仔稚魚短期群聚組成變動及採樣設計相關性研究─以淡水河口為例

Abstract

本研究有兩目的，其一為透過密集採樣，探討仔稚魚短期群聚變化及樣本是否能代表當季組成(採樣代表性)；其二為利用採獲資料為母群體，進行二次取樣(subsampling)，探討仔稚魚物種豐富度、密度等參數與努力量之取樣相關性。於2012年6-7月，利用網口直徑90 cm及網目大小330μm之圓口仔稚魚網，在北台灣淡水河口海域共3個測站，設計相隔2~43天之5個採樣日別，並且單一日別內設計相隔2~12小時橫跨日夜之4個採樣時間點，記錄當時潮汐並進行仔稚魚採集，以生命條碼輔佐外部形態進行物種鑑定，並以ANOSIM及SIMPER探討仔稚魚在測站(Station)、採樣日別(Sampling date)、採樣時間點(Sampling hour)、日夜(Day/ night)及潮汐(Tidal phase)等分群因子(factors)下之群聚組成差異及差異來源物種。本研究共採獲5302尾仔稚魚，依形態可鑑定出189個形態型(morpho-type)，分屬49科42屬及23種，經由生命條碼可鑑定出131個分類單元(taxa)，分屬52科91屬及78種，顯示生命條碼有助於提高物種鑑定能力。組成方面，測站間因間距短及底質相同造成仔稚魚組成無顯著差異，可合併分析。仔稚魚組成在採樣日別間有差異，因各魚種入添日別及豐度不同所造成。仔稚魚在採樣時間點及日夜間具有組成差異，但僅7月9日日間群內採樣時間點間有組成差異，其他採樣日別以日夜分群為主，日間主要特徵類群為沙鮻科sp. 2 (Sillago sp. 2)、吉打副葉鰺(Alepes djedaba)及雙邊魚屬(Ambassius sp.)之仔稚魚，夜間主要為鯷科sp. 1 (Engraulidae sp. 1)、仰口鰏(Secutor ruconius)及鯷科sp. 2 (Engraulidae sp. 2)。漲退潮間僅7月9日仔稚魚有組成顯著差異，但主要係因漲潮時各物種豐度顯著提高而造成統計上差異。二次取樣模擬結果，物種豐富度百分比隨取樣次數增加呈漸近曲線，且仔稚魚平均豐度愈高，所需取樣次數愈低。 綜合所得之結果建議，未來採樣設計上，如需了解當地完整之仔稚魚種組成，則應優先涵蓋日夜時段，如需進行環評監測則應依人力物力增加日別及小時採樣頻度，降低測站密度，並於漲潮時採集，且依設定物種數採獲百分比進行最適努力量採樣，如此應能以較低的成本獲得最大的效益或樣本的代表性。

The objectives of this research were: (1) to evaluate short-term larval fish variations and discuss the sampling representatives, through high frequency sampling; (2) to discuss sampling relations between species parameters (species richness and density) and the sampling cost. Larval fish were collected from 3 stations in costal waters of Tamsui river estuary, Northern Taiwan, from June to July, 2012, depending on different sampling frequencies (2~43 days intervals, and 2~12 hours intervals crossing day and night) by using the round mouth ichthyoplankton net (90 cm in diameter and 330 μm in mesh size). Larval fish were identified by morphology and mitochondrial DNA COI sequence barcoding. We evaluated larval fish assemblage variations under set factors including station, sampling date, sampling hour, day/ night and tidal phase, and taxa that contributed to assemblage variations most by using ANOSIM and SIMPER analysis, respectively. Total of 5302 larval fish were collected. There were 189 taxa belonging to 49 famalies, 42 genera and 23 species by morphological identifications, but there were 131 taxa belonging to 52 families, 91 gerera and 78 species with the aid of DNA barcoding, indicating that larval fish identification ability could elevate by using DNA barcoding. No significant differences in larval fish assemblages were found among stations, because of the short distance between each and the same substrate. There were significant larval fish assemblage differences among sampling dates, mainly resulted from different recruitment patterns of larval fish species. Significant differences in larval fish assemblages were found in different sampling hours and day/ night factors, but chiefly resulted from the latter, comparisons between sampling hours belonging to daytime or nighttime. The main typifying species in daytime group were Sillago sp. 2, Alepes djedaba and Ambassius sp., and the main typifying species in nighttime group were Engraulidae sp. 1, Secutor ruconius and Engraulidae sp. 2. There was significant difference in larval fish assemblages between flood tide and ebb tide only on 9, July, but the statistical significance mainly came from high larval fish density for each species during flood tide. The subsampling results showed that the cumulative curve of species richness percentage reached an asymptote with the elevated sampling number. Moreover, the higher the larval fish density was, the lower the number of hauls needed. The results suggested that, to elevate sampling representatives and correctly reflect larval fish assemblages in the Tamsui river estuary with the most economical sampling effort invested in the future larval fish sampling designs, we must firstly cover the day-night interval, and secondarily increase daily- and hourly- sampling frequencies, decrease the number of sampling stations as well as sample during flood tide. Besides, it is better that we set a target species richness percentage in each single sampling time according to the sampling effort we could afford.