far the work has proceeded. The agenda was adopted (Appendix II); and the participants were introduced.

6. The list of documents presented to the meeting is given in Appendix III.

2. Introduction to selected stock assessment topics of hilsa assessment in BOBLME Region

2.1. Information needs for a defensible stock assessment – Dr Rishi Sharma 7. Dr Rishi Sharma gave an overview on the BOBLME Project and mandate as it related to Hilsa.

Crucial to this are the stock assessment components which are essential to the entire project success. This is why the meeting and Hilsa assessment is extremely important in the region. Essential elements of a stock assessment are CPUE and effort at the resolution that mimics the stock and the life-history of the species. A basic stock assessment was presented using a Surplus Production (SP) Model and how BMSY would be estimated using Bangladesh catch and effort data (elements of this approach are available in Appendix IV as the approach for management, and will be finalized and submitted to a peer reviewed journal soon). Methods that use different sources of data from different countries and gear-types could be integrated into the overall fitting procedure using Maximum Likelihood Estimation (MLE) techniques. Essential in this would be to stratify catch and effort by gear, sector and country. In this manner fleet catchability could be assessed as well so we could compare effort controls, and a desired outcome in fishery yield. This approach was elected as the common framework for stock assessment of hilsa in the region.

8. The working group discussed the adequacy (quality) of the data in the various countries. Alternative assessment options should be pursued, rather than a single approach taking into account data from uncertain sources. Life history based modelling approaches like Leslie Matrix models, age-structured integrated assessments, FISAT based assessments, Surplus Production based assessments and PSA approaches should be pursued simultaneously.

2.2. Hilsa fisheries assessment in Bay of Bengal (Bangladesh)– Dr M. Anisur Rahman 9. Hilsa is the national fish of Bangladesh and numerous people are involved in the fishery.


2

10. Habitat destruction and overfishing are major factors affecting the abundance of Hilsa. A recent survey indicated that Hilsa has disappeared from 35 out of 235 rivers. Other anthropogenic effects such as pollution and poor floodplain management are affecting the Hilsa population in a negative manner.

11. Freshwater catches have remained fairly stable, but in recent years catches from the marine sector have increased markedly. In 2010, the marine sector took 209 450 t, while the freshwater sector took 103 162 t. The catches of the marine and freshwater sectors are given in the table below.

Table 1 Catch by marine and inland sectors since 1984 (source M. A. Rahman, BFRI)

Year Freshwater Marine Total

1984 90,082 56,000 146,082

1985 73,388 71,050 144,438

1986 94,797 96,294 191,091

1987 91,167 103,814 194,981

1988 78,551 104,950 183,501

1989 81,641 110,311 191,952

1990 112,408 113,943 226,351

1991 66,809 115,358 182,167

1992 68,356 120,106 188,462

1993 74,715 123,115 197,830

1994 71,370 121,161 191,531

1995 84,420 129,115 213,535

1996 80,625 126,660 207,285

1997 83,230 131,204 214,434

1998 81,634 124,105 205,739

1999 73,809 140,710 214,519

2000 79,165 140,396 219,561

2001 75,060 154,654 229,714

2002 68,250 152,343 220,593

2003 62,944 136,088 199,032

2004 71,000 184,438 255,839

2005 77,500 198,363 275,860

2006 78,273 198,850 277,123

2007 80,453 199,875 280,328

2008 89,900 200,100 290,000

2009 95,507 202,951 298,458

2010 103,162 209,450 312,612

2011 340,000


3

12. Starch Gell Allozyme analysis performed by CSIRO indicated that Hilsa is a single stock in the region shared by Bangladesh, India and Myanmar. However, some Bangladesh scientists are of the opinion that a sub-stock structure may exist.

13. Gonadal Somatic Index (GSI) indicates that peak spawning occurs in September to October in 4 distinct areas in Bangladesh, namely Kalirchar (down of Sandwip), (2) Moulavirchar (south of Hatia), (3) surrounding of Monpura (east of Bhola) and (4) Dhalchar Island (Charfashion, Bhola). These areas were found as the most significant areas of Hilsa spawning. Nursery areas are both freshwater and estuarine and nursery areas have been established in major parts of the Meghna River (Shatnol, Chandpur to Char Alexander).

14. Results using the FISAT program indicate a large variation in exploitation on the stock between 1992-2003 (Table 2). However this analysis did not estimate what the available spawning biomass or targets should be and this could be one area of possible improvement in an integrated assessment. The results from 2009 were presented in this meeting, and indications are that the Exploitation levels have not dropped over the period, as initially thought.

Table 2 Results from FISAT of multiple years of length based data assessments

Parameters 1992 1995 1996 1997 1998 1999 2000 2002 2003 2009

Asymptotic length (L¥) cm

61.1 58.3 59.97

61.50

66.00

60.00

62.50

53.70

53.60

53.00

Growth constant (K)/yr)

0.74 0.74 0.99 0.83 0.67 0.82 0.72 0.86 0.60 0.83

Total mortality (Z/yr)

2.41 2.61 3.19 3.29 3.43 3.77 2.79 3.51 3.03 3.23

Natural mortality (M/yr)

1.16 1.18 1.41 1.28 1.25 1.28 1.17 1.36 1.09 1.36

Fishing mortality (F/yr)

1.25 1.43 1.78 2.01 2.18 2.49 1.62 2.15 1.94 1.87

Exploitation rate (E)

0.52 0.55 0.56 0.61 0.63 0.66 0.58 0.61 0.64 0.58

Maximum yld/recruit (Emax)

- - 0.71 0.69 0.60 0.59 0.46 0.58 0.63 0.57

Size at first capture (Lc) cm

35.0 30.0 30.34

30.25

27.06

22.80

13.12

19.87

21.21

26.00

Growth performance (Æ)

- 3.40 3.55 3.50 3.46 3.47 3.45 3.51 3.03 3.37

15. Based on the assessment conducted using data up to 2005 the conclusion Bangladesh scientists arrived at were the following:


4

i. Hilsa fishery is suffering from growth over-fishing (indiscriminate catching of jatka, i.e. juvenile Hilsa)

ii. There is recruitment over-fishing (indiscriminate killing of mature female Hilsa) iii. The fishing mortality has increased due to fishing pressure with decrease in size at

first capture . 16. DOF and BFRI have undertaken many research initiatives that are dealing with community

partnership programs such as giving the jatka fishers and alternative source of livelihood, as well as more community outreach programs.

17. However, indications from 2009 still show a slightly higher than optimal harvest rate that should be around 50%.

18. BFRI pointed to some short-comings in the current survey systems and designs, as being outdated in terms of the sample frame, as well as being short-staffed to do an adequate job.

2.3. Hilsa fisheries assessment in Bay of Bengal (India) – Dr Naskar 19. India evaluated whether the FISAT based approach is appropriate for Hilsa as it is not in

equilibrium conditions, nor is the sampling systematic over the species in time. In addition length based techniques are not as accurate as otolith based measurements and gave biased estimates of the Von-Bertalanfy growth parameters over time. Finally variation in growth as being attributed to change or difference in stocks was a misleading result as confidence intervals overlapped across these variations. India also, explored a similar approach recommended for an integrated assessment using a parameter estimation techniques and a Surplus production model, but added some statistical complexity in the model by using both process and observation error models within a Bayesian context.

20. Preliminary assessments looking at both autoregressive and time series models indicate that the stock is rebuilding over time. However, the quality of the data and approach needs to be assessed.

21. Finally, it appears that the optimal yield is 2880 to 4980 Tons depending on the statistical model used and the optimal Biomass is between 1380 to 1725 Tons. However, the quality of the data and approach needs to be assessed.

2.4. Hilsa fisheries assessment in bay of bengal (Myanmar) – Mr Soe 22. A general overview on Hilsa Fisheries migrations in Myanmar was presented. This was some

work done recently to understand the migratory patterns and spawning behavior of Hilsa in the Irrawaddy region.

23. Hilsa is found in numerous parts of the Rakhine coast on the Bay of Bengal, the Irrawaddy River and the Tanintharyi Region in the Andaman Sea. Fishing occurs year round though September to October and February to April are the peak and moderate fishing periods respectively. Catch distributions vary substantially by marine region with January to June being the primary timing in the Ayeyarwady regions, while June through December appeared higher in the Mon and Tanintharyi Region in southern Myanmar. In the Rakhine region most catch occurs in July through October corresponding to the peak spawning timings in Bangladesh and Indian Rivers. Some general patterns on migration along the coast were presented as hypothesis during this meeting.

24. Myanmar’s focus is on more life-history based studies to identify the migratory route, spawning site and nursery grounds. In addition they would want to build assessment models like Myanmar and India to estimate changes in catch, recruitment capacity, fishing pressures and consequential livelihoods of the resources users. This is not possible without building capacity for scientists to assess resources status, and would require a large effort with BOBLME and other countries help to do this.

25. A National Plan of Action for Hilsa Management was presented by Myanmar which included elements of the fishery management, migration and stock assessments. Overall objectives


5

are to improve the livelihoods of communities locally, and optimize the yield through sustainable management practices. Conservation of the stocks is a key component that could be managed through time-area closures and mesh-size restrictions. A National Task force would also be set up with members from the different departments, namely Department of Fisheries and the Marine Fisheries Federation. This body would manage the fisheries by different jurisdictions, namely the Rakhine, Ayeyerwady and Mon state.

26. Prior to 2005, landings information is poor, and the trend from 2006 to 2010 to current years indicate a declining trend in catches (16920 to 12000 Mt, though 2009 had a peak catch of 19,021 Mt). Total landings in 2011 also showed a similar catch of 19097 Mt.

2.5. Effort standardization and survey for hilsa – Mr Gulati 27. India presented some approaches using standard GLM (Generalized Linear Models) based

techniques to relate CPUE trends over time. The approach takes into account zero count data, and uses the delta distribution to account for this based on earlier work. Log-Normal errors and Poisson models could also be sued with regard to this and were also discussed at this meeting.

28. Swept area approaches to estimate Biomass of Hilsa in the Bay of Bengal using a survey vessel were discussed. A survey vessel could be deployed this year to assess whether this feasible at a 30 M trawl depth. In addition, using fisherman in the estuary and delta to assess the biomass in a random manner using gillnets were also discussed at the meeting.

3. Hilsa (Tenualosa ilisha) stock assessment workgroup recommendations and work plan

3.1.1. Capacity building in newer stock assessment techniques 29. There is a strong need for capacity building to support the implementation of a Fisheries

Management Plan for the region. While the status and current knowledge may be sufficient for India and Bangladesh, Myanmar will need help to develop its plan. The countries were all supportive of an Integrated Fisheries Stock Assessment Tool for the region that is developed to address the life-history and peculiarities of Hilsa in the region. This will work in conjunction with the existing FISAT and ELEFAN tools that the countries are already using in the region. A maximum likelihood based integrated modelling framework was presented and tested on Bangladesh data and data from India. This framework would be the scientific basis of a joint management plan in the region.

Recommend

• A regional body of experts (i.e. The Hilsa Assessment Working Group, (HAWG*)) that will share knowledge across the region.

• Encouraging exchange of regional transnational knowledge and sharing in management planning processes through the HAWG for transboundary fisheries, and/or fisheries with similar characteristics

• Provide workshops and trainings on newer stock assessment approaches for the region from BOBLME.

• Strengthen the knowledge of decision makers • Governments to initiate capacity building, but with responsibilities and capacity

development carried out at a decentralized level, and inclusive of many stakeholder groups

• Develop the framework that uses: a) common model and data, b) standardized output, and c) decisions based on output (BOBLME will develop). What kind of data, frequency and spatial resolution.


6

3.1.2. Develop a fishery management plan for the region in hilsa 30. All countries expressed concerns of unilateral implementation of management

recommendations that were not standardized in other areas. Bangladesh, Myanmar and India recommend using a standardized time area closure mechanism as well as standardized gear regulations on gillnet use. Enforcement must be an integral part of this approach if it is to succeed. In addition standardized regulations as well as standardized data assessment programs need to be developed for this region. Fishery Survey of India presented an approach on how they collect standardized data by gear/ fisheries sector, and by marine and inshore areas. These programs will be standardized across the region, and be the basis of an M&E system in the region.

Recommend

• BOBLME finish the draft document on the technical elements of a FMP (or FIP) for Hilsa Assessment in the BOB Region which consolidates stock status information and fishing regulatory measures for protection and rebuilding.

• FSI finalize the Stratified Random Sampling design to estimate effort and catch by sector and gear in the marine and freshwater areas, and have a report. This will also include ideas on how to standardize effort across the region. By 15th May, 2012, information will be sent by Myanmar (KMSoe), and BGD (Mome and Kamruzzaman, DOF will provide this). By 15th June, we will have a draft recommendation on sampling design for the region (Sajeevan, FSI, Mome, SOD, BGD and KMSoe, Myanmar)

• Trainings held by BOBLME and FSI to train people in the region to collect the right sort of information by sector. Maybe this can be done in Yangon with Myanmar staff in August, 2012.

• Standardized regulation package developed by an advisory body and implemented in the region.

3.1.3. Develop a standardized model framework for stock assessment 31. The integrated type approach presented by BOBLME that uses all sources of data in all

sectors estimating catchability, and vulnerability by sector, gear and area and fitting to indices of CPUE and Survey based indices is the approach that could be pursued in parallel with the exiting FISAT, and ELEFAN based approaches.

32. This modelling approach is now in draft format and will be extended across India and Bangladesh and written in final technical note and a formal journal paper to be submitted to a high impact factor international refereed journal.

33. The assessment approaches developed will need to consider stock structure issues, or at the very least management units that may consider geographical boundaries (India, Myanmar and Bangladesh stocks), and possibly even a finer sub-stock structure (e.g. river of origin).

34. The HAWG noted that there was an immediate need to increase stock assessment capacity in each country through future training on Fishery Statistics and Stock Assessment, including running some simplistic assessments such a surplus production models, length based models, etc. with regional technical teams.

Recommend

• BOBLME in collaboration with BGD, IND and MYA to start developing the framework for the model, and then provide guidelines on what data to collect for the region.

• Train people on the understanding of the model and the dynamics.


7

3.1.4. Develop a standardized short and long term strategic action plan (SAP) for ecosystem health and resource evaluation for the river systems that hilsa occupy in BOBLME region

35. Since the work on short-term hasn’t begun yet, the countries expressed a need to drop the tagging piece from the projects (adjust the budget accordingly), and make it a part of the long-term strategic action plan to assess movement and migrations of the various components of hilsa.

36. A long discussion occurred on the range and existence of hilsa in this region, and how the rivers have changed over the last 100 years. The three systems being studied have a wide range of activities that have impacted them and the group proposed to develop a joint study on the health of these river systems over the duration of the project. The three river systems that will have a cross-section time series analysis are the following:

i. Lower Meghna River (BGD) ii. Lower Ganga, i.e. Hoohly-Matla River (IND)

iii. Lower Irrawaddy River (MYA)

Recommend

• BOBLME/Alternative sources of funding including country based fund a longer more comprehensive study on health of the three ecosystems that Hilsa occupy to understand the following: 1. Water Quality across these river systems 2. Habitat Quality and quantity in the river systems. 3. Species Biotic integrity in the systems, along with Juvenile Hilsa measurements. 4. Surveys on economic livelihoods of locals in the lower rivers.

• Standardize methods and data to be collected on these attributes by the region. • Long term study that will capture ecological connections and functions in the estuary and

freshwater sections of this region that tie in with Hilsa Recruitment and resiliency over time.

• What is a properly functioning FW/estuary/marine ecosystem and how can we improve habitat to keep populations resilient. This will be the focus of the longer term study.

3.1.5. Other priority items for consideration 37. Other priorities were discussed at the meeting, and are outlined below:

i. Explore captive breeding techniques to supplement the natural population levels of Hilsa in the region. Coordinate/conduct a meeting for exchange of ideas on this subject. CIFRI. World Fish Center (BGD) to coordinate this meeting. Put in touch with experts on sensitive culture.

ii. Explore alternative livelihoods for effort controls. (Lead: Mome, DOF, BGD). iii. Explore consumer control tools for awareness building. (Lead: Sajeevan, FSI). iv. Explore length frequency techniques and corrections to ELEFAN estimates from

empirical data for tropical fish (Malays technique calibrated with Otolith data and size frequency data to address bias issues in parameter values). Lead: CIFRI, Naskar

v. Develop a standardize index of abundance survey for the region. Dedicated ship-time is only available to India, so India will have to take a lead on this to collect survey based CPUE indices for Hilsa during a peak migration month (probably September) in the region. (Lead Gulati FSI-Writeup of approach).

vi. Data Repository for the region collating information on catch and length than can be used for analysis. This repository will assemble all the age-length data over time for each country. BOBLME will be the repository. MOU between countries for data sharing. FSI/CMFRI. Uniform data entry forms for this essential.


8

vii. Report on a regional action plan (RPOA) for hilsa using the ASEAN framework that would force policy people to make management action plans in the region. (Lead: KM Soe, MYA)

3.2. Proposals for research and M&E work in 2011-2013 38. Proposals updates were given to the countries. The proposals were currently in the process

of FAO approval and should come through by May. Some common elements/themes suggested by BOBLME RCU and submitted by the countries include the following avenues of research:

i. Collection of length frequency data both from commercial landings and experimental fishing.

ii. Collection of time series data on catch, effort and recruitment. iii. Assessment of age, growth, mortality and stock with the help of length based

models, or other newer techniques. iv. Tagging studies have been deferred to a longer term SAP for the region. v. Habitat/ecosystem indicator monitoring for assessing water quality

vi. In addition standardized methods for assessing length and developing a manual for this and shared across the countries was a desired objective.

39. In addition to this country specific proposal would perform additional work identified by country below:

i. Bangladesh: Assess Habitat quality with respect to historic and current levels for prioritization and protection

ii. India: Estimate abundance through independent surveys, and identify present spawning grounds

iii. Myanmar: Identify current spawning grounds.

3.3. New framework with technical management reference point to provide fishing advice

40. In addition reference points for management targets and advice for the region will be given such as that demonstrated below (Table 3). Regardless of the model and approach used, once this is designated as a transboundary stock, policy makers will need to assess how to share the optimal yield and allocation across the region. Currently India take 5-10%, and Myanmar utilizes another 5-10% of the overall Bay of Bengal catch. The remaining catch is from Bangladesh between 80-90%.

Table 3 stock status and management advice for the region

Ratio of B(curr)/B(Optimal)

Stock status Management advice

0.9 or above Healthy Regular fishing

0.6 to 0.9 Slightly overfished

Institute some effort and time area controls, develop action plans, awareness building, identification of spawning and rearing areas, size and mesh limit restrictions.

0.2-0.6 Overfished Institute multiple effort and time area controls, and gear restrictions, including ban of mono-filament net, insititute selective gear techniques, protect breeding grounds through closures during peak spawning, establish more juvenile sanctuaries. PR movement for awareness of consumers to decrease consumption of fish.


9

<0.2 Severely overfished and extinction risks

No fishing for 1-2 years, enforcement should be strengthened, and IUU minimized. Alternative livelihoods/rehabilitation for fishers

41. Graphically this would look like this (Figure 1).

Figure 1 Target reference points for hilsa management based on recommendations from Hilsa FAWG (Mumbai, April 25 2012)

42. While this advice is critical for sustainable management the following will be taken into account when developing and implementing this advice:

• This advice needs to be cognizant of the realities of implementation of these policies on the ground and the overall impact of these recommendations on the people.

• Countries need to make changes in implementation of these projects that would give the right quality of data into the stock assessment models.

• Countries need to establish the institutional mechanism in the long run. This is difficult, but is a key step for success.

• The countries need to develop a sampling plan that goes along with the region in terms of monitoring catch and effort by gear, sector and region. It is imperative that costs for this are estimated accurately for implementation in the Strategic Action Plan for the region.


10

3.4. Stock status advice for hilsa in the BOBLME region 43. The workshop conducted in Mumbai, India considered the range of information available,

and adopted the following stock status advice for the regional, Hilsa fish-stock in the Bay of Bengal.

4. Adoption of the report

44. The Report of the second full Hilsa Fisheries Assessment Working Group in Mumbai, India was adopted by email [May 25th, 2012].

The stock status of hilsa (Tenualosa ilisha) in the Bay of Bengal Region remains uncertain.

Although the stock status of hilsa (Tenualosa ilisha) is uncertain due to poor data quality, nonetheless using statistical smoothing methods and fitting it to a dynamic surplus production model provides estimates that determine that the stock is below optimal yield by 15-30% in Bangladesh and substantially lower in India (less than 50% of optimal spawning stock size levels). This is termed as marginally overfished in Bangladesh and overfished in India by the HFAWG working group in Mumbai. A common assessment on both stock simultaneously is now being pursued by all the countries involved.

It appears possible growth and recruitment over-fishing patterns are discerned in the stock in Bangladesh, India and Myanmar. Consequential management measures have tried to control the selectivity of gear to delay the age of first capture (closure of the juvenile jatka fishery) in Bangladesh, though not in other countries. Both India and Myanmar are investigating possibilities of implementing such measures as well for a harmonized action plan for the region.

It is not clear whether the current level of catch is sustainable in the countries. In Bangladesh, current catch has averaged 207,000 t in the marine sector for the last 5 years, while the freshwater sector 97,000 t averaged in the last 5 years. It is not clear whether the current level of catch is sustainable, though Bangladesh scientists think that the jatka closures are primarily the reason for the continuous increase in catch and continued persistence of the stocks in their waters.

In India, current catch is around 60000 Mt in the marine sector (2004 CMFRI data), while the freshwater sector is not known well (5,118 Mt average in late 2000’s in Hooghly and Matla Rivers in India averaged over 2006-2010). It not clear whether there are sub-stocks being fished on in West Bengal (India) are distinct from Bangladesh. Furthermore, it is not clear whether the current level of catch is sustainable.

In Myanmar there is declining trend in overall landings and a drop in CPUE as well as indicators of growth overfishing. Landings over the last 5 years averaged 15181t.

While Hilsa, is a highly productive (thus resilient) species and this may protect it to some extent from overfishing, pollution and loss and degradation of habitat are affecting the distribution and probably the productivity of the stock.


11

Appendix I List of participants BANGLADESH Dr M. Anisur Rahman, Hilsa WG, BGD, Team Leader Senior Scientific Officer Riverine Station, BFRI Chandpur-3602 Mobile: 01711 45 85 20 [email protected]

Mr Md. Asraful Alam, Hilsa BGD.WG member Scientific Officer Riverine Station, BFRI Chandpur-3602 Mobile : 01717 36 74 36 [email protected]

Mrs Masud Ara Momy, Hilsa BGD.WG member Assistant Director Jatka Conservation & Fishers Alternative Income Activities Project Department of Fisheries MatshyaBhaban, Dhaka Mobile: 01711 13 68 45 [email protected]

Mr Md. Kamruzzaman Hossain, Hilsa Bgd. WG member Fisheries Extension Officer Fisheries Extension Section Department of Fisheries Mobile: 01720 56 79 44 [email protected]

INDIA Dr K. Vijaykumaran, BOBLME NC India FSI Director General, Mumbai, India [email protected]

Dr Malay Naskar, Hilsa India WG member CIFRI, Barrakpore [email protected]

Dr D. Panda, Hilsa India WG member CIFRI, Bangalore [email protected]

Dr V Suresh, Hilsa India WG member CIFRI, Barrakpore [email protected]

Mr D. K. Gulati, Hilsa India WG member FSI, Mumbai [email protected]

Mr M. K. Sajeevan, Hilsa India WG member FSI, Mumbai [email protected]; [email protected]

MYANMAR Mr Khin Muang Soe, Hilsa Myanmar WG Team Leader Department of Fisheries, Myanmar [email protected]

BOBLME Regional Coordination Unit Dr Rishi Sharma Stock Assessment Coordinator BOBLME [email protected]


12

Appendix II Agenda

BOBLME HILSA FISHERIES ASSESSMENT WORKING GROUP

Mumbai, India on 24-25th April 2012.

Programme- Day one, April 24th, 2012.

9:00 Registration

10.00 Ice breaking/Self introduction

10.10 Welcome Address & Introduction to Assessment Workshop on Hilsa Fisheries from BOBLME- Dr Rishi Sharma

Address by DG India, FSI, Dr Vijay Kumar welcoming participants

10.30 Country Assessment Techniques paper on Hilsa fisheries and its stock assessment in Bangladesh

Dr M. Anisur Rahman, Riverine Station, BFRI, Chandpur

11:00 Country Assessment Techniques paper on Hilsa fisheries and its stock assessment in India

CIFRI, Kolkata (Dr Malay Naskar/Dr Panda/Dr Suresh).

11.30 Tea break

11.45 Country Assessment Techniques paper on Hilsa fisheries and its stock assessment in Myanmar- [To be Identified. KM Soe]

12.30 Wrap up (Dr Sharma)

13.00 Lunch

14:00 Integrated Assessment for the Region-Present draft paper that will be circulated prior to the meeting (paper to submitted to journal in July): Dr Sharma

15:00 Excel use with VB (or r use)- Demonstrating how the model works.

16:00 Alternative Models to be used- Use of Computers and how to develop an integrated model- Dr Sharma

16:30 Wrap up for the day


13

Programme- Day two, April 25th, 2012.

9.00 Day 1 Discussion- Dr Sharma

10.00 Management Advice and Needs (Technical)-Group Discussion

11.00 Tea Break

11.15 Improving current sampling for integrated models- Getting good measures of effort: Group Discussion

12:00 Independent Surveys of Biomass: FSI

13.00 Lunch

14.00 Next steps- Follow-up on Proposals on Hilsa- Each country will present their revised proposals for BOBLME

15.00 FAO update on proposal funding process and timeline

15.30 Short and Long term strategies

16.00 Tea Break

16:15 Wrap Up- Concluding Remarks

16.30 End


14

Appendix III List of documents presented to the meeting Presenter Title

Dr Rishi Sharma Hilsa Stock Assessment Framework: Current Status and Future Management Approaches

Dr Anisur Rahman Country Assessment Paper on Hilsa Fisheries and its Stock Assessment in Bangladesh

Dr Naskar with Dr D. Panda and Dr V Suresh

Country Assessment Techniques paper on Hilsa fisheries and its stock assessment in India

Mr Soe,Mr Htay Win and Mr Myint Thein

Possible Grounds for Spawning and Nursing of Hilsa Shad in Myanmar

Mr Gulati, Dr Vijaykumaran, Mr Sajeevan

Survey Catch Rate Standardization, using GLMs with Lognormal Error Assumption and Delta Distribution Mean and Variance Estimator

BOBLME-2012-HFAWG

16

Appendix IV Draft document outlining the stock assessment and management framework

SIMPLE APPROACH TO COMPLEX PROBLEMS

BOBLME-2012-HFAWG

17

Contents Introduction ..................................................................................................................................... 18

Hilsa Life History ........................................................................................................................ 18

Hilsa Fishery ............................................................................................................................... 19 Materials and Methods ................................................................................................................... 20

Results ............................................................................................................................................. 21

Catchability (q)............................................................................................................................ 21

Intrinsic growth rate (r), Carrying Capacity (K) and optimal yield (SMSY) ................................. 22

Sensitivity to Initial value of depletion and Model choice .......................................................... 24

Dynamic Model output using CPUE and q ................................................................................. 24

Effort Controls in Rebuilding Trajectory .................................................................................... 25

Model Fit ..................................................................................................................................... 26 Discussion ........................................................................................................................................ 26

Existing Knowledge and Management ....................................................................................... 26

New Framework and Improvements ........................................................................................... 27 Conclusions ...................................................................................................................................... 28

Acknowledgements ......................................................................................................................... 28

References ....................................................................................................................................... 29

BOBLME-2012-HFAWG

18

Introduction

Hilsa Life History Hilsa (Tenualosa ilisha) is the national fish of Bangladesh, and is an important fish consumed in the Indian states of West Bengal, Orissa, Tripura and Assam (Figure 1 below), as well as Myanmar. Unusual among tropical fish, the Hilsa is anadromous in nature, spending most of its life at sea but migrates upriver to spawn. The lower Ganges, lower Brahmaputra, Godavari, and Rupnarayan are the main spawning rivers (Figure 1 below) in India, and the Irrawaddy River is another river in Myanmar. Hilsa rear in the freshwater environment (Figure 2) before heading out to sea to feed and increase in size, and the returning to spawn as mature adults in much the way that Pacific salmon do (Sharma et. al. 2005). Migration upstream, however, is sometimes restricted by man-made barriers. Hilsa feed on plankton, mainly by filtering, but apparently also by grubbing on muddy bottoms. The spawning season occurs in rivers during the southwest monsoon and also from January to March (Fish BASE http://www.fishbase.org/ Summary/speciesSummary.php?id=1596&lang=english).

Figure 1: Current extent and catches of Hilsa (Tenualosa ilisha) Populations in the Bay of Bengal Region Hilsa, being anadromous, live in the sea for most of its life but migrate up to 1,200 km inland through rivers in the Indian sub-continent for spawning. Distances of 50-100 km are usually normal in the Bangladesh rivers. When mature, adult hilsa may reach up to 60 cm in total length. A large-sized hilsa weighs about 2.5 kg. The hilsa is known to be a fast swimmer (Southwell and Prashad 1918). Some conflicting views have been expressed on the minimum size of hilsa at first maturity. Day (1873) observed that the hilsa may attain first maturity at the end of the first year or at the beginning of the second year.

http://www.fishbase.org/

BOBLME-2012-HFAWG

19

Figure 2: Movement pattern of Tenualosa ilisha (hilsa) into different habitats (from Haldar 2005) In Bangladesh waters (Meghna River), Shafi et al. (1976) observed that, the size at first maturity is 21 cm in the case of males and 32 cm in the case of females. In Bangladesh, hilsa is available almost throughout the year in the major rivers. Hilsa starts spawning migration to upstream (Figure 2 above) during the southwest monsoon and consequent flooding of all the rivers (Rahman 2005). The eggs are deposited in fresh water and hatching takes place within 23 to 26 hours at an average temperature of 23ºC. The larvae and juveniles make their way downstream to the sea during a period of 5-6 months. They feed and grow on the way. In about 6-10 weeks the fry grow to about 12-20 cm and become known as jatka. At this stage they start migration to the sea for further growth and maturity. After growing for 1 year in the sea, hilsa become mature and undertake their spawning migration towards inland rivers to spawn again (Haroon 1998). Hilsa is a highly fecund species that may reproduce multiple times in its life (if allowed). Numbers of eggs are found to be 144 000 in 28 cm length fish up to 2.3 M in 44.5 cm length fish. The peak-breeding period of hilsa is placed during the full moon in the month of October (Halder 2004).

Hilsa Fishery The Hilsa fishery has occurred in Indian Sub-continent and Myanmar for hundreds of years (Day

1873). Culturally, the hilsa fish is an icon of the region and is used in Bengali marriage ceremonies as a sign of prosperity and fertility for the newly married couple in India. Traditionally, non-mechanized boats using gillnets would fish on this species in the mainstem of the Hooghly, Padma and Irrawaddy rivers in India, Bangladesh and Myanmar respectively. However, recent trends indicate that the fishery is moving to mechanized boats and moving more into the ocean and estuary locations to fish. In addition, In Bangladesh a juvenile fishery on hilsa fry (Jatka) has occurred for the last 50 years. Intense fishing pressure across the regions has seen the average size of the fish decline as well as overall biomass reduce in time, making the government make some tough decisions like closing areas and restricting fisheries in certain zones in Bangladesh, and to a lesser extent India and Myanmar (Halder 2004, Halder 2005).

BOBLME-2012-HFAWG

20

Current assessment techniques rely on methods developed by Pauly and David (1981) which essentially apply Elefan based methods to estimate fishing mortality on the stock, assuming natural mortality across different length classes. While this method provides an estimate of whether the fish are over or underexploited, target biomass levels and active management measures are rarely proposed in these fisheries using these methods. The methods presented here provide an alternative measure to assess the overall biomass and maximum sustainable yield levels for the stock using dynamic surplus production models, estimating the intrinsic growth rate, percent depleted in the fishery and catchability by different sectors. The data used is from 1986 to 2006 from Bangladesh (best information available), and the approach developed is proposed as the framework for a joint management model for the region. The model frameworks is flexible enough to incorporate size-based information in the estimating routines and thereby showing how the optimal yield targets may change based on better information used in the management.

Materials and Methods The model developed is a simple SP model (Logistic Model), and estimates two parameters r and

K (eq 1, Haddon 2011) fit to catch (observed) and Effort data (derived) in Table 1.

𝐵𝑡+1 = 𝑟𝐵𝑡 �1 − 𝐵𝑡𝐾�

(1)

𝐶𝑡 = 𝑞𝑡𝐸𝑡𝐵𝑡 (2)

q, r and depletion rate (the amount of Biomass available as a function of K when the dataset was compile, i.e. 1987 Biomass) can be estimated by fitting the modeled CPUE to the observed CPUE based on the data and effort standardization techniques.

The Likelihood Equations used a log-normal error structure

(3)

Where CPUE is catch per unit effort in the fishery, θ is the set of parameters, namely (r, K, and q) that are estimated to get the best fit by minimizing the negative log-likelihood function (eq. 3 above).

𝐵𝑡 = 𝐶𝑡𝐸𝑡𝑞𝑡

(4)

Table 1: Catch by sector and effort by marine sector in Bangladesh from 1987 to 2006.

( ) ( ) ( )( )∑=

−+=−

n

f f

ftftfft

PUECCPUECPUEL

12

2

,,, 2

ˆlnlnln)|(ln

σσθ

BOBLME-2012-HFAWG

21

Starting Biomass is a function of the depletion rate of the un-fished Biomass (K, this is assumed

currently to be between 2 and 5 M Tons in Bangladesh Rivers in the region),i.e. 𝐾~𝑈[2𝑀 𝑇, 5𝑀 𝑇]

Results

Catchability (q) Catchability was split into two time periods, one prior to the 2000 and the other after it. Based on

the fleet efficiencies, we estimated catchability by the different sectors, namely the motorized (MB) and non-motorized (NM) marine sectors. Uncertainty in the two sectors by time periods is shown in Figure 3. It appears that the NM sector is almost two times more efficient in catching Hilsa than the MB sectors. The probable reasons for these differences are the extensive experience of this sector in fisheries (100’s of years) and that the MB sector is probably targeting other species as well. In addition, the non-motorized sectors are probably fishing when the fish are migrating and easily vulnerable to gillnets used in the fishery. Another apparent feature is that the non-motorized fleet catchability is lowering over time, whereas in the motorized fleet this appears to be getting better.

YearInland Catch (mt)

Marine Sector (Non Motorized)

Marine Sector (Motorized)

Total Marine Catch (mt)

Catch Total (mt)

Effort (Non Motorized)-1000's*

Effort (Motorized)-1000's*

1987 91167 94851 8963 103814 298795 4.037 4.0431988 78551 91723 13227 104950 288451 5.186 4.2841989 81651 94990 15321 110311 302273 6.336 4.5251990 112408 95285 18658 113943 340294 7.486 4.7661991 66809 97573 17785 115358 297525 8.635 5.0071992 68356 102036 18070 120106 308568 9.785 5.2481993 74715 105128 17997 123115 320955 10.935 5.4901994 71370 103839 17322 121161 313692 12.084 5.7311995 84420 111475 17640 129115 342650 13.234 5.9721996 80625 109282 17378 126660 333945 14.383 6.2131997 83230 114921 16283 131204 345638 15.533 6.4541998 81634 110440 13059 124105 329238 16.683 6.6951999 73809 121909 18761 140710 355189 17.832 6.9362000 79165 119295 21072 140367 359899 18.982 7.1772001 75060 131254 23400 154654 384368 20.132 6.3772002 68250 131619 20724 152343 372936 21.281 6.6182003 62944 114274 21814 136088 335120 22.431 6.8592004 71001 157570 27268 184838 440677 23.581 7.1002005 77499 170756 27607 198363 474225 24.730 7.3412006 78273 170945 27905 198850 475973 25.880 7.582

* Estimated from Linear Interpolation

BOBLME-2012-HFAWG

22

Figure 3: Hilsa catchability by sector and period in Bangladesh

Intrinsic growth rate (r), Carrying Capacity (K) and optimal yield (SMSY)

As expected, Hilsa having a high fecundity rate coupled with a low age at maturation, the intrinsic growth rate of the population is extremely high (r between 0.4 and 0.6, Figure 4 a) below). This helps relieve some of the high levels of fishing pressure on this stock, but at the same time if for any reason there is recruitment failure on this stock, the stock would be severely threatened by the high fishing levels currently observed.

Based on our understanding of depletion levels of the stock, the carrying capacity is in the realm of 2-3 Million T. Further studies understanding hilsa life-history and relating it to suitable habitat types in the river could verify these estimates. In that case, we could estimate depletion if carrying capacity is fixed on this stock. While there are some Bayesian techniques to estimate both, we preferred to do as sensitivity analysis on depletion and compare model fits across different scenarios (Table 2 below).

In the logistic model presented, optimal spawning stock size is essentially a function of the virgin biomass (K), and hence optimal Biomass levels are in the vicinity of 1-1.3 Million Tons, with an optimal yield of 300, 000 T.

0

0.5

1

0 0.002 0.004 0.006 0.008 0.01 0.012

Resc

aled

Like

lihoo

d

Catchability by sector and Period

Catchability by Sector and Period

q-Motorized Period 2(01-06) q Motorized (86-2000)

q- Non Motorized (01-06) q-Non Motorized (86-00)

BOBLME-2012-HFAWG

23

Figure 4: Likelihood profiles of r,k, and optimal Biomass (spawning) for Hilsa

BOBLME-2012-HFAWG

24

Sensitivity to Initial value of depletion and Model choice Table 2: Initial Depletion levels and

Depletion Level

Parameters 40% 50% 53% 60% R 0.44 0.42 0.48 0.48

K 265782

1 2923102 2443082 2516272 q1nm 0.008 0.008 0.009 0.008 q2nm 0.006 0.006 0.007 0.006 q1m 0.002 0.002 0.003 0.002 q2m 0.003 0.004 0.004 0.003 -lglikihd 54.527 55.71 55.97 54.638

SMSY 132891

0 1461551 1221541 1258136 Optimal yield 294792 303546 293109 301128 2006 Bio (CPUE

1) 111716

4 1032958 927477 1049389 Level optimal 0.84 0.71 0.76 0.83 2006 Bio (CPUE

2) 112946

3 1027340 915512 1059844 Level Optimal 0.85 0.70 0.75 0.84 2006 Yield 277123 277123 277123 277123 Probability True 0.41 0.13 0.10 0.37

The table above displays four alternative models, and compares their likelihood values (criteria of

best fit) across them. If we use this as our set of possible models to describe the dynamics of Hilsa, then the probability that any one of them is true is shown below on the last row. Regardless of the depletion level thought to be on the stock in 1986 (when the time series of data begins), the 2006 levels of yield shows that the stock is below optimal yield targets by about 15 to 30%. While this is not highly undesirable for the stock, some reduction in effort would help manage the stock in a more optimal manner.

Dynamic Model output using CPUE and q Since the fit isn’t perfect from the logistic model, we will get an estimated catch that will not

exactly match the observed CPUE. However, if we use the estimates of q and use the continuous catch equation formulation (eq. 3, Figure 5), we can estimate another times series of abundance data of this stock. Adding the selectivity function across different age combinations will give us a different outcome, but since age data is not currently available, we looked at this on the whole population. In reality only a proportion of this population would be spawning biomass, and as such these estimates would need to be accordingly corrected.

BOBLME-2012-HFAWG

25

Figure 5: Fit to the data and abundance trends over time using the different CPUE data sets and the

Surplus Production Model

Effort Controls in Rebuilding Trajectory

Figure 6: Rebuilding Trajectories based on effort reduction Finally, as Figure 6 illustrates (projection), controlling effort will gradually bring the stock back to

optimal levels. If effort on the other hand increased, severe depletion could occur on the stocks presented here.

0

500000

1000000

1500000

2000000

2500000

0

50000

100000

150000

200000

250000

300000

350000

1985 1990 1995 2000 2005 2010

Cat

ch (M

T)

Year

Biomass Dynamic Model-Estimated

C atchTotal (mt)Estimated hr1Pre fishing Abundance based on r and KAb based on CPUE1Ab based on CPUE2

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

0

500000

1000000

1500000

2000000

2500000

1985 1990 1995 2000 2005 2010 2015

Rat

io o

f B/B

0

Bio

mas

s

Year

Ab based on CPUE1

Ab based on CPUE2

B/B0

Projection:

Reducing effort by 5% a year

Best Fit ModelBiomass Trends

BOBLME-2012-HFAWG

26

Model Fit Based on the residual diagnostics of the 1st model shown in Table 2 (Figure 7), we can see

there is no particular pattern of over or under-predicting the data observed in the model. In addition, the model appears to catch the trends in the two sectors quite well. It is apparent that the motorized sector maybe getting more efficient at catching Hilsa (as is illustrated by the catchability, Figure 3, and the apparent jump in CPUE in 2000’s as shown below in Figure 7).

Figure 7: Model Fit to different sector of CPUE data

Discussion

Existing Knowledge and Management Current assessment techniques have been conducted independently in Bangladesh and India, and use length at age based models assuming stable equilibrium. The results indicate that:

i. Hilsa fishery is suffering from serious growth over-fishing (indiscriminate catching of jatka, i.e. juvenile hilsa)

ii. There is recruitment over-fishing (indiscriminate killing of mature female hilsa)

iii. The fishing mortality has increased due to fishing pressure with decrease in size at first capture .

Based on these assessments, authorities in Bangladesh instituted both time/area closures as well as seasonal closures. This was primarily to minimize the juvenile (jatka) fishery on Hilsa Shad in the short-term. In addition spawning time closures were instituted to protect females during spawning. Closures instituted in 2005 appear to be reaping benefits (from Dr Rahman, BFRI, personal communication) as indicated in Table 3 (below).

BOBLME-2012-HFAWG

27

Table 3: Pre and post juvenile Hilsa abundance from 2005-2010 indicating benefits of management actions (from Dr Rahman, BFRI, Bangladesh).

Year Caught

jatka/100 m net/hour (Kg)

Decrease (%)

Increase (%) Comment

2005 0.94 - - Sanctuary 2006 0.61 35 - Sanctuary 2007 0.72 23 - Sanctuary

2008 1.89 - 101 Sanctuary + 10 days fishing ban



Simple measures such as these can be modeled in the framework presented by changing effort. As shown with this a 5% reduction in effort over time (assuming it corresponds to the 10 day fishing ban and other area closures restricting effort), a recovery trajectory closer to optimal could be realized (Figure 6 above). Catch information from Bangladesh from 2007 to 2010 confirms an increasing trajectory in yield (280,328 T in 2007, 290,000 T in 2008, 298,458 T in 2009 and 312,628 in 2010 respectively), though currently it is uncertain if this is due to effort restrictions as those are not available currently.

New Framework and Improvements However, there remain critical uncertainties in the overall performance of the stock structure as similar measures are not instituted in India, or Myanmar. If this is indeed a common stock shared by all three countries (Blaber 2009, Milton 2010), standard regulator measures need to be instituted across countries and enforced. Some basic sub-stock structure of the species is missing as is the overall catch information by country and area. In addition, even the quality of the information in Bangladesh needs to be verified as the sample design used to estimate catch is outdated. Finally, FISAT (age-length) models analyzed so far while fairly consistent over time, are difficult to use to assess optimally yield or optimal spawning biomass, and to demonstrate changes over time such as we have shown here. A progressive management plan is currently being developed (BOBLME Hilsa Working Group Report 2011) where data will be shared by all individuals involved along with key information such as:

i) index of abundance ii) age structure/length structure of the catch by jurisdiction iii) total catch information iv) Total effort and CPUE tabulation by country and sector that can be incorporated into the

modeling framework presented above

This approach is being collaboratively developed for the region currently. In addition reference points for management targets and advice for the region will be given such as that demonstrated below (Table 4).

Table 4: Stock Status and Management Advice for the region

BOBLME-2012-HFAWG

28

Ratio of current estimates of Biomass to Optimal Estimates of Biomass Stock Status Management Advice

1 or above Healthy Regular Fishing

0.7 to 1 Slightly

Overfished Institute some Effort and Time Area

Controls

0.4-0.6 Overfished Institute multiple Effort and Time Area

Controls, and gear restrictions

0.2-0.4 Severely

Overfished Restrictive Fishing

<0.2

Severely Overfished and Extinction Risks No fishing

Conclusions While the information base is not entirely complete, even current estimates of catch and effort over time coupled with some assumptions of depletions and using a simplistic surplus production model gives us a measure of stock status for the region (Bangladesh). If this is indeed a common stock across the region and shares the same recruitment pattern and fishing dynamics for the region, it may be safe to say that the Hilsa stock in the Bay of Bengal system is marginally overfished to overfished (between 0.5-0.8 of the optimal yield levels). However, simple approaches such as closing juvenile fishery areas, and restricting fisheries to certain time and areas results in effort controls which can rebuild the stock over time. Standardized measures and advice for the region are currently being developed and these will help in managing the stock sustainably over the long-run for these countries. This is important as Hilsa is the mainstay source of protein in the region, and is extremely important source of subsistence and livelihoods to Millions of fishers in the region.

Acknowledgements Chris O’ Brien, and Rudolph Hermes, provided constructive comments and directed work on earlier versions of this manuscript. This was a collaborative study and would not have occurred without the support of the governments of India, Bangladesh and Myanmar.

BOBLME-2012-HFAWG

29

References Blaber, S.J.M. 2009. Socioeconomic and biopolitical linkages in the management of tropical

shads. American Fisheries Society Symposium 69:677-690. Day, F. 1873. Report on the Freshwater Fish and Fisheries of India and Burma. Calcutta: India. Halder, G.C. 2004. Present status of the hilsa fishery in Bangladesh. Completion report of the studies

conducted under the ARDMCS, GEF component; and FFP. Report No. 38.8, Department of Fisheries, Dhaka, Bangladesh.

Halder, G.C. 2005. Strategy Report for Introduction of Selective Fishing of Hilsa, Bangladesh Fisheries Research Institute, Chandpur, Bangladesh.

Haroon, Y, 1998. Hilsa shad: Fish for the teeming millions, new management alternatives needed for the hilsa young. Shad Journal, 3:7

Milton, D. 2010. Status of Hilsa (Tenualosa ilisha) management in the bay of Bengal:An Assessment of Population Risk and Data Gaps for more effective regional management. Report to FAO Bay of Bengal LME Project.

Pauly, D. and N. David. 1981. ELEFAN-I a basic program for the objective extraction of growth

parameters from Length frequency data. Meeresforschung/Rep.Mar. Res.28(4):205-211. Rahman, M. 2005. The story of the Hilsa shad. [12 November 2007] <http://asiaarecipe.com/banchad.html> Shafi, M., Quddus, M.M.A and Islam N. 1976. Maturation, spawning, sex-ratio and fecundity of Hilsa

ilisha (Hamilton-Buchanan) of the river Padma. Proc. First Bangladesh Sci. Conf., 45. Sharma, R., Cooper, A., Hilborn, R. 2005. A quantitative framework for the analysis of habitat and

hatchery practices on Pacific Salmon. Ecological Modelling. 183: 231-250. Southwell, T., Prashad, B. 1918. Notes from the Bengal Fisheries Laboratory No. 4 Cestode parasites

of Hilsa. Rec. Ind. Mus., 15, Part II(9), 77-78.

BOBLME-2012-HFAWG

31

boblme-2012-ecology-10 · report of the hilsa fisheries assessment working group ii . 1 . 1....

Documents