New TPR paper: Impact of mid- to late Holocene precipitation changes on vegetation across lowland tropical South America: a paleo-data synthesis

By Richard Smith

Hello, and happy new year! In this blog post I’d like to briefly tell you about my first paper that was accepted and published in the journal Quaternary Research in late 2017.  It’s entitled “Impact of mid- to late Holocene precipitation changes on vegetation across lowland tropical South America: a paleo-data synthesis” and can be found here.

In late 2016, Frank was approached by Mark Bush who wanted to prepare a collection of papers for a ‘special issue’ to honour the lives of two prominent palaeoecologists, Prof. Paul Colinvaux (1930-2016) and Prof. Daniel Livingstone (1927-2016).  Both Paul and Daniel were leading figures in tropical palaeoecology, with Paul focusing predominantly on tropical South America and Daniel in tropical Africa.  Throughout their careers, they were at the forefront of much of the breakthroughs in tropical palaeoecology, as well as inspiring and training many graduate students to create an international legacy of palaeoecologists. More can be read about their amazing careers in Mark Bush and Will Gosling’s tribute article: “In search of the ice age tropics, a tribute to Prof. Daniel Livingstone and Prof. Paul Colinvaux”. 

(a) Dan Livingstone (photo courtesy of Duke University). (b) Paul Colinvaux at El Junco Crater Lake, Galapagos Islands (photo courtesy of Miriam Steinitz-Kannan). [From Bush & Gosling, 2018]

Given the focus as Paul and Daniel’s careers, the special issue was to focus on South American and Africa palaeoecology.  Frank suggested that I take this opportunity to write a paper to submit to this special issue on some work I had completed previously, synthesising existing palaeoecological data from across tropical South America from the mid-Holocene to the present. So that is what I did!

The main aims of this paper were: 1. To provide a more up-to-date mid-to late Holocene palaeoecological synthesis for this study region, as previous syntheses were a few years old or focused more on palaeoclimate inferences, 2. To present the palaeoecological data alongside key independent palaeoclimate records to assess any vegetation changes in the context of long-term climate change.  By independent, we mean that the palaeoclimate reconstruction has not been inferred from the same proxy as the palaeoecological data – as if this is the case, any causal links between climate and vegetation is impossible to disentangle.  

To this end, I collected data from 110 paleoecological sites from 87 previously published papers that reconstructed vegetation from 6,000 years ago (mid Holocene) to the present from across the Southern Hemisphere neotropics.  These reconstructions were based on the examination of a variety of proxy types, including: fossil pollen, stable carbon isotopes, phytoliths, and other geochemical analyses.  This was done through literature searches, or what I call paper mining – getting some key papers and mining their references for more relevant papers, rinse and repeat! Some other repositories of references helped, including the Neotoma database, and the Latin American Pollen Database.  

For each record, I recorded what the reconstructed vegetation was at 500-year time slices from 6,000 years ago to present so we can see a sort of timeline of how the vegetation changed over this period.  The independent palaeoclimate records that I used came predominantly from speleothems (cave drip deposits, most commonly known as stalagmites/stalactites). I picked 8 palaeoclimate records that covered different regions of the research area so we can gain a better idea of how climate change was the same/different across these regions, and how that affected the vegetation there.   

Now, although I don’t want this blog post to descend into a re-hash of the paper, I will give you a sneak peek at the figures of vegetation reconstruction to entice you to go have a read of the full paper! The figure above shows an animation of the mid-to late Holocene vegetation changes, with each coloured circle the reconstructed vegetation from a particular palaeoecological site.  The grey background just refers to the present-day biomes.  The size of the coloured circles represents at what scale the vegetation reconstruction represents, and is typically defined by the size of the lake or basin from where the record was taken from, e.g. a large circle indicates that this is a ‘regional scale’ reconstruction of vegetation (typically a record taken from a very large lake >5km2), whereas a small circle indicates a reconstruction that is very localised (typically a record taken from small bogs or soil pits).  This helps in interpreting the spatial scale of any vegetation changes.

This figure shows the palaeoclimate records.  Generally, values on the left of the x-axis are drier conditions, whereas values on the right of the x-axis are wetter conditions.  

With these figures, we can start to get a better understanding of how the vegetation in each region has responded to long term precipitation changes.  For instance we see, in the drier mid-Holocene climate, the more open savannah/grassland and dry forest vegetation at the south-west Amazonian ecotone and across Southern Brazil. There’s also the expansion of humid forests through the late Holocene in these areas (e.g. evergreen tropical forest in southwest Amazonia, Araucaria forests in southeast Brazil), and the ‘antiphased’ climate of north-east Brazil (which had a wetter mid-Holocene and drier late-Holocene). And throughout these changes, we can see the stability of evergreen tropical forests in central Amazonia.  

What this synthesis also highlights is the fact that, even though the number of palaeoecological sites in this research area has grown dramatically in the past couple of decades, there are still large areas where we need more! The ‘Cerrado’ savannah region between Amazonia and the Atlantic forests has hardly any archives, as well as the ‘Nordeste’ region of Brazil.  The major problem in these regions is the lack of suitable basins (lakes or bogs) because the climate is particularly dry! However, increasing the use of phytoliths could help to address this issue, as they can be extracted from other ‘basins’ such as soil pits, where pollen can’t survive. Some recent studies have shown the effectiveness of phytolith based vegetation reconstructions, and the work of fellow TPR members Heather and James will add to this in the future!   

Full reference to paper:

Smith, R., & Mayle, F. (2018). Impact of mid- to late Holocene precipitation changes on vegetation across lowland tropical South America: A paleo-data synthesis. Quaternary Research, 89(1), 134-155. doi:10.1017/qua.2017.89

Leave a Reply

Your email address will not be published. Required fields are marked *