Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Ocean Engineering and Marine Sciences

First Advisor

Mark B. Bush

Second Advisor

Robert Van Woesik

Third Advisor

Andrew Palmer

Fourth Advisor

Eraldo Ribeiro


By using the natural laboratory of past climate change I ask: How resilient are extant tropical forest ecosystems to climatic forcing? Do tropical forests systems show gradualistic or threshold responses to past climate change? How do climate forcings change floristic composition? And do humans amplify or suppress the effects of natural climate variability through land use? My study is set in western Amazonia and in the eastern flank of the Andes, regions renowned for high endemism and cultural importance. The time scale of the study spans the Mid-Holocene Dry Event (MHDE, c. 9000 to 4000 years ago), arguably the most drought-prone period of the last 100,000 years. By analyzing the vegetation response to the MHDE, obtained from fossil pollen of lake sediments from different settings, I investigate the direction and intensity of vegetation change at regional and local scales. This period was formative in animal domestication and in the onset of crop cultivation, yet is poorly studied compared with the Late Holocene. I present a 12,670-yr limnological history from Lake Miski, a high elevation lake in the Peruvian Andes. While many shallow Andean lakes dried up during the mid-Holocene, Lake Miski was a constant feature in the landscape. Fluctuations in the diatom communities tracked insolation, but interactions with the Pacific Ocean may have played a role in structuring local climate. Due to the presence of a forest microrefugium, Lake Miski appeared unusually climatically stable during the MHDE. My new limnological information provides additional insights as it shows that the presence of the forest withstood fluctuations that induced state changes in the lake and in the diatom flora during the MHDE, underscoring that microrefugia do not equate to ‘unchanging’ hydrologies or climates. I also present a high-resolution 8500-yr paleoecological history of Lake Llaviucu, a mid-elevation lake at a wet section of the Ecuadorian. The MHDE was expressed at the lake setting, but surprisingly, more strongly in the vegetation than in the diatoms. A trajectory of increasing use by humans seems to have peaked in terms of maize production between c. 2900 and 800 years ago, and thereafter pastoralism was of increased importance until European arrival disrupted the system. There is little doubt that the Llaviucu Valley is a manufactured landscape, but the changes observed were primarily in the balance of species abundances rather than in the obvious change in forest structure or through the addition of species. Maize was introduced, but this was of such local occurrence that it is not considered to be part of the larger assemblage. The manufactured landscape does not appear to be either a no-analog assemblage or a novel community. The same is true for the diatom assemblages, although the recent planktonic-dominated system is without parallel in the prior 9000 years. Lastly, I present a high-resolution core spanning 7600 cal yr BP from the wettest section of Amazonia. At Lake Pata, Brazil, the geochemistry of sediments indicated that between c. 6500 and 3600 years ago lake levels dropped. Despite apparent increases in drought stress, only a few species replacements and minor fluctuations in abundance were observed. This vegetation stability indicates that even in this edaphically dry setting, where soils cannot buffer dry events, vegetation does not change linearly with geochemistry. This observation needs to be incorporated into climate models that include soil-vegetation feedbacks, as Amazonia may be more resistant to climate change than is currently assumed. Much of the long-term resilience observed in these systems was attributed to a lack of human influence. The apparent capacity to withstand climate change is encouraging, but these systems are sensitive to fire, and if human-set fires were to be introduced, the forest destruction from fire would almost instantly overwhelm any resilience to climate change.


Copyright held by author

Included in

Biology Commons