Thushani Suleka Madhubhashini ELEPATHAGE 唐丹玲 王素芬

摘要馬林魚(yú)（Makira sp.）是斯里蘭卡一種有名的海洋魚(yú)類.為了降低在近海區(qū)域捕獲馬林魚(yú)的成本，研究馬林魚(yú)的豐度是否受溫度的影響十分重要.因此，對(duì)2～13.5°N，76.5～88°E區(qū)域內(nèi)2006年的Aqua/MODIS海表溫度數(shù)據(jù)和從斯里蘭卡國(guó)家水產(chǎn)資源研究和發(fā)展機(jī)構(gòu)獲取的2006年的馬林魚(yú)捕獲數(shù)據(jù)進(jìn)行了分析.通過(guò)對(duì)東北、東南、西北和西南海域的數(shù)據(jù)分別繪圖，分析海表溫度和魚(yú)類供應(yīng)的時(shí)空變化以及它們之間的關(guān)系.研究結(jié)果顯示：馬林魚(yú)最大的魚(yú)獲產(chǎn)量區(qū)域位于斯里蘭卡海域的西部海域；漁業(yè)單位捕撈量隨著海表溫度變化而變化；捕獲馬林魚(yú)最大頻率的海域海表溫度為27～28 ℃.根據(jù)經(jīng)驗(yàn)累積分布頻率（ECDF）分析的結(jié)果，漁獲率與22～31 ℃海表溫度之間有著顯著的關(guān)系，尤其與26～31 ℃海表溫度的關(guān)系最為顯著.對(duì)于不同的區(qū)域，捕撈馬林魚(yú)最合適的時(shí)間也不同，東北海域的3—5月和7—10月、西北海域的2—6月和8—9月、東南海域的2—7月和9—12月以及西南海域的3月—次年1月均適合馬林魚(yú)的捕撈.

關(guān)鍵詞海表溫度；Aqua / MODIS；地理信息系統(tǒng)；遙感；經(jīng)驗(yàn)累積分布頻率分析

中圖分類號(hào)S951.4；S934

文獻(xiàn)標(biāo)志碼A

0 導(dǎo)讀

本文原文為英文，希望感興趣的讀者進(jìn)一步關(guān)注原文.

馬林魚(yú)（Makira sp.）是一種具有很高經(jīng)濟(jì)價(jià)值的魚(yú)類，它在熱帶斯里蘭卡海域中十分常見(jiàn)，但是由于馬林魚(yú)在此海域的分布較為分散，捕撈成本較高.因此預(yù)測(cè)馬林魚(yú)可能出現(xiàn)的海域并且了解其出現(xiàn)的時(shí)間空間分布對(duì)于降低馬林魚(yú)捕撈成本十分重要.海表溫度可影響海洋物種的地理分布，在對(duì)北美西海岸的中上層物種的研究中發(fā)現(xiàn)，物種會(huì)隨著溫度的變化而發(fā)生遷徙.研究表明海水溫度是影響馬林魚(yú)繁殖的一個(gè)十分重要的因素.斯里蘭卡海域由于受到季風(fēng)的影響，水溫存在明顯的季節(jié)變化.西南季風(fēng)從5—9月，東北季風(fēng)從11月—次年3月.在這2個(gè)季節(jié)的過(guò)渡時(shí)期，由于海洋動(dòng)力環(huán)境變化較弱，從而導(dǎo)致海面溫度升高.斯里蘭卡海域海水溫度由于季風(fēng)影響而產(chǎn)生的季節(jié)變化可以從衛(wèi)星遙感圖像中看出.因此本文利用Aqua/MODIS衛(wèi)星遙感水溫?cái)?shù)據(jù)和馬林魚(yú)漁業(yè)捕獲數(shù)據(jù)，研究馬林魚(yú)的分布與海水表面溫度的關(guān)系，并根據(jù)海水溫度確定馬林魚(yú)的分布區(qū)域，盡量減少馬林魚(yú)漁業(yè)捕獲成本，并確定適合馬林魚(yú)的保護(hù)區(qū).

本文通過(guò)分析MODIS衛(wèi)星遙感圖像獲得的海表溫度數(shù)據(jù)以及實(shí)際的馬林魚(yú)捕獲數(shù)據(jù)，并應(yīng)用經(jīng)驗(yàn)累積分布頻率分析方法研究了兩者之間的關(guān)系.同時(shí)為了更好地了解不同海域的馬林魚(yú)的時(shí)空分布，研究海域被分為東北、東南、西北和西南4個(gè)海域.結(jié)果表明馬林魚(yú)在斯里蘭卡海域西部捕獲量最多，且漁獲率隨著海表溫度變化而變化.漁獲率與22～31 ℃海表溫度之間有著顯著的關(guān)系，尤其是與26～31 ℃海表溫度的關(guān)系最為顯著.根據(jù)海表溫度的變化規(guī)律和通過(guò)經(jīng)驗(yàn)累積分布頻率分析得到的合適海表溫度表明，在東北海域的3—5月和7—10月、西北海域的2—6月和8—9月、東南海域的2—7月和9—12月以及西南海域的3月—次年1月均適合馬林魚(yú)的捕撈.

馬林魚(yú)的潛在棲息地可以通過(guò)海表溫度的變化而確定.衛(wèi)星圖像在提取大面積的海面溫度方面提供了數(shù)據(jù)支持，而這一點(diǎn)是無(wú)法通過(guò)樣品分析進(jìn)行實(shí)際分析的.本研究表明，海洋表面溫度為馬林魚(yú)的分布提供了一個(gè)很好的指標(biāo)，可以用于初步檢測(cè)斯里蘭卡海域的潛在馬林魚(yú)漁場(chǎng).

Abstract Marlin （Makira sp.） is one of the famous oceanic fish species in Sri Lankan fisheries.To reduce the risk of redistribution of Marlins in the future with the temperature change，it is important to find out the relationship of Marlins with sea surface temperature.Moreover，by identifying this relationship the cost to find the Marlin abundant places in order to catch them can be reduced.In this study，SST values of the year 2006 from Aqua/MODIS images within the region of latitudes 2.0-13.5°N and longitudes 76.5-88.0°E and the simultaneous Marlin catch data obtained from National Aquatic Resources Research & Development Agency were obtained for the analysis.To describe the relationship between SST and Marlins catch per unit effort （CPUE），the Empirical Cumulative Distribution Frequency （ECDF） analysis was used.The temporal variations of SST and fish availability and the relationship between them were analyzed by plotting the data separately for Northeastern，Southeastern，Northwestern and Southwestern areas.Most of the catches could be seen in the western area off the Sri Lankan waters.The fish CPUE changes could be observed with the changes of the SST.Highest frequency of Marlin catches were within the areas with SST of 27-28 ℃.According to the ECDF analysis，there was a significant relationship between the changes of CPUE and SST values of 22-31 ℃.The maximum abundance could be seen within 26-31 ℃.According to the SST changing pattern and the suitable SST values obtained from ECDF analysis，periods of March to May and July to October in Northeastern area，F(xiàn)ebruary to June and August to September in Northwestern area，F(xiàn)ebruary to July and September to December in Southeastern area，and March to January in Southwestern area are suitable for Marlin fishery.

Key words sea surface temperature；Aqua/MODIS；geographic information system；remote sensing；empirical cumulative distribution frequency analysis

1 Introduction

Marlins are found in tropical Sri Lankan seas and they are one of the highly valuable food fish [1].These fish are oceanic and highly migratory [2].Marlins have been caught for sporting purpose since 1930s in Pacific，Atlantic，etc.Though Marlin fishing is considered as a kind of off-shore sport fishing activity，in Sri Lanka Marlins are caught for food purposes.Usage of long-lines for Marlin fishery has been popular，since it preserves the quality of fish flesh [3].

Wide distribution of Marlins takes long searching time which increases the fuel consumption and fishing cost.So it is important to predict the fishable aggregations of Marlins to reduce the cost of Marlin fishing.To identify the potential fishing grounds，it is essential to identify the temporal and spatial distribution patterns of this fish.

Sea surface temperature （SST） can influence the geographical range of marine species[4].Several studies have been carried out on the migration of pelagic species off the North American west coast （yellowtail，Seriolalalandei；Pacific bonito，Sardachiliensis；Pacific barracuda，Sphyraenaargentea；white seabass，Atractoscionnobilis；and skipjack tuna，Euthynnuspelamis） during years of anomalously warm sea temperatures [5].Meanwhile，the temperature is one of the important parameters that affect Marlin distribution too [6].

Sri Lankan waters experience seasonal temperature changes with the monsoon effects.The southwest monsoon extends from May to September and northeast monsoon from November to March.Between the two monsoons，there are two transition periods known as inter-monsoons when the dynamic conditions become weak and，as a result the sea surface gets heated [3].Remote sensing data can be easily used to analyze these spatial and temporal changes of ocean sea surface temperature.

The purpose of this study is to examine the relationship between Marlin distribution and sea surface temperature in Sri Lankan waters using Aqua/ MODIS remote sensing data and Marlins fishery data.According to their relationship，the potential Marlin abundant habitats can be demarcated to minimize the fishery cost and the conservation zones suitable for Marlins can also be identified.

2 Materials and methods

2.1 Study area

The area within latitudes of 2.0-13.5°N and longitudes of 76.5-88.0°E is the target area of this study.This area covers the Sri Lankan exclusive economic zone.

2.2 Satellite data

Satellite remote sensing data of 2006 Aqua/MODIS images were used to analyze the sea surface temperature （SST）.MODIS （Moderate-resolution Imaging Spectroradiometer） that was launched in 2002 by the United States provides ideal remote sensing images used for fisheries oceanography[7].

2.3 Fishery data

2.4 Processing of AQUA/MODIS images

The daily and monthly AQUA/MODIS level 2 images were downloaded from NASA website，and processed by softwares of SeaDAS，ArcGIS 9.3，and ILWIS 3.3.The images on relatively cloud-free days of area within latitudes 2.0-13.5°N and longitudes 76.5-88.0°E were projected using SeaDAS software to produce sea surface temperature maps.The study area was divided into four major regions for detailed analysis.

3 Results

Figure 2 shows the positions that Marlins are mostly aggregated.The fishery boats target these places to catch Marlins.The empirical distribution frequency analysis has been used to graphically indicate how the cumulative frequency of Marlin catches change with the sea surface temperature （SST） （Fig.3）.Moreover，the histogram in Figure 4 clearly indicates the SSTs which support the aggregation of the Marlins.

According to Figure 3 and Figure 4，it is clear that the ideal temperature range for Marlin is within 26-31 ℃.In order to find the monsoonal seasons that are ideal for the Marlin catch， seasonal changes of the sea surface temperature has been mapped in Figure 5.

The following Figure 6 shows the maps prepared to identify the locations and the months that are supporting the Marlin distribution in surface waters according to the variations of the SST.They can be used to identify the potential fishing grounds of Marlins as well as to demarcate the bio conservative hotspots.

4 Discussion

Data presented in this study support providing evidence of preferred temperature ranges for Marlins in order to manage the Marlin fisheries better in the future.

4.1 Marlin spatial variations with temperature

The distribution of Marline organisms can be attributed and analyzed according to the various environment conditions which we identify as “hydro-biological variables”. Development and sustainability of Marline related industries requires better understanding of these factors[10].Blue Marlins are highly distributed in tropics among all billfish[11] and are the most common billfish in the catches of longline fisheries in the tropical Pacific Ocean[12].The distributions of blue Marlin density vary seasonally and annually.According to the results of statistical study done with the remote sensing and fish catch data in this study，it is clearly proven that there is a strong relationship between sea surface temperature and Marlin fish distribution.The probable SST when Marlins could be seen is within 26-31 ℃.

According to the above results，the highest CPUE （15.00±0.00） in Northeastern area is available in waters with 27.26±4.44 ℃ SST.The highest mean CPUE in Southeastern area could be seen in months such as July and November with 28.91±0.83 ℃ and 27.46±1.27 ℃ SST，respectively.In Southwestern area the highest CPUE is found in October with a mean SST of 28.02±4.42 ℃.Marlins have been frequently caught from waters with 27.00-28.00 ℃.Hence，with the temperature change occurred with the ENSO events，Marlins shifting can be seen to the preferable temperatures[13].

4.2 SST seasonal variations

According to the mean SST values calculated for each area separately，the suitable temperatures for Marlins fishing are distributed as follows：March to May，July to October and December in Northeastern area，F(xiàn)ebruary to June and August to September in Northwestern area，F(xiàn)ebruary to July and September to December in Southeastern area，and March to January in Southwestern area.

According to the seasonal variations of the SST，the suitable habitats for the Marlins are mostly in Southwestern region in Northeast monsoon，Northwestern region in the first inter-monsoon，Northeastern region in Southwest monsoon，and again in Southern area in the second inter-monsoon.

The study of Goodyear et al.[6] clearly shows that the habitat utilization of blue Marlins depends on temperature-depth rather than the feeding behavior.They have observed blue Marlin as daylight and sight feeders that regularly descend into the water column to feed or possibly view forage above[6].Furthermore，according to this study the thermocline is a particular factor influencing Marlins.Therefore further studies are needed to investigate the relationship of the Marlin behavioral patterns with the thermocline distribution in ocean around Sri Lanka.

Moreover，SST directly affects the upwelling which brings nutrients to surface waters[14-15] and leads to high primary production attracting food items of Marlins to the surface waters.Hence，SST directly as well as indirectly affects the gathering of Marlins.

4.3 Implications for management

The spatial distribution of the species has to be considered when making effective strategies for managing marine ecosystems.Time-area closures may be the best approach to manage the fisheries and protect species such as bill fish that have pelagic larvae and highly migratory adults[16]，because limiting the fishing effort in large areas is not practical[17].Demarcating reserves for Marlin like species should be based on habitat type，ocean conditions，migratory patterns，and prey distributions in order to protect their spawning and nursery grounds.

Global phenomena such as El Nino make the changes in ocean environmental conditions which may affect large-scale shifts in distribution of migratory species.Therefore，identification of the distribution of migratory fish should be combined with the interannual variability.The relationships between SST and Marlin abundance derived in this study can be used to predict the spatial distribution of this fish and hence provide a basis to plan the closed areas.

Moreover，there may be several other factors affect the distribution of Marlins.If the relationship of Marlins with those other oceanographic conditions be identified，the prediction of the potential Marlin fishing grounds will be much easier and cost effective.The results of these studies can be used not only to identify the potential fishing grounds but also for the conservation activities.Most suitable habitats for the Marlins can be identified to demarcate the no-fishing zones and special areas for conservation （SACs） to escalate the abundance of Marlins.

5 Conclusion

The potential fishing grounds for Marlin may well correspond to the dynamics of temperature which could be identified synoptically by sea surface temperature.Satellite images provide a distinctive support in extracting the sea surface temperatures in large areas which cannot be practically analyzed by sample analysis.This study suggests that the preferred range of the sea surface temperature provides a good indicator of initially detecting the potential fishing grounds for Marlin in Sri Lankan ocean.According to the ECDF analysis，there was a significant relationship between the changes of CPUE and SST values of 22-31 ℃.The maximum abundance could be seen within 26-31 ℃.The behavior of Marlin species may change with their maturity stages too.A detailed analysis has to be carried out to identify the changes of Marlin behavior with their maturity stages and considering other oceanographic conditions.

References

[1] Wright C P B.Five of the best blue marlin spots[EB/OL].[2013-09-12].www.marlinmag.com/travel/international/five of the best blue marlin

[2] Prince E D，Ortiz M，Venizelos A，et al.In-water tagging techniques developed by the cooperative tagging center for large，highly migratory species[J].American Fisheries Society Symposium，2002，30：155-171

[3] Rajapaksha J K，Nishida T，Samarakoon L.Environmental preferences of yellowfin tuna （Thunnus albacores） in the northeast Indian Ocean：an application of remote sensing data to long line catches[J].2010

[4] Laurs R M，F(xiàn)ielder P C，Montgomery D R.Albacore tuna catch distributions relative to environmental features observed from satellites[J].Deep-Sea Res，1984，31（9）：1085-1099

[5] Hubbs C L，Schultz L P.The northward occurrence of southern forms of marine life along the Pacific coast in 1926[J].Calif Fish Game，1929，15（3）：234-241

[6] Goodyear C P，Luo J，Prince E D，et al.Temperature-depth habitat utilization of blue marlin monitored with PSAT tags in the context of simulation modeling of pelagic longline CPUE[J].Science of Social Psychology，2006，59（1）：224-237

[7] Yapa K K A S.Upwelling phenomena in the southern coastal waters of Sri Lanka during southwest monsoon period as seen from MODIS[J].Sri Lankan Journal of Physics，2011，10：15-23

[8] Andrade H A，Garcia C A E.Skipjack tuna fishery in relation to sea surface temperature off the southern Brazilian coast[J].Fisheries Oceanography，1999，8（4）：245-254

[9] Zainuddin M，Saitoh K，Saitoh S.Albacore tuna in relation to oceanographic condition in the northwestern North Pacific using remotely sensed satellite data[J].Journal of Fisheries Oceanography，2008，17（2）：61-73

[10] Uda M.A consideration on the long years trend of the fisheries fluctuation in relation to sea conditions[J].Nippon Suisan Gakkaishi，1957，23：368-372

[11] Molony B. Summary of the biology，ecology and stock status of billfishes in the WCPFC，with a review of major variables influencing longline fishery performance[C]∥The 1st Meeting of the Scientific Committee of the Western and Central Pacific Fisheries Commission，2005：8-19

[12] Hinton M G.Status of blue marlin in the Pacific Ocean，status of Tuna and Billsh stocks in 1999[R].Inter-American Tropical Tuna Commission，La Jolla，CA，USA，Stock Assessment Report 1，2001：284-319

[13] Ward P，Myers R A.Inferring the depth distribution of catchability for pelagic fishes and correcting for variations in the depth of longline fishing gear[J].Journal Canadien Des Sciences Halieutiques Et Aquatiques，2005，62（62）：1130-1142

[14] Yu J，Tang D L，Chen G B，et al.The positive effects of typhoons on the fish CPUE in the South China Sea[J].Continental Shelf Research，2014，84（4）：1-12

[15] Yu J，Tang D L，Li Y Z，et al.Increase in fish abundance during two typhoons in the South China Sea[J].Advances in Space Research，2013，51（9）：1734-1749

[16] Goodyear C P.An analysis of the possible utility of time-area closures to minimize billfish bycatch by U.S. pelagic longlines[J].Fishery Bulletin-National Oceanic and Atmospheric Administration，1999，97（2）：243-255

[17] Jensen O P，Ortega-Garcia S，Martell S J D，et al.Local management of a “highly migratory species”：the effects of long-line closures and recreational catch-and-release for Baja California striped marlin fisheries[J].Progress in Oceanography，2010，86（1/2）：176-186