我的主要研究是使用高多樣性昆蟲類群，探討不同生態層級如何適應不斷變化的環境。我的研究材料主要是鱗翅目昆蟲（即蝶與蛾，全世界約16萬種，臺灣約5000種），我對分布在東亞和東南亞的物種有相當的掌握和豐富的生物學知識，同時在博物館館藏管理、田野調查、數位典藏、生物多樣性資訊處理和廣泛的國際合作網絡方面有實際經驗。在這些的基礎上，來持續開展多樣的議題。以下針對我所關注的兩個主要議題作介紹：
My main researches revolve around the application of highly diverse insect taxa to explore how different ecological levels adapt to changing environments. My research group is primarily Lepidoptera (butterflies and moths, about 160,000 species worldwide and 5,000 species in Taiwan). I have a considerable grasp and richness of biology of species distributed throughout East and Southeast Asia, as well as practical experience in museum collection management, fieldwork, digital archiving, biodiversity information processing, and broad international collaborative networks. Building on these basic competencies, diverse topics can be undertaken sustainably. The following is an introduction to the development of two main topics that I have focused on:

1. 巨觀生態與氣候變遷生物學 / MACROECOLOGY AND CLIMATE CHANG BIOLOGY

氣候變化正在導致生態群聚的轉變。了解生物體面對的環境和生物壓力，可以預測族群動態、物種的交互關係反應、群落的重組。
Climate change is leading to a shift in ecological communities. Understanding how environmental and biological pressures that organisms face can forecast population dynamics, interdependent species responses, reorganized communities.

(1) 巨觀生理適應 / MACROPHYSIOLOGICAL ADAPTATION

全球氣候變化已經成為影響物種適應和分布的關鍵因素。氣候變異度假說（CVH）提出環境梯度下氣候變異如何形塑物種的生理適應與分布範圍，半個世紀以來，影響了我們對變化環境中生物脆弱性程度的關注和解釋。然而，過往的後設分析表明，分類群取樣密度和生理耐受方法的侷限皆可能影響能否對此假說進行完整的檢測。

Global climate change has become a critical influence on the adaptation and distribution of species. For half a century, the Climate Variability Hypothesis (CVH) has influenced our concern and interpretation of the degree of biological vulnerability in changing environments, the hypothesis proposes how the physiological adaptations and distribution ranges of species are effected by climate variability across environmental gradients. However, previous meta-analyses have shown that limitations in taxon sampling density and physiological tolerance methods may affect the complete testing of this hypothesis.

我在中央研究院沈聖峰老師研究室進行博士後研究的期間，我和我的同事帶領一個團隊在臺灣（亞熱帶地區）、中國西部（溫帶-亞熱帶地區）和馬來西亞西部（熱帶地區）的山區進行野外工作，測量這些不同生物地理區域的7,165個個體和1,033個蛾類物種的形態和對低溫與高溫的生理耐受特徵，並沿著海拔梯度進行分布範圍的採樣，以了解無脊椎動物對氣候變化的反應。我很期待此大尺度研究的成果能對氣候變化影響生物適應提出新的詮釋。

During my postdoctoral research in Dr. Sheng-Feng Shen's lab (Biodiversity Research Center, Academia Sinica), I have led a team with my colleagues to conduct field works in mountain regions of Taiwan (subtropical region), Western China (temperate-subtropical region) and Western Malaysia (tropical region) to measure the morphological and cold/hot physiological tolerance traits and sample the distribution data of 7,165 individuals and 1,033 moth species along elevational gradients in these different biogeographic regions, to understand how invertebrate’s response the climate change. I am very much looking forward to the results of this large-scale study that can provide a new interpretation of the impact of climate change on biological adaptation.

(2) 表型適應 / PHENOTYPIC ADAPTATION

生命的色彩多樣性是否有一個普遍的原則，這是一個迷人的問題。然而，在廣泛的地理範圍內解釋動物之間的顏色變化仍然具有挑戰性。我在博士後期間與合作者發表的研究（Wu et al., 2019, Nature Communications）展示了深度學習--一種人工智慧的形式--如何揭示沿著生態梯度的顏色特徵變化的微妙但明顯的模式，以及幫助確定產生這種生物地理模式的基本機制。我們使用屬於臺灣近2000個蛾類物種的2萬多張帶有精確GPS定位信息的圖像，深度學習模型產生了一個2048維的特徵向量，根據顏色和形狀特徵準確預測了物種的平均海拔。利用這個多維特徵向量，我們發現，在高海拔的群聚中，群聚內部的影像特徵變化較小。結構方程模型表明，這種圖像特徵多樣性的減少可能是寒冷環境選擇深色的結果，這限制了高海拔地區群聚的色彩多樣性。最終，在深度學習的幫助下，我們將能夠在前所未有的深度一步步探索自然形態變化的無窮形式。
The question of whether there is a universal principle for the colorful diversity of life is fascinating one. However, explaining color variation among animals at broad geographic scales remains challenging. My recent postdoctoral research (Wu et al., 2019, Nature Communications) demonstrates how deep learning—a form of artificial intelligence—can reveal subtle but robust patterns of color feature variation along an ecological gradient, as well as help identify the underlying mechanisms generating this biogeographic pattern. Using over 20,000 images with precise GPS locality information belonging to nearly 2,000 moth species from Taiwan, the deep learning model generates a 2048-dimension feature vector that accurately predicts each species’ mean elevation based on color and shape features. Using this multidimensional feature vector, we found that within-assemblage image feature variation is smaller in high elevation assemblages. Structural equation modeling suggests that this reduced image feature diversity is likely the result of colder environments selecting for darker coloration, which limits the color diversity of assemblages at high elevations. Ultimately, with the help of deep learning, we will be able to explore the endless forms of natural morphological variation at unpreceded depths step by step.

這項研究具有開創性，因為它使用了大量的公民科學數據，並結合了人工智慧來分析以前難以量化的生物特徵，如翅膀圖案和顏色，以獲得生物體如何面對環境和適應不斷變化的世界的一般原則。

This research is groundbreaking in that it uses a large amount of citizen science data and incorporates artificial intelligence to analyze biological features previously hard to be quantified, such as wing patterns and coloration, in order to obtain general principles of how organisms face their environment and adapt to the changing world.

此研究的新聞稿可由參考：連結一 、連結二

The press release for this study can be found here

2. 系統分類學、動物相和生物多樣性資訊 / SYSTEMATICS, FAUNA AND BIODIVERSITY INFORMATION

生物體如何適應其環境的變化？它們的演化歷史和多樣性是需要研究的重要課題之一。大規模的生態學研究往往需要大量的樣本採集，而物種的界定和鑑定對這類學科來說幾乎是必不可少的。我在系統分類研究專注於大異角類（Macroheterocera），這是一個相當大的單系群，包含了超過55%的鱗翅目（蝶與蛾）。在過去的十年中，我致力於國際博物館和研究機構的模式標本檢視，在資料庫中收錄它們的原始發表和可用的分布範圍，與國際分類學家保持良好的聯繫，並進行跨國的野外採集。這些努力部分體現在我多產的分類學出版物（超過50篇文章，包括21篇SCI），以及對東亞和東南亞鱗翅目生物相漸趨的掌握，例如作為重要生物相研究的主要作者之一，即2013年的《合歡山的蛾》和2020年的《南橫的蛾 第一冊》，其中包括大量的臺灣高山蛾，在變化的世界中牠們可能會受到較大的生存威脅。此外，在同儕與我所建置的《台灣產蝶蛾圖鑑（DearLep）》中，我們也積累了大量完善的公民科學資料用於進一步研究，並在其中頻繁更新綜合生物多樣性數資訊，用於學術交流和教育推廣。
How do organisms adapt to changes in their environment? Their evolutionary history and diversity are among the important topics to be studied. Large-scale ecological studies often require large collecting samples, and species delimitation and identification are essential to virtually such disciplines. I focus on Macroheterocera (macro moths), the largest monophyletic group harboring more than 55% of Lepidoptera (butterflies and moths). Over the past decade, I have dedicated myself to examining type specimens in international museums and research institutions, documenting their original publications and available distribution range in databases, maintaining good contacts with international taxonomists, and conducting transnational field collections. These efforts are partially evident in my prolific taxonomic publications (more than 45 articles, including 19 SCI’s), and the growing conviction of the East and Southeast Asian Lepidoptera fauna, e.g. as one of the main authors of important faunistic studies, Moths of Hehuanshan in 2013 and Moths of Nanheng in 2020, comprising large amounts of Taiwanese alpine moths that are potentially threatened in a changing world. In addition, the accumulation of a wealth of sound citizen science data for further research, as well as the frequently updated, comprehensive biodiversity database, DearLep established by colleagues and me, for academic communication and educational outreach.

我長期從事鱗翅目分類學和生物多樣性信息學的工作，於2019年參加了全球生物多樣性資訊機構（GBIF）的全球鱗翅目明錄工作坊，以促進生物多樣性信息可持續利用的整合。在臺灣，我作為中研院生物多樣性研究中心所正在開發的學名管理工具開發的諮詢者之一。這些努力可以讓生物多樣性數據的管理、學習和共享更加有效，同時，這些也為氣候變化下高多樣性類群的長期監測提供了基礎。
I handle Lepidoptera taxonomy and biodiversity informatics for long and just in last year, I participated in the workshop of Global Lepidoptera Checklist, Global Biodiversity Information Facility (GBIF) to promote integration for sustainable biodiversity information use. In Taiwan, I am one of the consultants for the development of the scientific name management tool being developed by the Center for Biodiversity Research, Academia Sinica. The efforts can let biodiversity data management, learning and sharing more efficiently, as well, these also provide the basis for long-term monitoring of high-diversity taxa under climate change.

綜上所述，臺灣及其周邊地區豐富的鱗翅目生物相和功能性狀資料被我們團隊良好地收錄，以開發多樣化的生態演化課題。結構良好的生物多樣性資訊網絡工具可以潛在地促進學生和不同生物群體的成員，來有效地理解和使用分類學和生態學數據進行進一步研究。
In summary, the abundant faunistic and functional trait data of Lepidoptera in Taiwan and its neighboring regions are well documented by our team to develop diverse ecological and evolutionary topics. The well-structured biodiversity information web tools can potentially facilitate students and members of different biological groups to efficiently understand and use taxonomic and ecological data for further study.