Effects of changes in sea surface temperature on migratory spanish mackerel of the coastal current in nha trang bay, 1996 2015

MINISTRY OF EDUCATION AND TRAINING NHA TRANG UNIVERSITY NGUYEN Y VANG EFFECTS OF CHANGES IN SEA SURFACE TEMPERATURE ON MIGRATORY SPANISH MACKEREL OF THE COASTAL CURRENT IN NHA TRANG

MINISTRY OF EDUCATION AND TRAINING

NHA TRANG UNIVERSITY

NGUYEN Y VANG

EFFECTS OF CHANGES IN SEA SURFACE TEMPERATURE

ON MIGRATORY SPANISH MACKEREL OF THE COASTAL

CURRENT IN NHA TRANG BAY, 1996 – 2015

MASTER THESIS

NHA TRANG - 2017

MINISTRY OF EDUCATION AND TRAINING

NHA TRANG UNIVERSITY

Management and Climate Change

Topic Allocation Decision

Decision on establishing the Committee:

Faculty of Graduate Student

QUACH HOAI NAM

NHA TRANG - 2017

i

UNDERTAKING

I undertake that the thesis entitled: “Effects of changes in Sea Surface Temperature

on migratory Spanish mackerel of the coastal current in Nha Trang Bay, 1996 – 2015” is my own work The work has not been presented elsewhere for assessment until the time this thesis is submitted

Khanh Hoa, 28th June 2017

Nguyen Y Vang

ii

ACKNOWLEDGMENT

The author acknowledges supporting from the Norwegian Agency for Development Cooperation (NORAD) for the NORHED Master's Program in Marine Ecosystem Management and Climate Change

I acknowledge funding from the Norwegian Agency for Development Cooperation (NORAD) for my scholarship to follow this master's program

I would like to express the deepest appreciation to the Faculty of graduate students of Nha Trang University for helping and giving best conditions as well as scholarship to finish my thesis My special thanks go to Prof Nguyen Thi Kim Anh, Prof Henrik Glenner, Prof Sigurd Stefansson and Prof Audrey Geffen for the continuous support of my study andresearch, for their patience, motivation, enthusiasm, and immense knowledge Their guidance helped me in all the time of research and writing of this thesis

I deeply thank the Fishing Cooperative Bich Hai and Vietnam National Centre for Hydrometeorological Forecasting and RDA (http://dss.ucar.edu) for their offers in collecting data We are thankful to Mr Huan in using of MapInfo Professional Software

to figure the map of study area Our gratitude is also extended to Mr Tan, leader of Bich Hai set-net fihsery for assisting with data collection without their substantial co-operation and assistance this paper would not have been successful

Last but not least, I would like to thank my family: my parents and to mybrothers and sister for supporting me spiritually throughout writing this thesis

Thank you!

Khanh Hoa, 28th June 2017

Author

Nguyen Y Vang

iii

TABLE OF CONTENTS

UNDERTAKING i

ACKNOWLEDGMENT ii

LIST OF SYMBOLS v

LIST OF ABBREVIATIONS vi

LIST OF TABLES vii

LIST OF FIGURES AND GRAPHS viii

ABSTRACT xi

CHAPTER 1 INTRODUCTION 1

CHAPTER 2 LITERATURE REVIEW 5

CHAPTER 3 RESEARCH METHODOLOGY 16

3.2 Data collection 17

3.2.1 Secondary data 17

3.2.2 Primary data 18

3.2.3 Data analysis 19

CHAPTER 4 RESULTS AND DISCUSSIONS 22

4.1 RESULTS 22

4.1.1 Variability of SST during period of 1996 – 2015 22

4.1.2 Changes in landings of Bich Hai Set-net fishery during 1996 to 2015 24

a Fluctuation in annual landings 24

b Inter-annual fluctuations in average landings calculated for fishing months 25 4.1.3 Variations in weight individual of Mackerel during study period 27

a Fluctuations in annual average weight individual 27

b Occurrence of several individual sizes in weight in the catches 29

iv

4.1.4 Correlation between SST and landings of Spanish Mackerel 30

a Between SST and landings based on annual calculation 30

b Between SST and Landings calculated for several months 31

4.1.5 Effect of changes in SST on Spanish Mackerel body-sizes 33

a Effect of changes in inter-annual average SST on average individual weight 33

c Effects of changes in annual average SST on different mackerel body sizes 34 4.2 DISCUSSIONS 36

CHAPTER 4 CONCLUSIONS AND RECOMMENDATIONS 43

4.1 CONCLUSIONS 43

4.2 RECOMMENDATIONS 43

REFERENCES 45

APPENDICES 50

v

LIST OF SYMBOLS

r: coefficient of correlation Pearson

p-value: Probability value (<0,05)

Kg: Kilogram

D: coefficient of net

vi

LIST OF ABBREVIATIONS

AVHRR: Advanced Very High Resolution Radiometer

CPUE: Catch Per unit Effort

EDF: Environmental Defense Fund

EPA: Environmental Protection Agency of United States FAO: Food and Agriculture Organization

MODIS: Moderate Resolution Imaging Spectroradiomater NASA: National Aeronautics and Space Administration

NEFSC: Northeast Fisheries Science Center

NSW: New South Wales

NOAA: National Oceanic and Atmospheric Administration PE: Polyethylene

SEAFDEC: Southeast Asian Fisheries Development center US: United States

RDA: Research Data Archive

SST: Sea Surface Temperature

vii

LIST OF TABLES

Table 4.1 The parameters of ANOVA statistic for trends of eight months from Feb to

Sep in Nha Trang Bay 23

Table 4.2 Parameters of the trends calculated for eight months (from Feb to Sep) 26 Table 4.3 Basically statistics of SPSS analysis for weight individual in details present

by months (from Feb to Sep) 28

Table 4.4 The coefficients (r, p) of regression analysis for two variables consist of

average SST and landings based month (from Feb to Sep) 31

Table 4.5 Parameters (Function, r and p-value) of correlation analysis for variability of

monthly average SST and %IRI of each body-sizes 34

viii

LIST OF FIGURES AND GRAPHS

Figure 1.1 Map of Spanish mackerel shared stock in Southeast Asia 2

Figure 1.2 Narrow Spanish mackerel 3

Figure 2.1 Structure of set-net fisheries used to capture in Nha Trang Bay 5

Figure 2.2 Catch processing of set-net fishery at Bich Dam Island 6

Figure 2.3 Average Global SST, 1880 – 2014 7

Figure 2.4 Change in SST, 1901 – 2014 7

Figure 3.1 Map of Nha Trang Bay where puts the Bich Hai set-net fishery 16

Figure 3.2 Map of whole coastal water that received monthly average SST data from RDA covered grid of 10 x 10 resolution 18

Figure 3.3 Devices to directly collect SST data on Bich Hai set-net fishery 19

Figure 4.1 Distribution of SST at set-net fishery are lower than the SST in outside 41

Figure 4.2 Distribution of SST at set-net fishery are higher than the SST in outside 42 Graph 3.1 The temperature recorded at three water levels: surface, middle and bottom. 19

Graph 4.1 Long term (1996 - 2015) mean SST variability (0C) recorded in Nha Trang Bay These data were recorded at Station of Meteorology Center which located in Nha Trang Bay adjacent Bich Hai set-net fishery 22

Graph 4.2 The upward trends of SST several months of year experienced during 1996 to 2015 Figure (a) illustrated the upward trend of Feb; upward trend of Mar (b); upward trend of April (c); upward trend of May (d); upward trend of June (e); upward trend of July (f); upward trend of Aug (g); upward trend of Sep (h) 24

Graph 4.3 Long term (1996 - 2015) mean landings variations (Kg) of Bich Hai set-net fishery located in Nha Trang Bay 25

Graph 4.4 The downward trends of landing of several months experienced during 1996

to 2015 Figure (a) illustrated the downward trend of Feb; downward trend of Mar (b);

ix

downward trend of April (c); downward trend of May (d); downward trend of June (e); downward trend of July (f); downward trend of Aug (g); downward trend of Sep (h).27

Graph 4.5 Yearly fluctuation of average weight individual during 1996 to 2015 The

downward trend is present by the regression equation: y = -0,0373x + 2,7807, p = 0,0005 28

Graph 4.6 The proportions of three Mackerel body-size in catches of set-net fishery,

including the small size, medium size and large size 29

Graph 4.7 Annual variability of average landings and average SST 30 Graph 4.8 The cross-plot illustrated the relationship between landings and SST in a 20-

year period The regression equation: Y1 = -17548,7*X1 + 505791,7 and correlation coefficients of r = -0,657; p=0,001 31

Graph 4.9 The Correlation between landings and SST exanimated for several months

during 1996 to 2015 the negative correlation illustrated in Feb (a) and May (d) and Sep (h); No correlation in Mar (b) and April (c); negative correlation in; Significant negative correlation in June (e), July (f) and Aug (g); 32

Graph 4.10 The variability of inter-annual average SST (Nha Trang station) and

average weight individual in the 20-year period 33

Graph 4.11 Significant negative relationship between annual average SST and average

weight individual based on yearly during study period Regression equation: Y3 = 0,9077*X1 + 29,189 and coefficients of r = -0,796; p < 0.001 34

-Graph 4.12 Significant positive relationship between monthly average SST and

frequency of mackerel body-size of <2,0kgs Regression equation: y = 25,644X1 – 671,823, and r = 0,458; p = 0,021 35

Graph 4.13 Significant negative correlation between monthly average SST and

frequency of mackerel body-size ranged from 2,0kgs to 4,0kgs Regression equation: y

= -21,570X1 + 654,638 and r = -0,413, p=0,035 35

Graph 4.14 The Significant negative relationship between monthly average SST and

frequency of mackerel body-size ranged > 4,0kgs Regression equation: Y6 = -4,072*X1 + 117,150 and r = - 0,413, p=0,035 35

x

Graph 4.15 The data of SST obtained from two sources The red line illustrated the

variability of SST derived from RDA website for entire Coastal water of Nha Trang Another line present those of Nha Trang station 36

Graph 4.16 Relationship between mean SST of RDA’s dataset and mean SST of Nha

Trang station in a 20-year period Regression equation: y = 0,5104x + 13,907, coefficients of r = 0,576 and p=0,004 37

Graph 4.17 Frequency of several sizes occurred in monthly average landings 39

xi

ABSTRACT

Climate change has been affecting marine ecosystems over the world in recent decades One of migratory pelagic species, Spanish mackerel of which distribution changed under changes in Sea Surface Temperature (SST), had a decline in landings observed in Nha Trang Bay in recent years This study investigated the reduction in annual landings of mackerel during 1996 to 2015 in relation to SST with two hypotheses: (1) A rise in sea surface temperature reduces the landings of Spanish mackerel; (2) The effect of a rise SST on Spanish mackerel is body size dependent Data

collections included catches in corresponding to SST in recent 20 years as well as catch samples in the field A Simple Linear Regression Model was used to test the SST-landings associations The results showed a significant negative correlation between

annual average SST and annual landings (r= -0,681; p-value = 0,001) In relationship

between SST and various mackerel body sizes, the annual average SST has a positive

correlation with small size (less than 2,0kg) with coefficient of r = 0,508 and p-value =

0,01, while the medium (2,0-4,0kg) and large size (higher than 4,0kg) showed the

significant negative correlation with the same coefficient of r = -0,413 and

p-value=0,035 Results provided the prediction that the impacts of changes in SST might

make more difficult for set-net fisheries in Nha Trang Bay to find and catch Spanish mackerel

Keywords: Climate change, Set-net fisheries, SST, Migratory Spanish mackerel

1

CHAPTER 1 INTRODUCTION

Climate change have been affecting on marine ecosystem very clear and dramatic (EDF, 2013) In term of climate driven, increases in Sea Surface Temperature (SST) had the significant impacts on coastal waters where have high biology such as the distribution and abundant of species, respect to pelagic species (Radlinski, 2013) Migratory species, for instance, which have a long distance moving are very sensitive

to the changes in upon temperature (Smith, 1985) One of whom, migratory mackerel species has strong response under increases in SST as was the shifts to preferred temperature areas (NOAA, 2015) In northeast coast of United States, Atlantic mackerel dispersed when SST rose (Radlinski, 2013) and occurred in the areas with SST ranged from 7 – 80C (Goode, 1884), while this species appeared in waters had about 40C in Northern Gulf of St Lawrence (Castonguay et al., 1992) By contrast with Senegalese waters, the landings of horse mackerel were highest in winter when average SST was lowest value (Diankha et al., 2015)

Spanish mackerel (Scomberomorus commerson) mainly distributed where

include inshore waters of Indo-Pacific with two moving circles, one way is from Red Sea and South Africa to Southeast Asia, another way is from North to China and Japan

(Status of fisheries resources in NSW, 2008) (Figure 1.1) This species moves on coastal

waters of Southeast Asia as predator of small fish such as anchovies, squid as well as prawns from Feb to Sep when have higher intensity of prey (Queensland fisheries, 2008;

Vo Dieu, 2009) This moving provides the vast majority of commercial production for countries, including Malaysia, Singapore, Brunei, Thailand Gulf, Vietnam and China and annual estimated quantity is up to 25,000 tones Because of narrow berried species (FAO/SEAFDEC, 1985), even though FAO and SEAFDEC had regulation to manage this species, stock assessment is also lack of information as well as effecting of environment conditions

In Vietnam, by fisherman monitoring who directly capture Spanish mackerel stock in coastal waters, this species moves along coastal waters at the first time of several years after a long migration in recent decades Amount of quantity bycatch of gillnets, purse seines net and set-net fisheries typically ranged the length from 620mm to 100mm

2

with three particular sizes, including 660mm, 760mm and 800 – 100mm and the second size is the most popular in annual catch with the ages estimated from 2 to 3 years old (Phan et al, 2003)

Figure 1.1. Map of Spanish mackerel shared stock in Southeast Asia

(Source: FAO/SEAFDEC Workshop On Shared Stock In Southeast Asia, 1985)

In Khanh Hoa province, Spanish mackerel stock annually typically occurs from Feb to Sep with high density This species was mainly captured by set-net fisheries which have been existing in a long time unchanged about site, traditional structure, mess sizes and fishing processing The main species which were captured by set-net fisheries

in Khanh Hoa province included Spanish mackerel, swordfish, skipjack, therein, the landings of Spanish mackerel has high proportion and provided about 150 tones per year (Khanh Hoa fisheries, 2010) These contributions are not only playing an important role

in fishing community as was improving their livelihood, these are also the main part of food chain of biodiversity in coastal waters in order to develop the friendly fisheries One of five set-net fisheries in Nha Trang Bay since 1960s named Bich Hai fishery

The track of migratory Spanish mackerel moving on coastal waters of Vietnam

3

which was created and operated by fisherman group who voluntarily offer their capital together This is like other set-net fisheries in coastal water of Khanh Hoa

Figure 1.2 Narrow Spanish mackerel

(Source: Queensland fisheries, 2008)

From the temporary observations on Bich Hai set-net fishery, however, it is the fact that the landings of Spanish mackerel had significant decrease during recent five years in the catches, from 29,4 tons in 2011 to 2,6 tons in 2015 (Khanh Hoa fisheries, 2015; Nguyen Y Vang, 2013) In comparison with another species in the catches, most

of whom were increased in different degrees, such as skipjack, swordfish and others While there are a lot of evidence from previous studies indicated a significant correlation between distribution of mackerel species and changes in SST as well as forecasting the fishing ground to catch this species almost coastal waters, there is a lack of information for those situation in catches of set-net fisheries Because of which is one of mackerel species, its behavior is sensitive to environmental conditions, respect to SST Therefore,

it is necessary to know that whether variability of SST affected the landings of Spanish mackerel in Nha Trang Bay as was shifting their life history where they typically were captured in a long time to other places in which they were unavailable to set-net fisheries

or other environmental conditions

Due to both changes in distribution and body size of fish have impacts on fish stock (Overholtz et al., 2011) In particular, the purpose of this research is to determine the possible relationship between Sea Surface Temperature variations and inter-annual

4

variability in landings of mackerel Furthermore, we also tested the reaction of Spanish mackerel in different body sizes to changes in SST Those were totally answered by

which two primary hypotheses were tested: (1) A rise in sea surface temperature

reduces the annual landings of Spanish mackerel; (2) The effect of a rise SST on Spanish mackerel vary among mackerel body sizes Relevant to hypothesis (1), from the

phenomenon occurred in Nha Trang Bay above, the finding is expected to the fact that the SST rise in recent 20 years is the cause of decline in annual landings of Spanish mackerel For hypothesis (2), it further assesses the frequency of each body size varies among different sizes of Spanish mackerel These findings are the cornerstone to continue to research the distributions of mackerel when this species moves out of Nha Trang Bay By doing this, the forecasting of Spanish mackerel distribution will be improved as well as the seasonal fishing and appropriated fishing gears to catch this migrant species

(Figure 2.1)

Figure 3.1 Structure of set-net fisheries used to capture in Nha Trang Bay

(1-Main leader; 2-sub-leader; 3-chamber; 4-Main Buoy; 5-anchor)

1

5

6

The principle of set-net fishery is to block the migratory fish stock and tend their moving along the main leader on chamber to catch them Typically, the sites of set-net fisheries are far from beach areas about 12 nautical miles with geographic characteristic

of environment conditions that attracted migratory species (Laongmanee et al., 2005)

In Khanh Hoa province, Vietnam, set-net fisheries typically set of five sites around Nha Trang Bay within coastal waters of Khanh Hoa province far from beach from 5 to 7 nautical miles where migratory species annually move on, including skipjack, swordfish, Spanish mackerel and other tuna species without fishing effort and quota regulations (Phan et al., 2003; Phu, 2017) These sites are located in depth ranged from

20 to 30m and current with speed of 2 to 4m/sec Set-net fishery activities occur from Feb to Sep, excepted three months at the end of the year had considerably storms or strong Northeast monsoon Since 1990s, this fishing gear was encouraged to develop in Nha Trang Bay by Khanh Hoa’ fisheries management as a way to protect environment and coastal resources

Figure 4.2 Catch processing of set-net fishery at Bich Dam Island

Sea Surface Temperature

Sea Surface Temperature which is an important physical component of the ocean has been consistently increasing during 1880 to 2015 with average rate of 0.170C per decade (NOAA, 2015) These changes are due either to long-term changes in heat input associated with global and regional warming (Levitus et al., 2012), or due to changes in circulation patterns (e.g strength and location of the North Atlantic sub-polar gyre; Hatun et al., 2009) For example, Trenberth et al., (2007) observed from 1901 to 2005,

7

the average ocean surface temperature increased 0.74°C In Southern East China Sea, Liu and Zhang (2013) indicated that the SST had an increase of 0.56±0.11°C per decade during period of 1982 – 2001 Supporting to future of SST scenario, the ocean water temperature upper 80m depth is expected to rise by 0.5°C in 2035, especially the size of Warm Pool (as defined by the 29°C isotherm) is expected to increase by 230 – 250% by

2035 under some places in Tropical Pacific Ocean

General Impacts of SST on marine organisms

Depending on different geographical regions, SST can affect on marine organisms including distributions, assemblage and composition, body-size of fish, especially with respect to coastal habitat (Cushing, 1982; Glantz, 1992; Bell et al., 2011; Ganachaud et al., 2012; Cheung et al., 2013; Hsueh-Jung Lu et al., 2014) Furthermore, the effects of climate change were also found on the population and body size of migratory species, therein mackerel which are very sensitive with changes in SST on worldwide (Radlinski et al., 2013)

Figure 5.3 Average Global SST, 1880 – 2014 Figure 6.4 Change in SST, 1901 – 2014

(Source: www.epa.gov/climatechange/indicators - Updated June 2015)

Prior to changes in distribution of marine organism in response to changes in SST, having said that ocean temperature rise changed their first living areas, coastal waters assemblage to different ways such as shifting to the north, east, various depth and both of those (Overholtz et al., 2011) Those biogeography shifting were highly depended on behavior of several species in responding to the range of temperature (Nye

et al., 2011) Taking example is one study in Northeast United State continental shelf where has shown a highlighted change in distributed marine fish with 36 species under

8

oceanographic changes with respect to SST changes (Nye et al., 2011; Overholtz et al., 2011) The study revealed that while some stocks located at increasing depths, many of those species were showed the poleward shifts These results are also supported by Southward et al (1995), a case study for distribution of fish under global warming in Southwest Britain This finding indicated that the latitudinal moving of zooplankton is the cause of pole ward shifting of pelagic fish ranged 120 miles over seventy years from early 1920s to 1980s in responding to fluctuations of climatic warming These patterns were also predicted that in next 50 years, the rising of temperature of 20C will be expected to latitudinal shift of 200 – 400 miles Another case study of Greenland water, Mackenzie et al., (2014) reveals that most of species shifts were be so far involved in small species at lower trophic level as much as the entering of large predator fish species

into high latitude subpolar including Bluefin tuna (Thunnus thynnus) and Atlantic mackerel (Scomber scombrus) long time ago

In contrast to tropical Pacific Island countries, the shift of pelagic fish in corresponding to changes in SST are different between regions In the expansion of the warm pool and contraction of the Pacific Equatorial Divergence (PEQD) province, tuna

is one of popular pelagic species had shift to the East under changes in ocean current and temperature (Lehodey et al., 1997; Bell et al., 2013) This case is due to the changes

in environmental conditions altered distributed place of zooplankton which attracted the small fish as were the main feeding of Tuna to offshore where has preferred temperature

In comparison with the Northeast Pacific, the pattern is reversed From observation the moving of zooplankton to North under warming, Zhang Da-juan et al., (2010) found the northward extending of both of pelagic and benthic – pelagic species These findings were the fundamental research for Cheung et al., (2014) to use the three different Earth System Models illustrated for the future changes in distribution of 28 species in this area These indicated that from 2000 to 2050, these study species are expected to shift poleward with average distance rate of 30.1±2.34 (S.E) km per decade Scientists also resulted that in scenario of distribution, the structure of pelagic species may change considerably in different regions in corresponding to changes in SST altered the fishing seasons This finding is similar to the result of studies being carried out by Ching-Hsien

Ho et al., (2014) and Hsueh-Jung Lu et al., (2014) for Tung Ao Bay in Taiwan coastal

9

waters They showed that the sample catches by set-net fishery changed in both distribution and structure of cold-water and warm water fish and in timing of fishing seasons

In addition to impacts of changes in sea surface temperature on marine fish, the assemblage, composition and seasonality of occurrence of fish were also changed in different ways, especially with respect to coastal waters On the one side, having said that those changes are likely positive, on the other side, these are inverse Supporting the first case, in the inshore areas of Mexican Central Pacific, the assemblage of some species was sudden increased by SST anomaly during El Nino event occurred from 1998

to 2000 (Godínez-Domínguez et al., 2000) By observations from small Gill net fishery,

at the end of an El Nino episode the seasonality of occurrence of inshore communities were changed and some uncommon species were more unusually abundant Furthermore, this study resulted that during El Nino episode, the species richness was higher than the non-anomalous year Another case found that the current regimes which bring the typical abundance of coastal species have been effected by ocean warming This leads to impacts of SST variations on assemblage, seasonality and structure of fish because of changes in food web and nutrient processing which attracted species richness (Kuang-Yao Hsia et al., 2004) This is supported by a case study in Japan Sea where has the Tsushima Warm Current that build the high biodiversity in coastal water based on cold and warm species By experiment of CPUE on two bottom trawl fisheries, Tian et al., (2011) indicated that meanwhile SST rise had negative impact on cold water demersal species, the warm demersal species were preferred when have increased in biomass and distribution during warming 1990s

In relation to impacts of warming on seasonality of coastal species, Ching-Hsien

Ho et al., (2016) had monitored the coastal stock that changed their attention between seasons in coastal water of Taiwan As a result of increase in SST, the winter and spring species decreased dramatically year after year, while summer and autumn species became dominant in the sample catches Together with Hsueh-Jung Lu et al., (2012), this study agreed that changes in seasonal fishing in coastal areas are likely not effected by anthropogenic activities These were coincided with the trend of SST variations

10

Effects of changes in SST on mackerel

Among the species which have been effected by annual SST changes is mackerel With a wide range moving this species is very vulnerable to water temperature variations (Overholtz et al., 2011; Radlinski et al., 2013; Diankha et al., 2015) By different ways, many scientists reveal that the response of mackerel to climate variability varied between different species and different regions all over the world SST variations obviously impact on seasonal migration, changes in distribution, preferred temperature, fluctuated landings and body-sizes of mackerel (Studholme et al., 1999) These, in turn, suggested that both positive and negative relationship between SST variations and presence of mackerel

SST and distribution of mackerel

Priority of SST variations impacts on mackerel is changes in distribution This leads to change the historical life either to mean latitude or depth or both (Perry et al., 2005) In 2013, Harding et al., carried out in the Northern Ocean to conduct that under effects of warming, the horse mackerel declined in average length at all significant latitudes over 30 years (1977 - 2007) From the collected data in the field during that period, horse mackerel displayed a shift toward higher proportions of individuals in 56%

of cells, whilst it was only 22% of no response and 22% of cells shifted towards higher proportions of larger individual In the inshore waters of Canada, Atlantic mackerel is also recorded its occurrence in period from 2002 to 2013 under water warming by Jørgen Berge et al., (2013) This finding indicated that in Svalbar, the occurrence of Atlantic mackerel has changes following northward into the Arctic in each summer In addition,

in term of changes in environmental gradients impacts on abundance and assemblage of fish in the Archipelago areas were also concerned when there were different occurrence

of fish in these areas with their various temperatures Martin Snickars et al., (2009) conducted that in coast of Sweden with different thresholds of sea surface temperature, the fish assemblage is also different in nine selected habitats

SST and changes in preferred temperature

Those distributions of mackerel might be associated with changes in the ranges

of preferred SST That means under SST variations, mackerel distributed to the places

Nøttestad et al., (2015) resulted that in Southwest region of Norway Sea, there were majority of mackerel school swam above the depth where temperature stood by thermocline of 70C and all mackerel stock occurred in warm water than 60C Furthermore, in Northwest region of Norway Sea, the preferred temperature of mackerel was expected to 80C and optimal value is 6,80C Another study by Castillo et al., (1996)

in Chile showed the different seasons in corresponding to ranges of SST, the frequency

of mackerel varied in catches

Relationship between SST and landing of mackerel

Changes in distribution and body size of mackerel by changes in SST are able to change their landings with different degrees Given suggestion that there were a significant relationship between SST and landings of mackerel This is due to either the moving of this species on different places to avoid the SST anomaly or replacing its recruitment areas respected to spawning to preferred temperature (Studholme et al., 1999; Jansen, 2014) In Senegal waters of Indonesia Ocean, Diankha et al., (2015) figured the correlation between SST and landings of mackerel during 1999 to 2009 As

a result of study, SST had significant negative relationship with landings of horse mackerel That means the decreasing in landings coincided with increasing in SST during period study Furthermore, these researchers also compared between landings in summer in which had highest average SST and winter with lowest SST The result indicated that landings by summer were always higher than those in winter between

1999 and 2009 Another study by Radlinski et al., (2013) for northwest coast of United States also determined that landings of Atlantic mackerel had negative correlations with

in SST anomaly from 1985 to 1999 focused on several Marchs For example, Landings

12

of Atlantic mackerel was caught in deeper water when SST higher than average value, while those was landed in shallower waters corresponding to SST lower than the average Fluctuations of ladings are also related to potential of recruitment effected by environmental conditions through changes in spawning areas (Jansen and Gislason, 2013) A reduction in spawning of mackerel in North Sea was because of unfavorable temperature variations including warming and cooling period leading to changes in spawning distribution (Jansen, 2014)

In order to analyze those relationships, recent studies suggested that the main reasons of those might be changes in zooplankton and some species as main feeding resources of mackerel associated with SST variations (Nøttestad et al.,2015; Plourde et al., 2015) These environmental condition controlled the spatial distributions of prey species where had high abundant attracted such pelagic predations stock, especially with respect to mackerel (Snickars et al., 2014; Robert et al., 2009) Nøttestad et al., (2015),

a study in Norway Sea, for example, generally found the swimming course of mackerel

is directly toward active feeding moving This supported that in all of regional samples, the largest fish stock was caught in which had the highest concentration of zooplankton associated with lowest SST Trenkel et al., (2014) also reported spawning mackerel stock changed their life history from spawning grounds to nursery and feeding areas as well as the opportunist feeders mackerel generally needed such as amphipods, pelagic bivalves and others (Olafsdottir et al., 2015)

Additionally, other research groups had investigated the depending of mackerel assemblage on chlorophyll-A concentration related to impacts of SST variation This is due to that SST variability had inverse relationship with chlorophyll-A associated with monsoons, current and depth (Nurdin et al., 2016) This negative relationship is also illustrated that during Southeast monsoon in the Archipelagic waters of Spermonde Indonesia, the Chlorophyll-A concentrations reached highest value by Aug in corresponding to lowest SST, while in Northwest monsoon, this pattern is inverse Semedi et al., (2009) used MODIS data to monitor the Chlorophyll-A distributions under effects of SST variability in Makassar Strait during 2000 to 2007 This provided the information of ranges of SST appropriated with distributing of Chlorophyll-A These results supported implementing of recent studies to investigate the mackerel fishing

13

ground As a result of monitoring of relationship between SST and Chlorophyll-A, for example, Semedi et al., (2009) predicted the fishing ground of Short mackerel which is commercial species of Indonesia fisheries Forecasting mackerel concentration by

monitoring occurrence of Chlorophyll-A is to figure the potential fishing ground of R

Kanagurta, a kind of mackerel species by Nurdin et al., (2016)

Fluctuations of mackerel landings are also related to changes in body size of mackerel Environmental conditions, SST variations might have various impacts on different body sizes Canales et al., 2015 resulted that there is a decreased trend of mackerel with larger sizes in catches during 1990 to 2008 From the information of catches in North Sea during recent 30 years, Harding et al., (2013) showed the higher proportion of smaller size mackerel, while the larger size of mackerel had higher rate in the catch landed in the Northern North Sea That means the large size is vulnerable to increases in SST (Flourde et al., 2015; Olafsdottir et al., 2015) Mackerel length and weight were declined during the most recent decade (Olafsdottir et al., 2015) Picking

up the sample of average individual in 2013 in comparison with those in 2002, there is

a decreased pattern in body size during this time The negative correlation between SST anomaly and large mackerel was also found by Radlinski et al., 2013 in Northeast coast

of the United States In some divided regions along this coast, the catch weighted of large mackerel was lower when SST increased and those were not significant correlation with smaller size

In order to use techniques to test the relationship between landing of Mackerel and SST, some technique models were flexibly used to indicate the expected information In priority of using of techniques, the Simple Linear Regression Model (lm) was prior to test the fish assemblage in relation to changes in SST anomalies (Ousmane Diankha et al., 2015; Mary K Radlinski et al., 2013; Bambang Semedi et al., 2009, Canales et al., 2015) In relation to SST data collecting, while the Moderate Resolution Imaging Spectroradiomater (MODIS) was used by Semedi et al (2009) to observe in Makassar Strait of Indonesia, Lu et al., (2014) obtained the SST data in Taiwan by using the Advanced Very-High Resolution Radiometer (AVHRR) database with input information from NOAA series of satellites Another ways related to survey distribution

of fish, Overholtz et al (1991) used Northeast Fisheries Science Center (NEFSC)

14

surveys model for Eastern US continental shelf indicated that the mean large length mackerel distributed to lower temperature area where were mostly attracted by smaller size of mackerel In comparison with this concept, Radlinski et al (2013) obtained SST data in the same area from satellite remote sensing technology to reveal that the catch

of mackerel was landed in deeper water offshore if there SST was warmer, reversely, the landing of mackerel was in shallow waters inshore These data analysis have been becoming familiar tools in research of impacts of nature variations on fish assemblage and distribution with various applications

These studies bring a conclusion that marine organisms, especially with respect

to mackerel species, has changed their distribution and body size either by effects of SST variability or direct following the feeding species which has flexible moving under changes in temperature While some pelagic and demersal species such as tuna, anchovy and herring had investigated their distribution and landing in coastal waters of Vietnam

in recent decades, there are lack of information in both annual migration, landings and distribution of Spanish mackerel as much as those changes associated with SST variability (Bui Dinh Chung et al., 1990; Do Cong Thung, 2011; Nguyen Thi Huong Thao et all., 2012) These works had done in North Gulf of Vietnam to forecast the fishing ground and figure fishing season (Bui Dinh Chung et al., 1990) For demersal species monitoring, while recent studies indicated their preferred temperature in spawning and landing, given results for effects of sea water temperature on quantity of catch of demersal fish in Northern Gulf of Vietnam by Nguyen Van Huong and Doan Bo (2013) was unclear that increased quantity of demersal fish in catch of bottom trawl in relation to range of SST from 260C to 290C

In comparison with the light of previous studies in mind, in Vietnam, lacking of information about relationship between environmental factors and migratory Spanish mackerel lead to congestion in accounting and predicting quantity of fishing activities

It is necessary, therefore, to observe that whether Spanish mackerel has respond to SST anomaly when they move on Nha Trang Bay caused the reduction of landings in recent decades Following the previous studies, this work tested the relationship between SST variability and annual landings of Spanish mackerel being similar to works of studies of Studholme et al., (1999); Jansen, (2014); Diankha et al., (2015); Radlinski et al., (2013),

15

Nøttestad et al., (2015); Plourde et al., (2015) Nevertheless, because of limitation in research processing, the landings of mackerel were mainly observed from landings of set-net fishery for 20 years, a traditional fishing gear is not changes in structure and position (Jung Lu et al., 2014) These data were, in turn, separated into three individual size classes following both Vo Dieu et al., 2009 and Radlinski et al., (2013), except that here we adjust these sizes by individual weight sizes from limited samples on set-net fishery represented in whole catches The daily SST variability were obtained from in situ station in Nha Trang Bay adjacent to location of set-net fishery as long as recent 20 years In term of requirement to test the difference between in situ station and whole Nha Trang Bay, these data are compared to SST data derived from RDA website being gathered from MODIS, AVHRR and NOAA with grid of 10 x 10 in advance for coastal water of Khanh Hoa, Vietnam Finally, this study also uses the simple linear regression

to test the correlation between SST variations and fluctuation landings, SST variability and changes in body sizes in catches

16

CHAPTER 3 RESEARCH METHODOLOGY

This study was totally employed by quantitative approach method, meaning that the formal, objectives, statistical model processing in which data collecting were utilized

to test study’s hypotheses Data collecting includes secondary and primary data observed in study area Quantitatively, secondary data consist of landings and SST during from 1996 to 2015, and primary data observed in study area including daily catch corresponding to SST in several catches These data were then analyzed to test hypotheses by Simple Linear Progression Model which resulted the quantitative findings were presented by two main coefficients of r (values from -1 to 1) and p (<0,05)

3.1 Study area

Figure 7.1 Map of Nha Trang Bay where puts the Bich Hai set-net fishery

The landings, SST in situ as well as SST derived from RDA and catch survey

data were investigated at Bich Hai set-net fishery which located at site within Bich Dam Island water ranged 24 to 27 meters in depth Bich Dam Island is geographically in the central of Nha Trang Bay, Khanh Hoa province of Vietnam where fishing activities is limited by Nha Trang MPA regulation, situated at 6,2 nautical miles away from Nha

17

Trang fishing harbor This area has the prevailing current regimes following the two monsoon, namely Northeast and Southwest monsoon In association with Southwest monsoon, in which the current of coastal water which moves along coastal water annually brings mackerel stock to Nha Trang Bay from Feb to Sep, is available to fish

by set – net fishing (Figure 3.1)

3.2 Data collection

3.2.1 Secondary data

Fish landed catch data

In term of this study, the Spanish mackerel landings in weight and number of fish were collected in order to test the relationship between SST and both of inter-annual landings, average monthly landings and body sizes in weight These data were obtained from Bich Hai set-net fishery during from 1996 to 2015 The total number of catches is appropriately 1738 catches typically distributed on eight months between Feb and Sep

in daily details These data were then subjected to annual landings, monthly landings

analysis (Appendix 2) as well as averaged for annual and monthly body size in weight

Those individual weight data were after separated to three sizes consist of small size (less than 2,0kg), medium size (from 2,0 to 4,0kg) and large size (higher than 4,0kg) in each catch These sizes were approximated with length groups of 620-700mm (2 years old), 700-850mm (three years old) and higher than 850mm (higher than 3 years old) (Vo

Dieu et al., 2003)

SST data recorded from 1996 to 2015

In requirement to test the relation to presences of mackerel landings, a 20-year series of daily SST was used in this study This data was provided by Meteorological Center of Central, Nha Trang city during 1996 to 2015 This SST data was after annually averaged based on between Feb and Sep in corresponding to timing of set-net fisheries

In corresponding to monthly average landings, annually averaged SST for several

months were also calculated (Appendix 1) Before using those data, we generally tested

the trend of average annual SST in situ at Nha Trang station with those at whole of coastal water surrounded Nha Trang Bay To do that, the second SST source was

obtained from Research Data Archive (RDA) consisted of data from NOAA (National

of Nha Trang Bay (Figure 3.2) These data covered the time period from 1841 to the

end of 2016 in monthly average value However, because of limited landings mackerel recorded from 1996 to 2015, those SST data were only picked up in recent 20 years

Figure 8.2 Map of whole coastal water that received monthly average SST data from RDA

covered grid of 10 x 10 resolution

(Source: Research Data Archive)

3.2.2 Primary data

Two variables were directly recorded on set-net fishery included number of fish

in catches in corresponding to SST The samples consist of four times almostly carried out at the first five days of several months from June to September, 2016 and total number of survey’s day is 20 days For catch survey, we recorded total 22 catches including quantities and number of fish in each catch These data were later calculated the average weight individual in order to test the impacts of SST on body size The SST

variable were also collected by thermal sensor device (16-Bit Digital SPI Temperature,

mode: ADT7320, Figure 3.3) at three water levels: surface (0,5m in-depth), middle

Red point is Nha Trang survey station adjacent to where Bich Hai set-net fishery were located; yellow square is whole of coastal water of Nha Trang Bay which specified SST obtained from website: rda.ucar.edu

19

(15m far from surface) where mackerel stock swims before captured (Phan et al., 2003) and bottom coincided with each catch

Figure 9.3 Devices to directly collect SST data on Bich Hai set-net fishery

However, those catches collected in the last three months of seasonal set-net fisheries, these cannot represent for entire of landings over fishing months of which landings concentration was between Mar and June We thus used the sampling temperature at three water levels to compare with SST as the testing the changes in temperature in corresponding to middle water level The results of testing indicated that the SST is appropriated proxy to use to test impacts of its changes in relation to mackerel (r=0,973 and p-value=0,00)

Graph 1.1 . The temperature recorded at three water levels: surface, middle and bottom

3.2.3 Data analysis

The Simple Linear Regression Model was typically used for all statistical

analyses and plotting in this study There were three relationships that were tested in

20

order to testing two hypotheses, including relationship between annual average SST annual mackerel landings during recent 20 years; relationship between annual average SST and fluctuations of inter-annual average mackerel body size in weight; the correlations between annual average SST and three sizes consist of small, medium and large size The SPSS analysis was used as a main tool in data processing to result to the necessary parameters as well as scatter plots presented in this study

Relationship between annual average SST and annual landings variations

The annual landings were calculated during the 20-year period from 1996 to 2015 from secondary data collecting focus on 8 months from Feb to Sep as well as those for

annual average SST In order to determine the relationship those variables, the simple

linear regression model was used following the function (1) {Y1 = α1 + β*X1 + ε1 (1-1) Where: Y1 is annual landings and X1 is annual average SST (Appendix 2 and Appendix

3) In addition, the relationship between SST and Landings basis on monthly values was

also carried out by function (2) {Y2 = α2 + β*X2 + ε2 (1-2) Where: Y2 means annual landings and X2 means annual average SST calculated for several selected months

Relationship between annual average SST and three different body sizes in catches

To test the impacts of SST variations on mackerel body size, this study carried out with two cases, including impacts of inter-annual average SST on annual average body size, and effects of annual average SST vary among three different body sizes in catches during recent 20 years As the first case, the relationship of those variables was tested by Simple Linear Regression Model following the function (3): {Y3 = α3 + β*X1

+ ε3 Where Y3 is annual average mackerel body size and X1 is annual average SST during

study period (Appendix 1 and Appendix 3)

For another case, the relationship between annual average SST and three different body sizes of mackerel in catches (small, medium and large size) are represented by those SST and the proportion of frequency (%Freq) of each mackerel body size in annual

landings (Radlinski et al., 2013; Julie Roux et al., 2016) %Freq values were calculated

by function (4):

21

%𝐹𝑟𝑒𝑞𝑖 = 𝐹𝑟𝑒𝑞𝑖

𝐹𝑟𝑒𝑞𝑇 ∗ 100 (4) Where 𝐹𝑟𝑒𝑞𝑖 is proportion of annual occurrence at size (i) and Freq T is total

of occurrence in catch calculated for size (i) during 20 years

These %Freq values were later tested the relationship with annual average SST by

correlation analysis again (Radlinski et al., 2013), following three functions:

Where Y4, Y5, and Y6 are %𝐹𝑟𝑒𝑞 𝑠𝑚𝑎𝑙𝑙 𝑠𝑖𝑧𝑒 , %𝐹𝑟𝑒𝑞 𝑚𝑒𝑑𝑖𝑢𝑚 𝑠𝑖𝑧𝑒 and

%𝐹𝑟𝑒𝑞 𝑙𝑎𝑟𝑔𝑒 𝑠𝑖𝑧𝑒 , respectively X1 is annual average SST These values were calculated during study period

All those analyses were performed by SPSS analysis with the results being

calculated at a confidence level of p-value less than 0,05 To figure the distribution of SST

compared between inside and outside of location of set-net fishery Panoply 4 Software getting from NASA website were used in this study MapInfo Professional Software was

also used to import the study area and the structure of set-net fishery was drawn by

AutoCAD software

4.1 Although there is a significant fluctuation, the average annual SST has increased

with range of 1,50C from 26,20C to 27,70C, a mean value of 27,030C and standard deviation of 0,3540C Separately, changes in SST were experienced in four scenes Beginning at lowest temperature (26,20C) in 1996, the annual SST average sharply increased in the first three years of the period reaching at 27,40C in 1998 before steeply dropped to 26,70C in 2000 After that, the production of set-net fishery rose again getting short-term fluctuation with average temperature of 270C in 2006, suddenly declined again from 2006 to 2008 to a historical low (26,60C) The landings recorded in remaining times from 2009 to 2015 had experienced as a strong signal in changes and then reached

a warmest value of 27,70C in 2015

Graph 2.1 Long term (1996 - 2015) mean SST variability (0C) recorded by Nha Trang station

of Meteorology Center adjacent Bich Hai set-net fishery

23

b Variations in average SST per month

Based on monthly calculation, there were increases in average SST with different

ranges during study period (Table 4.1) The dramatic fluctuation in SST of the first 5

months from Feb to June induced the higher ranges than others selected months The highest range was in April, which was a 3,20C spread in mean value, ranged from a low

of 26,40C in 1996 to a high of 29,60C in 2015 (Graph 4.2c) Following by the figure in

Feb, SST which fluctuated in range of 2,80C, were lowest in 1996 (23,50C) and highest

in 2015 (26,30C) (Graph 4.2a) March (Graph 4.2b) and May (Graph 4.2d) had less

than the first two factors with the same level of range of 2,60C, followed smaller range

in June (2,30C) In comparison with those calculated for July, Aug and Sep, the SST ranges were relatively narrow Especially with respected to pattern in Sep, there was bottom-up range of 1,20C, equally calculated by lowest in 2002 (27,50C) and highest in

2010 (28,70C) (Graph 4.2h) In July and Aug, the ranges of SST were tiny larger than

those in Sep getting higher ranges of 1,90C and 1,70C, respectively

Table 1.1 The parameters of ANOVA statistic for trends of eight months from Feb to Sep in Nha Trang Bay