Wood Cellular Dendroclimatology: Testing New Proxies in Great Basin Bristlecone Pine

Abstract

Dendroclimatic proxies can be generated from the analysis of wood cellular structures, allowing for a more complete understanding of the physiological mechanisms that control the climatic response of tree species. Century-long (1870-2013) time series of anatomical parameters were developed for Great Basin bristlecone pine (Pinus longaeva D.K. Bailey) by capturing strongly contrasted microscopic images through a Confocal Laser Scanning Microscope. Environmental information embedded in wood anatomical series was analyzed in comparison with ring-width series using measures of empirical signal strength. Response functions were calculated against monthly climatic variables to evaluate climate sensitivity of cellular features (e.g., lumen area; lumen diameter) for the period 1950-2013. Calibration-verification tests were used to determine the potential to generate long climate reconstructions from these anatomical proxies. A total of eight tree-ring parameters (two ring-width and six chronologies of xylem anatomical parameters) were analyzed. Synchronous variability among samples varied among tree-ring parameters, usually decreasing from ring-width to anatomical features. Cellular parameters linked to plant hydraulic performance (e.g., tracheid lumen area and radial lumen diameter) showed empirical signal strength similar to ring-width series, while noise was predominant in chronologies of lumen tangential width and cell wall thickness. Climatic signals were different between anatomical and ring-width chronologies, revealing a positive and temporally stable correlation of tracheid size (i.e., lumen and cell diameter) with monthly (i.e., March) and seasonal precipitation. In particular, tracheid lumen diameter emerged as a reliable moisture indicator and was then used to reconstruct total March-August precipitation from 1870 to 2013. Wood anatomy holds great potential to refine and expand dendroclimatic records by allowing estimates of plant physiological adaptations to external stressors. Integrating xylem cellular features with ring-width chronologies can widen our understanding of past climatic variability (including annual extreme events) and improve the evaluation of long-term plant response to drought, especially in connection with future warming scenarios.

Keywords: climatic variability; conifers; lumen diameter; multi-proxy; palaeoclimatic reconstruction; tree-rings; wood anatomy.

FIGURE 1

(A) Geographic location of the study area (red square) within the North American Great Basin (red shaded area) and the state of Nevada (NV). Sampled bristlecone pines were located in the Subalpine West site (gray circle) of the Nevada Climate-ecohydrological Assessment Network (NevCAN; other sites in this instrumented transect are shown by white circles), which is located on private land near the border of Great Basin National Park. (B) Walther–Lieth monthly climate diagram for the study area derived from PRISM data for the 1895–2013 period.

FIGURE 2

(A) Example of composite wood anatomy image (100-× magnification) acquired with the Confocal Laser Scanner Microscope. (B) Example of image with radial and tangential measurements performed on each cell within an annual tree-ring by means of the WinCELL software.

FIGURE 3

Standardized chronologies (1870–2013) of annual ring-width and anatomical parameters of bristlecone pine (light gray lines, individual series; shaded area, ± 1 standard deviation; red line, 30-year cubic spline; dashed line, number of cores). (A) Ring-width (20 cores); (B) ring-width (10 cores); (C) lumen area (LA); (D) LA of the 30% largest cells; (E) lumen radial diameter; (F) cell radial diameter; (G) lumen tangential width; (H) double wall thickness. Note the different scales of the y-axes.

FIGURE 4

Static (upper graph; significant months are shown in black) and moving (lower graph; significant months are marked by black asterisks) response functions of annual tree-ring chronologies against minimum air temperature and monthly precipitation for the period 1950-2013. (A) Ring-width (20 cores); (B) ring-width (10 cores); (C) lumen radial diameter; (D) cell radial diameter; (E) lumen area (LA); (F) LA of the 30% largest cells. Both temperature and precipitation, from previous October to current September (lowercase letters, previous year months; uppercase letters, current year months), were used in the upper graph (vertical bars indicate 95% confidence interval), while precipitation from March to September and a 30-year moving interval was used in the lower graph. For acronyms see Table 1.

FIGURE 5

Plot of seasonal correlations between annual tree-ring parameters and total precipitation for 3, 4, 5, and 6-month seasons. (A) Ring-width (20 cores); (B) ring-width (10 cores); (C) lumen radial diameter; (D) cell radial diameter; (E) lumen area (LA); (F) LA of the 30% largest cells. Dark gray bars represent significant coefficients using a 95% confidence interval. Lowercase letters, previous year months; uppercase letters, current year months. Note the different scales of the y-axes. For acronyms see Table 1.

FIGURE 6

Reconstructed total March–August precipitation from 1870 to 2013 using lumen diameter. Gray solid line, PRISM data; black dashed line, reconstructed series. Vertical lines define the two periods (1950–1981 and 1982–2013) used for the split calibration/verification exercise.