赤尾青竹絲攝食行為與溫度選擇之研究

Feeding and thermal selection of Chinese green tree viper, Trimeresurus s. stejnegeri

蔡添順(Tein-Shun Tsai)

指導教授 ： 杜銘章

國立臺灣師範大學/理學院/生命科學研究所/博士(2004年)

摘要

赤尾青竹絲為夜行性的樹棲頰窩蝮蛇。在野外主要以埋伏攝食的方式捕捉蛙類（主要食物）、鼠類、鼩鼱、蜥蜴或鳥類為食。先前已有許多研究報告以本種蛇為探討對象，但都未針對其攝食行為與生理進行深入探討。本論文的各個研究主題均與攝食相關，文中將探討其與樹棲生態、夜行性以及埋伏獵食等特性之關聯性。
在第一章中，我發現赤尾青竹絲捕捉到獵物後會緊咬不放，這樣的行為應發生於捕食較不具攻擊性、個體較小的獵物時，以及避免於樹枝上釋放獵物後不易再尋獲的問題。我在實驗室內（22 ± 1 oC）餵食小白鼠（Mus musculus）及澤蛙（Rana limnocharis），以錄影或肉眼觀察方式記錄其於完整感覺或視覺、紅外線熱感覺無法作用時的攝食行為。赤尾青竹絲在咬住獵物後會移動到樹枝上並將獵物咬離地面，此時獵物懸空並且身體因重力作用而傾斜，就像是以蛇咬處為支點的天平一般。待獵物停止掙扎後，便開始逐漸移動上下頜至獵物身體較上方的一端，然後開始吞食。由此可見，獵物被攻擊的部位決定了其被吞食的方位；而獵物被吞食的方位在攻擊前即已間接被決定。
如同其他大部份蛇類一般，赤尾青竹絲主要從老鼠頭部開始吞食，但是這種傾向在吞食澤蛙時並不明顯。與老鼠相比，有較多比例的澤蛙是從身體後端（後腳）被咬住。在完整的感覺狀況下，老鼠與澤蛙被從身體前方與後方吞入的樣本數比率分別為55：19及29：22。當視覺與紅外線熱感覺被阻斷時，吞食方位的比例則分別為14：4及14：9；此比率與完整感覺者並無顯著差別。赤尾青竹絲攝食澤蛙時所需時間明顯大於吞食老鼠者。在攝食老鼠時，從獵物剛被放入實驗箱中到被攻擊咬住的時間（T1）會在完整感覺狀況下達最低值。當視覺或熱感覺被阻斷時，T1則明顯增加。然而，在攝食澤蛙時，T1值在各個感覺處理組別中並無顯著差別。視覺與紅外線熱感覺在攝食澤蛙時似乎是較不重要的。此外，在攝食澤蛙時，從獵物前方吞食者其從咬住獵物到開始移動上下頜的時間（T2）明顯較從獵物後方吞食者來得久。在攝食老鼠時，從獵物後方吞食者，其從移動上下頜到完全吞食獵物而開始有伸舌頭動作所經過時間（T3）明顯較從獵物前方吞食者來得久。當視覺或熱感覺被阻斷時，攝食時間（T2, T3）會減少。
在第二及第三章中，我測量成熟雄蛇與雌蛇在選溫槽中的偏好溫度，並且同時探討實驗方法或裝置對選溫行為之干擾情形。我設計三個子實驗來探討熱電偶線，隱蔽處與水份提供與否對蛇類選溫行為的影響。實驗發現插入泄殖腔測量體溫的熱電偶線會促使赤尾青竹絲尾部纏繞熱電偶線並抬離溫度底板，或者阻礙其活動。舉尾的行為會造成溫度測量值偏離實際體溫。我發現在不使用熱電偶線而以錄影方式記錄蛇在選溫槽中的位置以代表其體溫的方法是可行的。選溫槽中未提供隱蔽處與水時，雄蛇攝食前後選溫值分別為23.0 ± 1.2 oC及24.7 ± 1.2 oC。當提供隱蔽處及水時，雄蛇攝食前後選溫值分別為22.5 ± 1.0 oC（選溫設定範圍Tset = 20.3 ~ 24.3 oC）及27.8 ± 0.6 oC（Tset = 26.5 ~ 28.8 oC）。而非生殖群雌蛇攝食前後選溫值分別為21.2 ± 1.4 oC（Tset = 20.6 ~ 23.8 oC）及24.8 ± 1.5 oC（Tset = 25.0 ~ 26.3 oC）。此結果顯示只有當選溫槽中有提供隱蔽處與水時，蛇類攝食後會表現趨高溫行為。雌蛇於攝食後或懷孕期間亦有趨高溫行為；懷孕雌蛇（N = 5）的偏好溫度為27.4 ± 2.0 oC。
第四章主要在研究食物大小（小於30%）與溫度（15~35 oC）對赤尾青竹絲有氧代謝、消化效率及消化速率的影響。結果發現其與一些坐等型覓食蛇類同樣具有較低的休息時代謝率（0.033 ± 0.002 ml O2/g/hr at 25 oC）。其休息時代謝率在20 oC與25 oC間差異不大（Q10 = 1.58）。除了15 oC外，其呼吸商值會在消化前期顯著增加。消化過程中的特殊動態耗能（SDA）、顛峰代謝率及顛峰代謝率增輻均隨著食物量而上升，而SDA及SDA係數則不受溫度影響。與其他響尾蛇亞科蛇類相似的是其具有較低的SDA係數值（15.8 ± 0.6%）。利用複迴歸分析以與其他坐等型覓食蛇類相比較時，可發現赤尾青竹絲之SDA反應似乎較慢且較平緩。其代謝能量係數，即消化效率（= 0.66~0.89）在15 oC達最低，而傾向在其攝食後Tset值時達高峰。此外，本研究從三方面來探討消化速率，包括排遺時間，食物骨骼消化所需時間（Tbone）以及與SDA相關的時間。最後一次排遺時間與Tbone會隨著食物量而增加，但第一次排遺時間則否。最後一次排遺時間在25 oC以上時會短於二週，但在15 oC時則會大於一個月。消化過程中耗氧量達顛峰所需時間、Tbone以及SDA完成時間分別約為最後一次排遺時間的20%，50%及80%。
第五章目的在進一步檢測赤尾青竹絲是否能藉由選溫行為來獲得最大的淨獲能值。由上一章的複線性迴歸分析中可獲知蛇重、老鼠重以及溫度，與多個消化相關因子之間的迴歸方程式。利用這些方程式，我嘗試來模擬計算赤尾青竹絲在特定攝食頻度及活動性時，每月的最大淨獲能值（Enet）。模擬的結果顯示出一些明顯變化趨勢。當食物量較小時，Enet在較低溫時方可達最大值。當攝食頻度增大時，在其攝食後Tset值時可達最高的Enet值。當攝食頻度減低時，攝食後選高溫的優點與必要性將消失，此時若只攝食10%食物量，Enet將會降為負值。由於赤尾青竹絲野外的攝食頻度偏低，故預測其將無法藉由攝食後趨高溫行為來獲得顯著較高的Enet值。我在文中提出樹棲蛇類在較低的能量效益下仍選擇高溫的可能原因。
由以上研究可發現許多嶄新的結果與推論。赤尾青竹絲夜行性、樹棲生活以及坐等型覓食的特性，對於本研究之結果有著重要影響。這些結果與推論是否同樣在野外以及其他樹棲或坐等型覓食蛇類中發生，是未來研究值得探討的主題之一。

關鍵字

攝食 ； 溫度選擇 ； 樹棲 ； 頰窩蝮蛇 ； 能量 ； 迴歸 ； 代謝率 ； 消化

並列摘要

The Chinese green tree vipers (Trimeresurus s. stejnegeri) are nocturnal and ambush prey in arboreal habitats. They feed mainly on amphibians in field, but rodents, shrews, lizards, geckos, and birds were also consumed. Several topics about this species have been studied, but no studies have focused on the behavior and physiology of feeding. In this dissertation, all subjects were associated with the feeding of this snake, and the relations of these subjects to arboreality, nocturnality, and ambush-feeding habitat were concerned.
In Chapter 1, I found Trimeresurus s. stejnegeri held on the prey after capturing it, which should be adapted to the less aggressive prey (i.e., frog), lower prey size, and avoidance the inconvenience for tracing the released prey from a twig. I fed the snake with frog (Rana limnocharis) and mouse (Mus musculus) in the laboratory (22 ± 1 oC), under three sensory deprivation conditions (intact cues, visual cue blocked, visual and infrared thermal cues blocked). The feeding behavior was recorded by videotape-recording method or naked eyes. The snake withdrawal and pulled up the prey from the ground following catching it. The prey was hanged and the body tilted, like a lever with the snake-biting site as the fulcrum. Mostly, the snake gradually moved its jaw to the upper end of the prey body and ingested it. That is, the capturing site, which should be decided before attacking the prey, determined the ingestion direction
The snake ingested mouse mainly from head, as in other snakes; but the pattern was not dominant in feeding frog. More proportion of frogs was stroked on the posterior end than that of mice. The ratio of prey ingested from anterior to posterior side was 55 : 19 and 29 : 22 for mouse and frog, respectively, under intact sensory condition. When the visual and infrared cues were blocked, the ingested direction ratio, 14 : 4 and 14 : 9 for mouse and frog respectively, did not shift significantly from above. The snakes spent more time feeding on frogs than on mice. On feeding mice, the spending time to launch successful strikes (T1) was lowest at intact cues groups and significantly increased when the visual and/or thermal cues were blocked. On feeding frogs, however, T1 did not differ significantly among the sensory deprivation conditions. Visual and thermosensory cues seem to be less important on feeding frogs (but not on feeding mice) for T. s. stejnegeri. Besides, on feeding frogs but not mice, the time from capturing prey to start moving the jaws (T2) was significantly longer when ingesting prey from the anterior end. On feeding mice but not frogs, the time from moving the jaws to start flicking the tongue following the ingestion (T3) was longer when feeding prey from the posterior end. The feeding time (T2, T3) decreased when the sensory cues were blocked.
In Chapter 2 and 3, I measured the temperature selection of adult males and females in a linear thigmothermal gradient and checked the degree of instrumental interferences. I conducted three experiments to study the possible effect of thermocouples, the influence of seclusion, and the presence of water on the temperature-selecting behavior of the snake. Thermocouples might change a snake’s preferred temperature (Tp) by causing it to lift its prehensile tail from the gradient floor or affecting its movement. With the videotape-recording method, the snake presented postprandial thermopily only when seclusion sites and water were provided in the gradient. In the absence of seclusion sites and water, the fasting and postprandial body temperature (Tb) of males was 23.0 ± 1.2 oC and 24.7 ± 1.2 oC, respectively. With seclusion sites and water, the fasting and postprandial Tb of males was 22.5 ± 1.0 oC (the set point Tset = 20.3 ~ 24.3 oC) and 27.8 ± 0.6 oC (Tset = 26.5 ~ 28.8 oC), respectively. The fasting and postprandial Tb of non-reproductive females (N = 16) was 21.2 ± 1.4 oC (Tset = 20.6 ~ 23.8 oC) and 24.8 ± 1.5 oC (Tset = 25.0 ~ 26.3 oC), respectively. Preferred temperature of females was higher after feeding or during pregnancy. Preferred temperature of pregnant snakes (N = 5) was 27.4 ± 2.0 oC.
In Chapter 4, I investigated the combined effect of meal size (below 30%) and temperature (15~35 oC) on the aerobic metabolism, digestive efficiency, and digestive rate in this study. As other sit-and-wait foraging species, the snake had lower resting metabolic rate (0.033 ± 0.002 ml O2/g/hr at 25 oC). But it showed less difference at 20 and 25 oC (Q10 = 1.58). Respiratory quotient significantly increased at the anterior part of digestion period except at 15 oC. Specific dynamic action (SDA), peak VO2 and scope of peak VO2 increased with meal size, while temperature had little effect on SDA and SDA coefficient. Similar with other crotalids, the SDA coefficient was lower (15.8 ± 0.6%) in this study. With regression analysis, I found SDA in T. s. stejnegeri responded latterly and less sharply than other sit-and-wait feeding snakes. The metabolizable energy coefficient (= 0.66~0.89) was lowest at 15 oC and tended to peak at the postprandial Tset of the snake. In addition, I investigated the digestive rate from three aspects, included gut passage time (gut movement), gastric digesting time (Tbone; chemical digestion), and the timing of SDA (digestive metabolism). The final defecation time (PTe) and Tbone, but not the first defecation time, increased with food ration. PTe was less than two weeks at above 25 oC, but was larger than one month at 15 oC. The time to peak VO2, Tbone, and duration of SDA represented about 20%, 50%, and 80% of total digestion process (PTe), respectively.
In Chapter 5, I tested whether temperature selection makes maximal net energy gain in Chinese green tree viper. I used multiple linear regressions, which expressing the effects of snake mass, mouse mass as well as temperature on digestion-associated variables, to simulate the monthly maximal net energy gain (Enet) under certain feeding frequencies and activity levels. With the energy budget model, I found some dominant trends. Enet peaked at lower temperature when snakes had less food ration. At high feeding frequency, selecting high body temperature, which closes to the postprandial Tp of the snake, could get higher Enet. When feeding frequencies lower down, Enet could be negative at 10% meal size, and the superiority or necessity to select high postprandial temperature disappear because of broad range of B80 at 20 ~ 30% meal size. Owing to the potential low feeding frequency in wild, T. s. stejnegeri may not get relative maximal Enet by postprandial behaviors. I have suggested potential reasons about why the snake selects high postprandial temperatures still under less energetic benefits.
In conclusion, many innovated ideas and investigations have been mentioned in this dissertation. Several characters (nocturnality, arboreal habitats and ambush-feeding behavior) of Chinese green tree vipers may have important influences on the results. Whether these attributes occurred in wild as well as in other arboreal and/or ambush-feeding snakes is worth to study in future researches.

並列關鍵字

Feeding ； Thermal selection ； Arboreality ； Pitviper ； Energy ； Regression ； Metabolic rate ； Digestion