于瑞宏,張笑欣,劉廷璽,郝艷玲

1 內(nèi)蒙古大學(xué)環(huán)境與資源學(xué)院, 呼和浩特 010021 2 內(nèi)蒙古農(nóng)業(yè)大學(xué)水利與土木建筑工程學(xué)院, 呼和浩特 010018

?

淺水湖泊穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別方法局限與展望

于瑞宏1,*,張笑欣1,劉廷璽2,郝艷玲1

1 內(nèi)蒙古大學(xué)環(huán)境與資源學(xué)院, 呼和浩特 010021 2 內(nèi)蒙古農(nóng)業(yè)大學(xué)水利與土木建筑工程學(xué)院, 呼和浩特 010018

淺水湖泊水體底泥交換強(qiáng)烈,極易受人類活動(dòng)干擾,超過(guò)一定閾值即可能發(fā)生災(zāi)難性的穩(wěn)態(tài)轉(zhuǎn)換,對(duì)其有效識(shí)別有助于湖泊富營(yíng)養(yǎng)化的及時(shí)防控與修復(fù)。淺水湖泊穩(wěn)態(tài)轉(zhuǎn)換可通過(guò)系統(tǒng)關(guān)鍵變量(葉綠素、溶解氧、浮游動(dòng)物、魚(yú)類等)的時(shí)間序列(判別不同穩(wěn)態(tài))、預(yù)警信號(hào)及閾值等進(jìn)行識(shí)別,其中預(yù)警識(shí)別可為湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換提供預(yù)判信息,有利于早預(yù)警早行動(dòng)。目前,淺水湖泊穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別因子(方差及自相關(guān)性等)主要用于“臨界慢化”現(xiàn)象,但在強(qiáng)大外力作用、強(qiáng)烈隨機(jī)擾動(dòng)及極端事件下,這些“臨界慢化”因子則可能出現(xiàn)誤用或錯(cuò)用?；跍\水湖泊基本特征,針對(duì)穩(wěn)態(tài)轉(zhuǎn)換的不同驅(qū)動(dòng)機(jī)制,探討“臨界慢化”因子的適用性與局限性,并展望其未來(lái)發(fā)展方向,旨在為湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別提供科學(xué)參考。

淺水湖泊；穩(wěn)態(tài)轉(zhuǎn)換；預(yù)警識(shí)別；臨界慢化；驅(qū)動(dòng)機(jī)制；局限；展望

淺水湖泊生態(tài)系統(tǒng)對(duì)外界干擾的反應(yīng)會(huì)隨著干擾強(qiáng)度的增強(qiáng)而出現(xiàn)結(jié)構(gòu)或功能的突然變化,即穩(wěn)態(tài)轉(zhuǎn)換[1],這種轉(zhuǎn)換通常具有突發(fā)性與難預(yù)知性,同時(shí)兼具非線性、多閾值、多穩(wěn)態(tài)以及遲滯效應(yīng)等特征。目前,國(guó)內(nèi)外學(xué)者主要圍繞淺水湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換的基本理論[2- 4]、驅(qū)動(dòng)因子[5- 6]、預(yù)警識(shí)別[7- 13]等開(kāi)展了大量研究,其中,預(yù)警識(shí)別是近年來(lái)的研究熱點(diǎn),也是湖泊富營(yíng)養(yǎng)化防控的有效方法,且已在“臨界慢化”(Critical slowing down)現(xiàn)象中證明了其有效性。所謂“臨界慢化”,就是接近臨界點(diǎn)時(shí),即使很小的外力擾動(dòng),生態(tài)系統(tǒng)也趨于緩慢恢復(fù)[14-15],該現(xiàn)象可直接通過(guò)擾動(dòng)實(shí)驗(yàn)[16]或間接通過(guò)“臨界慢化”因子(簡(jiǎn)稱CSD因子,如方差、自相關(guān)性、偏度、峰度及條件異方差等)的異常變化來(lái)進(jìn)行識(shí)別[17-18]。盡管CSD因子具有堅(jiān)實(shí)的理論基礎(chǔ)及許多的應(yīng)用經(jīng)驗(yàn),但其并不是預(yù)測(cè)所有類型穩(wěn)態(tài)轉(zhuǎn)換的靈丹妙藥。在外部驅(qū)動(dòng)(外源性氮磷負(fù)荷、氣候變化、風(fēng)浪、湖泊水位等)及內(nèi)部驅(qū)動(dòng)(魚(yú)類、水生植物等)共同作用下,除“臨界慢化”機(jī)制外,淺水湖泊生態(tài)系統(tǒng)還存在如慢-快循環(huán)轉(zhuǎn)換[19]、閃變[20- 21]、隨機(jī)共振[22]、極端瞬變[18]、驅(qū)動(dòng)力階躍變化[18]等多種機(jī)制,其中某些機(jī)制單獨(dú)或聯(lián)合作用下,則可能會(huì)導(dǎo)致CSD因子的誤用或錯(cuò)用。目前,淺水湖泊穩(wěn)態(tài)轉(zhuǎn)換預(yù)警研究大多是針對(duì)已發(fā)生轉(zhuǎn)換生態(tài)系統(tǒng)的回顧式反演,而非前進(jìn)式預(yù)測(cè),因此,如何準(zhǔn)確判斷特定淺水湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換的驅(qū)動(dòng)機(jī)制,并采取適宜的預(yù)警因子進(jìn)行識(shí)別,仍是國(guó)際研究難點(diǎn)。為此,本文針對(duì)淺水湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換的不同驅(qū)動(dòng)機(jī)制,揭示CSD因子的適用性與局限性,并展望其未來(lái)發(fā)展趨勢(shì),旨在為淺水湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別提供科學(xué)參考。

1 淺水湖泊的基本特征

淺水湖泊是相對(duì)于深水湖泊而言的湖泊范疇,目前并無(wú)通用的定義,本文采用Scheffer(1998)所著“Ecology of Shallow Lake”一書(shū)中的描述,作為淺水湖泊定義的參考[23],所謂淺水湖泊就是：(1)光線可穿透水體進(jìn)入湖底,即光補(bǔ)償深度(透明層深度)超過(guò)水深的水體；(2)平均水深小于3m,且夏季不存在熱力分層的湖泊。通常而言,淺水湖泊具有以下特征：(1)生態(tài)系統(tǒng)較為脆弱,具有較低的污染負(fù)荷能力[23]；(2)水生植物對(duì)淺水湖泊功能存在極大影響,可使湖體出現(xiàn)復(fù)雜的生態(tài)過(guò)程和反饋機(jī)制[2]；(3)水土界面常由于動(dòng)力擾動(dòng)處于不穩(wěn)定狀態(tài),物質(zhì)交換強(qiáng)烈,湖底沉積物內(nèi)源釋放對(duì)上覆水產(chǎn)生顯著影響[24]；(4)位于高強(qiáng)度農(nóng)業(yè)區(qū)的淺水湖泊,不能從根本上控制外源,當(dāng)沉積物中氮磷營(yíng)養(yǎng)鹽的生物地球化學(xué)循環(huán)、食物網(wǎng)結(jié)構(gòu)和生態(tài)環(huán)境破壞后,水動(dòng)力條件、表層底泥生物及理化性質(zhì)的變化等都會(huì)通過(guò)反饋機(jī)制阻礙生態(tài)恢復(fù)進(jìn)程[25]。

2 淺水湖泊“穩(wěn)態(tài)轉(zhuǎn)換”預(yù)警識(shí)別方法

生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換識(shí)別包括不同穩(wěn)態(tài)識(shí)別、預(yù)警識(shí)別及閾值識(shí)別等[26],常用識(shí)別方法見(jiàn)表1。就淺水湖泊生態(tài)系統(tǒng)而言,穩(wěn)態(tài)轉(zhuǎn)換預(yù)警及閾值識(shí)別定量方法主要包括：實(shí)驗(yàn)觀測(cè)、模型模擬及統(tǒng)計(jì)因子分析等3種方法[27]。(1)實(shí)驗(yàn)觀測(cè)法：側(cè)重于物種結(jié)構(gòu)和功能的監(jiān)測(cè),主要用于穩(wěn)態(tài)轉(zhuǎn)換理論的驗(yàn)證及生物操控對(duì)湖泊穩(wěn)態(tài)影響的評(píng)估[28],尤其將實(shí)驗(yàn)監(jiān)測(cè)數(shù)據(jù)與CSD因子相結(jié)合可有效用于穩(wěn)態(tài)轉(zhuǎn)換預(yù)警與預(yù)測(cè)[9,13,29]；然而,實(shí)驗(yàn)觀測(cè)法的條件限制較多,目前應(yīng)用實(shí)例極少,這些條件主要包括：1) 盡可能同時(shí)選取生物操控湖泊與對(duì)照湖泊,二者處于相同氣候及流域條件下,且環(huán)境變化不會(huì)對(duì)湖泊穩(wěn)態(tài)轉(zhuǎn)換產(chǎn)生影響；2) 需要大量實(shí)驗(yàn)觀測(cè)數(shù)據(jù),實(shí)驗(yàn)控制要適度,不能太快或太慢；如果外力增加太快，驅(qū)動(dòng)系統(tǒng)快速通過(guò)轉(zhuǎn)換點(diǎn),會(huì)出現(xiàn)大的觀測(cè)錯(cuò)誤或預(yù)警信號(hào)被多元非線性過(guò)程的相互作用所抑制；若取樣頻率太低，則可能錯(cuò)過(guò)穩(wěn)態(tài)轉(zhuǎn)換時(shí)段,導(dǎo)致無(wú)法識(shí)別；3) 外力擾動(dòng)盡量要小,強(qiáng)烈擾動(dòng)及線性過(guò)程累加可能會(huì)導(dǎo)致錯(cuò)誤預(yù)警；4) 需明晰生態(tài)系統(tǒng)復(fù)雜機(jī)理,否則難于遴選監(jiān)測(cè)指標(biāo)。(2)動(dòng)力學(xué)模型,如沉水植物模擬模型[30],營(yíng)養(yǎng)鹽動(dòng)力學(xué)模型[31],生態(tài)動(dòng)力學(xué)模型[32]、動(dòng)態(tài)食物鏈模型[33]等,主要用于穩(wěn)態(tài)轉(zhuǎn)換閾值識(shí)別,但難于預(yù)警穩(wěn)態(tài)轉(zhuǎn)換。(3)統(tǒng)計(jì)因子分析是湖泊穩(wěn)態(tài)轉(zhuǎn)換預(yù)警中最為廣泛應(yīng)用的方法,其可揭示長(zhǎng)時(shí)間序列監(jiān)測(cè)數(shù)據(jù)的規(guī)律,通過(guò)識(shí)別穩(wěn)態(tài)轉(zhuǎn)換發(fā)生前CSD因子出現(xiàn)的異常變化,即可判定湖泊生態(tài)系統(tǒng)是否趨近臨界點(diǎn),進(jìn)而確定系統(tǒng)變量是否發(fā)生突變,借以預(yù)警穩(wěn)態(tài)是否發(fā)生轉(zhuǎn)換。鑒于CSD因子分析可為預(yù)測(cè)和預(yù)防生態(tài)系統(tǒng)災(zāi)變提供可行有效的識(shí)別手段[12, 26],本文將著重對(duì)預(yù)警識(shí)別統(tǒng)計(jì)分析法進(jìn)行深入探討。

表1 湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換常用識(shí)別方法[26]

淺水湖泊生態(tài)系統(tǒng)預(yù)警識(shí)別統(tǒng)計(jì)分析法包括基于度量因子及基于模型等兩種識(shí)別方法[34]。其中,度量因子主要包括：方差[7],自相關(guān)性[8],偏度[11],峰度[12],條件異方差[13]等,以上因子通稱為CSD因子；識(shí)別模型則主要包括：時(shí)變自回歸模型[35],非參數(shù)漂移-擴(kuò)散模型[9]等。許多專家學(xué)者對(duì)于穩(wěn)態(tài)轉(zhuǎn)換預(yù)警統(tǒng)計(jì)分析方法已經(jīng)進(jìn)行了全面系統(tǒng)的回顧,其中,李玉照等[27]及Dakos等[34]對(duì)預(yù)警識(shí)別統(tǒng)計(jì)方法的基本理論、計(jì)算公式及文獻(xiàn)應(yīng)用進(jìn)行了系統(tǒng)完整的綜述,Andersen[36]則對(duì)預(yù)警識(shí)別的相關(guān)統(tǒng)計(jì)軟件及其應(yīng)用進(jìn)行了全面細(xì)致的總結(jié)。因此,本文不再贅述,僅給出淺水湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別步驟(圖1),以及CSD因子異常變化及其在各步驟中的影響因素(表2)。

圖1 湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別流程Fig.1 Steps for detecting early warning signals of regime shift in shallow lake ecosystems

3 “穩(wěn)態(tài)轉(zhuǎn)換”預(yù)警識(shí)別方法的適用性與局限性

生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換的驅(qū)動(dòng)機(jī)制可簡(jiǎn)單劃分為6種類型[18],主要包括慢速環(huán)境驅(qū)動(dòng)、慢-快循環(huán)轉(zhuǎn)換、閃變、隨機(jī)共振、極端瞬變、驅(qū)動(dòng)力階躍變化等(表3)。不難看出,外力作用及隨機(jī)擾動(dòng)決定著穩(wěn)態(tài)轉(zhuǎn)換機(jī)制及其被識(shí)別的可能性,而CSD因子是否有效則可反過(guò)來(lái)用于穩(wěn)態(tài)轉(zhuǎn)換機(jī)制類型的判斷。以下分別就“穩(wěn)態(tài)轉(zhuǎn)換”預(yù)警識(shí)別方法的適用性與局限性進(jìn)行詳細(xì)闡述及說(shuō)明。

表2 湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別方法

√: 出現(xiàn)相應(yīng)異?，F(xiàn)象；y: 缺失數(shù)據(jù)太多時(shí)需要內(nèi)插；n: 不需要內(nèi)插；d: 是否需要進(jìn)行數(shù)據(jù)變換取決于數(shù)據(jù)質(zhì)量；log: 對(duì)數(shù)變換；+: 敏感；-： 不敏感

表3 湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換驅(qū)動(dòng)機(jī)制

3.1 “穩(wěn)態(tài)轉(zhuǎn)換”預(yù)警識(shí)別方法的適用性

統(tǒng)計(jì)因子分析是“穩(wěn)態(tài)轉(zhuǎn)換”預(yù)警識(shí)別的常用方法[27],具有堅(jiān)實(shí)理論基礎(chǔ),廣泛應(yīng)用于富營(yíng)養(yǎng)化湖泊穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別,其優(yōu)勢(shì)在于長(zhǎng)時(shí)間序列CSD因子會(huì)在穩(wěn)態(tài)轉(zhuǎn)換前呈現(xiàn)顯著變化,但無(wú)需掌握湖泊生態(tài)系統(tǒng)的復(fù)雜動(dòng)態(tài)機(jī)制和過(guò)程。然而,由于生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換可能發(fā)生于分,小時(shí),日,月,年等不同時(shí)間尺度,甚至轉(zhuǎn)換過(guò)程中就直接消失,而CSD因子的變化則需要一定的時(shí)間段進(jìn)行識(shí)別。因此,CSD因子適用于“慢速環(huán)境驅(qū)動(dòng)”及“慢-快周期轉(zhuǎn)換”機(jī)制下穩(wěn)態(tài)轉(zhuǎn)換的識(shí)別。

(1)慢速環(huán)境驅(qū)動(dòng)：在慢速環(huán)境驅(qū)動(dòng)作用下,淺水湖泊生態(tài)系統(tǒng)完全符合“臨界慢化”機(jī)制,即在幾乎接近臨界閾值時(shí)發(fā)生穩(wěn)態(tài)轉(zhuǎn)換,可用于識(shí)別浮游植物、浮游動(dòng)物或全湖生物群等關(guān)鍵監(jiān)測(cè)變量的突變。Drake[17]采用方差、偏度、時(shí)間及空間自相關(guān)等統(tǒng)計(jì)因子的突變,用于判定以浮游動(dòng)物(大型水蚤)從減少到滅絕為代表的淺水湖泊穩(wěn)態(tài)轉(zhuǎn)換,研究指出,綜合考慮以上4種統(tǒng)計(jì)因子的異常變化可有效降低穩(wěn)態(tài)轉(zhuǎn)換判定的不確定性,且在轉(zhuǎn)換發(fā)生前110 d即可進(jìn)行預(yù)警[17]。Carpenter[9]等基于全湖野外實(shí)驗(yàn),連續(xù)3年人為引進(jìn)大嘴鱸魚(yú)改變湖泊原有食物網(wǎng),并以毗鄰相似湖泊為參照,采用方差、返回率、偏度及光譜比率等統(tǒng)計(jì)因子,識(shí)別葉綠素濃度、浮游動(dòng)物生物量、食浮游生物動(dòng)物密度等關(guān)鍵變量的突變,結(jié)果表明,在環(huán)境驅(qū)動(dòng)力緩慢作用下,湖泊生態(tài)系統(tǒng)物種循環(huán)和能量流動(dòng)路徑及效率受到擾動(dòng),從而導(dǎo)致穩(wěn)態(tài)轉(zhuǎn)換[9]。

(2)慢-快循環(huán)轉(zhuǎn)換：淺水湖泊系統(tǒng)中,慢-快循環(huán)轉(zhuǎn)換的發(fā)生需滿足某些特定條件[19],這些條件包括：1) 水生植物產(chǎn)生的內(nèi)源釋放效應(yīng)較大；2) 湖泊不太淺且水生植物對(duì)透明度無(wú)強(qiáng)烈影響,系統(tǒng)遲滯現(xiàn)象存在,但前后閾值相差不大。鑒于以上條件,循環(huán)轉(zhuǎn)換在現(xiàn)實(shí)中發(fā)生概率相對(duì)較小,僅見(jiàn)于英國(guó)的Alderfen Broad湖和荷蘭的Botshol湖。van Nes[19]基于慢-快理論對(duì)Botshol湖的研究表明,淺水湖泊中,營(yíng)養(yǎng)鹽負(fù)荷為慢速變量,濁度為快速變量,沉水植物為振蕩變量,快慢變量均與振蕩變量相作用產(chǎn)生正負(fù)反饋,短期內(nèi)沉水植物通過(guò)凈化水體產(chǎn)生正反饋,但長(zhǎng)遠(yuǎn)來(lái)看則會(huì)由于沉積物營(yíng)養(yǎng)鹽的內(nèi)源釋放產(chǎn)生負(fù)反饋,正負(fù)反饋共同作用形成湖泊循環(huán)轉(zhuǎn)換。盡管湖泊穩(wěn)態(tài)循環(huán)轉(zhuǎn)換過(guò)程中無(wú)臨界點(diǎn)出現(xiàn),但轉(zhuǎn)換前仍可從快速變量(濁度)動(dòng)態(tài)中觀測(cè)到CSD因子的突然變化[19]。

3.2 穩(wěn)態(tài)轉(zhuǎn)換”預(yù)警識(shí)別方法的局限性

由于湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換驅(qū)動(dòng)機(jī)制不盡相同,CSD因子在概念、應(yīng)用與方法層面均存在一定局限性[18],主要體現(xiàn)在：(1)概念層面，CSD因子能否對(duì)穩(wěn)態(tài)轉(zhuǎn)換進(jìn)行明確識(shí)別,很大程度上取決于這種轉(zhuǎn)換是否為臨界轉(zhuǎn)換(存在臨界點(diǎn))及是否逐漸趨近,也就是說(shuō),識(shí)別會(huì)受到驅(qū)動(dòng)機(jī)制的條件約束,若不能滿足適用條件,則會(huì)產(chǎn)生誤用或錯(cuò)用；(2)應(yīng)用層面，CSD因子的應(yīng)用與生態(tài)系統(tǒng)時(shí)空尺度和關(guān)鍵變量(葉綠素、浮游動(dòng)物、魚(yú)類等)監(jiān)測(cè)能力緊密相關(guān),強(qiáng)環(huán)境擾動(dòng)或不充分監(jiān)測(cè)均會(huì)降低其在適宜時(shí)間尺度預(yù)測(cè)適宜變量的能力；(3)方法層面，CSD因子的敏感性及顯著性不僅依賴于數(shù)據(jù)的數(shù)量和質(zhì)量,同時(shí)也取決于對(duì)所使用的各種統(tǒng)計(jì)工具的基本假設(shè)[34,37]。

3.2.1 CSD因子識(shí)別穩(wěn)態(tài)轉(zhuǎn)換的局限性

表3所示6種轉(zhuǎn)換機(jī)制中,除慢速環(huán)境驅(qū)動(dòng)、慢-快周期轉(zhuǎn)換等2種機(jī)制可由CSD因子有效識(shí)別外,閃變、隨機(jī)共振、極端瞬變、驅(qū)動(dòng)力階躍變化等4種驅(qū)動(dòng)機(jī)制下,穩(wěn)態(tài)轉(zhuǎn)換并不跨越臨界點(diǎn),無(wú)“臨界慢化”現(xiàn)象出現(xiàn),因此,難于使用CSD因子進(jìn)行有效識(shí)別。

(1)閃變 強(qiáng)烈擾動(dòng)可導(dǎo)致“閃變”現(xiàn)象,即系統(tǒng)在遠(yuǎn)離臨界點(diǎn)時(shí)就發(fā)生不同狀態(tài)間的往復(fù)躍遷,這種現(xiàn)象是系統(tǒng)離開(kāi)“安全運(yùn)行空間”(Safe operating space)的直接信號(hào)[20]。在湖泊生態(tài)系統(tǒng)中,當(dāng)“閃變”發(fā)生時(shí),方差將升高[7- 8, 20],自相關(guān)性則可能升高[34,37]或降低[20]；但在極端隨機(jī)或混沌動(dòng)態(tài)下,方差和自相關(guān)性均不能用于穩(wěn)態(tài)轉(zhuǎn)換預(yù)警[38- 39],此時(shí)非參數(shù)模型[34]與時(shí)變自回歸模型通?？商娲鶦SD因子,較好地識(shí)別該種“閃變”機(jī)制下的穩(wěn)態(tài)轉(zhuǎn)換。Dakos[21]的研究則進(jìn)一步表明,通過(guò)時(shí)間序列概率密度分布重建吸引域,并將重建潛力與統(tǒng)計(jì)因子分異特征相結(jié)合,是識(shí)別“閃變”機(jī)制下穩(wěn)態(tài)轉(zhuǎn)換的最好方法[21]。該種驅(qū)動(dòng)機(jī)制下,需要依據(jù)擾動(dòng)強(qiáng)度判定CSD因子可否用于穩(wěn)態(tài)轉(zhuǎn)換識(shí)別。

(2)隨機(jī)共振 最初由Benzi等人提出并用于解釋第四紀(jì)冰川變化[40],其后逐漸被用于描述非線性系統(tǒng)中由于內(nèi)、外噪聲存在而增加系統(tǒng)輸出的響應(yīng)[41]。就湖泊生態(tài)系統(tǒng)而言,隨機(jī)共振效應(yīng)通常發(fā)生于生物群落穩(wěn)定邊緣,因而輕微隨機(jī)環(huán)境噪聲就可能導(dǎo)致巨大的種群豐度變化。Beninca[22]研究了佛羅里達(dá)Tarpon湖浮游生物群落與氣溫隨機(jī)變化間的共振關(guān)系,結(jié)果表明,輪蟲(chóng)、水蚤類等浮游生物與溫度波動(dòng)幾乎具有相同的時(shí)間尺度變動(dòng)范圍,其生長(zhǎng)速率周期性與溫度紅噪聲時(shí)間尺度之間的共振,可引發(fā)浮游生物的強(qiáng)烈波動(dòng),從而導(dǎo)致湖泊物種發(fā)生轉(zhuǎn)換。當(dāng)該種驅(qū)動(dòng)機(jī)制作用時(shí),穩(wěn)態(tài)轉(zhuǎn)換前難于觀察到CSD因子(方差及自相關(guān)性)的增加。

(3)極端瞬變 強(qiáng)烈的外界擾動(dòng)會(huì)驅(qū)動(dòng)生態(tài)系統(tǒng)遠(yuǎn)離當(dāng)前狀態(tài),但并未真正轉(zhuǎn)變?yōu)榱硪环N穩(wěn)態(tài),其表現(xiàn)為極端事件下的短時(shí)間穩(wěn)態(tài)轉(zhuǎn)換,如低氧事件或病原體爆發(fā)導(dǎo)致魚(yú)群消失,后期回彈,但并未真正發(fā)生穩(wěn)態(tài)轉(zhuǎn)換[18]。CSD因子不能對(duì)該種驅(qū)動(dòng)機(jī)制下的穩(wěn)態(tài)轉(zhuǎn)換進(jìn)行識(shí)別。

(4)驅(qū)動(dòng)力階躍變化 恒定驅(qū)動(dòng)力突然發(fā)生階躍,并持續(xù)作用于生態(tài)系統(tǒng),促使生態(tài)系統(tǒng)發(fā)生永久的穩(wěn)態(tài)轉(zhuǎn)換,CSD因子不能對(duì)該種驅(qū)動(dòng)機(jī)制下的穩(wěn)態(tài)轉(zhuǎn)換進(jìn)行識(shí)別。驅(qū)動(dòng)力階躍變化機(jī)制尚未見(jiàn)于淺水湖泊的應(yīng)用實(shí)例,有待進(jìn)一步探討與研究。

3.2.2 CSD因子錯(cuò)誤預(yù)警及無(wú)預(yù)警的局限性

穩(wěn)態(tài)轉(zhuǎn)換預(yù)警研究通常會(huì)選擇相對(duì)理想的時(shí)間序列進(jìn)行分析,然而,當(dāng)環(huán)境因子隨機(jī)波動(dòng)較大、系統(tǒng)對(duì)環(huán)境因子不敏感或跟隨環(huán)境高頻波動(dòng)的能力降低時(shí),CSD因子則可能出現(xiàn)錯(cuò)誤預(yù)警及穩(wěn)態(tài)轉(zhuǎn)換前無(wú)預(yù)警信號(hào)的情形。其中,錯(cuò)誤預(yù)警通常源自外界隨機(jī)擾動(dòng)變化,CSD因子(方差及自相關(guān)性)的升高可能是由環(huán)境沖擊驅(qū)動(dòng),而非趨近臨界轉(zhuǎn)換[42]；而無(wú)預(yù)警信號(hào)則具有多種來(lái)源,主要包括：

(1)數(shù)據(jù)序列太短 已有研究多遴選長(zhǎng)序列、高質(zhì)量、低噪聲的湖泊監(jiān)測(cè)數(shù)據(jù)進(jìn)行穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別[7,11,12],但遇大波動(dòng)噪聲、驅(qū)動(dòng)力快速變化、短時(shí)序低質(zhì)量數(shù)據(jù)等情形時(shí),CSD因子則表現(xiàn)欠佳,難于在穩(wěn)態(tài)轉(zhuǎn)換前進(jìn)行預(yù)警[39]。鑒于此,為了提升短期密集時(shí)間序列數(shù)據(jù)用于穩(wěn)態(tài)轉(zhuǎn)換預(yù)警的可行性,從識(shí)別CSD因子異常轉(zhuǎn)向定義臨界預(yù)警水平可為預(yù)警識(shí)別提供新思路[12]。

(2)觀測(cè)頻率不適宜 增加觀測(cè)頻次有助于準(zhǔn)確估算方差,但增加觀測(cè)間隔、減少頻次有利于準(zhǔn)確估算自相關(guān)性[34, 37],二者互相牽制。因此,選擇適宜觀測(cè)頻率(不能過(guò)高或過(guò)低),保證湖泊生態(tài)系統(tǒng)過(guò)程與取樣過(guò)程時(shí)間尺度相匹配,有利于CSD因子的有效識(shí)別,否則就會(huì)出現(xiàn)無(wú)預(yù)警情形。

(3)空間過(guò)程擾動(dòng) 大空間擾動(dòng)、主要物種的不規(guī)律運(yùn)動(dòng)、空間異質(zhì)性或擴(kuò)散等空間交互作用,均會(huì)導(dǎo)致淺水湖泊系統(tǒng)關(guān)鍵變量時(shí)間序列CSD因子在穩(wěn)態(tài)轉(zhuǎn)換前無(wú)預(yù)警信號(hào)[43]；然而,空間CSD預(yù)警因子(如空間方差)通?？煽朔r(shí)間CSD因子的局限性,呈現(xiàn)升高趨勢(shì),有效提升空間過(guò)程擾動(dòng)情形下的預(yù)警識(shí)別能力[44]。

(4)多源噪聲共同作用 淺水湖泊生態(tài)系統(tǒng)通常會(huì)受到多源噪聲影響,當(dāng)環(huán)境隨機(jī)性同時(shí)體現(xiàn)在系統(tǒng)狀態(tài)和過(guò)程中時(shí),噪聲對(duì)方差的影響就不能僅采用簡(jiǎn)單方程進(jìn)行估算,而需借助多維方差矩陣[34],否則系統(tǒng)關(guān)鍵變量動(dòng)態(tài)變化就可能會(huì)被放大或抑制,從而導(dǎo)致突變前方差預(yù)警失真[45]。

(5)關(guān)鍵變量快速趨近臨界閾值 CSD因子的變化需要一定時(shí)間長(zhǎng)度進(jìn)行識(shí)別,當(dāng)外力緩慢作用于湖泊生態(tài)系統(tǒng)時(shí),CSD因子可在穩(wěn)態(tài)轉(zhuǎn)換前有效預(yù)警；然而,當(dāng)生態(tài)系統(tǒng)關(guān)鍵變量快速趨近臨界閾值,CSD因子通常在穩(wěn)態(tài)轉(zhuǎn)換后才會(huì)呈現(xiàn)升高趨勢(shì),無(wú)法提前進(jìn)行識(shí)別[12]。

(6)關(guān)鍵變量遴選錯(cuò)誤 并非所有關(guān)鍵變量對(duì)湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換都同等敏感,相同驅(qū)動(dòng)力作用下,不同變量的表現(xiàn)不盡相同,不恰當(dāng)?shù)淖兞窟x擇可能會(huì)導(dǎo)致轉(zhuǎn)換前無(wú)預(yù)警信號(hào)。Batt[46]的研究表明,相對(duì)于估算變量(總初級(jí)生產(chǎn)力、呼吸作用、生態(tài)系統(tǒng)凈生產(chǎn)量),直接監(jiān)測(cè)變量(溶解氧,pH,葉綠素a濃度)能獲得更好的預(yù)警效果,原因在于避免了生態(tài)模型用于變量估算的不確定性[46]。因此,基于淺水湖泊系統(tǒng)理論及實(shí)驗(yàn),選擇正確的系統(tǒng)變量至關(guān)重要。

(7)同步發(fā)生多種穩(wěn)態(tài)轉(zhuǎn)換 淺水湖泊食物鏈可能會(huì)同時(shí)經(jīng)歷多個(gè)不同但相關(guān)聯(lián)的穩(wěn)態(tài)轉(zhuǎn)換(如魚(yú)類、浮游動(dòng)物、浮游植物等)[15],此時(shí)系統(tǒng)行為時(shí)間序列通常是多種轉(zhuǎn)換共同作用的結(jié)果,這會(huì)直接導(dǎo)致湖泊生態(tài)系統(tǒng)響應(yīng)被抑制或放大,當(dāng)響應(yīng)被抑制時(shí),穩(wěn)態(tài)轉(zhuǎn)換前則無(wú)預(yù)警信號(hào)發(fā)生[47]。

4 “穩(wěn)態(tài)轉(zhuǎn)換”預(yù)警識(shí)別展望

盡管統(tǒng)計(jì)因子分析應(yīng)用于湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別具有一定的局限性,在強(qiáng)擾動(dòng)作用下還會(huì)出現(xiàn)誤用或錯(cuò)用,但大量成功實(shí)例證明了其潛力及有效性,未來(lái)可從多源數(shù)據(jù)綜合利用、監(jiān)測(cè)方案創(chuàng)新及多方法聯(lián)合使用等方面不斷提升與完善。

(1)遙感數(shù)據(jù)與實(shí)驗(yàn)監(jiān)測(cè)數(shù)據(jù)有機(jī)結(jié)合

理論需求與可獲取高頻監(jiān)測(cè)數(shù)據(jù)的不匹配問(wèn)題是淺水湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換研究的挑戰(zhàn),而自動(dòng)水質(zhì)監(jiān)測(cè)與遙感技術(shù)的有機(jī)結(jié)合則為穩(wěn)態(tài)轉(zhuǎn)換預(yù)警提供了機(jī)遇。遙感技術(shù)可提供實(shí)時(shí)、高頻率、長(zhǎng)時(shí)間、大空間尺度浮游藻類群的實(shí)時(shí)影像,將航片和衛(wèi)星影像獲取的空間信息與長(zhǎng)時(shí)序水質(zhì)、植被監(jiān)測(cè)數(shù)據(jù)相結(jié)合,能夠?yàn)樯鷳B(tài)系統(tǒng)動(dòng)態(tài)時(shí)空特征提取提供新途徑,從而遠(yuǎn)離預(yù)警錯(cuò)誤信息及不明信號(hào),為穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別奠定基礎(chǔ)。

(2)機(jī)理模型與統(tǒng)計(jì)模型有機(jī)結(jié)合

目前使用的生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換模型包括統(tǒng)計(jì)模型、動(dòng)力模型、均衡模型及智能模型[26],其中統(tǒng)計(jì)模型可揭示長(zhǎng)時(shí)間序列監(jiān)測(cè)數(shù)據(jù)的規(guī)律,機(jī)理模型可從多尺度闡明穩(wěn)態(tài)轉(zhuǎn)化理論,若將統(tǒng)計(jì)模型與機(jī)理模型相結(jié)合,就可以形成多種組合用于穩(wěn)態(tài)轉(zhuǎn)換識(shí)別,很大程度上克服單獨(dú)使用統(tǒng)計(jì)模型的局限性。此外,聯(lián)合使用方差、偏度、峰度及自相關(guān)等CSD因子的異常變化用于穩(wěn)態(tài)轉(zhuǎn)換的判定,可有效提高“閃變”機(jī)制下穩(wěn)態(tài)轉(zhuǎn)換識(shí)別效果。

(3)從突變預(yù)測(cè)到恢復(fù)力圖解

穩(wěn)態(tài)轉(zhuǎn)換的先驗(yàn)知識(shí)可能會(huì)導(dǎo)致穩(wěn)態(tài)轉(zhuǎn)換預(yù)警的偏差[48],已有淺水湖泊研究通常局限于定點(diǎn)監(jiān)測(cè),而對(duì)不同時(shí)空尺度恢復(fù)力的大小不得而知,如果能夠繪制不同時(shí)間不同地理位置的恢復(fù)力分布圖,則能有效估算不同時(shí)間不同位置的恢復(fù)力及穩(wěn)態(tài)轉(zhuǎn)換發(fā)生的可能性,以便于不同區(qū)域優(yōu)先順序管理。

(4)期望穩(wěn)態(tài)CSD因子安全允許范圍的確定

Carpenter的研究表明[49],方差升高可作為臨界慢化現(xiàn)象穩(wěn)態(tài)轉(zhuǎn)換的預(yù)警信號(hào),然而為避免生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換,短時(shí)間內(nèi)快速降低方差,不僅不能阻止穩(wěn)態(tài)轉(zhuǎn)換發(fā)生,還會(huì)增加系統(tǒng)跨越臨界閾值的風(fēng)險(xiǎn),導(dǎo)致生態(tài)系統(tǒng)向相反穩(wěn)態(tài)發(fā)展,可見(jiàn)找出期望穩(wěn)態(tài)CSD因子安全允許范圍是有效預(yù)警的重要條件。

[1] Carpenter S R. Regime shifts in lake ecosystems: pattern and variation// Kinne O, ed. Excellence in Ecology. Book 15. Oldendorf/ Luhe: International Ecology Institute, 2003: 10- 10.

[2] Scheffer M, Hosper S H, Meijer M L, Moss B, Jeppesen E. Alternative equilibria in shallow lakes. Trends in Ecology & Evolution, 1993, 8(8): 275- 279.

[3] Scheffer M, Carpenter S R, Foley J A, Folke C, Walker B. Catastrophic shifts in ecosystems. Nature, 2001, 413(6856): 591- 596.

[4] Scheffer M, Jeppesen E. Regime shifts in shallow lakes. Ecosystems, 2007, 10(1): 1- 3.

[5] Zhang S Y, Liu A F, Ma J M, Zhou Q H, Xu D, Cheng S P, Zhao Q, Wu Z B. Changes in physicochemical and biological factors during regime shifts in a restoration demonstration of macrophytes in a small hypereutrophic Chinese lake. Ecological Engineering, 2010, 36(12): 1611- 1619.

[6] Janssen A B G, Teurlincx S, An S Q, Janse J H, Paerl H W, Mooij W M. Alternative stable states in large shallow lakes?. Journal of Great Lakes Research, 2014, 40(4): 813- 826.

[7] Carpenter S R, Brock W A. Rising variance: a leading indicator of ecological transition. Ecology Letters, 2006, 9(3): 311- 318.

[8] Carpenter S R, Brock W A, Cole J J, Kitchell J F, Pace M L. Leading indicators of trophic cascades. Ecology Letters, 2008, 11(2): 128- 138.

[9] Carpenter S R, Cole J J, Pace M L, Batt R, Brock W A, Cline T, Coloso J, Hodgson J R, Kitchell J F, Seekell D A, Smith L, Weidel B. Early warnings of regime shifts: a whole-ecosystem experiment. Science, 2011, 332(6033): 1079- 1082.

[10] Kleinen T, Held H, Petschel-Held G. The potential role of spectral properties in detecting thresholds in the Earth system: application to the thermohaline circulation. Ocean Dynamics, 2003, 53(2): 53- 63.

[11] Guttal V, Jayaprakash C. Changing skewness: an early warning signal of regime shifts in ecosystems. Ecology Letters, 2008, 11(5): 450- 460.

[12] Biggs R, Carpenter S R, Brock W A. Turning back from the brink: detecting an impending regime shift in time to avert it. Proceedings of the National Academy of Sciences of the United States of America, 2009, 106(3): 826- 831.

[13] Seekell D A, Carpenter S R, Pace M L. Conditional heteroscedasticity as a leading indicator of ecological regime shifts. The American Naturalist, 2011, 178(4): 442- 451.

[14] Wissel C. A universal law of the characteristic return time near thresholds. Oecologia, 1984, 65(1): 101- 107.

[15] Scheffer M, Bascompte J, Brock W A, Brovkin V, Carpenter S R, Dakos V, Held H, van Nes E H, Rietkerk M, Sugihara G. Early-warning signals for critical transitions. Nature, 2009, 461(7260): 53- 59.

[16] Veraart A J, Faassen E J, Dakos V, van Nes E H, Lürling M, Scheffer M. Recovery rates reflect distance to a tipping point in a living system. Nature, 2012, 481(7381): 357- 359.

[17] Drake J M, Griffen B D. Early warning signals of extinction in deteriorating environments. Nature, 2010, 467(7314): 456- 459.

[18] Dakos V, Carpenter S R, van Nes E H, Scheffer M. Resilience indicators: prospects and limitations for early warnings of regime shifts. Philosophical Transactions of the Royal Society B, 2015, 370(1659): 20130263.

[19] van Nes E H, Rip W J, Scheffer M. A theory for cyclic shifts between alternative states in shallow lakes. Ecosystems, 2007, 10(1): 17- 28.

[20] Wang R, Dearing J A, Langdon P G, Zhang E L, Yang X D, Dakos V, Scheffer M. Flickering gives early warning signals of a critical transition to a eutrophic lake state. Nature, 2012, 492(7429): 419- 422.

[21] Dakos V, van Nes E H, Scheffer M. Flickering as an early warning signal. Theoretical Ecology, 2013, 6(3): 309- 317.

[22] Benincà E, Dakos V, van Nes E H, Huisman J, Scheffer M. Resonance of plankton communities with temperature fluctuations. The American Naturalist, 2011, 178(4): E85-E95.

[23] Scheffer M. Ecology of Shallow Lakes. London: Chapman and Hall, 1998: xiv-xiv.

[24] 曾海鰲, 吳敬祿. 近50年來(lái)?yè)嵯珊亟饘傥廴镜某练e記錄. 第四紀(jì)研究, 2007, 27(1): 128- 132.

[25] Moss B, Madgewick J, Phillips G. A Guide to the Restoration of Nutrient-enriched Shallow Lakes. Norfolk: Broads Authority/ Environment Agency, 1996: 21- 122.

[26] Filatova T, Polhill J G, van Ewijk S. Regime shifts in coupled socio-environmental systems: review of modelling challenges and approaches. Environmental Modelling & Software, 2016, 75: 333- 347.

[27] 李玉照, 劉永, 趙磊, 鄒銳, 王翠榆, 郭懷成. 淺水湖泊生態(tài)系統(tǒng)穩(wěn)態(tài)轉(zhuǎn)換的閾值判定方法. 生態(tài)學(xué)報(bào), 2013, 33(11): 3280- 3290.

[28] Meijer M L, de Boois I, Scheffer M, Portielje R, Hosper H. Biomanipulation in shallow lakes in the Netherlands: an evaluation of 18 case studies. Hydrobiologia, 1999, 408- 409: 13- 30.

[29] Pace M L, Carpenter S R, Johnson R A, Kurtzweil J. Zooplankton provide early warnings of a regime shift in a whole lake manipulation. Limnology and Oceanography, 2013, 58(2): 525- 532.

[30] van Nes E H, Scheffer M, van den Berg M S, Coops H. Dominance of charophytes in eutrophic shallow lakes-when should we expect it to be an alternative stable state?. Aquatic Botany, 2002, 72(3/4): 275- 296.

[31] Kuiper J J, van Altena C, de Ruiter P C, van Gerven L P A, Janse J H, Mooij W M. Food-web stability signals critical transitions in temperate shallow lakes. Nature Communications, 2015, 6: 7727.

[32] Kong X Z, Dong L, He W, Wang Q M, Mooij W M, Xu F L. Estimation of the long-term nutrient budget and thresholds of regime shift for a large shallow lake in China. Ecological Indicators, 2015, 52: 231- 244.

[33] Ofir E, Gal G, Goren M, Shapiro J, Spanier E. Detecting changes to the functioning of a lake ecosystem following a regime shift based on static food-web models. Ecological Modelling, 2016, 320: 145- 157.

[34] Dakos V, Carpenter S R, Brock W A, Ellison A M, Guttal V, Ives A R, Kéfi S, Livina V, Seekell D A, van Nes E H, Scheffer M. Methods for detecting early warnings of critical transitions in time series illustrated using simulated ecological data. PLoS One, 2012, 7(7): e41010.

[35] Ives A R, Dakos V. Detecting dynamical changes in nonlinear time series using locally linear state-space models. Ecosphere, 2012, 3(6): 1- 15.

[36] Andersen T, Carstensen J, Hernández-García E, Duarte C M. Ecological thresholds and regime shifts: approaches to identification. Trends in Ecology & Evolution, 2009, 24(1): 49- 57.

[37] Lenton T M, Livina V N, Dakos V, van Nes E H, Scheffer M. Early warning of climate tipping points from critical slowing down: comparing methods to improve robustness. Philosophical Transactions of the Royal Society A, 2012, 370(1962): 1185- 1204.

[38] Contamin R, Ellison A M. Indicators of regime shifts in ecological systems: what do we need to know and when do we need to know it. Ecological Applications, 2009, 19(3): 799- 816.

[39] Perretti C T, Munch S B. Regime shift indicators fail under noise levels commonly observed in ecological systems. Ecological Applications, 2012, 22(6): 1772- 1779.

[40] Benzi R, Parisi G, Sutera A, Vulpiani A. Stochastic resonance in climatic change. Tellus, 1982, 34(1): 10- 16.

[41] McDonnell M D, Abbott D. What is stochastic resonance? Definitions, misconceptions, debates, and its relevance to biology. PLoS Computational Biology, 2009, 5(5): e1000348.

[42] Dakos V, van Nes E H, D′Odorico P, Scheffer M. Robustness of variance and autocorrelation as indicators of critical slowing down. Ecology, 2012, 93(2): 264- 271.

[43] Dakos V, van Nes E H, Donangelo R, Fort H, Scheffer M. Spatial correlation as leading indicator of catastrophic shifts. Theoretical Ecology, 2010, 3(3): 163- 174.

[44] Carpenter S R, Brock W A. Early warnings of regime shifts in spatial dynamics using the discrete Fourier transform. Ecosphere, 2010, 1(5): 1- 15.

[45] Brock W A, Carpenter S R. Early warnings of regime shift when the ecosystem structure is unknown. PLoS One, 2012, 7(9): e45586.

[46] Batt R D, Carpenter S R, Cole J J, Pace M L, Johnson R A. Changes in ecosystem resilience detected in automated measures of ecosystem metabolism during a whole-lake manipulation. Proceedings of the National Academy of Sciences of the United States of America, 2013, 110(43): 17398- 17403.

[47] Brock W A, Carpenter S R. Interacting regime shifts in ecosystems: implication for early warnings. Ecological Monographs, 2010, 80(3): 353- 367.

[48] Boettiger C, Hastings A. Early warning signals and the prosecutor′s fallacy. Proceedings of the Royal Society B, 2012, 279(1748): 4734- 4739.

[49] Carpenter S R, Brock W A, Folke C, van Nes E H, Scheffer M. Allowing variance may enlarge the safe operating space for exploited ecosystems. Proceedings of the National Academy of Sciences of the United States of America, 2015, 112(46): 14384- 14389.

Forewarned is forearmed: limitations and prospects of early warning indicators of regime shifts in shallow lakes

YU Ruihong1,*, ZHANG Xiaoxin1, LIU Tingxi2, HAO Yanling1

1CollegeofEnvironmentandResources,InnerMongoliaUniversity,Hohhot010021,China2CollegeofConservancyandCivilEngineering,InnerMongoliaAgriculturalUniversity,Hohhot010018,China

In shallow lake ecosystems, the exchange rate between water and sediment is high and can be easily disturbed by anthropogenic pressures. Additionally, catastrophic shifts occur when a threshold is reached; however, effective methods for detecting regime shifts can contribute to the timely control and restoration of eutrophication. Regime shifts of key variables, including chlorophyll, dissolved oxygen, and biomass of zooplankton and fishes, may be detected either through time series analysis, early warning signals provided by the abrupt changes of indicators, or identification of thresholds of explanatory variables, in which early warning signals can provide the most useful forecasting information. Thus far, the most widely used early warning indicators, such as autocorrelation and variance, are usually applied to lake ecosystems subject to the phenomenon of critical slowing down (CSD). However, under conditions of high stochasticity, strong external perturbation, and extreme events, these measures may underperform or subject to misinterpretation. Thus, in terms of driving mechanisms, the applicability and limitations of CSD indicators are herein discussed, as well as their prospects. The aim of this study is to synthesize what is currently known about the early warning detection of regime shifts in shallow lake ecosystems.

shallow lake ecosystems; regime shift; early warning; critical slowing down; driving mechanisms; limitations; prospects

國(guó)家自然科學(xué)基金項(xiàng)目(51469018, 61461034); 水利部公益性行業(yè)科研專項(xiàng)(20150104); 內(nèi)蒙古科技廳科技引導(dǎo)項(xiàng)目(20140707); 內(nèi)蒙古自然科學(xué)基金項(xiàng)目(2014MS0403); 內(nèi)蒙古科技廳應(yīng)用研發(fā)項(xiàng)目(20130428)

2016- 04- 01; 網(wǎng)絡(luò)出版日期:2017- 02- 22

10.5846/stxb201604010594

*通訊作者Corresponding author.E-mail: yrh0108@163.com

于瑞宏,張笑欣,劉廷璽,郝艷玲.淺水湖泊穩(wěn)態(tài)轉(zhuǎn)換預(yù)警識(shí)別方法局限與展望.生態(tài)學(xué)報(bào),2017,37(11):3619- 3627.

Yu R H, Zhang X X, Liu T X, Hao Y L.Forewarned is forearmed: limitations and prospects of early warning indicators of regime shifts in shallow lakes.Acta Ecologica Sinica,2017,37(11):3619- 3627.