Guest Post by Willis Eschenbach
Anthony highlighted a study called “Recent Plant Diversity Changes on Europe’s Mountain Peaks” (paywalled here , hereinafter Pauli2012). The Supplementary Online Information (SOI) is here.
The study concluded that the plants had moved vertically up the mountain by 2.7 metres, and Anthony was justifiably amused by the accuracy of the number, calculated to the nearest tenth of a metre. But that hardly begins to touch the oddity of the study. Here is their Figure S1, from their SOI:
Figure 1, from the Supplemental Online Information from Pauli2012. ORIGINAL CAPTION: Fig. S1. Basic design of a study summit, divided into eight summit area sections (SAS) that were used as sampling areas; the upper four areas extend from the summit point down to the 5-m contour line, the four lower from the 5-m to the 10-m contour line.
See what I mean about the oddity?
The crazy thing to me is, that they are only studying the area right at the very tippy-top of the summit of the mountains. They are solely and only looking at the top ten metres (33 feet) of vertical elevation of the mountain … and from that tiny vertical slice off of the mountaintop, they profess to be able to tell if the plants are moving uphill …
Now, I’ve spent a bit of time at the tops of mountains. They are subject to many variations in weather. The biggest one is the wind. Wind is a huge factor up at the mountaintops, and even a slight change in the average wind direction can turn a warm spot into a cold spot, or turn a wet spot into a dry spot.
So when (not if but when) there is a change in the composition of the plants eking out a living at the very mountaintop, my first suspect would be a change in the prevailing wind.
So, what do they have to say about the wind as a confounding factor in their study? Well … nothing. The wind doesn’t even get mentioned.
Next, the claim is made that a change in the warmth is allowing or encouraging the plants to move uphill. This presumes, of course, that the plants are near the top of their temperature range.
But these suckers are living at the very top of the mountain. Are we supposed to believe that somehow, in the mere ten vertical metres of the mountaintop that are being studied, the top limit of plants’ ability to resist cold temperatures just happens to fall in that very narrow range?
Next, we have to consider the difference in temperature due to a vertical move of 2.7 metres. The adiabatic lapse rate is 1°C per hundred metres vertical movement. That means the inherent temperature difference would be about 0.03°C …
Let’s be realistic. Plants that live on mountaintops live in cold, windy, dry conditions. Even the slightest change in any of those can easily stunt or kill off the plants that have a tenuous foothold there. Their range is constantly shifting and changing as those factors shift and change.
As a result, the only way to study the question would be with lots of temperature and wind and humidity and precipitation sensors scattered all around the mountaintop. The downwind side of the peak will be different from the upwind side. The sunrise side will be different from the sunset side. The side that gets the mist and clouds will be different from the dryer side.
Without those kinds of detailed measurements of those variables, any study done on this basis, of the top ten metres of mountain summits, will show us exactly nothing. There are too many confounding variables, and we cannot account for them without the necessary measurements.
I suppose I shouldn’t be surprised that this kind of rubbish gets published, but hope springs eternal … and in climate science, hope gets frustrated about as often …
In any case, I have long held that the quality of a scientific paper is inversely proportional to the square of the number of authors. This study, which is about four pages long, has 32 authors … just sayin’ …
w.
PS—Did you notice in Figure 1 that this is the gang that couldn’t draw straight? The inner box doesn’t line up with the outer box. So many authors … so few artists …
PPS—In researching this, I looked at a number of photos of mountain summits … many of them are steep, up to about 30° or so. The slope of many of them seemed to be somewhere around 10°. If the slope is 10°, the total study area is about 1,600 square metres. That’s less than a fifth of a hectare, or less than half an acre, a tiny area for such a study … so figuratively they are arguing not only about how many plants can dance on the head of a pin, but about just exactly where on the head of the pin they happen to be dancing this year, compared to where they were dancing seven years ago …
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Willis,
I managed to download more temperature logging files from the gloria project simply by guessing their structured URI (I can provide the 25MB zip file on demand). I calculated the anomalies and found them to be highly autocorrelated (as can be expected from experimental design). Thus to reduce the amount of data I computed daily averages and used GLS w/ AR1 to calculate slope and CI (library ‘nlme’).
The results are interesting (w/ respect to the science paper) since there are no significant temperature trends to be found nowhere. The smallest p-value obtained was 0.3 but that very time series just ran about one year…
what do you figure the temperature differance is a whole 2.7 metres higher? If it really made a differance then the top of the plant would be differant from the bottom of the plant.
Wolfgang Flamme says:
April 27, 2012 at 11:33 am
Very well done, Wolfgang, nice work. I was pretty sure that there was no significant temperature change, if there had been they would have trumpeted it …
w.
Willis,
actually I’m suspicious especially about the p-values obtained.You haven’t possibly been able to replicate some of them? Slopes seem to match with what your partial analysis says here (April 26, 2012 at 2:27 pm).
Hope I can try a slightly different ansatz this weekend for verification.
What I was taught about tree line was that it was the highest point on a slope where a new seedling could mature sufficiently from germination to the onset of winter to survive until the next spring. If you’ve ever really LOOKED at a mountain, you’ll see that the “tree line” isn’t static. The tree line climbs and dips along the side of a mountain, depending on conditions. Sometimes it’s higher on the north side of a mountain, simply because that’s the side that gets the most moisture, while the south side is perpetually in a dry zone. Soil nutrients, wind flow, moisture, sunlight, shadow (and what part of the day the shadow is present), can alter tree line. You also find that every once in a while there’s a tree growing “above timberline” in a given area, when for some unexplained reason it managed to survive conditions that killed off all other seedlings at that elevation. Unless this “study” (which from what I’ve read here occurred not every year for seven years, but once in 2001 and once in 2008) accounted for each of those variables, it’s worthless as a check of anything.
Also of note is that mountain tops tend to have highly variable populations of herbivores, year to year, (as populations are prone to predator / prey unstable oscillations) and they get a very wide range of lightning strikes. So is that plant moving uphill from warmth, or from ‘return to lighting burn’ area?
It’s also the case that we just had a dramatic shortening of atmospheric height as the sun went to near zero aa and UV plunged. THAT has put the cold layers at much lower altitude. (My local mountain, Mount Hamilton, has had MUCH more snow this year than the recent past.)
The simple fact is that “range” will be limited by the “very cold year” that comes every 60 years or so. Then the plants spend the next 59 crawling back up until the next extreme freeze kills ’em again… So in addition to “way too small” a study area they have “way to short a time window”. Needs at lest 60 years (and perhaps even more. There is that 1470 year Bond Event deep freeze cycle too…)