Leiden University logo.

nl en

Forests and human-plant histories viewed from the island grounds: implications of top soil modern phytolith behavior for Caribbean paleoethnobotanical and paleoenvironmental studies

Niels Koning, a Bachelor student of the Faculty of Archaeology, Leiden University, conducted a research internship on paleoethnobotany and paleoecology under the supervision of Dr. Jaime R. Pagán-Jiménez in the context of the ERC Synergy Project NEXUS1492 led by Prof. Dr. Corinne L. Hofman. His research delved into phytolith analysis of modern top soils from four plots that correspond to different forest types in the southern foothills of the Northern Range of the Dominican Republic. Niels compared modern phytolith results with vegetation currently known to grow in the selected areas. This research was aimed at assessing if meaningful correlations exist between recovered modern phytoliths and observed plant communities. The main goal was to test if ancient phytolith assemblages recovered by different means in buried contexts are reliable indicators of the past floristic communities associated with the archaeological sites studied by Nexus 1492 in the Dominican Republic.

Sample acquisition and modern vegetation

Phytoliths are microscopic silica bodies that form in many plants and have the capacity to preserve in soils for thousands or millions of years after the decay of the vegetal organs. They can be used to identify plant families, genus, and even species because of their varying shapes and sizes (see also Santiago-Marrero’s post on this topic). Pagán-Jiménez took the four soil samples used for this study in January 2018 in the northern Dominican Republic and also obtained qualitative data on the existing floristic communities at each forest plot (Figure 1). Studied samples came from the surface of four different forest plots and were taken by carefully scraping between 3 to 5 millimeters of the uppermost layer of each plot with a clean hand shovel. The first sample (Mph-1) was taken in a subtropical evergreen forest consisting only of trees and medium sized shrubs. The second sample (Mph-2) was taken from a palm cluster located in a subtropical, semi-arid forest in which mostly palms and some short shrubs were present. The third sample (Mph-3) was taken from a grassland in the same forest as the second sample. This grassland consisted of a single grass species of the Chloridoideae sub-family. The last sample (Mph-4) was taken from a subtropical, partially evergreen forest, which consisted mostly of shrubs and trees and a low amount of herbs. However, when Pagán-Jiménez visited this same plot in June 2016, the area was covered by a combination of grasses, shrubs, herbs, and trees.

Figure 1. Pictures of the sampled areas in January of 2018: (1) subtropical evergreen forest (Mph-1); (2) palm cluster in a subtropical semi-arid forest (Mph-2); (3) grassland in a subtropical semi-arid forest (Mph-3); and (4) subtropical partially evergreen forest (Mph-4).

Processing and analysis

Before the samples could be formally analyzed, soils were processed following Pagán-Jiménez’ protocol for phytolith analysis at the Faculty of Archaeology, Leiden University. The protocol includes the elimination of carbonates and oxides, the digestion of organic matter and the chemical flotation and extraction of phytoliths. Once the phytoliths have been extracted and dried, each sample is homogenized in the respective tubes and mounted on new microscope slides with PermountTM. Recovered phytoliths were identified and counted up to 250 specimens per sample on the basis of a list of key morphotypes and a counting form created both by Pagán-Jiménez, and using two microscopes: Leica DM750-Pol and Leica DM2700-Pol.


After the identification and counting of phytoliths, they were grouped into five major categories: Herbaceous, Arboreal (palm), Arboreal (others), Poaceae (=grasses) and Economic Plants. Arboreal phytoliths were divided into "palms" and "other arboreal" because palms have an environmental and cultural significance and are easily distinguishable from other arboreal phytoliths.

Figure 2. Phytolith assemblages of the four forest plots compared to the registered vegetation growing at them in January of 2018 (Mph-1 to Mph-4) and also in June of 2016 (Mph-4).

In Figure 2, the percentage of documented phytoliths are distributed into the five major categories as is indicated by the blue line. The percentage of the current vegetation registered in January of 2018 is indicated with an orange line for each forest plot. These graphs show that there is a strong correlation between modern phytolith assemblages from the soil surface and the observed vegetation growing at the studied plots. There are also some discordances as well (Figure 2, see graph Mph-4). Phytoliths from plants that were not growing there at the time of sampling were registered in all four forest plots. It seems feasible to think that some displacement/movement of phytoliths through aeolian activities during dry periods might be one reason for this, but the main reason could be that those phytoliths came from plants that existed there in the recent past. This is also made evident out of sample Mph-4 (partially evergreen forest), where the observed vegetation had changed a lot in the last two years. In this case, earlier observations (2016, green line in Figure 2, Mph-4) on the modern vegetation of that plot show more correlation with the registered phytolith assemblage than the vegetation data raised in 2018.


The results of this study show a strong correlation between modern surface soil phytoliths and recent, but longer vegetation history of the studied forest plots. Modern dominant plant populations registered through qualitative observation are well represented in the phytolith record at each studied plot, though minor discordances here documented seems to attest to vegetational changes occurring at the plots in the recent past, perhaps in the last couple of decades prior to this study. In principle, we believe this would also be the case with ancient soil samples where the phytolith assemblage from a specific spot will reflect an accumulation of vegetation over a more extended period of time. This study also suggests that low amounts of registered phytoliths may not have come from the local vegetation, but because of aeolian activities or similar processes (natural or anthropogenic) occurring in the surroundings. These are things that might be important to keep in mind when interpreting phytolith assemblages from studies dealing with ancient human-plant-environment interrelationships.

By Niels Koning

This website uses cookies. Read more