2025 Theses Doctoral

Disentangling tree growth and climate response of Andean-Amazon forests in Bolivia

Oelkers, Rose Christine

The tropical Andes region of South America (~5º-24ºS) is one of the most unique and biodiverse regions of the world. It is under-sampled for dendrochronological studies, as most trees in this area lack defined annual rings and instead feature continuous growth in their wood due to a general lack of seasonality. However, annual growth rings can form in some tropical tree species in areas with seasonal precipitation, such as the low latitude Andes. Thus, they could be very valuable for understanding past climate variability in the region.

This dissertation describes newly developed tree-ring chronologies from wood samples collected in forests in the Andes-Amazon corridor of the Madidi National Park (MNP), a biodiversity hotspot in northern Bolivia. These are the first robust tree-ring chronologies developed for this region of Bolivia. They are part of a broader effort to increase the spatial coverage of forest and climate data in the tropics, where tree-ring records are short and sparse. The sites and species discussed herein are i.) a high elevation (4000-4400 m.a.s.l.) tropical treeline Polylepis pepei BB Simpson (Simpson, 1979) network, ii.) Juglans boliviana C. DC. Dode (Dode, 1906), a new species for dendrochronology, from a tropical wet forest (1100-1300 m.a.s.l.) and iii.) low elevation (895-1000 m.a.s.l.) Amburana cearensis Fr. Allem. (Smith, 1940) from a seasonally dry valley near the Tuichi River. Tree-ring anatomy, ring width (RW) and radiocarbon (14C) series were used to provide new information on growth dynamics and climate sensitivity for these tropical tree species. Results are highly relevant for determining the role that tropical forests will have as a carbon source and/or sink in this era of global warming.

Chapter 1 highlights the challenges inherent in traditional tree-ring cross-dating in one of the selected tree species, J. boliviana. These challenges include radial eccentricity in the stem, complex wood anatomy, and variable controls on tree growth. J. boliviana is a key indicator species of the humid lower montane Yungas that has not been previously studied for dendrochronology. Several methods for verifying annual periodicity (i.e. dendrochronology, histological slices, radiocarbon) and the building of robust chronologies for tropical tree-ring sites were employed. Tree growth, longevity, and dendroclimatic potential were assessed for this species, which is used structurally (e.g. for building of bridges) and medically (e.g. treatment of diabetes) by local indigenous communities. The RW chronology, which spans from 1814 to 2017, reflected patterns of low and high frequency variability and an increase in growth since 1979. Climate-growth analyses revealed J. boliviana RW is negatively impacted by maximum temperatures and positively correlated to mean precipitation during the dry months of the year (May-Oct). These results agree with other Juglans-related (J. australis and J. neotropica) dendroclimatic studies from Argentina and Peru and illustrate the importance of dry-season precipitation on tropical tree growth; a dynamic topic in recent literature.

Chapter 2 builds on the analyses in Chapter 1 by comparing the growth variability and climate sensitivity from the humid J. boliviana forest to the seasonally dry P. pepei, and A. cearensis sites. Despite the proximity of the forest sites in MNP (14°33´-14°43´ S, 68°41´-69°04´ W), the species inhabit distinct vegetation zones shaped by orography along the eastern Andean foothills. This chapter seeks to understand diverging trends in radial growth, land-use impacts, and climate-growth relationships in MNP using gridded climate products (CRU, CHIRPS) and tropical sea surface temperature (SST) indices (Hadley1 SST). Raw RW at the upper-montane P. pepei site, sensitive to summer (Nov-Jan) precipitation changes, has been declining since 1995, likely due to increasing minimum temperatures and decreasing spring (SON) precipitation in the central Andes. Conversely, A. cearensis RW, which is strongly linked to peak wet-season rainfall (Dec-Apr), has been growing exceptionally well since the late 1970s, at the same time that late-summer rainfall has increased. Flood events related to anomalously warm SST in the tropical Atlantic and/or colder (La Niña) conditions in the Pacific lead to higher growth in the lower montane sites (J. boliviana and A. cearensis), while El Nino-related droughts were associated with narrower RW overall. A. cearensis RW also recorded the strongest summer ENSO signal among the species, potentially due to the proximity of this site to the Tuichi River. These newly developed chronologies (1867-2018) provide long term, annually resolved RW information that may provide new insights into the future stability of these forests under the threats of deforestation and climate change.

Chapter 3 features a new annual (1950-2019) radiocarbon reconstruction for the Southern Hemisphere (SH) developed from a tree sample from the high elevation, remote P. pepei forest near the town of Keara. This is the first continuous post-1950 A.D. 14C bomb curve record for Bolivia and the lowest latitude site within the SH 14C radiocarbon Zone 1-2 (SH Zone 1-2). Radiocarbon variability both within and between individual trees of the same species was observed in several sites. This is particularly the case during the 1960s when atmospheric radiocarbon exponentially rose due to nuclear bomb testing in the Northern Hemisphere (NH). There is radiocarbon enrichment after 1973, as found for other tree-ring 14C chronologies from Brazilian Amazon forests. This may reflect the rate of carbon cycling in the Amazon Basin at this time and the influence of varying tropospheric circulation patterns within the tropical low-pressure belt (TLPB) and Intertropical Convergence Zone (ITCZ) boundaries. Air parcel backward trajectory models showed that most air masses arriving to the Keara forest during the wet season were derived from the Amazon Basin and Atlantic Ocean reservoirs. This finding confirms the influence of northern air masses on tree growth and isotopic variability in MNP and other forests of the Amazon-Andes ecotone in SH Zone 1-2.

These understudied forests play a critical role in the livelihoods of the local communities and their water supply from the Andes. These new results indicate significant progress in efforts to expand the South American tropical tree-ring network and improve our understanding of the causes of tree growth variability linked to climate and other environmental factors in the era of recent anthropogenic change.