台灣梅花鹿發情期吼叫行為之探討

Abstract

本研究以台灣梅花鹿(Cervus nippon taiouanus)為對象，探討其於發情期（1）兩種吼叫聲Moan與Howl之聲音特徵與（2）數量的日、月變化，（3）吼叫行為之功能，並以（4）回播實驗測試頻率(frequency)、吼叫率(rate, roars per minute)對吼叫聲功能之影響，（5）最後嘗試以吼叫聲辨識個體。結果顯示台灣梅花鹿此兩種聲音聲音頻率富有變化並有許多泛音，具備易於定位的特質，且音量亦大，在距離7 m測量下，Moan音量為81 – 88 dB(A)，Howl音量為92 – 96 dB(A)。2006年10月至2007年1月間，在社頂地區共進行8次觀察，記錄吼叫聲的類型與數量，發現6時與17時為Moan出現的兩個高峰，11時與0時另有兩個小高峰，而6時也是Howl出現的高峰，在20時與1時亦有兩個小高峰；兩種吼叫聲數量均在10月中旬開始漸漸上升，於11月中下旬達到最高峰，隨後迅速下降，至1月中旬結束。關於吼叫聲功能的探討，在小墾丁渡假村及台北市立動物園對圈養鹿群分別進行93.5小時與70小時的行為觀察，結果顯示絕大多數的吼叫聲都是由位序最高的雄鹿所發出(小墾丁渡假村：Moan = 100%，Howl無記錄；台北市立動物園：Moan = 97.3%，Howl = 75%)，且位序越高的雄鹿展現發情相關行為的次數也越多(小墾丁渡假村：r = 0.90, P<0.01；台北市立動物園：r = 0.26, P = 0.74)；由吼叫行為及其伴隨行為顯示，Moan為位序高的雄鹿驅離其他雄鹿時發出，Howl則與雄鹿間互相競爭有關，當雄鹿朝雌鹿吼叫Moan時雌鹿並無反應，因此無法確認是否有吸引雌鹿之功能。以回播實驗測試頻率與吼叫率的變化是否會影響吼叫聲之功能，結果顯示雄鹿對吼叫率較高的Moan投以顯著較多的注意力(P = 0.027)，推測吼叫率高的Moan可能更令其他雄鹿感到威脅。分析各項聲音特徵於個體間之差異，再以區別分析(discriminant analysis)作個體辨識，結果以Moan Part I可正確辨識97.1%的聲音，以Moan Part II可正確辨識80.0%的聲音，以Howl可正確辨識95.2%的聲音，顯示不同個體的聲音有明顯的差異且能正確的被區分。This study focused on two kinds of vocalizations, moan and howl, and their behavior of Formosan sika deer (Cervus nippon taiouanus) during rut for their (1) sound characters, (2) diurnal and seasonal variations of activity, (3) functions, (4) influence of frequency and rate (roars per minute) to the functions of vocalizations by playback experiments, and (5) individual acoustic variation and identification. The results showed that the frequency of moan and howl varied and consisted of many harmonics, which was located easily. The sound levels were high about 81-88 dB(A) in moan and 92-96 dB(A) in howl at a distance from 7 m. From October 2006 to January 2007, 8 observations were done to record the type and amount of vocalizations. The amount of moans reached to two peaks around 6:00 and 17:00 hr respectively and two smaller peaks were observed around 0:00 and 11:00 hr. The amount of howls reached to a peak around 6:00 and two smaller peaks around 1:00 and 20:00. The amounts of both moan and howl began in the middle of October and gradually rose to a peak in the middle and end of November. Afterwards, few vocalizations were recorded until mid January 2007. To study the functions of the vocalization and its behavior, 93.5-hour and 70-hour behavior observations on farm-raised deer were conducted in Kenting resort and Taipei zoo. Most vocalizations were vocalized by dominant males (Kenting resort: moan = 100%, howl was not observed; Taipei zoo: moan = 97.3%, howl = 75%) which processed more sexual contact with the female (Kenting resort: r = 0.90, P< 0.01 ; Taipei zoo: r = 0.26, P = 0.74.) The results showed that moan was used by dominant males to expel other males, and howl was related to intrasexual competition. There were moans toward females, but females had never responded to them, so we did not verify that the function of moans wa concerned with intersexual attraction. To measure the effects of frequency and rate to the functions of the vocalizations, we performed playback experiments with moan and howl separately. The results showedthat the male significantly paid more attention to other moans with higher rate (P = 0.027), which suggested that the male felt more threatened while the rate of moans was higher. Moreover we tested the difference of sound characters among individuals, using discriminant analysis to identify individuals. We found that 97.1% of moan part I, 80.0% of moan part II, and 95.2% of howl could be identified on individual bases. It revealed that sounds among individuals were significantly different and could be correctly identified.