Proxy Climate Data and Models of the 6 ka Time Interval
FILE COpy / RETURN TO:
6 ka Mean July Temperature
in Eastern Canada from
Bartlein and Webb's (1985)
Pollen Transfer Functions:
Comments and Illustrations
J.H. McAndrews
Department of Botany, Royal Ontario Museum, 100
Queens Park, Toronto, Ontario M5S 2C6 and
Departments of Botany and Geology, University of
Toronto
I.D. Campbell
Forestry Canada, 5320-122 St., Edmonton, Alberta
T6H3S5
Fossil pollen are the most abundant form of proxy data for
testing postglacial climate models. A large data set of pollen
surface samples linked with climate stations has been used to
calculate temperature and precipitation as transfer functions
(Bartlein and Webb 1985, Gajewski 1988) and as response
surfaces (Prentice et al. 1991). Fossil pollen diagrams from
lake sediment permit climate reconstruction through postglacial time and specifically for the alleged hypsithermal.
The most ambitious effort using multiple-regression to derive
transfer functions was Bartlein and Webb (1985) who partitioned eastern North America into 13 calibration regions of
which 11 are relevant to Canada. Using 211 pollen diagrams
they identified the 6 ka pollen spectrum, calculated the mean
July temperature using the function from the region with the
best modern analogue and mapped the results. They found
that at 6 ka: 1) the steepest latitudinal temperature gradient
was roughly 100 km north of its modern position in the
latitude of the southern Great Lakes. 2) southern and southcentral Canada was over one degree warmer than today,
3) residual ice sheets kept the near-north over 2 degrees
cooler than today and 4) the southern prairie provinces were
only as warm or cooler than today.
PUBLICATIONS
NORTHERN FORESTRY CENTRE
5320 - 122 STREET
EDMONTON, ALBERTA
T6H 3S5
2) In the western states adjacent to the prairie provinces the
6 ka temperature was calculated to be cooler than today
which is inconsistent with evidence that it was warmer
than today (McAndrews 1966). We feel that this is because ragweed (Ambrosia), which has proliferated historically because it is adapted to European disturbance,
was not used to calibrate the regional transfer function.
Immediate prehistoric ragweed pollen abundance unrelated to human disturbance occurs southward in the relatively warm Mississippi valley region. but at 6 ka it was
more abundant northward, implying warmer summer
temperatures.
3) There is no modern pollen analogue for the high 6 ka
beech (Fagus) pollen percentage of southern Ontario. We
suspect that the temperature may be higher than calculated.
We suggest that new and better regional transfer functions be
derived: 1) by using only surface samples where the regional
pollen rain is expressed, i.e. lake surface sediment, 2) by using
calibration samples where pollen identifications are reliable,
and 3) by using immediately prehistoric spectra where vegetation is disturbed and not in equilibrium with the environFig. 1.
Map of northeastern North America showing
vegetation zones and surface sample sites used
by Bartlein and Webb (1985) to derive transfer
functions. Letters locate pollen diagram sites:
R . R Lake, U . Upper Mal/ot Lake, 0 . Lake OB,
T· Tonawa Lake. and H . Hams Lake.
We applied their transfer functions to the 6 ka pollen spectrum
in each of 41 pollen diagrams ranging westward from Newfoundland to Saskatchewan, Minnesota and North Dakota
(Fig. I). Figures 2-6 show five Ontario pollen diagrams
spanning the postglacial together with the calculated mean
July temperature. Our results generally confirm their results,
but three insights were derived.
1) Although Lake Superior generates a distinctively cold
July air mass today, at 6 ka the July temperature near the
lake was anomalously 2.5 degrees warmer than today
(Fig. 3), whereas a more distant site was only a half degree
warmer (Fig. 4). This implies that at 6 ka lake Superior
did not generate a distinctive summer air mass.
22
' 00 '
>0 '
6 ka Mean July Temperature in Eastern Canada
R Lake, Northern Ontario
B.T"*UUI 6.0. WlBilkll . OJ IIIy~
Jul y Mean Temperature
1
R>diocarbon
T
T
Dates
2000
2000
2370
c..
co
Ib
4000
4000
6000
6490
I.
• Tyrrell
Sea
8000
Ip
%0
30
600
300
300
o
o
IO
11
12
13
14
15
16
degr<es C
Pen:cotage of pollen swn.
Pollen diagram for R Lake (McAndrews et al. 1982) and derived mean July
temperature showing that 6 ka temperature was lower than today due to residual
glacier ice. The temperature curve was smoothed using an un weighted three-point
running mean of values derived from transfer function A of Bartlein and Webb (1985).
Triangles indicate the mean July temperature values at the two nearest climate
stations: Winisk on the coast is 100 km distant whereas Big Trout Lake is inland and
380 km distant. Pollen zones follow McAndrews (1993); subzone 1p contains recycled
pollen (ef. McAndrews 1984) and thus produces a spuriously high temperature.
Fig. 2.
Upper Mallot Lake, Central Ontario
I _ ~.."..
July Mean Temperature
1
2000
"
c.. 4000
4660
4000
CD
2'
~6000
,"'0
•
6000
>-
2.
8000
8000
l'
983S
Ja Ip
If.
30 0
0
)0 0
Percentage: or pollen SWD.
Fig. 3.
) 0600
30600
0
000
IS
16
\7
18
19
20 2 1
dccr=C
Pollen diagram for Upper Mallot Lake and smoothed mean July temperature using
transfer function J. Note that the hypsithermal at 6 ka is about 3 degrees higher than
today implying that Lake Superior did not have a cooling effect on local climate.
Searchmount climate station is 40 km inland from Lake Superior and 60 km distant.
23
Proxy Climate Data and Models of the 6 ka Time Interval
Lake QC. Central Ontario
R~iocarbon
Dales
~ 940
"c:l
2b
6000
~
6530
r"'"
8000
'600
10000
Ib
Ip
,,0
30
IS
30 6 0 0 3 0 0
0
Fig. 4.
16
17
18
19 20
21
devecs C
Pen:au..age of pollen sum.
Pol/en diagram for Lake QC and smoothed mean July temperature using transfer
function J. Note that the 6 ka temperature is similar to that of Upper Mal/ot Lake.
Sudbury climate station is 40 km distant from Lake QC.
Tonawa Lake. Southern Ontario
.
"
<OllO
,. """
IllQ)
I. I_
10000
II Ip
,0
30
0
30
60
0
0
0
0
0
0
30 0
0
0
0
0
Fig. 5.
_c
10 11
Percc:ntage of poIlcu sum .
12
\J
14
"
\6
11 11 19 20 21 22
Pol/en diagram for Tonawa Lake and smoothed transfer functions G, J, and A. Bancroft
climate station is 55 km distant.
Hams Lake. Southern Ontario
,
Rod""""",
D.~
I
,)0
14110
1<XX1
I
:!670
"'0
4270
:;;
l'170
~
' 500
1161 0
91)0
>-"
<OOO l
!
""' ~
i
I
""" i
-
10000 1..
10 400
... 0
;0
I
I
~
24
t
l1lOO
"
J.
~
~
~
J.
"""
"'"
)
11000
I
I
,.
ooo}O oo
luly Mean Tem~
3d
~
00)0600
Percentage of pollen sum.
Fig. 6.
.
I
2l>
"
I
,0000
lb
oo
o
:}()o
0
0
0
" ""..-C
" 11 " depaC
\6
19 20 :!I
!2 23
24
Pol/en diagram for Hams Lake adapted from Bennett (1 987) with additions and
smoothed transfer functions G and J. Brantford climate station is 15 km distant.
6 ka Mean July Temperature in Eastern Canada
men!. In addition statistical confidence intervals need to be
ca lculated.
We thank Z. Yu for his constructive comments and computer
programming.
References
Bartlein. PJ. and Webb, T., Ill. 1985. Mean July temperature at
6000 yr BP in eastern North America: regression equations for
es timates from fossil-pollen data. Syllogeus 55:301-342.
Bennett. K.D., 1987. Holocene history of forest trees in southern
Ontario. Canadian Journal of Botany 65 : 1792-1801.
Gajewski , K., 1988. Late Holocene climate changes in eastern North
America: estimates from pollen data. Quaternary Research
29:255-262.
McAndrews. J. H., 1966. Postglacial history of prairie, savanna and
forest in northwestern Minnesota. Memoirs of the Torrey
Botanical Club 22: 1-72.
McAndrews. J.H., 1984. Pollen analysis of the 1973 ice core from
Devon Island ice cap, Canada. Quaternary Research 22:68-76.
McAndrews, J.H. , 1993. Pollen diagrams from southern Ontario
applied to archaeology. In R. MacDonald and B. Warner, eds.
Great Lakes Archaeology and Paleoecology : Exploring
Interdisciplinary Initiatives for the Nineties. In press.
McAndrews, J.H., Riley, J.L., and Davis. A.M., 1982. Vegetation
history of the Hudson Bay Lowland: a postglacial pollen
diagram from the Sutton Ridge . Naturaliste canadien
109:597 -608.
Prentice, I.e., Bartlein, PJ., and Webb,T., III., 1991. Vegetation and
climate change in eastern North America since the last glacial
maximum. Ecology 72:2038-2056.
25
"
PROXY CLIMATE DATA and MODELS of the SIX THOUSAND
YEARS BEFORE PRESENT TIME INTERVAL:
The Canadian Perspective
Abstracts of a workshop
:.
"
Compiled by
Alice Telka
Terrain Sciences Division
Geological Survey of Canada
;\
Canadian Global Change Program
Incidental Report Series
NO. IR93-3
THE ROYAL SOCIETY OF CANADA
July 1993
I
~;.
~:;;';
. .
,«
\
.~
.
"
.
' ::~ 7