Simulating communication routes in Late Antique Mallorca (Balearic Islands, Spain)Patricia Murrieta Flores , Catalina Mas Florit and David Wheatley1 21

1

BackgroundThe purpose of this paper is to present the results of a GIS based spatial analysis, simulating possible communication in the East part of Mallorca (Balearic Islands, Spain) during Late Antiquity. This work is part of a larger project that aims to study the settlement patterns in the east of the island of Mallorca during this period, one of the least known periods in the history of the islands. Towards the end of 4th century, the Balearic Islands were an independent Roman province. In the year 455 AD, the Vandals conquered the islands, and later in 534 AD fell to the Byzantines. This last rule theoretically lasted until the Islamic conquest in Fig.1 Mallorca and the Iberic Peninsula 902-903 AD. A total of 1521 paths were then created between these nodes and, in order to characterise areas most likely to have been passed through, a grid of one kilometre cells was created and used to count how many paths crossed each cell. A map was then derived to show the natural areas of transit with respect to energetic cost of movement.

Archaeology Department, University of Southampton. 2 Departament de Prehistria, Histria Antiga i Arqueologia, Universitat de Barcelona.

It has also been interesting to observe the recorded settlements are predominantly located in natural pass areas, and therefore a statistical test of significance was carried out in order to explore the idea of a possible correlation between distance from the calculated LCPs and the settlement locations. The test allowed rejection of the null hypothesis (that the settlements are randomly distributed with respect to the calculated paths) at the 0.05 level of significance, suggesting there is a significant association.Fig. 3 Calculated least cost paths

Fig. 6 Kolmogorov-Smirnov test

ObjectivesThe archaeological evidence and the studies concerning the social dynamics of the period, have led to suggestions that Early Christian churches such as Son Pereto, Sa Carrotja and Son Fadrinet played a fundamental role in the organization of the territory, and therefore it was considered important to test their possible relationships with natural communication zones. Following this idea another objective was to investigate if the settlements of the same period presented a pattern of having a relationship to the areas of natural transit. To model possible communication routes in Late Roman was also an interesting objective because of the lack of archaeological evidence regarding historical paths in our study zone.

Although the test does not explain the nature of this relationship it supports in a more robust way the observation previously made and requires further investigation. Finally, comparing the LCP based on energetic cost against paths based on velocity, we could observe that the routes traced are totally different. Neither the churches nor the settlements seem to be related with paths that are calculated on the basis of velocity. It is considered essential to further investigate the reasons why churches and settlements are located in principal zones of natural transit in terms of energetic cost.

Fig. 4 Natural Areas of Trnnsits: Intersection of paths per kilometre

Additionally, a line density analysis was carried out in order to compare the results of both analyses.

MethodologyPedestrian energetic cost is a constant variable for movement in all societies. Consequently, for the Cost Surface Analyses it was decided to experiment with the formula developed for physiological studies by Pandolf et al. (1976), that calculates energetic expenditure in watts and to compare these results with those obtained by the same formula but modified by Van Leusen (2002). This author discusses the problem of anisotropy and the differential function of physiological expenditure when we move uphill and downhill, introducing the idea of the axis of cost Pandolf et al. symmetry at m=1.5w+2(w+l)(l/w)2+n(w+l)(1.5v2+0.35vg) Van Leusen -6 of slope to m=1.5w+2(w+l)(l/w)2+n(w+l)(1.5v2+0.35vg*abs(g+6)) the formula Where: m = the energy expended in watts developed by w = the total weight moved (weight of the individual in Kg) Pandolf et al. l = the weight of the load carried in Kg Comparing the cost surface results for Fig. 2 Formula developed by Pandolf et al., and later modified by Van Leusen both formulas we can conclude that the formula modified by Van Leusen more accurately represents the cost expended when travelling uphill or downhill. The resulting cost surfaces were used for the calculation of the Least Cost Paths. To establish the location of the nodes used as a sample for the calculation of the LCP, a characterization of the surface was carried out using the DEM of Mallorca through the analyses available in Landserf. From this analysis, the natural passes were identified and these locations were selected as nodes for the experiment. In addition, the historically known ports were also used as nodes due to the available archaeological evidence suggesting their importance as points of trade and commerce for the zones in which the churches were constructed.v = the speed of travel in m/s n = the terrain factor and g= the gradient (slope) in percent

Fig. 5 Line density analysis Fig. 7 Comparison between the least cost paths: Energy vs. Velocity

As a third experiment, an additional Cost Surface Analysis was carried out. This time the LCP were calculated with the formula developed by Gorenflo and Gale (1990), which calculates cost in terms of velocity. The purpose of this new analysis was to compare these results with the LCP resulting from the previous paths calculated using the function of the energetic cost of movement.

Discussion and ResultsWithin the results, it is notable that the churches of Son Pereto and Son Fadrinet seem to be located along an axis which appears to have been one of the easiest zones of transit in the eastern part of the island. This result is particularly interesting because of the social and economic dynamics of this period in which churches organized the surrounding territory. Furthermore, it may be possible to suggest that this axis could have served as communication hub, establishing a series of links to places at higher altitudes in which other communities are located. Due to its location in the port, it is possible to suggest that the church of Sa Carrotja exhibits a different dynamic to those of the inland churches. It is interesting to note, for example, that this church is also located within an area where many roads converge and which also seems to be a natural corridor on the coast.Fig. 8 Natural passage between the churches travelling from South to North-West