| Summary: | 碩士 === 東海大學 === 生命科學系 === 107 === Highly adaptable and long involved in human economic activities, wild boar (Sus scrofa) and domestic pig (Sus scrofa domesticus), two of the most representatives of the swine (Suidae) family, have emerged in a relatively short span of time as the most widespread and essential animals across the world. The main difficulties in the application of modern morphometric studies to archaeological pig-bone remains have not been totally tackled, owing to the former comparative basis are unsound and the latter are often fragmented. The objectives of this research are to assess the great morphological and size variabilities using criteria that could readily be applied to characterize these subspecies. 71 craniometric measurements have been made characterizing both the basic skull shapes and proportions with 5 non-morphological variables were carried to 1782 wild boar, 2955 domestic pig and 987 hybrids individuals. All pigs’ carcasses were bought from elite breeders, seedstock producers, meat markets, abattoirs and hunters. The skulls have been prepared for measurement by routine methods including removal of hide and soft tissues, maceration in water and degreasing in organic solvents. Based on the values of data used: 1) Curve fitting analysis has been made to define a best fit model for growth trajectory; 2) X-ray computed tomography scan analysis to complete dental chronology; 3) Cluster analysis to automatically specify the relationship of the observations with each other; 4) Principal component analysis to capture morphological variables that mainly affect clustering results; 5) Decision analyses (Decision Tree & Random Forest) to help identify a strategy to optimize expected utility decision that can be applied to those from archaeological remains; 6) Data partitioning for model development and validation.
It has been found out that growth rate is 3 times faster in domestic pig whereas teeth eruption time is 1.6 times slower comparing to wild boar. The elongated skull with low-profile as classical morphology for wild boar identification; however, the variation of cranial morphology associated with different housing types, it could be easily confused in past literatures. These subspecies could be told apart by 1) Mandible: length from the angle to anterior-most point of symphysis (LA), length of mandible (UL) and length from the condyle to anterior-most point of symphysis (LC); 2) Teeth: arch length between M2 and interdentale of the mandible (LM2IL) & length of the lower third molar (M3L); 3) Cranium: total length of the skull, internasal point to foramen magnum (CTL) & length of maxilla (LL); 4) Bone mineral concentration: mandibular bone fluoride concentration (BF) & enamel fluoride concentration (EF). The identification of these specimens only needs two steps, by the angle between lower border and anterior couture of mandible (UBACA) and LR, both highly correlated to the size of masseter muscle, which represent the importance of mastication utility difference in identification. According to the archaeological remains tested from Benghsichueh area of Gubengang site of Taiwan in the 1860s, our methods can provide a quantifiable, clear classification among wild, domestic and hybrid Sus scrofa, with significant implications for zooarchaeological research, and their application a stronger baseline for the interpretation of ancient materials.
|
|---|
