People of the Swamp: LiDAR and "Invisible" Structures at El Pilar

Methods

Summary

The first step in this project, which has already been completed, involved the visual analysis of LiDAR imagery from El Pilar in a geographic information system (GIS). We identified and cataloged anomalies that may represent ancient structures, analyzed their locations on the landscape to look for spatial patterns in settlement, and flagged areas of apparently lower settlement density for further study.

Fieldwork will be divided into three research facets: targeted visits to LiDAR features, systematic survey of seasonal swamps and areas with few identified features, and detailed investigations of previously identified "swamp structures." Some aspects of data collection and recording will be common to each facet, but each will require slightly different methods to successfully complete.

We will obtain coordinates from the GIS and load them into handheld GPS units to guide the survey team to features identified on the LiDAR map, most of which are located in upland areas. Using other information in the GIS, such as the routes of paths through the jungle and the topography of the area, we will plan each day to maximize the number of features we visit. Once the team reaches a feature location, we will first determine if the feature is cultural or natural. If the feature is a man-made structure, we will map the structure with a measuring tape, compass, and GPS, and conduct a reconnaissance of the surrounding area to locate any additional structures, storage pits, or landscape modifications such as quarries. Any structures encountered en route to the targeted features, but not visible on the LiDAR map, will be recorded in the same way. This will continue the program of field-checking the results of LiDAR survey and will expand our understanding of settlement at El Pilar.

Low-lying areas with few LiDAR features will be systematically surveyed. The survey team will impose an artificial grid over these areas in the GIS and load grid-square coordinates into the handheld GPS units. The team will then cut paths through the forest in these areas, using maps and GPS units, that will allow us to explore them for structures the LiDAR may not have detected. Any structures encountered will be recorded with the methods described above. This survey will help fill in any gaps in the LiDAR map and will potentially identify more "swamp structures" for intensive analysis.

"Swamp structures" identified in previous surveys had GPS coordinates recorded and can be revisited in a way similar to explorations of unknown LiDAR features. Once these structures are relocated, they will be mapped with greater precision and investigated more closely for evidence of their function. Artifacts on the surface will be collected for later analysis, and their building materials will be examined and sampled. This will help answer the question: what were people doing in these areas where our models suggest they would not live?

All data collected in the field will be uploaded into the project GIS at the end of each day and examined for spatial patterns. Statistical analyses of structure location, labor investment, and proximity to natural resources will be conducted after the data have been processed. These analyses will be used to examine social organization, settlement preferences, and human-environment interactions at El Pilar.

Challenges

Two major challenges of this project will be locating structures not visible on the LiDAR map and, in the case of the mysterious "swamp structures," determining their function. Systematic survey was used this year at El Pilar to explore about 1.5 square kilometers of seasonal swamp and was successful at locating "invisible" structures, so we have confidence that it will prove useful again. Large-scale excavation remains the best way to understand how ancient people used structures, and that is not an option at this stage of the project. Surface collection of artifacts, more precise mapping, closer inspection of building materials, and analyses of spatial relationships among structures and natural landscape features will provide multiple clues to structure function. If this evidence does not allow us to definitively say how the "swamp structures" were used, it will provide a solid basis for hypotheses that can be tested by future research.

Pre Analysis Plan

Our current settlement model suggests the ancient Maya preferred to build their residences in well-drained upland areas over low-lying seasonal wetlands. This hypothesis makes intuitive sense - people would rather live on a breezy hilltop than a mosquito-infested swamp - and finds support in previous survey data and newer LiDAR imagery. More recent survey in low-lying areas indicates this picture is incomplete, however, because a number of small and irregularly shaped structures were discovered in the swamps that must be integrated into the broader picture of ancient settlement at El Pilar. LiDAR anomalies must also be visited to determine if they are cultural features.

Preliminary assessments of structure function are the first step in updating the settlement model. Archaeologists know quite a bit about ancient Maya residences due to intensive research over the past 40 years, and the majority of structures visible on LiDAR imagery usually represent households or residential groups. The mapping procedures described above and field observations of associated artifacts will provide a the basis for determining the function of LiDAR-visible features. The function of smaller, irregular structures in seasonally swampy areas will be more difficult to determine, and this will require more intensive mapping, laboratory analysis of associated artifacts, and a detailed examination of their relationship to the landscape. As a hypothetical example, if one of the "swamp structures" is long, narrow, and low to the ground, and no pottery or other domestic trash is found nearby, it is probably something other than house. We would need to examine its spatial relationship to nearby seasonal watercourses or other natural features to determine its function, and we may find it had something to do with managing the flow of water into, or away from, some other area during the rainy season. If another low structure has a roughly rectangular layout with bits of pottery and stone tool fragments on its surface or near its edges, we may determine it was a residence.

We will explore variability in settlement and land use by integrating all data from the field into the Maya Forest GIS. This allows us to compare the size and layout of structures, the labor invested in architecture, and the spatial relationships (for example, average distance) between man-made and natural features across the 20 square kilometers of the El Pilar Archaeological Reserve for Flora and Fauna. We can use geospatial statistics to measure, for example, the density of different types of structures in certain areas and compare this to documented soil types and other natural features. This will give us a better sense of where people were living, which will allow us to generate hypotheses of why they chose to live there and how they were using other areas. This is the basis for understanding the nature of ancient Maya urbanism and the social organization that fostered the growth and florescence of Classic-period civilization.