Physiological sensitivity of yellow-cedar to certain climate conditions appears to be range wide

Guess which site has mass yellow-cedar mortality?

(Mostly) range-wide climate data, from Amphitrite Point in Canada (about 48 degrees N) to Cannery Creek in Alaska (about 60 degrees north).  The sites in red have seen mass mortality as a result of hanging out in a climate transition zone to which yellow-cedar is uniquely maladapted.

(Mostly) range-wide climate data, from Amphitrite Point in Canada (about 48 degrees N) to Cannery Creek in Alaska (about 60 degrees north).  The sites in red have seen mass mortality as a result of hanging out in a climate transition zone to which yellow-cedar is uniquely maladapted.

Yellow-cedar mortality is well described, resulting from a physiological adaptation which takes advantage of historically reliable climatic cues for its phenology - specifically, cedar de-cold hardens early in the spring to take advantage of post-winter nitrogen availability.  Historically, deep snows have protected it from cold snaps and root freezing.  The lack of winter snow resulting from 1) emerging from the Little Ice Age and 2) anthropogenic warming is making those phenological stages vulnerable to freeze damage and mortality results.

Random forest modeling, conducted at the rangewide scale, identifies a distinct zone of mortality - shown here as a relative probability.  Higher values indicate more likely mortality - it's pretty clear that from about -5 to -1 or so is a pretty bad place to be.

Random forest modeling, conducted at the rangewide scale, identifies a distinct zone of mortality - shown here as a relative probability.  Higher values indicate more likely mortality - it's pretty clear that from about -5 to -1 or so is a pretty bad place to be.

Lab and greenhouse experiments have found that -5 C soil temperatures are, more or less, the point at which damage occurs in non-hardened individuals (they are quite cold tolerant earlier in the winter).  This requires a combination of cold air masses and a lack of snow, which generally only occur in areas where the mean winter temperatures are near zero - random forest modeling suggests that the 0 to -5 C mean temperature of the coldest month is the best predictor of where mortality occurs.

 

But climate change is, well, changing.  For more typical climate-induced mortality, like traditional physiological tolerance thresholds, a climate shifts and a whole landscape is changed - everything crosses.  But if the mortality and phenological mismatch is tied to a BAND of climate, like it appears here, then we have some interesting potential implications.

Specifically, this suggests that elevated rates of climate-induced mortality is temporary, assuming the climate keeps changing.  Eventually, as in the figure, you'll come out "on the other side."  Then mortality rates should decline.  Since mortality is usually triggered by proximal events (in this case, low snow + cold snap), it won't happen every year - so faster warming may, surprisingly, result in lower mortality overall because less time is spent in the transitional mortality zone (again, see figure).  

A paper detailing this, using yellow-cedar as a model organism and successfully predicting observed mortality rates based on weather station and climate data, is in review.

The transitional mortality zone hypothesis, which states that increased variability around a specific threshold drives mortality - not necessarily the threshold itself - holds up well in test cases.  One implication is that faster climate change may result in less severe mortality because less time is spent in the highly variable, exposed "danger zone."

The transitional mortality zone hypothesis, which states that increased variability around a specific threshold drives mortality - not necessarily the threshold itself - holds up well in test cases.  One implication is that faster climate change may result in less severe mortality because less time is spent in the highly variable, exposed "danger zone."