A Geospatial Analysis of Mounds in Savai‘i and ‘Upolu, Sāmoa: An Evolutionary Ecology Approach

Abstract

The presence of monumental architecture in the archaeological record has long been a focus of archaeological research. The construction of large and elaborate structures is a costly endeavour which often seems to impart no benefit to those building the monuments. Costly signalling and bethedging are two explanations from evolutionary ecology which can be applied to this problem by determining the contexts in which monument construction would carry a selective advantage. This thesis evaluates these two explanations by investigating stone and earthen mounds and platforms from Savai‘i and ‘Upolu in Sāmoa. These are amongst the most common type of archaeological feature in the landscape, yet targeted research that explains why these costly mounds were built is lacking. One of the primary expectations of costly signalling is that monuments will be located near areas where contested resources are abundant, whilst this is unlikely to be the case if bet-hedging is an appropriate explanation. In order to test some of the expectations of costly signalling and bet-hedging in Sāmoa, a semi-automated feature identification method was used to identify mounds in LiDAR data, and a model of agricultural productivity was produced to map locations where the cultivation of dryland taro should have been successful. Ultimately, mounds were found to be preferentially located on or near locations identified as having high agricultural potential. This suggests that costly signalling may be an appropriate explanation for mound-building behaviours on Sāmoa, where mounds served to demonstrate resource holding potential and competitive ability.