TECHNICAL NOTETesting different Earlywood/Latewood delimitations for the establishment of Blue Intensity data: A case study based on Alpine Picea abies samples
Introduction
Tree rings in conifers of the mid and high latitudes are usually characterized by a transition of large-lumen and thin-walled cells in the first section (the so-called earlywood) to narrow-lumen and thick-walled cells in the second and last part of a ring (latewood) (Schweingruber, 1988). Earlywood (EW) cells form in the beginning of the growth season and ensure efficient transport of water and nutrients, but are prone to cavitation (Pratt et al., 2007). Latewood (LW) cells appear distinctly darker, have higher densities and provide more mechanic stability, but are less efficient in water transport (Sperry et al., 2006; Björklund et al., 2017). As a response to the local and regional climate, a treeʾs allocation of resources to the formation of either EW or LW cells, and the properties of those so formed cells, vary between and within years resulting in different characteristics of each annual increment and the cells therein (Schweingruber, 1988; Björklund et al., 2017). Dendroclimatologists measure those variations, summarize them in different so-called parameters (e.g. ring width) and thus draw conclusions about the climate of the past. Owing to the different times of formation, EW and LW comprise different signals which can be used to reconstruct different climatic variables of different seasons (Schweingruber, 1988; Wood and Smith, 2015).
A wood property exhibiting climate induced inter- and intra-annual variability, is density. If there is one growth limiting factor (e.g. temperature at the altitudinal tree line), it determines the density characteristics of each annually formed tree ring to a high degree (Schweingruber, 1996). Therefore, wood density is a good climate proxy and has successfully been applied to a wide range of paleoclimatic studies (e.g. Cleaveland, 1986; Briffa et al., 1988; Chen et al., 2012). However, the establishment of density values is time consuming and the costs for acquiring required X-ray measurement devices are considerable (McCarroll et al., 2002). This has impeded establishment of high-quality tree ring based climate reconstruction where access to such a device was not available.
The development of the novel Blue Intensity (BI) method (McCarroll et al., 2002), which can serve as a surrogate for density measurements, thus means methodological progress for many tree ring researchers worldwide. BI works on the principle that the absorption of light in the blue spectrum of a tree ring has been shown to be highly correlated with X-ray wood density measurements, and thus is similarly an attractive proxy for temperature (McCarroll et al., 2002; Campbell et al., 2011; Rydval et al., 2014; Björklund et al., 2019). However, it has also been reported that the signals in BI and density are not entirely the same (Kaczka et al., 2018). To date, it is still not entirely clear, what exactly the BI values represent (Björklund et al., 2019). Nonetheless, BI has recently been applied successfully to create long chronologies (Fuentes et al., 2018), improve archaeological dating (Wilson et al., 2017), investigate drought periods (Babst et al., 2016) and address the growth divergence problem (Buras et al., 2018). A growing body of methodological literature deals with continuously refining and developing the method (e.g. Kaczka et al., 2018).
The intra-annual variations of density/BI in tree rings allow the establishment of different parameter time series, e.g. mean densities of EW or LW. Due to the gradual transition from EW to LW in most coniferous species (e.g. Picea abies), however, a defined delimitation is a prerequisite for the establishment of EW or LW based time series. Different approaches exist to delimit the two. First, the EW/LW boundary is set where the double cell wall thickness is greater than the radial lumen diameter (Mork, 1928). This method can only be used on wood anatomical micro sections and is therefore time-consuming. Second, one fixed density/BI threshold (usually between 0.40 and 0.55 g/cm3 for density (Koubaaa et al., 2002)) is defined which considers those parts of the tree ring with higher values than the threshold as LW and the ones with lower values as EW (Schweingruber, 1983; Cleaveland, 1986). This can be problematic if inter-annual density/BI fluctuations within a tree ring series are high (Björklund et al., 2017; Samusevich et al., 2020). And third, a percental value k is set which calculates an EW/LW boundary threshold T for each individual ring based on the difference between maximum density/BI (MXBI; acronyms in this paper always referring to the BI parameters) and minimum density/BI (MNBI) of each ring according to the formula (Schweingruber, 1983):
This ‘floating threshold’ approach is a more dynamic method that is now commonly implemented in standard analysis programs such as CooRecorder (Cybis Electronics, Sweden), WinDendro™ (Regent Instruments, Canada) and LignoVision™ (Rinntech, Germany). Given the subtle transition of EW to LW, all approaches are arbitrary. Nonetheless, when undertaking a dendroclimatic study that involves a parameter based on EW or LW information, it is desirable to choose a k value that can extract as much information as possible. Parameters of interest with dependence on the EW/LW boundary, that have tradition in both density and BI studies, are: Earlywood width (EWW), Earlywood absorption/density (EWBI), Latewood width (LWW) and Latewood absorption/density (LWBI). Additionally, one can calculate ΔBI as introduced by Björklund et al. (2014) as the difference between MXBI and EWBI for each tree ring. This parameter is widely used alongside its refinement, adjusted ΔBI (Björklund et al., 2015). Originally, ΔBI was proposed for Scots pine (Pinus sylvestris) with the goal of mitigating the influence of a color transition occurring between heartwood and sapwood in that species and which, if not considered, would bias the obtained BI values. Even though the investigated species in this study, spruce (Picea abies), does not feature such a marked transition, the parameter was calculated as it, too, is potentially influenced by changes in the definition of the EW/LW boundary.
The previously mentioned parameters MXBI and MNBI, the average absorption/density of the whole tree ring (TRBI) and the total ring width (TRW) are also standard tree rings parameters computed automatically by most software programs. They do not change with different k values, but still were calculated in this study for comparison.
Different laboratories and authors have used different k values to obtain dendroclimatic parameter time series, e.g. k = 30 % (Österreicher et al., 2015) and k = 50 % (Björklund et al., 2017), but the reasoning behind this choice is not explained. At the same time, k = 80 % has been shown to align best with visual and wood anatomical definitions (Samusevich et al., 2020). We hypothesize that the k value has a large influence on the climatic signal of obtained time series, which thus impacts the potential of tree ring studies to derive climate information. Indeed, previous studies have shown that different methodological approaches in the preparation and analysis of samples can impact the climatic signal of derived parameter time series (Björklund et al., 2019). Therefore, the goal of this study is to test different k values on the EW/LW boundary depended tree ring parameters and correlate the obtained series with instrumental climate data. Thus, we aim to explore whether and how different definitions of EW and LW, based on the k value, influence the retained climatic signal of a range of dendroclimatic parameter time series.
Section snippets
Methods
Two cores per tree were extracted from 16 living spruce trees (Picea abies) growing at a west-facing slope near the local tree line at 1640–1740 m a.s.l. in the Kitzbühel Alps, Austria (Fig. 1, site code: KARB; 12°'28'16" E, 47°23'28" N). The site is beneath a ridge with a peak at 1812 m a.s.l., and the surrounding topography is characterized by rounded mountains of the Northern Alpine Greywacke Zone that were subjected to mining in prehistoric times (Pichler et al., 2009). Primary rock soils
Results
The 39 chronologies cover the period 1832 to 2008, sample depth is at least 7 from 1853 (first year of the temperature data set) and full replication (16 trees) is available from 1876 onwards (Fig. 3a). The TRW chronology shows high crossdating results, e.g. t-values up to 18 and Gleichläufigkeit values up to 80 %, with several spruce chronologies from comparable elevated sites in the Alps (not published) demonstrating a coherent variability of this local chronology. The individually
Discussion
Here, we present a data set that shows the dendroclimatic potential of several tree ring parameters as a function of the setting of the EW/LW boundary. Hence, we want to contribute to informed decision making regarding the choice of the k value in future studies. It is, however, important to point out that our results rely on one data set of spruce from a high-elevated site in the Alps, and thus, our work can be considered a singular case study. For other locations and with different species -
Conclusions
Different laboratories and authors have used different definitions of EW and LW to derive tree ring chronologies based on EW or LW information. Here, we analyze how such variations influence the dendroclimatic potential of BI based tree ring parameters. We performed a correlation test between temperature and tree ring chronologies where the EW/LW proportion was varied by setting different values for the percental threshold k which defines the EW/LW demarcation. We show that absorption intensity
Declaration of Competing Interest
The authors report no declarations of interest.
Acknowledgements
We thank the Austrian meteorological office (ZAMG), especially Ingeborg Auer, for providing the temperature data.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
References (42)
- et al.
Summer temperature patterns over Europe: a reconstruction from 1750 AD based on maximum latewood density indices of conifers
Quat. Res.
(1988) A dendrochronology program library in R (dplR)
Dendrochronologia
(2008)- et al.
Divergent growth of Norway spruce on Babia Góra Mountain in the western Carpathians
Dendrochronologia
(2018) - et al.
Characterization and climate response patterns of a high-elevation, multi-species tree-ring network in the European Alps
Dendrochronologia
(2005) - et al.
The core-microtome: a new tool for surface preparation on cores and time series analysis of varying cell parameters
Dendrochronologia
(2010) - et al.
Different maximum latewood density and blue intensity measurements techniques reveal similar results
Dendrochronologia
(2018) - et al.
A “signal-free” approach to dendroclimatic standardisation
Dendrochronologia
(2008) - et al.
Dendrochronological analysis and dating of wooden artefacts from the prehistoric copper mine Kelchalm/Kitzbühel (Austria)
Dendrochronologia
(2009) - et al.
Blue intensity for dendroclimatology: should we have the blues? Experiments from Scotland
Dendrochronologia
(2014) - et al.
Comparison of methods for the demarcation between earlywood and latewood in tree rings of Norway spruce
Dendrochronologia
(2020)
Facilitating tree-ring dating of historic conifer timbers using Blue Intensity
J. Archaeol. Sci.
HISTALP—Historical instrumental climatological surface time series of the Greater Alpine Region
Int. J. Climatol.
Blue intensity parameters derived from Ponderosa pine tree rings characterize intra-annual density fluctuations and reveal seasonally divergent water limitations
Trees
Blue intensity and density from northern Fennoscandian tree rings, exploring the potential to improve summer temperature reconstructions with earlywood information
Clim. Past
Using adjusted blue intensity data to attain high-quality summer temperature information: a case study from Central Scandinavia
Holocene
Cell size and wall dimensions drive distinct variability of earlywood and latewood density in Northern Hemisphere conifers
New Phytol.
Scientific merits and analytical challenges of tree-ring densitometry
Rev. Geophys.
Alterstrend bei Jahrringdichten und Jahrringbreiten von Nadelholzern und sein Ausgleich
Tree-ring width and density data around the Northern Hemisphere: part 1, local and regional climate signals
Holocene
Testing for tree-ring divergence in the European Alps
Glob. Change Biol.
Blue intensity in Pinus sylvestris tree rings: a manual for a new palaeoclimate proxy
Tree. Res.
Cited by (6)
A definition and standardised terminology for Blue Intensity from Conifers
2024, DendrochronologiaCharacteristics of a multi-species conifer network of wood properties chronologies from Southern Australia
2022, DendrochronologiaCitation Excerpt :Additionally, earlywood and latewood definitions in these species have thus far relied on arbitrary algorithms applied to Silviscan (see below) output and are likely to be suboptimal. Useful methodologies for identifying early- and late-wood in a limited number of species have been discussed by Samusevich et al. (2020) and Frank and Nicolussi (2021) but have not yet been adequately tested for the Tasmanian species. Ring width chronologies were first developed for all sites following standard techniques of visual crossdating after which ring widths were measured using a Velmex measuring stage attached to digital encoder and computer.
I-BIND: International Blue intensity network development working group
2021, DendrochronologiaCitation Excerpt :The existence of x-ray measurements additionally helped to improve the understanding of BI, solidifying the notion that BI can be employed as a stand-alone tree-ring parameter. The constant effort to improve the understanding of both the biological and physical principals behind BI is continuing by a growing number of labs experimenting with these parameters which is reflected in a steady growth of publications dedicated fully or partially to such improvements (Shepard et al. 1996; McCarroll et al., 2002; Campbell et al., 2007, 2011; Babst et al., 2009; Rydval et al., 2014; Wilson et al., 2014; Björklund et al., 2014; Kaczka et al., 2018; Björklund et al., 2019; Reid and Wilson, 2020; Björklund et al., 2020; Wang et al., 2020b; Frank and Nicolussi, 2020; Wilson et al., 2021). The development of new tools, especially the incorporation of BI measurement into standard tree-ring software packages such as Windendro (Regent Instruments, Canada), Cdendro & CooRecorder (Cybis Elektronik & Data AB, Sweden), and LignoVision (Rinntech-Metriwerk GmbH & Co.
Tree-ring hydrological research in the Himalaya: State of the art and future directions
2024, Progress in Physical GeographyBlue intensity of Swiss stone pine as a high-frequency temperature proxy in the Alps
2023, European Journal of Forest ResearchDendroclimatic Potential of Blue Intensity-Based Chronologies of Northern Fennoscandia Scots Pine
2022, Journal of Siberian Federal University - Biology