金門緬甸蟒(Python bivittatus bivittatus)的活動模式、棲地利用與體溫調節

全文

(1)國立臺灣師範大學生命科學系碩士論文. 金門緬甸蟒(Python bivittatus bivittatus)的活動模式、棲地利用與體溫調節 Movements, habitat use and thermoregulation of Burmese python (Python bivittatus bivittatus) in Kinmen island. 研究生：周時平 Shih-Ping Chou 指導教授：林思民博士 Si-Min Lin 中華民國 102 年 6 月.

(2)

(3) 致謝我的研究必須仰賴許多人的協助才能完成，首先感謝林思民老師的指導，總是全力的支持我的研究，不厭其煩的與我討論，使我能夠順利完成論文；感謝賴燕雪獸醫能撥空為蟒蛇開刀植入發報器，要是我們自己來開刀，蛇應該都回不去金門了。感謝實驗室成員的協助：開啟緬甸蟒研究之路的龍哥、共患難的彥博大學長、俊文、致維、無聊大師展蔚、阿傑、恰恰、一直都很有趣的書書、芳神、李昱、夯賀、曾威、閣桓、非常聒噪的王蟲子、品如、阿寶、浥璋。若是沒有你們與我討論實驗方法與分析，幫我寄蛇或是七手八的腳的固定蛇，提供各種支援以及做一些蠢事當笑料，我想在金門的研究不會如此順利，也不會這麼精彩。感謝吳聲海老師以及大學時一起在實驗室的學長姊們，因為有你們我才能夠延續著我對自然的好奇心以及野生動物研究的熱情，直到完成這個研究。特別是蔡圃、伶樺、小六、朝仁和帝溶，是我遇到各式瓶頸時諮詢的好對象。感謝林務局的經費支持以及金門縣政府建設處的各位，尤其是立偉、光耀、元達以及怕蛇的愛瓊姊，有了你們的幫忙使實驗能更順利的進行。最後要謝謝家人以及雅涵支持我繼續念研究所，使我在求學期間沒有後顧之憂，能夠全心全意完成我的研究，我愛你們。希望我的研究能夠為科學盡一份心力，讓更多人了解並愛護緬甸蟒，感謝這些一直被我跟蹤的蟒蛇，希望你們能夠平平安安的在金門生活下去。得之於人者太多，出之於己者太少，要感謝的人太多了，就謝天吧！.

(4) Table of contents 中文摘要…………………………………………………………………3 Abstract…………………………………...………………………………4 Introduction………………………………………………………………5 Materials and methods…………..…………………………….………….8 Results…………………………..……………..…………………..……14 Discussion………………………………………………………….……16 References……………………………………………………...….……22 Table 1 Gender, morphometric data and tracking details ………………26 Table 2 Home range size of the pythons in the four seasons…...……….27 Table 3 Dunn’s test values on daily movements……………...………....28 Table 4 Macrohabitat use of the pythons…………………….....……….29 Table 5 Adjusted standardized residuals tests of macrohabitat use...…...29 Table 6 Microhabitat use of the pythons………………...………....…...30 Table 7 Adjusted standardized residuals tests of microhabitat use……...30 Table 8 Dunn’s test value of canopy cover……………………..……….31 Table 9 Air temperature and pythons’ body temperature..........................32 Table 10 Home range comparison from literature………………………33 Figure 1 Python records and military forces……………………………34 Figure 2 Phylogeny and gene genealogy of pythons…………….……...35 Figure 3 Home range sizes of pythons during the tracking period…..….36 Figure 4 Home range sizes of all pythons in different seasons……….37 Figure 5 Moving distance per day in all seasons…………….………….38 Figure 6 Moving distance per day during daytime and nighttime………39 Figure 7 Percentage of macrohabitat use of the pythons………………..40 1.

(5) Figure 8 Percentage of microhabitat use of the pythons………………..40 Figure 9 Canopy cover where the python stayed……………………….41 Figure 10 Pythons’ body temperatures in daytime……………………42 Figure 11 Pythons’ body temperatures in nighttime………..…………...43 Figure 12 Reduced body condition of a python…………….....………..44 Figure 13 Body temperature of pythons during a cold spell...………….45 Appendix 1 Definition of macrohabitats……….……….………………46 Appendix 2 Home range of all pythons..…………….………………….48 Appendix 3 Moving route of all pythons………...……………………..51. 2.

(6) 摘要緬甸蟒(Python bivittatus bivittatus)是世界上最大型的蛇類之一，因為在美國為外來入侵種而惡名昭彰。由於緬甸蟒的入侵，許多關於緬甸蟒的研究受到高度的重視；然而針對於原生地族群的研究目前仍舊極度缺乏。在本研究中，我們將針對於原生地緬甸蟒的基礎生態資料進行調查，例如：活動模式、活動範圍、棲地利用以及體溫調節，以利未來為原生地緬甸蟒保育方面提供有效的管理策略。本研究自 2010 年五月至 2013 年二月以無線電追蹤進行資料收集，總共追蹤了 12 隻個體。結果顯示，緬甸蟒在夏季及秋季的夜晚頻繁活動，在溫暖的季節裡一個晚上的移動距離可以達到至少 1.2 公里，而一年內的活動範圍可達到 892.7 公頃。隨著季節的變化，緬甸蟒也有季節性的棲地利用，同時藉由改變棲地利用來調節體溫。在夏季會選擇高覆蓋度的棲地，並頻繁的使用沼澤、灌叢、草地以及森林的環境，避免白天體溫過高。而冬季則會選擇廢棄的軍方地下通道或是洞穴來躲藏，以躲避寒冬的低溫。. 關鍵詞：棲地利用、活動範圍、金門、蟒科、無線電追蹤. 3.

(7) Abstract The Burmese python Python bivittatus bivittatus, one of the largest snake species in the world, is famous due to its invasion to USA. Although quite a few studies have been conducted on the invasive population, scientific information of in the native range is still extremely scarce. In this study, we aim to investigate the basic ecological information of the python such as activity pattern, home range, habitat use, and thermoregulation in order to provide a proper management strategy for their conservation. From May 2010 to February 2013, radio telemetry was applied to track 12 individuals. The results showed that the pythons have a highest activity in summer and autumn nights, with the potential to move more than 1.2 km within a single night in warm seasons. The home range size may exceed 892.7 hectare within a year. The results also showed that seasonal habitat use could effectively help regulating their body temperature to avoid from extreme temperatures. The pythons chose high canopy cover habitat and frequently used marshes, shrubs, grasslands, forests in summers, while underground tunnels and caves were commonly used as refugia in winters.. Keywords: habitat use, home range, Kinmen, Pythonidae, radio telemetry. 4.

(8) Introduction The Burmese python Python molurus bivittatus Kuhl, 1820, sometimes evaluated as a valid species P. bivittatus (Jacobs et al. 2009), is one of the largest snake species in the world. With its rareness, potential commercial value, risk of poaching, and ongoing population decline, this snake is listed under Category II in CITES (Stuart et al. 2012) and is listed as a Class I protected species in China (Zhao 1998; Zhao et al. 1998). The vast majority of the habitat and range of this species is below 200 m in elevation (Barker and Barker 2008). Primary diets are mammals and birds (Dorcas and Willson 2011; Dove et al. 2011; Dorcas et al. 2012). Although an 8.22 m record in total length was widely reported from the literature, the maximal length of Burmese python seems to be less than 5.74 m after reevaluation on these uncorroborated records (Barker et al. 2012). The natural distribution of the Burmese python includes South and Southeast Asia, including Indo-China Peninsula, southern China, and in Indonesia on Java, southern Sulawesi, Bali and Sumbawa. However, the invasive population of this species in Florida of the US has become a focal issue in conservation biology and invasive species managements. In recent years, concerning research were published including the distribution of pythons in the past and future (Pyron et al. 2008; Rodda et al. 2009; Rodda et al. 2011); their potential impacts on native species (Dove et al. 2011; Dorcas et al. 2012); dispersal and establishment of this invasive population (Willson et al. 2010; Hart et al. 2012); thermoregulation (Mazzotti et al. 2010; Avery et al. 2010; Snow et al. 5.

(9) 2010; Dorcas et al. 2011); and remove attempts (Reed et al. 2011). However, some conclusions have caused controversy among scientists because most of these deductions were based on indefinite inferences, since research on native Burmese python population is extremely rare (Pyron et al. 2008; Barker and Barker 2010; Rodda et al. 2011). Before the invasive population established, most studies on Burmese python were based on laboratory experiments or captive breeding individuals (Hutchison et al. 1966; Wang et al. 2003; Secor and White 2010). Under this circumstance, research on native python population is critical (Dorcas and Willson 2011). Research on spatial resource use is important to species' life history. Habitat use can provide critical information on species conservation or remove of invasive species (Rainbolt and Coblentz 1999; Harvay and Weatherhead 2006; Lavoie et al. 2007; Oh et al 2010). Past research of snakes confirmed the seasonal habitat use, which was influenced by sex, seasonally physical factors, prey abundance, reproductive condition, and the condition of macrohabitat and microhabitat (Shine and Fitzgerald 1996; Webb and Shine 1997; Wund et al. 2007; Slip and Shine 1988a). Seasonal habitat use usually accompanied by seasonal thermoregulation (Shine and Lambeck 1985; Slip and Shine 1988b; Heard et al. 2004; Row and Blouin-Demers 2006). A critical factor for the spread of the python population in the US is the temperature tolerance to overcome the cold winters. As ectotherms, most reptiles are reliant on specific components of their habitat in order to maintain appropriate body temperatures (Huey 1982). Body temperature also affects on physiological and developmental processes (Peterson et al. 6.

(10) 1993), and consequently contributes to fitness (Christian and Tracy 1981; Huey and Kingsolver 1989). Quite a few studies have been conducted on the python of Florida. The predominant habitats of python in Florida included tree islands, marl prairie, mangrove forest, sawgrass marsh, hardwood hammock, and artificial habitats such as levees, canals, and roads (Snow et al. 2006; Dorcas and Willson 2011; Mazzotti et al. 2010). In the cooler months, pythons hibernates under rock or grass heaps and in burrows, caves, stone ruins, holes in riverbanks, or hollow trees (Snow et al. 2006). Despite of these information, detailed and quantifiable research about Burmese python is still scarce in their native range, including home range, moving distance, habitat use and thermoregulation. In order to provide a better evaluation on the invasive population in Florida, research on their native habitat is critical. The demilitarized zone of Kinmen provides a valuable chance for the python study. The Kinmen islands, comprising Kinmen (134.3 km2) and Lesser Kinmen (14.9 km2), is located roughly 2 km from the southeastern coastline of China. During the Cold War period, bombardments by China destroyed most of the buildings and vegetation on the island, possibly extirpating the python population. In recent years, political tensions have subsided, and 95% of the military forces have been withdrawn from the islands, allowing pythons and other wildlife populations to recover 40 years after being extirpated (Fig. 1). Increased python predation on small livestock has recently become a nuisance to farmers, villagers, and aboriginal people, who are unaware of old python records from the island and therefore regard them as invasive pests. However, molecular phylogeny and 7.

(11) haplotype networks showed a close relationship between Kinmen and Chinese populations, rejecting the speculation from local people that pythons were introduced from Southeast Asia (Fig. 2). Since Kinmen Island has similar latitude and climatic factors as the Everglades National Park in southern Florida (USA), ecological and physiological research on the Kinmen population has potential value for comparative studies of habitat selection and niche modeling. In this study, we aim to study the basic ecological information from the Burmese python by means of radio telemetry with temperature data logger implantation. We aim to answer the following questions including: (1) Is there any differences on home range size and movements among different seasons? (2) Is there any differences on microhabitat and macrohabitat utilization among different seasons? (3) How do the pythons regulate their body temperature to avoid from extreme body temperature on the island with high climate fluctuation in Kinmen?. Materials and methods Study site We conducted this study from May 2010 to February 2013 in Kinmen, a coastal island located at the southeast of China (longitude 118.210°E to 118.473°E, latitude 24.386°N to 24.529°N). The area of this island is 135.7 square kilometers, with a highest elevation of 253 meters. Seasonal climatic cycles are characterized by hot summers (mean maxium temperature of 35°C in July) and cool winters (mean minimum temperature of 6.5°C in January). Annual precipitation is approximately 8.

(12) 1047 mm. Kinmen is a highly developed island, with large areas occupied by agricultural farmlands and abandoned military camps. Because of the highly populated soldiers and the frequent military activities during the Cold War, there is no native forest remained on the island. Secondary forests, shrubs, grasslands, ponds and swamps, and granitic hills are major landscapes on the island.. The Snakes and the surgeries Twelve snakes were used for radio trackings, including 5 males and 7 females. Serial numbers and body sizes of the snakes are listed in Table 1. In order to reduce stress to the snakes and the risk of injury to the investigators, all measurements were taken at their first capture under narcosis using a gas anesthesia system by a licensed veterinarian. Two temperature data loggers (iButton DS1922L, MAXIM Ltd, California, United States of America) with a radio transmitter (R1550B, ATS Ltd, Minnesota, United States of America) were implanted into the body of the snakes. The two data loggers started with a half-hour lag and alternately recorded body temperature of the pythons every 30 minutes lasting for more than 11 months. The data loggers and transmitter were implanted under the hypodermis in sterile conditions and isoflurane anaesthesia. After anaesthesia, all snakes were measured snout-vent length (SVL) to the nearest cm using a metric tape, weighed to the nearest kg using an electronic loadcell scale, and sexed by probing for hemipenes. The mass of all experimental pythons were heavy enough (transmitter < 0.7％ body mass) to carry the transmitters. The snakes were released to 9.

(13) the original habitat in Kinmen after the cut of the surgery has recovered.. Radio telemetry We tracked pythons two times a day (morning and evening) by radio receiver (SIKA Radio Tracking Receiver, Biotrack Ltd, Dorset, United Kingdom) and directional antenna (Yagi Antenna, Biotrack Ltd, Dorset, United Kingdom). Snakes were traced to a distance that could reliably locate their position and microhabitat, but with no disturbance. We recorded the global positioning system (GPS) data of all the python sites by GPS handhelds (Oregon 550t, GARMIN Ltd, New Taipei City, Taiwan). Macrohabitat and microhabitat types were recorded after the position of the snakes were founded. The percentage of canopy cover at each perching site was evaluated by using a spherical densitometer (Lemmon 1957) on the second day after the python has left the site. In some occasions, when the pythons entered into military protection areas where the scientists were not allowed to enter, we recorded the GPS data by triangulation. All the data were recorded a week after the pythons were released in order to prevent from the bias of a newly released snake. Location of the snakes were divided by season (spring: March-May, summer: JuneAugust, autumn: September-November, winter: December-February) and mapped onto a base map using software ArcGIS version 9. After the tracking period, we retrieved the data loggers by conducting a same surgery from the recovered snakes. The snakes were finally released to 10.

(14) the origin habitat with only a lifelong passive integrated transponder tag.. Home range and movements The GPS data was analysed by using the Geographic Information System (GIS) software ArcMap (ArcGIS 9, ESRI Ltd, California, United States of America). Hawth’s Analysis Tools (Beyer, 2004) were applied to calculate the distance of each movement from the four seasons, and to calculate home range of each snake in each season by minimum convex polygon (MCP) method (Mohr, 1947). The home range was log-transformed to meet the assumptions of analysis of variance (ANOVA). Since males and females did not represent statistic significance, the two sexes were combined in the following analyses. ANOVA was applied to determine if python home range differs among season. The Tukey-Kramer honestly significant difference (HSD) test was used to decide the rank of the seasons. We used Kruskal-Wallis test to detect whether there was significant difference on movements among the four seasons. After Kruskal-Wallis test, the Dunn’s test (Dunn, 1964) was used to do post hoc to decide rank of the seasons. Subsequent calculation of Dunn’s test in each season was made as follows:. Where. is the number of observations for the treatment. the number of observations for the treatment 11. , and. ,. is. is the number of.

(15) observations for total treatment.. indicated. significant difference between treatment and treatment , where the average of the ranks for the treatment , ranks for the treatment ,. is. is the average of the. is the number of treatments, and. the z-value from normal distribution significance of test at. is =0.05.. Movements in daytime and nighttime within each season was further separated and compared by Wilcoxon rank sum test to evaluate their activity pattern in different seasons.. Macrohabitat use and microhabitat use The categories of macrohabitat were defined as the following six types: secondary forest, planted forest, agricultural land, granitic woodland, wetland, and artificial habitat (more detail in supplementary data 1). Categories of microhabitat were classified into five types: in the shrub, on the grasslands, within log or litter, in a cave or a burrow, and in the water. For each season, we performed the. goodness of fit test to. determine whether the observedfrequency between macrohabitat types differed among seasons. The adjusted standardized residuals test was used to determine which of the macrohabitat types significantly favored or avoided in each season. Subsequent calculation of adjusted standardized residuals test in each season was applied as follows:. 12.

(16) Where ,. is the observed values in macrohabitat type. is expected values in macrohabitat type. sample size in macrohabitat type , and. is total sample size.. and season ,. and season is total. is total sample size in season ,. follows a normal distribution with mean. 0 and standard deviation 1, all absolute values over 2 indicate significant differences between the observed and the expected values. Identical calculation was applied in the microhabitat analysis.. Canopy cover Forest overstory density of the perching sites was shown by percentage calculated as follows:. To determine whether canopy cover differ among the seasons, Kruskal-Wallis test was applied. Dunn’s test was subsequently used to do post hoc after Kruskal-Wallis test to decide of the rank of seasonal canopy coverage.. Thermoregulation We picked the python body temperature data from 12 noon and 12 midnight records to symbolize the diurnal and night body temperatures. These temperatures were compared with air temperature from Kinmen Weather Station, Central Weather Bureau by pair-t test or Wilcoxson signed-rank tests. 13.

(17) We used JMP (JMP version 7.0, SAS Institute, USA) for statistical analyses, reported all means ± 1 SE and significance of test at =0.05.. Results Home range and movements There was no significant difference in home range size between males and females during the experimental duration (t=-2.00, d.f.=10, p=0.07, Table 1) (Fig. 3). When divided by seasons, home range significantly differed (F=10.57, d.f.=3,24, p<0.0001, Table 2). Home range of the pythons in summer was significantly higher than that in spring and winter, while that in spring was significantly higher than in winter (Fig. 4). There was no significant difference between summer and autumn (Tukey-Kramer HSD: spring vs. summer, difference=0.3447, p<0.05; spring vs. autumn, difference=-0.275, p>0.05; spring vs. winter, difference=-0.7288, p>0.05; summer vs. autumn, difference=-1.0462, p>0.05; summer vs. winter, difference=1.6422, p<0.05; autumn vs. winter, difference=1.0225, p<0.05). Movement distances (meters/day) significantly differed among the seasons (Kruskal-Wallis: χ2=62.56, d.f.=3, p<0.0001, Fig. 5). The average daily movements in spring, summer and autumn were significantly higher than that in winter (Table 3 and Fig. 5). When classified by time, movements at night was significantly higher than that at daytime (p<0.0001, Fig. 6) in summer and autumn. In contrast, movements represented no significant differences between day and night in winter and spring (p>0.05, Fig. 6). 14.

(18) Macrohabitat use Macrohabitat use of the python varied significantly among the seasons (χ2=650.57, d.f.=15, p<0.0001, Fig. 7). In winter, the pythons preferred planted forests, agricultural lands and granitic woodlands (Table 5). The habitat in spring retained the three major types used in winter, but represented a higher ratio to use artificial habitats. In contrast, macrohabit use in summer turned into secondary forests and ponds, and showed prominent avoidance to other habitat types (Table 5). Situation in autumn was similar to that in summer, except the higher usage for artificial habitat, planted forest, and agricultural lands (Table 5).. Microhabitat use Microhabitat use represented significant differences among the seasons (χ2=825.09, d.f.=12, p<0.0001, Fig. 8). In spring, pythons preferred to use the microhabitat of grasslands and shrubs and avoided staying in the water (Table 7). In contrast, the frequency to use water and shrubs significantly increased in summer (Table 7). In autumn, water bodies, log and litter, and grassland were frequently used (Table 7). The microhabitat usage pattern was worth of especially noted in winter: they spent most of the time in caves or burrows.. Canopy cover Forest overstory density of the pythons varied significantly between 15.

(19) the seasons (χ2=99.64, d.f.=3, p<0.0001, Fig. 9). The coverage in spring and summer were significantly higher than that in autumn and winter. There was no significantly difference between spring and summer, nor between autumn and winter (Table 8 and Fig. 9).. Thermoregulation Body temperatures of the pythons was significantly higher than air temperatures in winter and spring in both daytime and nighttime (p<0.0001, Table 9, Fig. 10 and 11). In summer, the body temperature was significantly higher at night (p<0.0001), but significantly lower at daytime (p=0.029, Table 9). In autumn, the body temperature was significantly higher at night (p<0.0001) and represented no difference at daytime (p=0.68).. Discussion Annual activity trends of the python Our results indicated that the pythons use a variety of macro and microhabitat to fit their physiological demands in different seasons. In spring, pythons hid under grassland and shrub in artificial habitat, planted forest, agricultural land and granitic woodland, with higher canopy coverage. The pythons used variable macrohabitat types possibly in the purpose to seek for its prey after fasted in cool winter. In this season, we found the pythons basking in daytime, showing a higher body temperature than air temperature. The pythons are probably taking the 16.

(20) advantage of basking behavior to increase their metabolism. As the temperature increased by season, the pythons gradually reduced the frequency of basking and chose microhabitat of higher canopy coverage to avoid hyperpyretic situation from sunlight. In summer, the pythons have a largest home range size among any other seasons (Fig. 4). They dispersed a wide range across the study area during summer, showing a most active mobility in summer nights. In this season, they frequently used secondary forest and wetland, and preferred microhabitats with higher canopy coverage such as shrubs. Some individuals stayed in the water during this period and might extended for quite a few day or weeks. The air temperature was so high that the python hid themselves to regulate body temperature during daytime. This pattern is similar to Southern African python (Python natalensis), which regulates body temperature by burrows, water, or Southern African Bushveld (Alexander 2006). Pythons are still active at autumn’s nighttime with an excellent mobility up to 1220 m movement within several nocturnal hours. In this season, the macrohabitat use was roughly the same as that in summer, but the microhabitat turned into habitats with low canopy cover age such as grasslands, or log and litter. Long-distance movements at this season are deduced to find the proper wintering refugia. The demand for sunlight increased with the air temperature decreased lead them to choose low canopy cover microhabitat. Owing to the comparatively stable temperature in the water, a few individuals still soaked in the water at this period. Not surprisingly, the home range size and moving distance was 17.

(21) smallest among all seasons in cool winter. The activity of snakes reduced as the air temperature decrease (Slip and Shine 1988a; Shine and Fitzgerald 1996; Webb and Shine 1997; Heard et al.2004; Row and Blouin-Demers 2006). The underground tunnels, caves and burrows were commonly used as refugia, and we found the microhabitat canopy cover was lower than warm season. The use of refugia could effectively regulate body temperature to avoid extreme temperature in winters (Slip and Shine 1988; Heard et al. 2004). We deduced that the pythons might choose microhabitat with lower canopy coverage in order to rise their body temperature in a more efficient manner, which was also shown from the invasive python population in Florida (Snow et al. 2006).. Characteristic of the python in Kinmen The python population in Kinmen seems to grow much smaller than the Southeast Asian and the Florida invasive populations. The maximum length record was only 358 cm from a mature female. Lack of intermediate-to large-sized prey items, as well as the violently seasonal climate change might have caused this phenomenon. Owing to the limitation in financial supports and man powers, were not able to trace sufficient samples providing powerful analyses between sexes and among season. However, the home range size of Burmese python has shown to be larger than any other snakes even members of family Pythonidae (Table 10) (Slip and Shine 1988b; Shine and Fitzgerald 1996; Webb and Shine 1997; Brito 2003; Pearson et al. 2005). The largest two home range sizes were 892.7 (PM35) and 768.1 hectares 18.

(22) (PM28), respectively (Table 1). Burmese python exhibits both diurnal and nocturnal activity in Florida and in its native range. Pythons in Florida was primarily nocturnal in summer and diurnal in winter (Snow et al. 2006). In Kinmen island, pythons showed nocturnal during summer and autumn, but showed both diurnal and nocturnal in spring and winter. The abandoned military structures such as bomb shelters, underground tunnels and blockhouses provided refugia for pythons. The abundance of such refugia may dictate python density (Dorcas and Willson 2011). In the study from Florida, pythons wintering in artificial habitats showed a higher survival rate than those in natural habitats (Mazzotti et al. 2010). However, python still suffered from the low and fluctuating temperature in Kinmen. One of the pythons (PM21) was found with serious respiratory infection during radio tracking (Fig. 12). Another example came from the death of another python (PM17) outside its refugium in winter. It showed that the climate in Kinmen might have reached their tolerance of this species. We didn’t find any female python breeding through our tracking duration. However, several juveniles were captured in this period. It was deduced that the pythons do not reproduce every year in Kinmen island. The prominent seasonal climate fluctuation, as well as food resource limitation, might have led to this result. This phenomenon might be congruent to previous study from the reticulate python Python reticulates (Shine et al. 1999), and possibly fit the situation in Florida (P. T. Andreadis, personal communication).. 19.

(23) How did the pythons survive in the cold climate? We found that the pythons could maintain their body temperature higher than air temperature in refugia during the cold spell (the record in 17 December 2010, air: 4℃; PM05: 23.1℃; PM13: 17.6℃; PM16: 21.7℃). We picked the body and the air temperature data during one of the most severe cold spell in 2010 winter (14 December 2010 to 20 December 2010, Fig. 13), and compared to most severe cold spell occurred in Florida, 2010 winter (Mazzotti et al. 2010). We found an interesting phenomenon: the body temperature from native pythons was stable when air temperature was reducing; on the contrary, the body temperature from invasive python fluctuated with air temperature, and finally resulted to mortality at that cases. The most debatable issue about Burmese pythons in the United States is whether the invasive pythons could disperse to other States north from Florida. Several studies proposed that the pythons will invade most States in south of America (Rodda et al. 2009; Rodda et al. 2011). However, some others proposed the reverse due to the severe limitation by ecological constraints (Pyron et al. 2008). Because pythons are strongly limited to the small area of suitable environmental conditions, the predicted distribution of the python will be strongly limited by their microhabitat availability (Pyron et al. 2008). It showed the physical factor in environment was possibly not the most important part of Burmese python distribution. Although the warm climate of the Florida probably makes underground shelter unnecessary, microhabitats that provide retreats during cold winter weather may be a limiting factor in cooler 20.

(24) regions (Dorcas and Willson 2011). In our research, we found python used refugia to avoid extreme body temperature in winters. This result demonstrated that the suitable habitat can protect pythons from cold-induced mortality. It’s possible that pythons expanding north of their current in South Florida will survive only in regions where large underground retreats exist (Dorcas and Willson 2011). In conclusion, the Burmese python in Kinmen showed prominent seasonal variation in movement pattern, home range, and habitat use, and temperature regulation. Seasonal habitat use could effectively regulate body temperature, avoided hyperpyretic situation from sunlight in summers and resisted harm from low temperature in winters. Although with limited sample size, this study provides the very first ecological information for this species, which is a focal target both in biological conservation and invasive managements.. 21.

(25) Reference Alexander, G. J. (2006). Thermal biology of the Southern African Python (Python natalensis): does temperature limit its distribution? In R. W. Henderson & R. Powell (Eds.), Biology of the boas and pythons. Eagle Mountain, USA: Eagle Mountain Publishing. Avery, M. L., Engeman, R. M., Keacher, K. L., Humphrey, J. S., Bruce, W. E., Mathies, T. C., & Mauldin, R. E. (2010). Cold weather and the potential range of invasive Burmese pythons. Biological Invasions, 12(11), 3649-3652. Barker, D. G., & Barker, T. M. (2008). The distribution of the Burmese Python, Python molurus bivittatus. Bulletin of the Chicago Herpetological Society, 43(3), 33-38. Barker, D. G., & Barker, T. M. (2010). A critique of the analysis used to predict the climate space of the Burmese python in the United States by Rodda et al. (2008, 2009) and Reed and Rodda (2009). Bulletin of the Chicago Herpetological Society, 45(6), 97-106. Barker, D. G., Barten, S. L., Ehrsam, J. P., & Daddono, L. (2012). The corrected lengths of two well-known giant pythons and the establishment of a new maximum length record for Burmese pythons, Python bivittatus. Bulletin of the Chicago Herpetological Society, 47(1), 1-6. Beyer, H. L. (2004). Hawth's Analysis Tools for ArcGIS. Available at http://www.spatialecology.com/htools. Brito, J. C. (2003). Seasonal variation in movements, home range, and habitat use by male Vipera latastei in northern Portugal. Journal of Herpetology, 37(1), 155-160. Christian, K. A., & Tracy, C. R. (1981). The effect of the thermal environment on the ability of hatchling Galapagos land iguanas to avoid predation during dispersal. Oecologia 49, 218-223. Dorcas, M. E., & Willson, J. D. (2011). Invasive pythons in the United States: ecology of an introduced predator: University of Georgia Press, USA. Dorcas, M. E., Willson, J. D., & Gibbons, J. W. (2010). Can invasive Burmese pythons inhabit temperate regions of the southeastern United States? Biological Invasions, 13(4), 793-802. Dorcas, M. E., Willson, J. D., Reed, R. N., Snow, R. W., Rochford, M. R., Miller, M. A., Meshaka, W. E., Andreadis, P. T., Mazzotti, F. J., Romagosa, C. M., & Hart, K. M. (2012). Severe mammal declines coincide with proliferation of invasive Burmese pythons in Everglades National Park. Proceedings of the National Academy of Sciences of the United States of America, 109, 2418-2422. 22.

(26) Dove, C. J., Snow, R. W., Rochford, M. R., & Mazzotti, F. J. (2011). Birds Consumed by the Invasive Burmese Python (Python molurus bivittatus) in Everglades National Park, Florida, USA. The Wilson Journal of Ornithology, 123(1), 126-131. Dunn, O. J. (1964). Multiple Comparisons Using Rank Sums. Technometrics, 6(3), 241-252. Hart, K. M., Schofield, P. J., & Gregoire, D. R. (2012). Experimentally derived salinity tolerance of hatchling Burmese pythons (Python molurus bivittatus) from the Everglades, Florida (USA). Journal of Experimental Marine Biology and Ecology, 413, 56-59. Harvey, D. S., & Weatherhead, P. J. (2006). A test of the hierarchical model of habitat selection using eastern massasauga rattlesnakes (Sistrurus c. catenatus). Biological Conservation, 130(2), 206-216. Heard, G. W., Black, D., & Robertson, P. (2004). Habitat use by the inland carpet python (Morelia spilota metcalfei: Pythonidae): Seasonal relationships with habitat structure and prey distribution in a rural landscape. Austral Ecology, 29, 446-460. Huey, R. B. (1982). Temperature, physiology and the ecology of reptiles. In C. Gans & F. H. Pough (Eds.), Biology of the reptilia, Ecology and Behaviour (Vol. 12). London: Academic Press. Huey, R. B., & Kingsolver, J. G. (1989). Evolution of thermal sensitivity of ectotherm performance. Trends in Ecology & Evolution 4, 131-135. Hutchison, V. H., Dowling, H. G., & Vinegar, A. (1966). Thermoregulation in a Brooding Female Indian Python, Python molurus bivittatus. Science, 151, 694-695. Jacobs, H. J., Auliya, M., & Böhme, W. (2009). Zur Taxonomie des Dunklen Tiger pythons, Python molurus bivittatus KUHL, 1820, speziell der Population von Sulawesi. Sauria, 31(3), 5-16. Lavoie, C., Donlan, C. J., Campbell, K., Cruz, F., & Carrion, G. V. (2006). Geographic tools for eradication programs of insular non-native mammals. Biological Invasions, 9(2), 139-148. Lemmon, P. E. (1957). A new instrument for measuring forest overstory density. Journal of Forestry, 55(9), 667-668. Macartney, J. M., Gregory, P. T., & Larsen, K. W. (1988). A tabular survey of data on movements and home ranges of snakes. Journal of Herpetology, 22(1), 61-73. Mazzotti, F. J., Cherkiss, M. S., Hart, K. M., Snow, R. W., Rochford, M. R., Dorcas, M. E., & Reed, R. N. (2011). Cold-induced mortality of invasive Burmese pythons in south Florida. Biological Invasions, 13(1), 143-151. 23.

(27) Mohr, C. O. (1947). Table of equivalent populations of North American small mammals. American Midland Naturalist, 37, 223-249. Oh, D. H., Moteki, S., Nakanish, N., & Izawa, M. (2010). Effects of human activities on home range size and habitat use of the Tsushima leopard cat Prionailurus bengalensis euptilurus in a suburban area on the Tsushima Islands, Japan. Journal of Ecology and Field Biology, 33(1), 3-13. Pearson, D., Shine, R., & Williams, A. (2005). Spatial ecology of a threatened python (Morelia spilota imbricata) and the effects of anthropogenic habitat change. Austral Ecology, 30, 261-274. Peterson, C. R., Gibson, R. A., & Dorcas, M. E. (1993). Snake thermal ecology: the causes and consequences of body temperature variation. In R. A. Seigel & J. T. Collins (Eds.), Snakes: ecology and behavior. Caldwell: The Blackburn. Pyron, R. A., Burbrink, F. T., & Guiher, T. J. (2008). Claims of potential expansion throughout the U.S. by invasive python species are contradicted by ecological niche models. PLoS One, 3(8), e2931. Rainbolt, R. E., & Coblentz, B. E. (1999). Restoration of insular ecosystems: control of feral goats on Aldabra Atoll, Republic of Seychelles. Biological Invasions, 1, 363-375. Reed, R. N., Hart, K. M., Rodda, G. H., Mazzotti, F. J., Snow, R. W., Cherkiss, M., Rozar, R., & Goetz, S. (2011). A field test of attractant traps for invasive Burmese pythons (Python molurus bivittatus) in southern Florida. Wildlife Research, 38, 114-121. Rodda, G. H., Jarnevich, C. S., & Reed, R. N. (2009). What parts of the US mainland are climatically suitable for invasive alien pythons spreading from Everglades National Park? Biological Invasions, 11(2), 241-252. Rodda, G. H., Jarnevich, C. S., & Reed, R. N. (2011). Challenges in identifying sites climatically matched to the native ranges of animal invaders. PLoS One, 6(2), e14670. Row, J. R., & Blouin-Demers, G. (2006). Thermal quality influences effectiveness of thermoregulation, habitat use, and behaviour in milk snakes. Oecologia, 148(1), 1-11. Secor, S. M., & White, S. E. (2010). Prioritizing blood flow: cardiovascular performance in response to the competing demands of locomotion and digestion for the Burmese python, Python molurus. J Exp Biol, 213(1), 78-88. Shine, R., Ambariyanto, Harlow, P. S., & Mumpuni. (1999). Reticulated pythons in Sumatra biology, harvesting and sustainability. Biological Conservation, 87, 349-357. Shine, R., & Fitzgerald, M. (1996). Large snakes in a mosaic rural landscape: the 24.

(28) ecology of carpet pythons Morelia spilota (Serpentes: Pythonidae) in coastal eastern Australia. Biological Conservation, 76, 113-122. Slip, D. J., & Shine, R. (1988a). Habitat Use, movements and activity patterns of free-ranging diamond pythons, Morelia spilota spilota (Serpentes : Boidae): a Radio telemetric Study. Australian wildlife research, 15, 515-531. Slip, D. J., & Shine, R. (1988b). Thermoregulation of free-ranging diamond pythons, Morelia spilota (Serpentes, Boidae). Copeia, 4, 984-995. Snow, R. W., Krysko, K. L., Enge, K. M., & Oberhofer, L. (2006).Introduced populations of Boa constrictor (Boidae) and Python molurus bivittatus (Pythonidae) in southern Florida. In R. W. Henderson & R. Powell (Eds.), Biology of the boas and pythons. Eagle Mountain, USA: Eagle Mountain Publishing. Snow, R. W., Wolf, A. J., Greeves, B. W., Cherkiss, M. S., Hill, R., & Mazzotti, F. J. (2010). Thermoregulation by a brooding Burmese python (Python molurus bivittatus) in Florida. Southeastern Naturalist, 9(2), 403-405. Stuart, B., Nguyen, T. Q., Thy, N., Grismer, L., Chan-Ard, T., Iskandar, D., Golynsky, E., Lau, M. W. N. (2012). Python bivittatus. IUCN 2012. IUCN Red List of Threatened Species, Version 2012.2. Wang, T., Zarr, M., Arvedsen, S., Vedel-Smith, C., & Overgaard, J. (2003). Effects of temperature on the metabolic response to feeding in Python molurus. Comparative Biochemistry and Physiology, 133, 519-527. Webb, J. K., & Shine, R. (1997). A field study of spatial ecology and movements of a threatened snake species, Hoplocephalus bungaroides. Biological Conservation, 82, 203-217. Willson, J. D., Dorcas, M. E., & Snow, R. W. (2010).Identifying plausible scenarios for the establishment of invasive Burmese pythons (Python molurus) in Southern Florida. Biological Invasions, 13(7), 1493-1504. Wund, M. A., Torocco, M. E., Zappalorti, R. T., & Reinert, H. K. (2007). Activity ranges and habitat use of Lampropeltis getula getula (Eastern Kingsnakes). Northeastern Naturalist, 14(3), 343-360. Zhao, E. M. (1998). China red data book of endangered animals–Amphibia and Reptilia. Beijing: Science Press (in Chinese). Zhao, E. M., Huang, M. H., &Zong, Y. (1998). Fauna Sinica, Reptilia, Squamata, Serpentes (Vol. 3). Beijing: Science Press (in Chinese).. 25.

(29) Table 1. Gender, morphometric data and tracking details of each python. Python number. Sex. SVL (cm). Weight (kg). Home range (ha). Tracking days. Number of location. Tracking period. PM01. F. 275.0. 9.85. 113.3. 41. 70. 25 May 2010-11 September 2010. PM05. M. 214.0. 7.12. 246.3. 69. 137. 10 July 2010-15 June 2011. PM12. F. 223.0. 8.22. 22.4. 20. 30. 25 May 2010-26 October 2010. PM13. M. 186.3. 4.22. 113.4. 15. 31. 27 October 2010-18 July 2011. PM15. F. 321.5. 17.15. 677.1. 269. 537. 14 April 2011-7 May 2012. PM16. F. 186.0. 3.94. 222.1. 27. 53. 27 October 2010-21 July 2011. PM17. F. 300.5. 24.74. 136.5. 210. 419. 14 April 2011-28 January 2012. PM21. F. 221.5. 8.54. 65.4. 31. 61. 14 August 2012-25 February 2013. PM23. M. 203.5. 5.39. 309.3. 154. 306. 27 August 2011-22 July 2012. PM28. M. 206.0. 5.10. 768.1. 169. 334. 11 August 2011-19 July 2012. PM30. F. 311.0. 21.60. 85.1. 62. 124. 27 August2011-28 November 2011. PM35. M. 294.0. 19.58. 892.7. 66. 130. 2 February 2012-2 December 2012. 26.

(30) Table 2. Home range size (hectare) of the pythons in the four seasons. Python number. Spring. Summer. Autumn. Winter. PM01. -. 113.3. -. -. PM05. 52.7. -. -. -. PM12. -. -. 22.5. -. PM13. -. 14.9. 8.4. -. PM15. 5.1. 212.8. 164.7. 5.9. PM16. -. 146.4. -. -. PM17. 0.1. 107.4. 9.7. 0. PM21. -. -. 39.2. 0.2. PM23. 30.6. -. 75.4. 21.4. PM28. 12.0. 29.8. 487.1. 0.7. PM30. -. -. 81.7. -. PM35. 8.1. 734.8. 46.8. 0.7. 27.

(31) Table 3. Dunn’s test values on daily movements,. indicated significant difference between two seasons.. i. Summer. Autumn. Winter. Summer. Autumn. Winter. Spring. 891.65. 411. 89.93. 87.74. 101.42. 68.26. 14.49. 154.04. Summer. 959.91. 496. -. 83.31. 97.62. -. 53.77. 222.30. Autumn. 906.14. 555. -. -. 95.60. -. -. 168.53. Winter. 737.61. 310. -. -. -. -. -. -. 28.

(32) Table 4. Macrohabitat usage of the Burmese python in the four seasons. Spring. Summer. Autumn. Winter. Artificial habitat. 24. 0. 10. 15. Planted forest. 110. 10. 74. 92. Secondary forest. 99. 319. 215. 76. Wetland. 0. 172. 110. 0. Agricultural land. 154. 118. 154. 151. Granitic woodland. 68. 26. 10. 90. Table 5. Adjusted standardized residuals tests of macrohabitat use in the four seasons. All absolute values over 2 indicated significant differences between the observed and the expected values. Spring. Summer. Autumn. Winter. Granitic woodland. 4.737. -5.499. -7.274. 8.896. Artificial habitat. 4.688. -4.721. -1.099. 1.599. Wetland. -9.502. 11.826. 4.732. -9.328. Secondary forest. -6.141. 10.095. 2.203. -8.404. Planted forest. 7.401. -10.750. -0.592. 4.794. Agricultural land. 3.417. -6.302. -0.402. 3.328. 29.

(33) Table 6. Microhabitat use of the Burmese python in the four seasons. Spring. Summer. Autumn. Winter. 0. 169. 91. 0. 102. 49. 99. 283. 61. 91. 91. 3. in grassland. 138. 179. 203. 36. in shrub. 86. 109. 67. 11. total. 387. 597. 551. 333. in water in cave or burrow in or on log and litter. Table 7. Adjusted standardized residuals tests of microhabitat use in the four seasons. All absolute values over 2 indicated significant differences between the observed and the expected values. Spring. Summer. Autumn. Winter. in shrub. 4.758. 3.055. -1.943. -6.446. in grassland. 2.848. 0.142. 4.327. -8.345. in or on log and litter. 1.694. 1.817. 2.766. -7.303. in cave or burrow. -1.065. -13.333. -6.541. 25.166. in water. -8.884. 12.314. 2.097. -8.095. 30.

(34) Table 8. Dunn’s test value of canopy cover of the perching sites,. indicated significant difference between two. seasons. i. Summer. Autumn. Winter. Summer. Autumn. Winter. Spring. 888.84. 366. 86.09. 87.08. 96.90. 72.21. 191.92. 177.49. Summer. 961.05. 499. -. 80.27. 90.83. -. 264.13. 249.70. Autumn. 696.92. 473. -. -. 91.77. -. -. 14.14. Winter. 711.35. 306. -. -. -. -. -. -. 31.

(35) Table 9. The records and analytic values of air temperature and python body temperature during daytime and nighttime in all seasons. Air temperature (means ± SE). Body temperature (means ± SE). n. t. d.f.. p-value. Analysis method. Spring daytime. 21.47±4.45. 25.40±5.46. 4. -17.72. 298. <0.0001. Pair-t test. Summer daytime. 29.62±3.07. 29.05±2.00. 5. 2.61. 358. 0.029. Pair-t test. Autumn daytime. 26.97±3.46. 26.24±2.78. 5. 548. 275. 0.68. Wilcoxon Signed Rank test. Winter daytime. 15.97±4.03. 20.87±4.09. 3. -15976.5. 269. <0.0001. Wilcoxon Signed Rank test. Spring nighttime. 16.64±4.05. 21.71±3.79. 4. -22262. 299. <0.0001. Wilcoxon Signed Rank test. Summer nighttime. 26.05±1.76. 28.38±2.01. 5. -31079. 358. <0.0001. Wilcoxon Signed Rank test. Autumn nighttime. 22.35±3.61. 25.10±2.50. 5. -17910.5. 276. <0.0001. Wilcoxon Signed Rank test. winter nighttime. 11.10±3.03. 19.72±2.94. 3. -18291.5. 269. <0.0001. Wilcoxon Signed Rank test. 32.

(36) Table 10. The home range (MCP method) of Burmese python compared to other snakes. Species Maximum home range (ha) Citation Python bivittatus bivittatus 892.7 This study Crotalus horridus 207.4 Macartney et al. 1988 Morelia spilota spilota 198 Slip and Shine 1988a Morelia spilota mcdowelli 124.16 Shine and Fitzgerald 1996 Pseudechis porphyriacus 46 Shine 1987 Lampropeltis getula getula 28.2 Wund et al. 2007 Hoplocephalus bungaroides 11.43 Webb and Shine 1997. 33.

(37) Records of pythons. A. Military force 100,000. 37. 100. 80. 30 60. 55,000. 20 40. 10. 25,000. 4. 20. 10,000. 0. 0. 0. 5,000. 0. Military force (# of soldiers). Records of pythons. 40. 0. 1950. 1960. 1970. 1980. 1990. 2000. 2010. B Records of pythons. 17. 30. 16. 25,000 12. 20. 8 10,000. 5. 10. 4. 5,000. 2 0. Military force (# of soldiers). 20. 0 2000. 2002. 2004. 2006. 2008. 2010. Fig. 1. Numbers of Python bivittatus bivittatus records from local newspapers compared to numbers of military forces (in numbers of soldiers) on Kinmen Island over larger (A) and smaller (B) time scales. 34.

(38) A. 90/86/100 FZ001 76/63/98 FZ004 70/63/87 PM019. B. Fuzhou. PM004 PM001. Kinmen. PM002 PM003. PM005 100/100/100. PM006 PM012 PM013 PM014 PM015. Fuzhou. PM016. PM017. Vietnam. PM018 88/87/91 PM020. Kinmen. PM021 PM022. C. PM023 PM024 PM025. PM026 PM027. Fuzhou. PM028 PM029. Kinmen. PM030 PM031 PM008. PM009 PM010 53/-/71 0.0005. PM007. Vietnam. PM011. Vietnam. Fig. 2. Phylogeny and gene genealogy among Python bivittatus bivittatus populations from Kinmen, Fuzhou (China), and Vietnam. (A) A maximum-likelihood tree of the concatenated dataset; (B) a minimum-spanning network of the mitochondrial cytochrome b gene; (C) a minimum-spanning network of the mitochondrial cytochrome oxidase subunit I (COI) gene.. 35.

(39) (a). (b). Fig. 3. Home range size of Burmese pythons in Kinmen island during the tracking period. There was no significant difference in home range size between males and females during the experimental duration (t=-2.00, d.f.=10, p=0.07). (a) Male python home range; 456.96±342.96 hectare, n=5; (b) female home range; 188.84±224.22 hectare, n=7. 36.

(40) Home range of each python (ha). 500 450. 400. a. a b. b c. c. 350. 300 250 200 150 100 50 0. Spring. Summer. Autumn. Winter. Fig. 4. Home range size of all pythons in different seasons (spring: 18.10±19.92, n=6; summer: 194.18±247.69, n=7; autumn: 103.94±151.62, n=9; winter: 4.84±8.43, n=6). The home range area of the pythons was significantly different among seasons (F=10.57, d.f.=(3,24), p<0.0001). Different letters indicate significant differences between seasons, a > b > c; least significant difference post hoc comparisons.. 37.

(41) Moving distance per day (m/day). 200. a. a. a b. 180. 160 140 120 100. 80 60 40 20. 0. Spring. Summer. Autumn. Winter. Fig. 5. The moving distance per day (m/day) showed significant difference among seasons (Kruskal-Wallis: χ2=62.56, d.f.=3, p<0.0001, spring: 16.75±56.21, n=6; summer: 54.50±132.90, n=6; autumn: 33.45±104.52, n=8; winter: 7.30±30.73, n=5). Different letters indicate significant differences between seasons, a >b; least significant difference post hoc comparisons.. 38.

(42) Moving distance per day (m/day). 300. ＊ 250. ＊. 200 150 100 50 0. Spring. Summer. Autumn. Winter. Fig. 6. Moving distance per day during daytime (open bars) and nighttime (solid bars). Movements in daytime were significantly higher than nighttime in summer and autumn (*). In contrast, movements represented no differences between day and night in winter and spring. (mean ± SE, spring: daytime 9.18±35.76 m/day, nighttime 24.59±70.72 m/day, n=6, z=-098, d.f.=1, p=0.33; summer: daytime 8.44±44.03 m/day, nighttime 102.46±171.99 m/day, n=6, z=8.20, d.f.=1, p<0.0001; autumn: daytime 6.61±33.61 m/day, nighttime 60.20±138.87 m/day, n=8, z=-6.50, d.f.=1, p<0.0001; winter: daytime 10.38±34.93 m/day, nighttime 4.42±25.96 m/day, n=5, z=1.71, d.f.=1, p=0.09). 39.

(43) 100% 90% 80% 70%. Granitic woodland. 60%. Agricultural land. 50%. Wetland. 40%. Secondary f orest. 30%. Planted f orest. 20%. Artif icial habitat. 10% 0%. Spring Summer Autumn Winter Fig.7. Percentage of macrohabitat use of the Burmese python in the four seasons. Python macrohabitat use varied significantly differed among seasons (χ2=650.57, d.f.=15, p<0.0001).. 100% 90% 80% 70% 60%. in shrub. 50%. in grassland. in or on log and litter. 40%. in cave or burrow. 30%. in water. 20% 10% 0%. Spring Summer Autumn Winter Fig. 8. Percentage of microhabitat use of the Burmese python in the four seasons. Python microhabitat use varied significantly differed among seasons (χ2=825.09, d.f.=12, p<0.0001). 40.

(44) Canopy cover of each season (%). 80. a. a. 70. b. b. Autumn. Winter. 60 50. 40 30 20 10 0. Spring. Summer. Fig. 9. The canopy cover (mean ± SE) where the python stayed showed significant difference among the four seasons (χ2=99.64, d.f.=3, p<0.0001, spring: 38.98±21.69, n=6; summer: 45.55±25.37, n=7; autumn: 29.74±24.06, n=7; winter: 29.42±20.10, n=5). Different letters indicate significant differences between seasons, a >b; least significant difference post hoc comparisons.. 41.

(45) 35. p<0.0001＊. p<0.0001＊. p=0.68. 30. Temperature (℃). p<0.0001＊ 25 20 15. 10 5 0. Spring. Summer. Autumn. Winter. Fig. 10. Mean temperature (± 1 SE) of the pythons’ body temperatures during daytime (solid bars) compared to contemporary air temperatures (open bars) in the four seasons. The body temperature was significantly higher than air temperature in spring and winter, but significantly lower in summer. (spring: body temperature 25.40±5.46 n=4, air temperature 21.47±4.45; summer: body temperature 29.05±2.00 n=5, air temperature 29.62±3.10; autumn: body temperature 26.24±2.78 n=5, air temperature 26.97±3.46; winter: body temperature 20.87±4.09 n=3, air temperature 15.97±4.03). 42.

(46) 35. Temperature (℃). 30. p<0.0001＊ p<0.0001＊. p<0.0001＊. p<0.0001＊. 25 20. 15 10 5 0. Spring. Summer. Autumn. Winter. Fig. 11. Mean temperature (± 1 SE) of the pythons’ body temperatures during nighttime (solid bars) compared to contemporary air temperatures (open bars) in the four seasons. The pythons kept their body temperature significantly higher than air temperature in all seasons. (spring: body temperature 21.71±3.79 n=4, air temperature 16.64±4.05; summer: body temperature 28.38±2.01 n=5, air temperature 26.05±1.76; autumn: body temperature 25.10±2.50 n=5, air temperature 22.35±3.61; winter: body temperature 19.72±2.94 n=3, air temperature 11.10±3.03).. 43.

(47) Fig. 12. Reduced body condition was recorded from some pythons in winter. Some abnormal secretion was observed on this python’s nose and jaw when it was basking.. 44.

(48) 40. 35 40 40 30 35. 35 25 30. PM13 PM16. 0 5. 0 14-Dec-10. 15-Dec-10. 16-Dec-10. 17-Dec-10. 18-Dec-10. 19-Dec-10. 20-Dec-10. 21-Dec 21-Dec 21-Dec. 50 10. 20-Dec 20-Dec 20-Dec. PM16 PM13 PM05. 19-Dec 19-Dec 19-Dec. 10 155. 18-Dec 18-Dec 18-Dec. PM05 PM16 Air. 17-Dec 17-Dec 17-Dec. 15 20 10. 16-Dec 16-Dec 16-Dec. 25 20 15. 15-Dec 15-Dec 15-Dec. Air. PM05 Air. 20101214 20101214 20101214. 30 20 25. PM13. 21-Dec-10. Fig. 13. Body temperature of three pythons compared to contemporary air temperature during a cold spell from 14 Dec. to 20 Dec., 2010. The air temperature (black line) was decreased by cold spell then fluctuated. In the mean while, the body temperatures of the pythons (red, blue, green line) showed a comparatively stable pattern.. 45.

(49) Appendix 1: Definition of Macrohabitats We couldn’t get topographic chart, aerial photograph, satellite image, land use or land cover profiles because Kinmen is a military island, it was very hard to calculate habitat availability in habitat range or python home range. There were a lot of army bases in habitat range or python home range, so it was impossible to measure habitat availability by walk. Most of the forest in Kinmen island was secondary forest and planted forest, which had been destroyed by shellfire and military development, now the forest in Kinmen was planted by army almost. To examine seasonal habitat associations displayed by these pythons, six macrohabitat types were described within the study area based on variation in topography, vegetation and artificial disturbance: 1.. Secondary forest: in secondary forest, there were complete forest structure and plentiful canopy by vegetation, the flora diversity was the most highest in all macrohabitat types. overstorey consists of Casuarina equisetifoliaand Leucaena leucocephala; stands of the shrubs Lantana camara; understorey vegetation is dominated by Ipomoea cairica, Dicranopteris linearis and Bidens pilosa.. 2.. Planted forest: some areas in natural python home range were planted factitious arbores became the windbreak forests. Regularly weeded in these areas result in insufficient vegetation on the ground. Vegetation is dominated by Melaleuca leucadendra, Sapium sebiferum, Palaquium formosanum and Casuarina equisetifolia.. 3.. Agricultural land: In this type artificial disturbance were middling, large scope field were planted broomcorn and wheat in Kinmen, small scope field were planted garlic, peanut or vegetable. Most of small scope agricultural land was surrounded by nature vegetation, in large scope field was surrounded by vegetation niggardly.. 4.. Granitic woodland: structurally complete remnant woodland found on steep slopes and hill crests; abundant crevices and rocky gaps in granitic woodland. The main base was constructed of granitic rocks, overstorey consists of Pinus massoniana and Acacia confuse; stands of the shrubs Rosa cymosa and Maytenus diversifolia.. 5.. Wetland: Including ponds, lakes and marsh, many ponds and lakes which natural or artificial were 46.

(50) supplies water to agriculture, livelihood and military in Kinmen. Marsh linked many ponds, creek between pond; became some zones of wet grassland were distributed in Kinmen island. Main vegetation is Typha orientalis and Kyllinga brevifolia. 6.. Artificial habitat: Including factitious building and strong human disturbance area, such as the farms of cattle and fowls, dump, village and town. There is no nature vegetation almost in this habitat type.. 47.

(51) (a). (b). (c). (d). Appendix 2: The home range of all pythons. (a)PM01, (b)PM05, (c)PM12, (d)PM13, (e)PM15, (f)PM16, (g)PM17, (h)PM21, (i)PM23, (j)PM28, (k)PM30, (l)PM35 48.

(52) (e). (f). (g). (h). 49.

(53) (i). (j). (k). (l). 50.

(54) (a). (b). (c). (d). Appendix 3: The moving route of all the pythons. (a)PM01, (b)PM05, (c)PM12, (d)PM13, (e)PM15, (f)PM16, (g)PM17, (h)PM21, (i)PM23, (j)PM28, (k)PM30, (l)PM35 51.

(55) (e). (f). (g). (h). 52.

(56) (i). (j). (k). (l). 53.

(57)