Radiocarbon and wood anatomy as complementary tools for generating tree-ring records in Bolivia

Abstract

The science of tropical dendrochronology is now emerging in regions where tree-ring dating had previously not been considered possible. Here, we combine wood anatomical microsectioning techniques and radiocarbon analysis to produce the first tree-ring chronology with verified annual periodicity for a new dendrochronological species, Neltuma alba (commonly known as "algarrobo blanco") in the tropical Andes of Bolivia. First, we generated a preliminary chronology composed of six trees using traditional dendrochronological methods (i.e., cross-dating). We then measured the ¹⁴C content on nine selected tree rings from two samples and compared them with the Southern Hemisphere (SH) atmospheric ¹⁴C curves, covering the period of the bomb ¹⁴C peak. We find consistent offsets of 5 and 12 years, respectively, in the calendar dates initially assigned, indicating that several tree rings were missing in the sequence. In order to identify the tree-ring boundaries of the unidentified rings we investigated further by analyzing stem wood microsections to examine anatomical characteristics. These anatomical microsections revealed the presence of very narrow terminal parenchyma defining several tree-ring boundaries within the sapwood, which was not visible in sanded samples under a stereomicroscope. Such newly identified tree rings were consistent with the offsets shown by the radiocarbon analysis and allowed us to correct the calendar dates of the initial chronology. Additional radiocarbon measurements over a new batch of rings of the corrected dated samples resulted in a perfect match between the dendrochronological calendar years and the ¹⁴C dating, which is based on good agreement between the tree-ring ¹⁴C content and the SH ¹⁴C curves. Correlations with prior season precipitation and temperature reveal a strong legacy effect of climate conditions prior to the current Neltuma alba growing season. Overall, our study highlights much potential to complement traditional dendrochronology in tree species with challenging tree-ring boundaries with wood anatomical methods and ¹⁴C analyses. Taken together, these approaches confirm that Neltuma alba can be accurately dated and thereby used in climatic and ecological studies in tropical and subtropical South America.

Keywords: Bolivia; Neltuma alba; South America; algarrobo blanco; bomb-pulse radiocarbon dating; tropical dendrochronology.

Figure 1

Neltuma alba species distribution map (A) and climograph (B) of the study site at La Ventolera, Bolivia.

Figure 2

Panoramic image of a histological cut of a sapwood section of Neltuma alba. detailing the main wood anatomical traits used to take quantitative measurements and for differentiation of tree ring boundaries.

Figure 3

Comparison of distinctive diagnostic traits of ring boundaries as seen on the same section of a transversal cut of N. alba sapwood prepared with the classical approach of fine sanding viewed under a stereoscope and a histological cut viewed under a light transmission microscope. Below, two zoom cases of these anatomical traits are shown to high detail using a 200X magnification.

Figure 4

Ring width chronologies of 23 radius from 6 individuals of Neltuma alba shown as detrended index [above (A)] and raw measurements [below (B)] covering the period 1876-2018, with a common climate study period from 1960-2015.

Figure 5

Radiocarbon test results of two sample trees [VTP01 (A, B) and VTP07(C, D)] before and after wood anatomy ring boundary analysis. Lines represent the calibration radiocarbon curve for the South Hemisphere while dots show single tree ring radiocarbon values (black circles show expected year values after classical crossdating and red dots represent actual calendar date values). Horizontal images show the original prepared wood material used for crossdating with the calendar dates assigned before the wood anatomy analysis (black) and after (red).

Figure 6

Climate–growth correlations between tree ring width and monthly climatic variables [Precipitation (A) and Tmax (B)] in N. alba. Correlations were calculated for a 24 months period starting in July before the onset of the previous growing season to June after the offset of the target growing season. Simple correlations are shown in a whisker format above a color matrix showing the dynamic correlations. Previous growing season is shown inside a discontinuous line rectangle (Oct – Apr) and the target growing season is surrounded by a continuous line rectangle (Oct – APR). Significance values where set at P< 0.05. Study period 1960-2015. *significant value.

Figure 7

Covariance removed correlation coefficients between tree ring width index and two monthly climatic variables: precipitation (green (A)) and Tmax (red (B)); covering from July before the previous growing season to June after the target growing season. Bars indicate partial monthly correlations where the correlation of tree-ring width with precipitation and temperature was calculated after removing the covariability between temperature and precipitation, significant correlations are signed with an asterisk. Continuous and discontinuous grey rectangle lines indicate current and previous growing season respectively.

Figure 8

Spatial correlations (1960-2015) of mean tree ring width vs. precipitation and maximum temperatures is plotted based on the highest seasonal correlations shown in figure 7. Correlations were performed using CRU TS4.06 Precipitation during the period July to September (A) and CRU TS4.06 Tmax during the month of December (B) both covering the period previous to the growing season. The red star signals the sampling site of the studied species and the white dotted background fillings signal non-significant spatial correlations. *Location of our site.