CLIMATOLOGICAL PROBLEMS IN THE ARCTIC AND SUB-ARCTIC 
in the coniferous trees that dominate the Boreal forest, 
at least in its northern parts. This does not mean 
that precipitation is unnecessary to the growth, but 
that the combined snow melt of May and July and 
the precipitation of June, July, August, and Septem- 
ber are everywhere enough to support the restricted 
growth these subarctic temperatures can sustain. 
This view has been challenged by Sanderson [50], 
who found that drought is a marked restricting factor 
in the Boreal forest of western Canada. By using 
Thornthwaite’s new system of climatic classification, 
she was able to show that widespread moisture defi- 
ciencies occur throughout the part of the Canadian 
Boreal forest lying west of Ontario; the entire region 
is dry sub-humid or semi-arid, according to her re- 
sults. She maintained that the forest is stunted by this 
summer drought, or may even be replaced by scrub 
or-prairie [48]. Soil acidity is much lower than one 
might expect for the typical Boreal forest podsols, 
pH determinations varying from 7.0 to 7.6 in northern 
Saskatchewan, and from 6.3 to 7.8 at Fort Simpson 
on the Mackenzie. These are values ranging from mildly 
acid to mildly alkaline, far different from the strongly 
acid reactions of soils from Quebec, Labrador, and 
the Finnish plateau. 
It is probable that the truth lies somewhere between 
these two opposed views. Hustich, Erlandsson, and the 
other workers of the European school have been ac- 
customed to the Boreal forest in its wetter phase— 
the Finnish and Lapp plateaus resembling Labrador 
and Northern Ontario in climate—and may have been 
misled by sheer lack of exposure to the drier types of 
microthermal climate. On the other hand Giddings’ 
tree-ring work on the Mackenzie [15] confirms Hus- 
tich’s view rather than Sanderson’s: he makes no men- 
tion of stunted rings due to drought years. Sanderson 
is also open to attack on the grounds that Thorn- 
thwaite’s scheme contains sweeping assumptions about 
available soil moisture that may not apply in perma- 
frost areas, or in areas of glacially disturbed drainage. 
As Clark [7] says, her view that drought restricts 
growth in the Canadian northwest is “‘... a somewhat 
startling conclusion for the many who have spent sum- 
mers there with almost perpetually wet feet! This 
apparent anomaly is explained as the result of poor 
drainage rather than of humid climate... .” 
It appears quite probable that there are certain 
areas of the Boreal forest where summer drought is 
effective in the manner Sanderson suggests, not only 
in Canada but in Siberia: salty or alkaline marshes 
and prairies occur in the Yakutsk basin of the Lena 
Valley, just as they do in the Peace River Valley [53]. 
But it also seems likely that once enough moisture is 
provided during the growing season, as it is in most areas, 
further moisture has little or no influence on the forest: 
it cannot induce greater luxuriance, nor can variations 
in available moisture above this quantity have any 
effect on the annual-ring record. In all but a few areas 
the effective climatic control within the Boreal forest 
appears to be thermal; the regional subdivisions of 
the forest are hence correlated with isopleths of tem- 
955 
perature or functions of temperature, and ignore the 
precipitation distribution. 
Zonal Divisions of Arctic and Sub-Arctic Vegetation. 
With this background, we can now examine the cli- 
matic correlations of the major divisions of natural 
vegetation. 
The Arctic Tree Line, the Northern Limit of Tree 
Growth. This line is often extremely sinuous, and is hard 
to map, but its course is now tolerably well known 
round the world. As it divides the truly Arctic tundras 
on the north from the Boreal forest formation, it is 
an extremely important boundary. 
Early attempts to find the poleward limits of tree 
growth stressed July temperature. Supan [52] observed 
that the isotherm of 10C (50F) for the warmest 
month followed the arctic tree line closely. Later K6p- 
pen [36] adopted this isotherm as the division between 
tundra (Ekistothermal) and microthermal climates. The 
general correspondence of the two lines can be seen 
on world maps, though in detail they may diverge 
considerably. The 10C isotherm has enjoyed a prestige 
among biologists and geographers far exceeding its 
merit. 
Vahl [62] differed from K6ppen in assigning some 
significance to winter temperatures. Forests are ob- 
served to flourish in certain maritime regions right up 
to or even beyond the 10C isotherm. Thus Tierra del 
Fuego, at the southern tip of South America, is forested, 
though the warmest month is only 8-9.8C (46-50F). 
Here the mean daily temperatures remain above 0C 
throughout the winter, and it is evident that the mild- 
ness of the cooler season is the permissive factor. In 
continental interiors, on the other hand, the tree line 
fails to reach the 10C isotherm for the warmest month. 
This isotherm passes just north of Baker Lake in 
Keewatin, for example, some 200 mi north of the tree 
line. Presumably the much greater winter cold of the 
continents is the operative factor. In short, the colder 
the winter, the warmer the summer has to be to sup- 
port tree growth. 
Nordenskjéld [45] tried to express this idea in a 
rough-and-ready substitute for the Képpen-Supan line. 
He suggested that the tree line was most closely fol- 
lowed by the isotherm 
W = 9 — 01k, (1) 
where W is the mean temperature of the warmest 
month and fk is the mean temperature of the coldest 
month, both in centigrade degrees. In Fahrenheit, this 
identity becomes 
W = 514 — O.1k. (2) 
The ‘‘Nordenskjéld line,” as it is often called, fits the 
arctic tree line better than any other purely climatic 
isopleth. Figure 1 shows its course. 
The attempts so far discussed are hit-and-miss affairs 
with no rational basis. No adequate discussion of tree- 
line climates is known to the present author. Thorn- 
thwaite has proposed a climatic classification supposedly 
“rational” in that it depends upon a reasoned considera- 
tion of the soil moisture cycle and the water needs of 
