劉海霞,李素霞,劉廣龍,黃凱旋,楊 斌,蘇靜君,王 杰,張晉誼,侯景耀

基于復(fù)合指紋圖譜和貝葉斯模型的茅尾海懸浮顆粒物源解析

劉海霞1,2,3,李素霞2,3*,劉廣龍2,4**,黃凱旋5,楊 斌3,蘇靜君5,王 杰4,張晉誼2,侯景耀2

(1.桂林理工大學(xué)環(huán)境科學(xué)與工程學(xué)院,廣西 桂林 541006；2.北部灣大學(xué)資源與環(huán)境學(xué)院,廣西 欽州 535011；3.北部灣大學(xué)，廣西北部灣海洋災(zāi)害研究重點(diǎn)實(shí)驗(yàn)室,廣西 欽州 535011；4.華中農(nóng)業(yè)大學(xué)資源與環(huán)境學(xué)院,湖北 武漢 430070；5.中國科學(xué)院生態(tài)環(huán)境研究中心,城市與區(qū)域生態(tài)國家重點(diǎn)實(shí)驗(yàn)室,北京 100085)

為明確茅尾海中懸浮顆粒物的來源,采集了茅尾海流域紅樹林土壤、堤岸土、河口顆粒物、茅尾海沉積物以及灣外顆粒物等懸浮顆粒物潛在源樣品.基于多元統(tǒng)計(jì)復(fù)合指紋圖譜方法,篩選出最佳指紋因子組合,進(jìn)而通過貝葉斯混合模型得出五種潛在源對茅尾海懸浮顆粒物的貢獻(xiàn)率.結(jié)果表明:Mg、Al、Mn、Pb、Fe五種指紋元素可作為最佳指紋因子組合,累計(jì)判別正確率為78%.貝葉斯混合模型結(jié)果顯示,茅尾海懸浮顆粒物主要來源于河口和灣外輸送,貢獻(xiàn)率最高達(dá)到58.9%和68.6%.其中,靠近河口區(qū)域主要受河流匯入影響,其貢獻(xiàn)率達(dá)到42.2%～58.9%;靠近灣外區(qū)域則以灣外顆粒物貢獻(xiàn)為主,貢獻(xiàn)率達(dá)到44.9%～68.6%.各點(diǎn)位的沉積物貢獻(xiàn)率均較低,紅樹林土壤和堤岸土的貢獻(xiàn)率都在10%左右.總的來說,由河口匯入和潮汐作用帶入的顆粒物是茅尾海懸浮顆粒物的主要來源.

貝葉斯模型；復(fù)合指紋圖譜；顆粒物；源解析

茅尾海位于廣西壯族自治區(qū)欽州市南部海域,是我國最大的近江牡蠣天然采苗區(qū).近年來,由于茅尾海海洋工程的建設(shè)、養(yǎng)殖業(yè)的過度發(fā)展,茅尾海的污染日益加重[1].據(jù)《廣西海洋環(huán)境狀況公報(bào)》[2]顯示,2017年經(jīng)由欽江、茅嶺江入海的化學(xué)需氧量分別為32960t和39000t,占污染物總量的近95%,是最主要的入海污染物,其次為氮磷營養(yǎng)鹽.研究表明:化學(xué)需氧量、無機(jī)氮以及磷酸鹽已成為影響欽州海域水質(zhì)狀況的主要因素[3].2003～2010年的水質(zhì)監(jiān)測數(shù)據(jù)表明,茅尾海的水質(zhì)從貧營養(yǎng)往中度富營養(yǎng)和重富營養(yǎng)化發(fā)展,已屬于輕度污染[4].盡管2008年在國務(wù)院批復(fù)的《廣西北部灣經(jīng)濟(jì)區(qū)發(fā)展規(guī)劃》[5]的要求下開展了污染整治和海域生態(tài)功能恢復(fù)行動,但是茅尾海灣的水體富營養(yǎng)化程度并未發(fā)生明顯降低,對海域的持續(xù)健康發(fā)展造成了極大的潛在隱患.

水體中懸浮顆粒物作為營養(yǎng)鹽的重要載體,其遷移轉(zhuǎn)化對水體營養(yǎng)鹽的循環(huán)過程及其生態(tài)效應(yīng)具有重要影響[6].已有研究表明,水體中懸浮的顆粒物可經(jīng)生物礦化或光化學(xué)分解釋放出溶解態(tài)無機(jī)鹽,進(jìn)而為浮游植物的生長提供營養(yǎng)[7-9].造成水體富營養(yǎng)化的顆粒物正磷酸鹽是顆粒物生物可利用性磷的重要來源[10].水體中的顆粒物也是重金屬的重要載體[11].在水體懸浮顆粒物遷移過程中的污染物吸附、貯存和解吸等環(huán)境行為對水生態(tài)系統(tǒng)造成潛在的巨大危害,由此引發(fā)的水體富營養(yǎng)化以及人體健康風(fēng)險(xiǎn)等問題已成為流域水環(huán)境研究的熱點(diǎn)[12-13].追溯懸浮顆粒物的來源是控制非點(diǎn)源污染以及更合理實(shí)施水土保持方針的重要前提.因此,追溯茅尾海顆粒物來源對控制水體中懸浮顆粒物濃度繼而提升茅尾海水質(zhì)具有重要意義.但由于海灣水體中懸浮顆粒物來源廣泛且認(rèn)知不清,致使其靶向控制舉措的提出缺乏科學(xué)依據(jù).

復(fù)合指紋技術(shù)作為研究泥沙及懸浮顆粒物來源的有效手段能反映土壤侵蝕、泥沙輸移和沉積特征.該技術(shù)基于侵蝕源地的土壤特征篩選具有診斷能力的復(fù)合指紋因子,利用質(zhì)量平衡模型建立源地土壤與懸浮顆粒物的定量關(guān)系,最終獲得各源地的相對貢獻(xiàn)率[14].近年來,復(fù)合指紋識別技術(shù)在流域泥沙溯源示蹤方面已經(jīng)得到了廣泛的應(yīng)用,尤其在英國[15-17]、美國[18-19]和澳大利亞[20-22]等發(fā)達(dá)國家和地區(qū).早在20世紀(jì)90年代,Collins等[23]運(yùn)用復(fù)合指紋圖譜結(jié)合多元混合模型,評估了英國Dart流域和Plynlimon試驗(yàn)小流域的泥沙來源.2003年,Krause等[24]篩選出137Cs和部分重金屬指標(biāo)作為復(fù)合指紋因子,在前人研究的基礎(chǔ)上,創(chuàng)新性地運(yùn)用FR2000混合模型計(jì)算了各泥沙源的貢獻(xiàn)比例及不確定度. 2009年,Poleto等[25]研究發(fā)現(xiàn)隨著復(fù)合指紋因子數(shù)量的增加,泥沙來源的不確定性水平有降低趨勢. 2013年,Gellis等[26]選擇樣品的營養(yǎng)元素(TOC、TN、TP)和碳氮穩(wěn)定同位素(δ13C、δ15N)以及金屬元素作為復(fù)合指紋因子,發(fā)現(xiàn)河岸帶的泥沙貢獻(xiàn)率最大,占比約53%,并且融雪期的河岸帶侵蝕速率最大.

國內(nèi)學(xué)者在長江流域[27-30]、黃土高原[31-33]、西南地區(qū)[34-35]、東北黑土區(qū)[36-37]以及福建紅壤和花崗巖崩崗區(qū)[38-40]等地區(qū)也采用指紋識別法開展泥沙溯源研究,在指紋因子的擴(kuò)選、復(fù)合指紋的應(yīng)用和溯源模型的選擇等方面均取得了良好的進(jìn)展.關(guān)于貝葉斯混合模型已廣泛地應(yīng)用于穩(wěn)定同位素以及植物水分來源等研究中,但卻鮮有運(yùn)用復(fù)合指紋識別技術(shù)與貝葉斯混合模型結(jié)合進(jìn)行懸浮顆粒物溯源的報(bào)道.據(jù)此,結(jié)合茅尾海這種特殊的地理環(huán)境,本研究選取茅尾海流域及其入海河口作為研究對象,基于復(fù)合指紋圖譜和貝葉斯溯源模型對茅尾海中懸浮顆粒物進(jìn)行來源解析,細(xì)化溯源結(jié)果,清楚茅尾海區(qū)域的主要懸浮顆粒物來源,從而能夠更加精準(zhǔn)地控制源地顆粒物流失,以便于水生態(tài)環(huán)境管理.以期為茅尾海水質(zhì)提升策略的制定提供科學(xué)依據(jù).

1 材料與方法

1.1 研究區(qū)域概況

茅尾海位于欽州灣北部,所轄海域面積約135km2,是一個(gè)典型內(nèi)寬口窄的橢圓形半封閉內(nèi)灣[41],該區(qū)域?qū)賮啛釒ШＱ笮约撅L(fēng)氣候,年均降雨量2104.2mm,高溫多雨、夏長冬短[42],潮汐為不規(guī)則全日潮,平均潮差為2.51m[43].欽江、大欖江、茅嶺江是流入茅尾海的主要河流[2],受欽江、大欖江、茅嶺江等主要入海河流的影響形成了獨(dú)特的河口—海灣—濕地多生態(tài)系統(tǒng)[44].紅樹林作為該海岸帶重要的生態(tài)系統(tǒng)之一,與堤岸土不同的是其具有維護(hù)生物多樣性、促淤固灘、凈化水質(zhì)等功能的同時(shí)對全球環(huán)境和氣候具有重要指示意義.茅尾海也是近海牡蠣的全球種質(zhì)資源保留地和我國最重要的養(yǎng)殖區(qū)與采苗區(qū)[44],是“中國大型牡蠣之鄉(xiāng)”,盛產(chǎn)大型牡蠣、青蟹、對蝦、石斑魚等,其主要養(yǎng)殖區(qū)域位于茅尾海北部區(qū)域.盡管欽州市政府發(fā)布了《茅尾海綜合整治規(guī)劃》,進(jìn)行了多項(xiàng)綜合修復(fù)和修復(fù)工程,但茅尾海的健康發(fā)展仍受到環(huán)境污染脅迫.

1.2 樣品采集與分析

1.2.1 潛在懸浮顆粒物源地劃分及樣點(diǎn)布設(shè) 收集研究區(qū)域及其周邊環(huán)境等資料,結(jié)合研究區(qū)域及周圍土地利用類型實(shí)際情況,將物源類型劃分為紅樹林土壤(S2)、堤岸土(S3)、河口顆粒物(S4)、茅尾海沉積物(S5)以及灣外顆粒物(S6)五種,沉積物樣點(diǎn)對應(yīng)點(diǎn)位為茅尾海水柱的樣點(diǎn)(C1～C14)再根據(jù)劃分好的物源類型布設(shè)采樣點(diǎn),并保證每個(gè)種類至少采集10個(gè)以上樣品以減小組內(nèi)變異.采樣點(diǎn)位置見圖1.

處理來自潛在源頭的懸浮顆粒物和土壤.利用采水器在樣點(diǎn)附近共采集60L水樣,通過高速冷凍離心獲得水樣中的懸浮顆粒物.利用懸移質(zhì)采泥器在樣點(diǎn)附近多點(diǎn)采集表層(0～2.5cm)沉積物樣品,樣品采集后放入透氣性好的聚乙烯袋中密封保存,直至樣品分析.土壤樣品利用洛陽鏟采集表層土壤(0～5cm)的樣品,放入透氣性好的聚乙烯袋中密封保存,直至樣品分析.將采集的物源土壤與懸浮顆粒物樣品經(jīng)過風(fēng)干和烘干處理后放入研缽中研磨,過100目篩.稱取0.1g(精確到±0.0001g)樣品,加入9mL濃硝酸和3mL氫氟酸,微波消解.使用電感耦合等離子體質(zhì)譜儀(ICP-MS)測定樣品中的地球化學(xué)元素,共13種,即Al、As、Ca、Cd、Cr、Cu、Fe、K、Mg、Mn、Ni、Pb、Zn.

圖1 采樣點(diǎn)位置

1.2.2 保守性檢驗(yàn) 一系列因素可以影響自然環(huán)境中示蹤劑的保守性,包括氧化還原電位、溫度、選擇性顆粒遷移、吸附/解吸或沉淀/溶解反應(yīng)等.因?yàn)檫@些因素,懸浮顆粒物在遷移過程中可能會發(fā)生指紋因子的改變,因此需要通過統(tǒng)計(jì)檢驗(yàn)的手段排除這些可能會影響最終溯源結(jié)果的干擾因子.為評估示蹤劑的保守性,使用范圍檢驗(yàn)來識別非保守示蹤劑,將懸浮顆粒物樣品中的示蹤劑濃度與源樣品最小和最大值之間的范圍進(jìn)行比較.

1.2.3 指紋因子篩選 采用復(fù)合指紋圖譜法進(jìn)行懸浮顆粒物來源示蹤,需篩選一組具有統(tǒng)計(jì)意義上最佳判別能力的復(fù)合指紋識別因子[29].采用了兩種統(tǒng)計(jì)方法篩選能夠區(qū)分源物質(zhì)的復(fù)合指紋組合:

(1)通過Kruskal-Wallis H檢驗(yàn)(K-W H),篩選各潛在物源之間差異顯著的指紋.

(2)利用K-W H檢驗(yàn)和判別函數(shù)分析(DFA)的組合,篩選一組判別能力最強(qiáng)的組合指紋因子.

利用SPSS計(jì)算K-W統(tǒng)計(jì)量H及其概率值進(jìn)行檢驗(yàn),并與顯著性水平(一般取0.05)比較.若<,表明該屬性在組間具有顯著性差異,可以作為潛在的指紋因子,進(jìn)入下一步檢驗(yàn).

通過K-W H檢驗(yàn)在潛在顆粒物來源之間顯示出統(tǒng)計(jì)學(xué)顯著性差異的示蹤劑進(jìn)一步用于判別分析.DFA的理論基礎(chǔ)是提供一組可以區(qū)分源物質(zhì)組別的權(quán)重.然后,這些權(quán)重可以被用于區(qū)分源物質(zhì),提供它們屬于每個(gè)可能的源組的概率.在每個(gè)判別步數(shù)中,選擇的顆粒物示蹤劑會最大程度地降低整體Wilks’lambda.當(dāng)所有樣本均已正確分類時(shí),或在給定步數(shù)中可用于包含的其余所有屬性均不具有改善源辨別力的能力時(shí),逐步判別過程將停止.

1.2.4 模型的構(gòu)建 根據(jù)采集顆粒物的指紋因子濃度[40],利用構(gòu)建多元混合模型的手段(例如IsoSource, MixSIR,SIAR,MixSIAR)來進(jìn)行溯源解析[45].其中,貝葉斯混合模型(Bayesian mixing models)包含3種模型(分別是MixSIR,SIAR和MixSIAR).對于不確定性的定量化分析,貝葉斯混合模型由于其在利用先驗(yàn)信息以及捕獲不確定性來源方面具有更為明顯的優(yōu)勢[46],因而在生態(tài)學(xué)中應(yīng)用較為廣泛,經(jīng)常應(yīng)用于流域泥沙溯源[47]、植物水分溯源[45-48]、污染物溯源[49-52]和食物網(wǎng)營養(yǎng)溯源[53-54]等領(lǐng)域的研究.貝葉斯混合模型通過引入貝葉斯統(tǒng)計(jì)理論,并考慮源數(shù)值的不確定性、分類協(xié)變量和連續(xù)協(xié)變量以及先驗(yàn)信息等,對簡單的線性混合模型進(jìn)行了改進(jìn).與MixSIR和SIAR等模型相比較,MixSIAR模型結(jié)合了多種來源,并基于先驗(yàn)信息的不確定性、連續(xù)協(xié)變量和乘法誤差結(jié)構(gòu),提高了“源”對“匯”貢獻(xiàn)率的計(jì)算準(zhǔn)確性[45].

根據(jù)貝葉斯理論,所有f(每個(gè)源對目標(biāo)懸浮顆粒物樣本的貢獻(xiàn)率)的后驗(yàn)概率分布與先驗(yàn)概率分布成正比,再乘以似然函數(shù),然后除以它們的總和,即:

式中:(data|f)是給定數(shù)據(jù)f的似然函數(shù);(f)表示基于先驗(yàn)信息的先驗(yàn)概率;f是個(gè)通過Dirichlet分布隨機(jī)生成的向量,表示顆粒物源貢獻(xiàn)比例.似然函數(shù)的計(jì)算公式如下:

式中:μ、σ分別為根據(jù)隨機(jī)抽取的f計(jì)算獲得的懸浮顆粒物樣品中第個(gè)指紋因子的均值和標(biāo)準(zhǔn)差;X代表第個(gè)顆粒物樣本的第個(gè)示蹤劑;為顆粒物源的個(gè)數(shù).其中,μ和σ的計(jì)算公式如下:

式中:mSourcei表示第個(gè)顆粒物來源的第個(gè)顆粒物示蹤劑的均值;2Sourcei表示第個(gè)顆粒物來源的第個(gè)顆粒物示蹤劑的方差.

2 結(jié)果與分析

2.1 最佳指紋因子篩選

2.1.1 示蹤劑的保守性檢驗(yàn) 表1比較了懸浮顆粒物源中示蹤劑的濃度,標(biāo)準(zhǔn)范圍檢驗(yàn)的結(jié)果和懸浮顆粒物示蹤劑濃度平均值與顆粒物源地樣點(diǎn)的平均值比較結(jié)果表明,所有指紋因子均保守.

表1 五種物源示蹤劑濃度平均值(ave)、最小值(min)、最大值(max)和標(biāo)準(zhǔn)偏差(SD)

2.1.2 K-W H檢驗(yàn) 根據(jù)分析步驟,將顆粒物源數(shù)據(jù)導(dǎo)入到SPSS 24中,利用Kruskal-Wallis非參數(shù)檢驗(yàn)對剔除了異常值的顆粒物源地?cái)?shù)據(jù)進(jìn)行檢驗(yàn),計(jì)算K-W統(tǒng)計(jì)量及其概率值(表2).根據(jù)計(jì)算結(jié)果對各溯源點(diǎn)對應(yīng)的源地土壤中13種待篩選的指紋因子進(jìn)行初步篩選,排除潛在源地種差異不顯著的因子,并剔除其中的非保守的指紋因子,通過檢驗(yàn)的指紋因子可進(jìn)入下一步的多元判別分析中.

表2 Kruskal-Wallis H檢驗(yàn)結(jié)果

Kruskal-Wallis H檢驗(yàn)的結(jié)果表明,絕大多數(shù)指紋因子組間差異顯著,僅有1個(gè)指紋因子As的值>0.05,未通過檢驗(yàn),其余十二種都可作為初步篩選的指紋因子,并進(jìn)入下一步的多元判別分析中.

2.1.3 DFA檢驗(yàn) 將Kruskal-Wallis H檢驗(yàn)篩選出的不同源地間差異顯著的指紋因子,進(jìn)入多元判別分析(DFA)[28],找出最佳復(fù)合指紋因子.在SPSS軟件中運(yùn)用逐步判別分析法,計(jì)算每個(gè)步數(shù)的Wilks’ lambda與指紋因子辨別正確率(表3).若Sig<0.05,則對應(yīng)的判別函數(shù)具有統(tǒng)計(jì)學(xué)意義,判別結(jié)果可靠.下中的Sig值均<0.05,說明本研究中建立的各判別函數(shù)均具有統(tǒng)計(jì)學(xué)意義,可用于懸浮顆粒物源地的辨別.

表3 DFA檢驗(yàn)結(jié)果

在上一步篩選后得到的12個(gè)指紋因子中,共有5個(gè)指紋因子(Mg、Al、Mn、Pb、Fe)具有判別能力,其累積指紋因子辨別正確率分別為48.8%、73.2%、63.4%、73.2%、78.0%,說明該組合能夠較好地判別懸浮顆粒物源地,可入選最佳指紋因子組合.其中的Mg和Al因子的單指紋因子辨別率最高,為48.8%和43.9%;其次是Fe、Mn和Pb.根據(jù)Carter等[55]的研究成果,累計(jì)指紋因子辨別正確率在70%以上,說明判別效果較好.

2.2 來源解析

各溯源地的懸浮顆粒物源貢獻(xiàn)比率平均值和標(biāo)準(zhǔn)差見表4.在摘要統(tǒng)計(jì)表中,為了能夠更加直觀地比較各懸浮顆粒物源的貢獻(xiàn)率大小,將其平均值作為源貢獻(xiàn)比率,制成橫向百分比條形圖(圖2),其中C0數(shù)據(jù)為各溯源樣點(diǎn)指紋因子平均值.

由圖2和表4可以看出,總體而言,灣外顆粒物和河口顆粒物是對懸浮顆粒物貢獻(xiàn)率最高的來源,C1和C3的沉積物貢獻(xiàn)率次之,這兩個(gè)點(diǎn)位比較靠近堤岸,周圍有濱海公園,受人為環(huán)境影響較大,造成沉積物貢獻(xiàn)率達(dá)到30.9%和22.1%,而其他點(diǎn)位沉積物的貢獻(xiàn)率最低僅為5%,各點(diǎn)位均值的貢獻(xiàn)率僅7.7%,是各類源中最低的.紅樹林土壤和堤岸土的貢獻(xiàn)率在各點(diǎn)位的貢獻(xiàn)率都較低,紅樹林在各點(diǎn)位的貢獻(xiàn)率最低為C8的5.7%,最高為C14的11.6%,各點(diǎn)位的含量平均值計(jì)算出來的貢獻(xiàn)率也僅有10%,堤岸土在各點(diǎn)位的貢獻(xiàn)率最低為C8的6.1%,最高為C1的13.4%,各點(diǎn)位均值得到的貢獻(xiàn)率為10.5%,相對于其他來源紅樹林和堤岸土的影響較小.河口顆粒物的貢獻(xiàn)率最高達(dá)到了58.9%,該點(diǎn)位最靠近茅嶺江匯入的河口,說明較于沉積物來說,該點(diǎn)顆粒物的遷移影響更大.灣外顆粒物對各點(diǎn)位的貢獻(xiàn)率最高達(dá)到了68.6%,其次是53.4%,這兩個(gè)點(diǎn)位均是靠近北部灣,由潮汐帶入的顆粒物給茅尾海帶來較大的影響.

各潛在源的貢獻(xiàn)表現(xiàn)出來的差異明顯,紅樹林作為海岸帶重要生態(tài)系統(tǒng),因其發(fā)達(dá)的根系起到了促淤固灘和凈化水質(zhì)的功能,使得紅樹林對茅尾海顆粒物的貢獻(xiàn)相對較低;堤岸土對茅尾海顆粒物的貢獻(xiàn)主要是通過地表徑流,由于茅尾海周圍有一條護(hù)岸帶,攔截了部分由堤岸帶入的顆粒物,也因距離相對較遠(yuǎn),運(yùn)輸能力不如懸浮顆粒物;另外茅尾海作為北部灣的內(nèi)海,內(nèi)寬口窄的形狀造成風(fēng)浪很小,這極大程度上降低了沉積物的再懸浮作用,從而導(dǎo)致沉積物對顆粒物的影響相對較低;養(yǎng)殖區(qū)主要集中在河流輸入影響區(qū),養(yǎng)殖過程中產(chǎn)生的廢棄物直接影響了茅尾海顆粒物的產(chǎn)生,造成河口的顆粒物對河流輸入影響區(qū)的貢獻(xiàn)較大;距離養(yǎng)殖區(qū)較遠(yuǎn)的茅尾海南部主要受到潮汐的影響,決定了灣外顆粒物對潮汐影響區(qū)的貢獻(xiàn).因此,需要特別針對河流采取措施減少河流輸送帶來的顆粒物影響.

表4 各溯源地的泥沙源貢獻(xiàn)比率平均值和標(biāo)準(zhǔn)差

注:*括號中的值表示不確定性范圍(90%置信度范圍:5%～95%).

圖2 各懸浮顆粒物源貢獻(xiàn)百分比

3 結(jié)論

3.1 利用SPSS進(jìn)行指紋因子篩選得出Mg、Al、Mn、Pb、Fe五種重金屬組合可作為最佳指紋因子,總判別正確率達(dá)到78%,能較好的分辨各潛在來源地.

3.2 貝葉斯模型結(jié)果表明,茅尾海懸浮顆粒物主要來源于河口顆粒物和灣外顆粒物,貢獻(xiàn)率最高達(dá)到58.9%和68.6%,說明水體中的懸浮顆粒物受到的水流擾動作用較強(qiáng);距離灣外近的點(diǎn)位灣外顆粒物貢獻(xiàn)較多,距離河口近的點(diǎn)位河口顆粒物貢獻(xiàn)較多.而沉積物、紅樹林土壤和堤岸土對各點(diǎn)位的貢獻(xiàn)率都相對較小.

3.3 鑒于茅尾海主要水產(chǎn)養(yǎng)殖區(qū)位于河流匯入影響區(qū),為了有效提升茅尾海水產(chǎn)養(yǎng)殖和水生態(tài)系統(tǒng)的健康發(fā)展,應(yīng)采取措施有效控制河流的顆粒物輸送.

[1] 吳文成,任露陸,蔡信德.茅尾海沉積物營養(yǎng)物質(zhì)空間分布特征研究[J]. 農(nóng)業(yè)環(huán)境科學(xué)學(xué)報(bào), 2014,33(1):162-171.

Wu WC, Ren L L, Cai X D. Spatial distribution of nutrients in sediments of the Maowei Sea [J]. Journal of Agricultural Environmental Sciences, 2014,33(1):162-171.

[2] 高金榮,趙則春.欽州茅尾海海域化學(xué)需氧量的分布特征及與富營養(yǎng)化的關(guān)系[J]. 海洋湖沼通報(bào), 2021,43(3):145-150.

Gao J R, Zhao Z C. Distribution characteristics of chemical oxygen demand in maowei sea area of Qinzhou and its relationship with eutrophication [J]. Bulletin of Ocean Limnology, 2021,43(3):145-150.

[3] 楊 斌,方懷義,鐘秋平,等.欽州灣夏季營養(yǎng)鹽的分布特征及富營養(yǎng)化評價(jià)[J]. 海洋通報(bào), 2012,31(6):640-645.

Yang B, Fang W Y, Zhong Q P, et al. Distribution characteristics of nutrients and eutrophication assessment in summer in Qinzhou Bay [J]. Ocean Bulletin, 2012,31(6):640-645.

[4] 藍(lán)文陸,彭小燕.茅尾海富營養(yǎng)化程度及其對浮游植物生物量的影響[J]. 廣西科學(xué)院學(xué)報(bào), 2011,27(2):109-112.

Lan W L, Peng X Y. Eutrophication status and its impact on phytoplankton biomass in the Maowei Sea [J]. Journal of Guangxi Academy of Sciences, 2011,27(2):109-112.

[5] 韋重霄,趙 爽,宋立榮,等.欽州灣內(nèi)灣茅尾海營養(yǎng)狀況分析與評價(jià)研究[J]. 環(huán)境科學(xué)與管理, 2017,42(9):148-153.

Wei C X, Zhao S, Song L R, et al. Analysis and assessment of nutrient status in Maowei Sea of Inner Qinzhou Bay [J]. Environmental Science and Management, 2017,42(9):148-153.

[6] 范成新,張 路,秦伯強(qiáng),等.風(fēng)浪作用下太湖懸浮態(tài)顆粒物中磷的動態(tài)釋放估算[J]. 中國科學(xué)(D輯:地球科學(xué)), 2003,33(8):760-768.

Fan C X, Zhang L, Qin B Q, et al. Estimation of phosphorus release in suspended particulate matter in Taihu Lake under the action of wind and waves [J]. [J]. Science in China (Series D: Earth Sciences), 2003, (8):760-768.

[7] Li X, Guo M, Duan X, et al. Distribution of organic phosphorus species in sediment profiles of shallow lakes and its effect on photo- release of phosphate during sediment resuspension [J]. Environment international. 2019,130:104916.

[8] Hu B, Wang P, Wang C, et al. Photogeochemistry of particulate organic matter in aquatic systems: A review [J]. Science of The Total Environment, 2022,806:150467.

[9] Guo M, Li X, Song C, et al. Photo-induced phosphate release during sediment resuspension in shallow lakes: A potential positive feedback mechanism of eutrophication [J]. Environmental pollution (1987), 2020,258:113679.

[10] 張毅敏,王 宇,楊 飛,等.太湖不同生態(tài)型湖區(qū)懸浮顆粒磷空間分布和降解速率[J]. 中國環(huán)境科學(xué), 2016,36(7):2128-2138.

Zhang Y M, Wang Y, Yang F, et al. Spatial distribution and degradation rate of phosphorus of suspended particles in different ecological lake areas of Taihu Lake [J]. Chinese Environmental Science, 2016,36(7):2128-2138.

[11] 王珊珊,潘存鴻,李 宏,等.杭州灣泥沙中重金屬元素的分布及影響因素[J]. 中國環(huán)境科學(xué), 2017,37(12):4701-4709.

Wang S S, Pan C H, Li H, et al. Distribution and influencing factors of heavy metal elements in sediment in Hangzhou Bay [J]. Chinese Environmental Science, 2017,37(12):4701-4709.

[12] 秦伯強(qiáng).長江中下游淺水湖泊富營養(yǎng)化發(fā)生機(jī)制與控制途徑初探[J]. 湖泊科學(xué), 2002,14(3):193-202.

Qin B Q. A Preliminary Study on the occurrence mechanism and control pathway of eutrophication in Shallow Lakes in the middle and lwer reaches of the Yangtze River [J]. Lake Science, 2002, 14(3):193-202.

[13] Yi Y, Yang Z, Zhang S. Ecological risk assessment of heavy metals in sediment and human health risk assessment of heavy metals in fishes in the middle and lower reaches of the Yangtze River basin [J]. Environmental Pollution, 2011,159(10):2575-2585.

[14] 周子柯,王永平,滕昊蔚,等.復(fù)合指紋技術(shù)示蹤泥沙來源研究進(jìn)展[J]. 泥沙研究, 2021,46(6):1-9.

Zhou Z K, Wang Y P, Teng H W, et al. Research progress on tracer sediment source by composite fingerprint technology [J]. Sediment Research. 2021,46(6):1-9.

[15] Russell M A, Walling D E, Hodgkinson R A. Suspended sediment sources in two small lowland agricultural catchments in the UK [J]. Journal of Hydrology, 2001,252(1):1-24.

[16] Walling D E, Collins A L, Stroud R W. Tracing suspended sediment and particulate phosphorus sources in catchments [J]. Journal of Hydrology, 2008,350(3):274-289.

[17] Collins A L, Walling D E, Webb L, et al. Apportioning catchment scale sediment sources using a modified composite fingerprinting technique incorporating property weightings and prior information [J]. Geoderma, 2010,155(3/4):249-261.

[18] Liu B, Storm D E, Zhang X J, et al. A new method for fingerprinting sediment source contributions using distances from discriminant function analysis [J]. CATENA, 2016,147:32-39.

[19] Zhang X C J, Liu B L. Using multiple composite fingerprints to quantify fine sediment source contributions: A new direction [J]. Geoderma, 2016,268:108-118.

[20] Motha J A, Wallbrink P J, Hairsine P B, et al. Determining the sources of suspended sediment in a forested catchment in southeastern Australia [J]. Water resources research, 2003,39(3):1056.

[21] Hughes A O, Olley J M, Croke J C, et al. Sediment source changes over the last 250 years in a dry-tropical catchment, central Queensland, Australia [J]. Geomorphology, 2009,104(3):262-275.

[22] Laceby J P, Olley J. An examination of geochemical modelling approaches to tracing sediment sources incorporating distribution mixing and elemental correlations [J]. Hydrological Processes, 2015, 29(6):1669-1685.

[23] Collins A L, Walling D E, Leeks G J L. Source type ascription for fluvial suspended sediment based on a quantitative composite fingerprinting technique [J]. Catena (Giessen), 1997,29(1):1-27.

[24] Krause A K, Franks S W, Kalma J D, et al. Multi-parameter fingerprinting of sediment deposition in a small gullied catchment in SE Australia [J]. CATENA, 2003,53(4):327-348.

[25] Poleto C, Merten G H, Minella J P. The identification of sediment sources in a small urban watershed in southern Brazil; an application of sediment fingerprinting [J]. Environmental technology, 2009,30 (11):1145-1153.

[26] Gellis A C, Noe G B, Gellis A C, et al. Sediment source analysis in the Linganore Creek watershed, Maryland, USA, using the sediment fingerprinting approach; 2008 to 2010 [J]. Journal of soils and sediments, 2013,13(10):1735-1753.

[27] 常維娜,周慧平,高 燕.基于復(fù)合指紋法的九鄉(xiāng)河小流域泥沙來源解析[J]. 水土保持學(xué)報(bào), 2014,28(6):106-110.

Chang W N, Zhou H P, Gao Y. Sediment source analysis of Jiuxiang river small watershed based on composite fingerprint method [J]. Journal of soil and water conservation, 2014,28(6):106-110.

[28] 陳太麗,史忠林,王永艷,等.三峽水庫典型支流消落帶泥沙顆粒態(tài)磷復(fù)合指紋示蹤研究[J]. 農(nóng)業(yè)工程學(xué)報(bào), 2019,35(20):118-124.

Chen T L, Shi Z L, Wang Y Y, et al. Fingerprinting particulate phosphorus absorbed by sediments for riparian zone deposits in tributary of Three Gorges Reservoir [J]. Transactions of the Chinese Society of Agricultural Engineering, 2019,35(20):118-124.

[29] 郭 進(jìn),文安邦,嚴(yán)冬春,等.復(fù)合指紋識別技術(shù)定量示蹤流域泥沙來源[J]. 農(nóng)業(yè)工程學(xué)報(bào), 2014,30(2):94-104.

Guo J, Wen A B, Yan D C, et al. Quantifying catchment scale sediment source using composite fingerprinting technique [J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2014,30(2):94-104.

[30] 周慧平,常維娜,張龍江.基于泥沙指紋識別的小流域顆粒態(tài)磷來源解析[J]. 農(nóng)業(yè)工程學(xué)報(bào), 2015,31(13):251-256.

Zhou H P, Chang W N, Zhang L J. Source analysis of particulate phosphorus in small watershed based on sediment fingerprint identification [J]. Journal of agricultural engineering, 2015,31(13): 251-256.

[31] 陳方鑫,張含玉,方怒放,等.利用兩種指紋因子判別小流域泥沙來源[J]. 水科學(xué)進(jìn)展, 2016,27(6):867-875.

Chen F X, Zhang HY, Fang N F, et al. Using two fingerprint factors to distinguish sediment sources in small watershed [J]. Advances in water science 2016,27(6):867-875.

[32] 楊明義,徐龍江.黃土高原小流域泥沙來源的復(fù)合指紋識別法分析[J]. 水土保持學(xué)報(bào), 2010,24(2):30-34.

Yang M Y, Xu L J. Analysis of sediment sources in small watershed of Loess Plateau by composite fingerprint identification method [J]. Journal of soil and water conservation, 2010,24(2):30-34.

[33] 趙恬茵.復(fù)合指紋識別法研究黃土高原小流域泥沙來源[D]. 楊凌:西北農(nóng)林科技大學(xué), 2017.

Zhao T Y. Study on sediment source of small watershed in Loess Plateau by composite fingerprint identification method [D]. Yangling: Northwest University of agriculture and forestry science and technology, 2017.

[34] 宓 瑩.云南大石壩水庫流域土壤侵蝕與沉積泥沙來源的關(guān)系研究[D]. 南京:南京師范大學(xué), 2015.

Mi Y. Study on the relationship between soil erosion and sediment sources in dashiba reservoir basin, Yunnan [D]. Nanjing: Nanjing Normal University, 2015.

[35] 汪 南.基于復(fù)合指紋分析法的元江干熱河谷區(qū)小流域泥沙來源研究[D]. 昆明:云南大學(xué), 2018.

Wang N. Study on sediment source of small watershed in dry and hot valley of Yuanjiang River Based on composite fingerprint analysis [D]. Kunming: Yunnan University, 2018.

[36] Huang D, Du P, Des E W, et al. Using reservoir deposits to reconstruct the impact of recent changes in land management on sediment yield and sediment sources for a small catchment in the Black soil region of northeast China [J]. Geoderma, 2019,343:139-154.

[37] 杜鵬飛,黃東浩,秦 偉,等.基于不同模型不同指紋因子的東北黑土區(qū)小流域泥沙來源分析[J]. 水土保持學(xué)報(bào), 2020,34(1):84-91.

Du P F, Huang D H, Qin W, et al. Sediment source analysis of small watershed in Northeast Black Soil Area Based on different models and different fingerprint factors [J]. Journal of soil and water conservation, 2020,34(1):84-91.

[38] Lin J, Huang Y, Wang M, et al. Assessing the sources of sediment transported in gully systems using a fingerprinting approach: An example from South-east China [J]. CATENA, 2015,129:9-17.

[39] 郝福星,黃炎和,林金石,等.指紋法研究花崗巖區(qū)典型崩崗小流域懸浮泥沙來源[J]. 水土保持學(xué)報(bào), 2017,31(2):45-49.

Hao F X, Huang Y H, Lin J S, et al. Study on the source of suspended sediment in typical collapse hill small watershed in granite area by fingerprint method [J]. Journal of soil and water conservation, 2017, 31(2):45-49.

[40] 周 曼,林嘉輝,黃炎和,等.復(fù)合指紋法分析紅壤區(qū)強(qiáng)度開發(fā)小流域泥沙來源[J]. 水土保持學(xué)報(bào), 2019,33(1):20-24.

Zhou M, Lin J H, Huang Y, et al. Using composite fingerprints to qualify sediment source in watershed with intensive exploitation on red soil region [J]. Journal of Soil and Water Conservation, 2019, 33(1):20-24.

[41] 賴廷和,何本茂,韋蔓新,等.欽州灣內(nèi)灣貝類養(yǎng)殖海區(qū)水環(huán)境特征及營養(yǎng)狀況初探[J]. 黃渤海海洋, 2001,19(4):51-55.

Lai T H, he B M, Wei M X, et al. Preliminary study on water environment characteristics and nutritional status of shellfish culture area in Inner Bay of Qinzhou Bay [J]. The Yellow Sea and the Bohai Sea, 2001,19(4):51-55.

[42] 劉永泉,凌博聞,徐鵬飛.談廣西欽州茅尾海紅樹林保護(hù)區(qū)的濕地生態(tài)保護(hù)[J]. 河北農(nóng)業(yè)科學(xué), 2009,13(4):97-99.

Liu Y Q, Ling B W, Xu P F. On the wetland ecological protection of Maowei Hai Mangrove Reserve in Qinzhou, Guangxi [J]. Hebei Agricultural Science, 2009,13(4):97-99.

[43] 張伯虎,陳沈良,谷國傳.廣西沿岸重點(diǎn)港灣的潮型與潮汐特征[J]. 海洋學(xué)研究, 2010,28(3):9-16.

Zhang B H, Chen S L, Gu G C. Tidal patterns and tidal characteristics of key coastal harbors in Guangxi [J]. Oceanographic research, 2010, 28(3):9-16.

[44] 楊 斌,方懷義,許麗莉,等.欽州灣水質(zhì)污染時(shí)空變化特征及驅(qū)動因素[J]. 海洋環(huán)境科學(xué), 2017,36(6):877-883.

Yang B, Fang H Y, Xu L L, et al. Spatio-temporal variation characteristics and driving factors of water pollution in Qinzhou bay [J]. Marine Environmental Science, 2017,36(6):877-883.

[45] Wang J, Lu N, Fu B. Inter-comparison of stable isotope mixing models for determining plant water source partitioning [J]. The Science of the total environment, 2019,666:685-693.

[46] 周慧平,陳玉東,常維娜.指紋技術(shù)識別泥沙來源:進(jìn)展與展望[J]. 水土保持學(xué)報(bào), 2018,32(5):1-7.

Zhou H P, Chen Y D, Chang W N. Sediment source fingerprinting: Progresses and prospects [J]. Journal of Soil and Water Conservation, 2018,32(5):1-7.

[47] Smith H G, Karam D S, Lennard A T. Evaluating tracer selection for catchment sediment fingerprinting [J]. Journal of Soils and Sediments, 2018,18(9):3005-3019.

[48] 杜俊杉,馬 英,胡曉農(nóng),等.基于雙穩(wěn)定同位素和MixSIAR模型的冬小麥根系吸水來源研究[J]. 生態(tài)學(xué)報(bào), 2018,38(18):6611-6622.

Du J S, Ma Y, Hu X N, et al. Applying dual stable isotopes and a MixSIAR model to determine root water uptake of winter wheat [J]. Journal of Ecology, 2018,38(18):6611-6622.

[49] Anezaki K, Nakano T, Kashiwagi N. Estimation of polychlorinated biphenyl sources in industrial port sediments using a Bayesian Semifactor Model considering unidentified sources [J]. Environmental Science & Technology, 2016,50(2):765-771.

[50] Longman J, Veres D, Ersek V, et al. Quantitative assessment of Pb sources in isotopic mixtures using a Bayesian mixing model [J]. Scientific reports. 2018,8(1):6154-6170.

[51] 張 妍,張秋英,李發(fā)東,等.基于穩(wěn)定同位素和貝葉斯模型的引黃灌區(qū)地下水硝酸鹽污染源解析[J]. 中國生態(tài)農(nóng)業(yè)學(xué)報(bào)(中英文), 2019,27(3):484-493.

Zhang Y, Zhang Q Y, Li Fadong, et al. Source identification of nitrate contamination of ground-water in Yellow River Irrigation Districts using stable isotopes and Bayesian model [J]. Chinese Journal of Eco-Agriculture, 2019,27(3):484-493.

[52] Liu C, Li Z, Dong Y, et al. Response of sedimentary organic matter source to rainfall events using stable carbon and nitrogen isotopes in a typical loess hilly-gully catchment of China [J]. Journal of hydrology, 2017,552:376-386.

[53] Moore J W, Semmens B X. Incorporating uncertainty and prior information into stable isotope mixing models [J]. Ecology Letters, 2008,11(5):470-480.

[54] Demopoulos A W J, Mcclain-Counts J P, Bourque J R, et al. Examination of Bathymodiolus childressi nutritional sources, isotopic niches, and food-web linkages at two seeps in the US Atlantic margin using stable isotope analysis and mixing models [J]. Deep-sea research, 2019,148:53-66.

[55] Carter J, Owens P, Walling D, et al. Fingerprinting suspended sediment sources in a large urban river system [J]. The Science of The Total Environment. 2003,314-316:513-534.

Analysis of the source of suspended particulate matter in the Maowei Sea based on composite fingerprint map and Bayesian model.

LIU Hai-xia1,2,3, LI Su-xia2,3*, LIU Guang-long2,4**, HUANG Kai-xuan5, YANG Bin3, SU Jing-jun5, WANG Jie4, ZHANG Jin-yi2, HOU Jing-yao2

(1.School of Environmental Science and Engineering, Guilin University of Technology, Guilin 541006, China；2.School of Resources and Environment, Beibu Gulf University, Qinzhou 535011, China；3.Guangxi Key laboratory of Marine Disaster in the Beibu Gulf University, Qinzhou 535011, China；4.School of Resources and Environment, Central China Agricultural University, Wuhan 430070, China；5.State Key Laboratory of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China)., 2022,42(6)：2844～2851

Toreveal the source of suspended particulate matter in Maoweisea, samples of potential sources of suspended particulate matter such as mangrove soil, embankment soil, estuarine particulate matter, Maowei Sea sediment and particulate matter outside the Bay in Maowei Sea Basin were collected. Based on multivariate statistical composite fingerprinting method, the optimal fingerprint factor combination was formatted, and then the contribution rates of five potential sources to suspended particulate matter in Maowei Sea are obtained by Bayesian mixed model. The results show that five fingerprint elements of Mg, Al, Mn, Pb and Fe can be used as the best combination of fingerprint factors, and the cumulative discrimination accuracy rate is 78%. The results of the Bayesian mixed model showed that the suspended particulate matter in Maowei Sea is mainly derived from the estuary and the outer bay transport, and the contribution rate reaches 58.9% and 68.6%. Among them, the area near the estuary is mainly affected by the river inflow, and the contribution rate of estuarine particulate matter reaches 42.2%～58.9%; In the areas close to the outer bay, the contribution rate of particulate matter outside the bay was 44.9%～68.6%. The contribution rate of sediment at each point was low, and the contribution rate of mangrove soil and embankment soil was about 10%. In general, particulate matter brought in by estuarine inflow and tidal action is the main source of suspended particulate matter in the Maowei Sea.

Bayesian model；composite fingerprinting；particulate matter；source resolution

X55

A

1000-6923(2022)06-2844-08

劉海霞(1998-),女,湖北仙桃人,桂林理工大學(xué)碩士研究生,主要從事水體污染控制與生態(tài)修復(fù)研究.

2021-11-12

國家自然科學(xué)基金資助項(xiàng)目(31960242);廣西自然科學(xué)基金資助項(xiàng)目(2020GXNSFAA297080,2016GXNSFAA380242)

* 責(zé)任作者, 教授, 284137449@qq.com