Ancient Trade Routes and Least Cost Path Analysis in the Maya Region
Lake Atitlan and the Pacific Coast, Guatemala
This study integrates archaeological survey and geospatial modeling to predict the possible pre-Hispanic trade routes between the south region of Lake Atitlán and the Pacific Coast of Guatemala (Area of Interest -AOI-). In the past, archaeologists have assumed that an exchange route existed between the people settled in the lake region and their neighbors settled in the piedmont and Pacific Coast located to the south of their territory. The archaeological record of both regions indicates that many settlements developed and occupied the AOI from the Early Preclassic (1500-900 BC) to the Postclassic (1524 AD) and that they were in contact and communication.
In February 2018, myself and a team of anthropologists were guided by three local men who had accompanied their father on walking trip to the coast before car roads were built after 1970’s. It was inferred that those walking paths might have been used for centuries thus, he trails were surveyed and recorded with a GPS. Later, the collected data was compared to the routes predicted with the Least Cost Path Analysis (LCPA) that generated a geospatial model of the possible paths between both regions based on the terrain slope.
The routes predicted with the LCPA compared to the routes surveyed on the field demonstrate that travel routes were not selected prioritizing to optimize the least effort across the mountainous terrain. Nevertheless, it shows that they intersect at some points and most importantly, the predicted models show that archaeological sites were settled along the route in pre-Hispanic times likely to meet strategic purposes such as access or/and trade routes control.
Hi Maria!
First off, this is a super interesting analysis and I’m amazed at the length of distance people were travelling on foot to reach the coast! You mentioned there is ceramic evidence of interaction between Lake Atitlan and the Pacific Coast. I was wondering if there has been any isotopic research done on these populations (particularly with respect to nitrogen, which is especially sensitive to consumption of marine proteins) that might suggest populations further inland at Lake Atitlan were eating coastal, marine foods?
Hello Sylvia
Yes, there is ceramic evidence of interaction between the two areas. I don’t know of any isotope analysis conducted between these two populations, but I think you have a great idea here. Today, there is a constant trade of food products from the coast to the lake. For example, every day there are men selling coconuts on the streets that come fresh from the coast. I am not sure about marine foods, but salt, cotton, and the famous cacao are some of the products that are mentioned in our ethnographic, ethnohistoric and archaeological sources. You might find this article interesting:
https://doi-org.proxy.library.vanderbilt.edu/10.1016/j.jas.2003.10.009
HI Maria, Nice seeing this poster, and I like the images of the field crew and the local landscape where you and your collaborators worked. Were the sites in your study all occupied at the same time? What is the estimated time that it takes to get from, for example, a Pacific coastal site to one near Lake Atitlan? I liked your observation that some of the other sites (not shown) were along those routes too– are they also contemporaneous, or do you see a shifting in site locations across time?
The sites that I chose were occupied at the same time, the ones along the route I have to check. This was the first time I realized this information and I definitely have to continue exploring my results and polishing my ideas. Nevertheless, some of those sites have been occupied continuously from the Late Preclassic, with strong evidence from Late Classic and; Postclassic (according to E.M.Shook site records that you can find in famsi.org). The longes route we walked took us about 11 hours from start to end. We began walking early in the morning and arrived at San Juan La Laguna at about 7 pm. I have to admit that our friends walked faster than us and they said it was a route they walked in 6-7 hours. It is very physically demanding to walk from south to north uphill in that steep mountains. The route we did to the Pacific coastal we did by car, I have to say. We were so tired and a little burn out from the other previous routes to go walk to the sea under that burning sun and humid weather.
Maria, you have an interesting start here at looking at logistical networks in the Classic Maya world. I wonder which of the routes is the ethnographically documented one (if any)? Did you run LCPs only between those sites with known exchange relationships (through ceramics or other indices of trade/interaction)? I wonder where routes would converge if you ran LCPs between all/several of the highland sites and the coast termini–that would be interesting because where they converge (and they would eventually converge) would seem to be a particularly strategic zone.
Hi professor
The Zapotitlan map from 1529 was our main reference and it shows some routes, but they marked the roads known that connect with Colonial towns, some of which are modern settlements today and some “lost towns” that we also survey (material for a different map). But the routes represented in this map were not attempted to be archaeologically surveyed before. The straight parallel trails from north to south follow a similar pattern to those represented in the Zapotitlan map.
I think that running that analysis would be a great idea, and I am also interested in trying out the mobility affordance analysis like the one conducted by Lauren Kohut.
I have a fragment of an old map from 1932 that I georeferenced to compared the roads represented in that generated map, and I notice blurry segmented lines that I thought might have been representing old walking paths from Guatemala. Two of those blurry trails coincide with the trails surveyed (pink ones). I’ll have to make a second poster!
Morato, M. (2011), El Mapa de la Relación Geografica de Zapotitlan, una Isla de Racionalidad en un Ocenno de Empirismo, in Journal of Latin American Geography, Vol 10, No. 2., 217-229, Conference of Latin Americanists Geographers.
Maria, this is a beautiful poster! I love the idea of comparing least cost paths with known community routes. Which routes on your map are the least cost paths, and which are the local paths that you have traveled?
Also, you mention that the LCPs and known routes could be different due to missing archaeological sites in the LCP calculation. What other factors could explain this divergence? Sentiment, tradition, certain areas with specific resources? I would love to hear more about this, and particularly if the Spanish paths described in the ethnohistorical sources followed different routes as well. Three different ways to get between two sites is a lot, but perhaps it is more common than I think!
Hello Sam, I glad you liked my poster.
The LCP routes are the ones that have different colors, each color represents the connection between two sites, and the pink trails are the ones that we walked. To calculate the LCP I chose only eight archaeological sites that are known to have had some interaction (Shook.et.al, 1970; Popenoe de Hatch.M. &C. Galindo, 2014, Paredes y Kaplan, 2004; Davies G., 2018), and that from the Colonial period to modern times have been occupied by the Tz’utujil Maya. There are more than fifty sites in the AOI, so for the purpose of this project, I did not conduct the analysis for each of the sites located within the AOI for practical reasons. I used three (San Pedro la Laguna, San Juan la Laguna y Santiago Atitlan) sites from the lake and five sites (Chocola, Palo Gordo, Cocales, Patulul, and Los Tiestos) from the Piedmont and Coast region. The LCP analysis has to be run from the start point to the endpoint in a single process for each site interaction, so for the lake sites each site was analyzed five times; the piedmont and Cost sites were analyzed three times each. I was aware that common sense tells that the same path that was used to go from the north to the south had to be used to go from south to north, nevertheless, my intuition (or empirical knowledge) told me that I had to run the tool in both directions to each site to find out if there was a different pattern. This is how I found there were actually different routes to go from the piedmont to the lake, for example, the red trail from Palo Gordo that goes all the way up through Chicago to San Pedro la Laguna, and the blue trails from Cocales to Palo Gordo that merges with the red trail (parallel to the surveyed trail that goes all the way uphill, not lower terrain), and from Palo Gordo to San Juan la Laguna (the turquoise trail that intersects with the walked trail).
The ethnohistorical documents, in particular El Mapa de la Relación Geográfica de Zapotitlan de 1529, were an important reference to survey these routes. The site Los Tiestos was integrated into the survey and this map because John Prybot, an ethnohistorian member of the team (second from right to left in the photo) remembered having seen archaeological mounds in a spot that in the map shows a town called Xicalapa. We found an archaeological site in the location but we called it Los Tiestos because it is the name given to the site by local people and because we are not sure if they are the same site.
Maria,
Thank you for sharing your research with us! As you know, I am also interested in Maya trade routes, and this is a really fascinating way to begin looking at that! I’m not surprised that the paths you recorded do not mimic the LCPs. As you’ve pointed out, there was possibly control to specific access routes that only permitted certain groups to pass. I also wonder if there was any sort of ritual component to which paths were taken? Because the paths overlap with specific archaeological sites, is there any evidence that those sites could have been places of exchange along those routes? It would be interesting to follow up this study with an analysis of related artifact assemblages between a few of those sites (although that would not be a small undertaking). Great work, and I can’t wait to hear more!
Hello Phyllis
Thank you for your comments and questions.
I found the communication routes very interesting too, there is so much to study from these interactions. I think that the reason to select the paths were many. First, there are several natural resources that were harvest from different econiches along the route from edible plants, medicinal plants, ornamental plants used in special events, cypress trees that were used to built canoes; a special tree bark with a flammable resin that is used to ignite fire essential in the household; a particular spike from a three used as a needle; bamboo and other different flexible canes for building constructions and hut roofs. Second, there was probably political tension between the sites and they were avoiding crossing through foreign territories, ore avoiding tribute to other sites. Gavin has been working hard in a book chapter where I am a collaborator about this research, and there we are going to expand on these ideas. I found it very interesting that the landscape is explained by stories with a moral message, for example, there is a high pick that can be seen on the horizon from the cost and is the trademark of the route to the south. It has seven picks and it is called “The Seven Brothers”, our friends from San Pedro told us those stories while we walked, and show me all those details about the products that were harvest along the route. Some of the towns along the routes are still considered Lugar Sagrado, and traditional Mayan ceremonies have been performed there for centuries. you might want to look at Lugar Ceremonial Indigena Santa Teresita, Palo Gordo Suchitepequez.
Wow! What incredible research! I am really impressed and excited to learn more! I’ll be looking for that book chapter when it’s published 🙂
Hi Maria- this poster is lovely, I really enjoyed seeing your personal photos alongside the map and information! I feel as though your project is a great reminder of the wonders of modern technology/GIS, and also the fact that you sometimes just can’t beat getting out in the field and doing research on the ground. I noticed you mentioned roads built for car travel in the 1970s, do you expect any of these roads or other more recent forms of infrastructure have concealed possible pathways, or is this area relatively rural?
Hello Alyssa
Thanks for your comments, I understand I miss the field so much!
I think the road did conceal the old routes, paved roads provided an easy and faster way to move between the two regions. People invested in trucks and busses, so the old trails became less efficient to transport goods.