回首頁
 
 
  美國加州大學柏克萊分校博士
   04-2359-0121轉 32410(研究室), 32411(實驗室)
   04-2355-2498
   sunif@thu.edu.tw   ifsun@hotmail.com
 
我的研究方向主要有下列兩個:
一、東南亞熱帶雨林植物大開花現象對森林更新機制之影響
以龍腦香科植物為主的東南亞熱帶雨林有一非常特別的現象,那就是不定期的大開花現象。許多樹種可能三年、五年,甚至十年才開花一次。一旦開花,森林中超過百分之五十的種類都會加入。許多學說試圖解釋大開花現象,其中最被接受的就是掠食飽和假說(Predator satiation hypothesis)。這個學說認為,以種子為食的動物族群,在平常年份因食物缺乏而被控制於低族群量。到了大開花時,這些族群量低的動物無法消耗突然大量出現的種子,許多植物種子因此逃過一劫,進而發芽長大。但是過去四年我們的研究發現,不同種類的昆蟲及動物會在種子發育的不同時期攻擊種子;而有些昆蟲的數量能在很短時間內加倍。此外,參與大開花的植物種類,個體數及開花強度等,均依氣候條件、生長環境不同而有極大不同。這些結果都與以往的推論和假設有相當大的出入。因此,我現在的研究重點之一即在釐清掠食飽和假說的許多問題及其適用性。

除了研究大開花本身現象外,另一個研究重點是探討大開花現象對森林更新機制之影響。熱帶雨林是全球生物多樣性最高的地區,這些具有相似需求的物種如何能共存而不產生競爭排除(Competitive exclusion),一直是生態學家期望解答的謎題。多年來許多理論試圖解釋熱帶雨林的高生物多樣性,其中以天敵學說(Enemies hypothesis or Janzen-Connell hypothesis)最受重視。此學說認為越接近母樹的區域,種子密度越高,因而吸引大量天敵的掠食及攻擊,導致該處種子的死亡率高,這個效應被稱為密度-距離依變效應(Density-distance dependent effect)。由於此效應的作用,成年母樹死後釋出的空間由同種小樹取代的機率遠低於其他種類的小樹,造成自我替代(Self replacement)機率的降低,而有助於樹種多樣性的維持或提高。

如果掠食飽和假說成立,則大開花時,參與樹種的種子及小苗數量和存活率將大大提高,對參與大開花樹種的更新將提供極大優勢。然而,密度-距離依變效應預測高種子密度將降低種子存活率,因此,大開花所帶來的優勢是否會使這些樹種族群競爭力提高,實為一大謎題。這個獨特的物候現象與密度-距離依變效應之間如何交互作用,及其與樹種多樣性的關係,即是本研究計畫試圖釐清的問題。我與美國喬治亞大學的Steve Hubbell教授、STRI的Joe Wright博士、以及馬來西亞森林研究所的Nur Supardi博士共同合作研究這些問題已超過四年。我們的研究地點在馬來西亞的帕索森林保護區(Pasoh forest reserve)。我們希望能繼續監測及研究這個題目十年,這樣不但可解決生態學上的多個疑問、更可提供當地政府作為保育政策擬定及林業經營之參考。

二、人工林森林生態系整合型研究
森林為台灣最重要的生態系。依第三次台灣森林資源及土地利用調查資料,台灣全島森林覆蓋度超過百分之五十八,其中超過百分之七十六為國有林地。而在國有林地中,約三分之一為人工造林地。由此角度來看,人工林生態系實為全台灣非常重要的生態系。由於過去造林偏重於造林木之木材利用,因此形成大面積單純林相之林分結構。為符合現今人工林永續發展-生態系經營之理念,實有必要對現存之人工林實施疏伐作業,配合林下人工間植或天然更新方式形成混淆或複層林,以增加人工林結構之異質度和生物多樣性,達到生態系經營之目的。然而該如何疏伐,不同疏伐處理對生物多樣性及森林功能的影響為何,對原生樹種更新及復育的影響又為何,現今這方面的基本資料極度不足。本研究主要目的即在瞭解人工林不同疏伐處理後,原生樹種之種子是否能到達,到達後是否能發芽長大,生長及存活。本研究將於林務局南投林管處巒大事業區74、75、76林班之柳杉人工林內,挑選五處造林年齡相似(約三十年)之柳杉林,以一公頃面積為單位設立試驗樣區,進行0%,12.5%,25%及50%四種不同的疏伐處理;並從種子、小苗及大樹各方面著手,進行長時間之監測,以探討不同疏伐強度處理對原生樹種更新之影響。
1. General flowering events in SE Asia lowland Dipterocarp forests
Trees of the family Dipterocarpaceae dominate the canopy and emergent layers of the hyper-diverse, aseasonal, humid lowland tropical forests of Southeast Asia.A striking feature of these forests is that many species, especially dipterocarps, flower gregariously and fruit highly synchronously at irregular multi-year intervals, a phe-nomenon known as general flowering (GF).Many tree species of other plant families also synchronize their reproduction with the dipterocarps during GF years.Such a high degree of reproductive synchrony in species-rich tropical forest tree communities is paradoxical because, in theoretical models, increasing the degree of reproductive synchrony among species actually decreases the possibilities for their coexis-tence.In part to resolve this paradox, we began a long-term study to quantify the actual degree of reproductive synchrony in a lowland dipterocarp forest in Malay-sia, monitoring the reproductive phenology not only of dipterocarps but also of other tree species as well.

Hypotheses to explain the GF phenomenon have been both proximate, about the environmental cues that trigger GF events, and ultimate, about the evolution of repro-ductive synchrony by natural selection.The most widely accepted ultimate explana-tion for GF is Janzen’s “predator satiation” hypothesis. There are two components to this hypothesis: (i) selection for flowering at long and irregular intervals to drive down populations of seed predators, and (ii) selection for tight synchrony in seed fall, such that too many seeds are dropped in too short a time to be all consumed by the reduced numbers of seed predators, allowing more seeds to escape by germinating and growing out of the predation-vulnerable stages. According to this hypothesis, selection acts against regular annual fruiting because it supports the uninterrupted year-to-year buildup of seed predator populations.

We carried out our study in the 50 ha Forest Dynamics Plot in Pasoh Forest Reserve, Negeri Simbilan, Malaysia.Our research project addresses several key questions: (i) Are GF events essentially “all-or-nothing” phenomena, or are they variable in perva-siveness and intensity?(ii) If the GF events are not all or nothing, how did flowering and fruiting among species and individuals in one event correlate with patterns in another mast?(iii) Were the levels of pre-dispersal and post-dispersal seed predation both consistent with the predator satiation hypothesis?


2. Managing forest plantations for biodiversity and environmental conservation, and timber production in Taiwan
Forest ecosystem is one of the most important ecosystems in Taiwan.According to the recently land survey, about 60% of the island’s total landmass is still covered by forests.Among these area, 14% or 310,000 ha are plantation forests.In the past, the management goal of plantation was to produce large diameter trees with high timber value. This focus, however, has been gradually changed.The new goal is timber production with consideration of its impacts on climate changes, biodiversity losses, ecosystem functions and public acceptance.We (TFRI, CTEB, and TFB) organize a multi-disciplinary and multi-institutional research team that with integrated approaches to pursue our goals.My research project specifically focuses on the regeneration of native species.I will establish 20 one hectare permanent plots in Cryptomerioid japonica plantation forest in central Taiwan.Within this plot, I will set up seed traps, seedling plot, as well as monitor the growth and survival of existing saplings and adult trees to investigate the effects of different thinning practices on the recruitment of native species.