The following analysis uses data published in W. Wyatt Oswald et al., “Subregional Variability in the Response of New England Vegetation to Postglacial Climate Change,” Journal of Biogeography 45, no. 10 (2018): 2375–88, https://doi.org/10.1111/jbi.13407. The spreadsheet I used is available upon request.
- High-resolution data permits reconstruction of Holocene forest cover changes
- Initial forestation after deglaciation in 12 000 BCE led by birch and pine
- Dramatic decline in forest canopy between 1630 and 1708; almost complete recovery by 2001
Berry Pond is an unimaginatively named site north of Boston, Massachusetts (figure 1). Its low elevation (42 m), regular precipitation (1236 mm per year), and soil (mostly glacial till) make it a site typical of southern New England. The authors of this study present an impressively detailed pollen count stretching back to 16 000 years before present (BP), or 14 072 BCE. The sampling gives us data at a very high resolution. This data is freely available through the Neotoma Paleoecology Database. I downloaded this data and here present a brief analysis and interpretation with an eye to tracing the Holocene forest history of New England.
The graph tells a remarkably coherent story of the forest’s response to disturbance (figure 2). The canopy tree count includes species such as maple, chestnut, hickory, oak, and hemlock — characteristic trees of a well-established forest in southern New England. In this category, I also included pioneer trees, namely pine and birch. These trees like open canopy, so they are the first to “pioneer” an area that has no other trees in it. Thus we see that the initial response to deglaciation at 12 000 BCE is a steep climb in the percent of canopy trees, from 56% to 97% in less than 2 000 years. This dramatic increase in forest cover is led by birch and pine, which rise to their all-time high of 75% in 10 800 BCE. Over the next 11 500 years, the relative pollen counts stay pretty similar, with canopy trees at 95–100% and the percent of flowering grasses (indicators of open land) below 5%. In other words, the landscape that native people of New England encountered was mostly forest, without much open land (at least in the area of Berry Pond).
The next great change appears with European colonization (figure 3). The remarkable thing about this data is that there are enough samples to give us a high-resolution picture of even the vegetation change over the past five hundred years — with eighteen samples just between 1630 and 2001. We can see a significant decline in the percent of canopy trees (to a low of 82% in 1788) accompanied by a significant rise in the percent of flowering grasses (to a truly remarkable high of 18% in the same year, 1708). Since 1630, the first settlement of the area, the settlers cut down a remarkable amount of forest in just over seventy years — one human lifetime. This change corresponds well to the historical literature on English colonization of New England. Settlers greatly valued timber; not only did they use the woods for all kinds of construction, furniture, and tools, but timber was also sought after for shipbuilding and export to England. In other words, settlers cleared woods not just to create land for agriculture and grazing, but also to create a variety of commercial products. It’s worth noting that the pollen count probably underrepresents the actual destruction of forest, since some areas saw the planting of new trees for use as timber; for instance, pine plantations might represent some of the increase in pine between 1677 and 1708 (softwoods and especially pine were especially valued for their timber). Indeed, there is a spike in pine from 14% in 1670 to 29% in 1708, even as the total forest canopy reaches its minimum and birch, another pioneer species, remains at around 14% (see figure 4).
I have just scratched the surface with this brief analysis of the data. For instance, I have completely left out cryptogams — ferns, mosses, etc. While these might be less appropriate indicators of land cover, they are useful measures of other terrestrial changes. For instance, Rumex is a good indicator of water-logged soil, or we could use the count of mosses to similarly indicate moisture. A different way to expand my analysis of this data would be to differentiate the category of “canopy trees.” This is a useful aggregate for the purposes of determining land cover change, but tells us little about the composition of the forests. While forest cover remained steady between deglaciation and colonization, the composition changed in response to climatic shifts. Disaggregating the data reveals, for instance, a forest of mostly spruce and jack pine up through 9 000 BCE. (The presence of spruce is particularly remarkable, since it reaches a high of 18% in 11 673 BCE only to remain below 0.5% from 11 000 BCE onwards; see figure 4.) From then until 4 000 BCE, the forest at Berry Pond is mostly oak, which is joined by hickory and chestnut up until the present. Finally, expanding beyond the pollen count would allow us to learn more about the site. For instance, there might be archaeological finds (tools, walls, ruins of buildings) around Berry Pond that help us clarify settler and native uses of the local forest. Similarly, examining the site in person might reveal stumps or other evidence of past disturbance on current trees that date back to the colonial era.