As you can see on my Current Research page, Kendra McLauchlan and I have been developing a collection of tree-ring chronologies throughout the entire contiguous United States (and maybe Alaska, if we’re feeling ambitious).
As you can see in the map, we currently have 9 sampling locations representing 7 states (CA, CO, KS, OK, MN, IN, NH), shown in green. An additional 12 states are currently being sampled by Novus RCN collaborators (OR, ID, WY, UT, NM, KY, WV, OH, NJ, NY, VT, FL) and are shown in yellow. The red dots represent locations where I am currently in the process of obtaining permission to sample, which represent 14 more states (NE, SD, ND, IA, MO, AR, TX, LA, TN, MS, AL, GA, NC, SC). Once all these regions are sampled, we will have a grand total of 33 states out of the anticipated 48, nearly 70% of our goal. Requests are out to Novus participants and Kendra/my personal networks to fill in the remaining 15 states.
The green locations, representing data currently in hand are datasets that are already published (see Poulson et al 1995, McLauchlan et al 2007, McLauchlan and Craine 2012, Wolfe et al 2013), or were collected with the intent to publish, but not yet having done so. These are easy to work with as all the hard (and expensive) work has already been done! In these cases, I simply receive a data file of δ15N values which I format to fit out meta data file.
The yellow locations primarily represent places where collaborators have agreed to sample on our behalf and send us tree cores. In these cases, Kendra and I will process and run the data ourselves. The process (seen in the photos below) consists of sanding a core to clarify the rings, imaging the core in order to electronically measure ring widths, splitting the core so each ring is separated from the others, then weighing out the appropriate amount of tissue for N-isotope analysis and wrapping it in a tiny tin-foil capsule to aid in combustion in the mass spectrometer.
This is where things get a bit tricky. Cores are small (the ones pictured above are 3-4 mm wide), but otherwise easy to use. Cores naturally split apart at the ring boundary (since the ‘ring’ you see is the result of denser tissue at the end of the growing season), and so splitting rings apart from cores is speedy and rather painless. The down side though is that their tiny size and rather fragile nature means they sometimes break easily, which can cause problems if sections are lost or mis-ordered when split. Tininess also means that narrow rings will not provide enough tissue to run N-isotope analysis (which requires 10-30 mg of tissue), so we will not be able to sample at an annual resolution.
Cross-sections (also known as ‘cookies’) are easier in that they provide ample amounts of tissue, and having the whole cross-section can aid in the identification of false rings (the appearance of a ring formed during the growing season, i.e. what looks like two rings is actually one growing season) and missing or locally absent rings (years in which a ring was not formed at all, or only formed in certain parts of the trunk, and therefore not visible across the entire cross-section). False and missing rings can seriously hinder cross-dating of ring chronologies, where a calendar year is assigned to each ring, so having a full cross-section can be very beneficial. The down side is that their large size (the ones pictured from Alaska are nearly a foot across and several inches thick) make it difficult to split tissue from each ring. Splitting rings from cookies often requires a hammer and fine-edged chisel to pry tissue apart.
You’ll also notice in the photos above that trees of the same species, living in the same region (those pictured being Pinus rigida from New Jersey), can be very different in appearance. In some, the ring boundaries are extremely faint. In others, they are very dark and contain wide late wood bands (the light portion of the ring is the early wood, laid down first in the growing season, while the rest is termed late wood), while others are very narrow. This variability is why the ratio of early- to late-wood in a chronology is a common dendrochronological metric. One core, which covered the entire cross-section of the tree shows much wider ring widths on one side of the center than the other. This pattern indicates the tree was not growing straight up, but leaning to the narrower-ringed side. Additional tissue is laid down on the ‘back’ side of the tree to balance out it’s weight.
Honestly, those differences are part of why I love dendro work. Every time I sand a new core, I don’t know what I’ll find – wide rings or narrow? Dark/obvious boundaries or very faint? Bright orange heartwood or pale tan? What drives these differences? In fact, much of that is not known – particularly when the differences occur within a species, in a small geographic range.
On the flip side, sometimes interesting patterns are visible within a group of trees. As I was processing the Idaho cores, I noticed they all had noticeably more narrow rings in the last ~50 years than the rest of their life span, and that almost every tree showed an unusually large ring in 2004. Why was 2004 an unusually productive year? The fact that it was every tree suggests a large-scale phenomenon. Perhaps it was unusually wet or warm – was 2004 an El Niño year? Does Idaho experience climatic differences in El Niño years? There have been many El Niño’s in the last ~50 years, why did they each not produce similar growth enhancements? If not a climatic factor, what else could have caused anomalously high growth in an isolated year? Why are the last several decades so low in productivity? Is this region experiencing some sort of stress or disturbance that is impacting tree health? Or perhaps this is a natural aging pattern in this species? I have tentative answers to most of these questions, but I look forward to diving into them in more detail. Each region will present it’s own set of similar questions, and hopefully, we can build some broad trends between sites.
I am currently working with samples mailed to us from Idaho, New Jersey, and Alaska and anticipating additional samples in the coming weeks from several more states. These will be the first data collected specifically for this project. We are also waiting to hear back on a funding request, which would enable the 10-day road trip to sample the fourteen red locations. Keep your fingers crossed!