Early Holocene climate at Castle Peak, southern Coast Mountains, British Columbia, Canada

Early Holocene climate at Castle Peak, southern Coast Mountains, British Columbia, Canada

Palaeogeography. Palaeoclimatology. Palaeoecology, 95 (1992): 153-167 Elsevier Science Publishers B.V., Amsterdam 153 Early Holocene climate at Cast...

3MB Sizes 4 Downloads 24 Views

Palaeogeography. Palaeoclimatology. Palaeoecology, 95 (1992): 153-167 Elsevier Science Publishers B.V., Amsterdam


Early Holocene climate at Castle Peak, southern Coast Mountains, British Columbia, Canada J o h n J. C l a g u e a, R . W . M a t h e w e s b, W . M . B u h a y c, a n d T . W . D . E d w a r d s ~

~Geological Survey of Canada, I00 West Pender St., Vancouver, B.C., V6B IR8, Canada hDepartment ¢ff Biological Sciences and Quaternary Research Institute, Simon Fraser UniversiO', Burnaby, B.C., V5A 1S6, Canada ~Department Of Earth Sciences, Quaternw3" Sciences Institute, and Waterloo Centre for Groundwater Research, University ¢~[Waterloo, WateHoo, Ont.. N2L 3G l, Canada (Received December 18, 1991: revised and accepted March 19, 1992)

ABSTRACT Clague, J.J., Mathewes, R.W.. Buhay, W.M. and Edwards, T.W.D.. 1992. Early Holocene climate at Castle Peak, southern Coast Mountains, British Columbia, Canada. Palaeogeogr., Palaeoclimatol.. Palaeoecol.. 95:[53 [67. New palynological, radiocarbon, and stable isotope data from Castle Peak m the southern Coast Mountains indicate that the earl}' Holocene climate of southern British Columbia was warmer and perhaps drier than today. Fossil wood fragments are common above timberline at Castle Peak and have yielded 13 radiocarbon ages ranging from 9.1 ka to 8. I ka. This evidence for higher timberline during the early Holocene is in agreement with pollen data indicating a warmer and possibly longer summer growing season in this area during the early Holocene. and is consistent with theoretical considerations based on Milankovitch forcing of climate change. The oxygen and hydrogen stable isotope composition of fossil wood cellulose is similar to that of living trees in the same area. Growing season relative humidity values derived from the isotopic data provide some support for episodes of dryness at times of elevated timberline, although the majority of paleohumidity estimates fall within the range of modern moisture levels.

Introduction P a l e o c n v i r o n m e n t a l d a t a from the s o u t h e r n Canadian Cordillera and adjacent northwestern United States suggest that climate d u r i n g at least part o f the early H o l o c e n e was w a r m e r and drier than t o d a y (see Baker, 1983; M a t h e w e s , 1985; B a r n o s k y et al., 1987: A n d e r s o n et al., 1989; C l a g u e and M a t h e w e s , 1989: a n d C a r r a r a et al., 1991, for reviews a n d references). These d a t a include pollen a n d p l a n t macrofossil records, occurrences o f fossil w o o d a b o v e present timberline, isotopic d e t e r m i n a t i o n s on such w o o d , a n d sediment records from s u b a l p i n e lakes. Fossil conifer w o o d has been f o u n d at a n d a b o v e

Corre.wondence to: J.J. Ctague. Geological Survey of Canada, 100 West Pender St.. Vancouver, B.C., V6B IR8. Canada. 0031-0182:92/$t)5.00

timberline in m a n y places in s o u t h e r n British C o l u m b i a and A l b e r t a ( K e a r n e y a n d L u c k m a n , 1983a, b: L u c k m a n a n d K e a r n e y , 1986; C l a g u e and M a t h e w e s , 1989). Such occurrences p r o v i d e direct evidence that climate in these areas has been w a r m e r in the past. We previously r e p o r t e d such an o c c u r r e n c e at Castle Peak, in the s o u t h e r n C o a s t M o u n t a i n s o f British C o l u m b i a (Fig. 1) and, on the basis o f the d i s t r i b u t i o n and age o f the fossil w o o d , suggested that m e a n g r o w i n g season t e m p e r a t u r e in this area m a y have been 0.4-0.8¢~C w a r m e r than at present from 9.1 ka to 8.2 ka (Clague and M a t h e w e s , 1989). In this paper, we present: (1) new p a l y n o l o g i c a l d a t a which s u p p o r t the c o n t e n t i o n that timberline at the Castle Peak site was higher d u r i n g the early H o l o c e n e than t o d a y ; a n d (2) oxygen a n d h y d r o g e n i s o t o p e d a t a which p r o v i d e i n f o r m a t i o n on paleoh u m i d i t y at this high elevation site. These d a t a

.{.!' 1992 --- Elsevier Science Publishers B.V. All rights reserved.



123000 /

1 2 2 ° 5 8 ,`

122o56 /








Fig. I. Map of the study area showing the locations of fossil wood sites. Pollen analysis was performed on sediments at site 1. Stipple= forested terrain. Topographiccontours in feet.

provide a better understanding of climate and vegetation in the mountains of southern British Columbia during what appears to have been the warmest part of the Holocene. Apart from its obvious scientific interest, the warm climate of the early Holocene, which can be inferred from data of the type presented here, may have importance as an analogue for the anticipated warmer climate of the next century.

Setting The Castle Peak cirque is located at the headwaters of Paradise Creek in the southeastern Coast Mountains (Figs. l and 2). It faces northwest and is bordered by peaks ranging in elevation from 2400 m to 2490 m. The cirque floor extends down to 2070 m elevation, which is about 70 m above timberline.

E A R L Y H O L O C E N E C L I M A T E AT C A S T L E P E A K , S O U T H E R N C O A S T M O U N T A I N S . B R I T I S H C O L U M B I A


Fig. 2. Floor of the Castle Peak cirque; view to the west from the vicinity of site 6 (Fig. 1). The cirque floor is covered by colluvium and alluvium.

Scattered clumps of krummholz Abies lasiocarpa (subalpine fir) occur above timberline on the lower slopes of the cirque, and isolated krummholz Pinus albicaulis (whitebark pine) and Picea engelrnannii (Engelmann spruce) grow to elevations of 2180 m on otherwise treeless north- and west-facing slopes and to 2320 m on south-facing slopes• Bare rock is exposed on steeper slopes of the Castle Peak cirque. The lower slopes and floor of the cirque are mantled by colluvium and alluvium of latest Pleistocene and Holocene age (Figs. 2 and 3). Much of this sediment is diamicton consisting of angular clasts of local provenance in a matrix

rich in silt and clay. The diamicton probably is reworked till deposited by creep or flow from adjacent slopes during or shortly after deglaciation. Wedge-shaped bodies of stratified silt, sand, and gravel locally overlie the diamicton unit and probably are slopewash deposits. The presence of peat beds and possible paleosols within these stratified sediments suggests that slopewash activity was episodic. These sediments, as well as talus adjacent to steep rock slopes, have been reworked into gravelly alluvium by Paradise Creek and its tributaries. The alluvium is locally inset into the diamicton (Fig. 3).

• , °

d ~'~ ~<--~o~ . . ~ v


o o


Sand. Gravel. Silt (Slopewash?)


t> C> A ~ A~,J v

3-0 ~ " = o : ' ~ ' ~ -

Gravel (Alluvium)

<3 L ~ - . . < 3 Diamicton (Colluvium)

Fig. 3. Generalized cross section of late Pleistocene and early Holocene ~diments at site l (see Figs. l and 2 for location). Pollen analysis was performed on the ,sediments directly overlying the arrowed log.


J.J. C L A G U E

Fossil wood Logs, branches, and other smaller woody detritus are c o m m o n at the surface and in colluvial and alluvial sediments within the Castle Peak cirque, well above present timberline (Figs. I and 4; Clague and Mathewes, 1989). The wood that has been identified is entirely white pine (probably Pinus albicaulis) and fir (probably Abies lasiocarpa). Pinus albicaulis and Abies lasiocarpa pres-


ently characterize the timberline at Castle Peak and in the adjacent eastern sections of the Coast Mountains (Selby and Pitt, 1984). The highest occurrences (2192-2237 m elevation) are branches and small logs, up to I1 cm in diameter, lying on the surface. These have been released from a thin layer of diamictic colluvium by solifluction and gullying. The small size and gnarled form of the fossil wood at these high sites suggest that timberline may have been


q9 !

~4P ,.

Fig. 4. Fossil wood at site 1, Castle Peak cirque. A. Pine log projecting from wedge of stratified sediments overlying diamicton. This log directly underlies sediments that were analyzed for pollen (arrow indicates peat bed) and has been radiocarbon dated at 9120 + 120 yr B.P. (Table I). B. Fir log lying on alluvium at the base of the section. The age of this log is 8580 + 90 yr B.P. (Table 1).


near this elevation at the time the trees were living. Much larger logs are present on the floor of the cirque, both on the surface and within colluvial and alluvial sediments (Clague and Mathewes, 1989), Of particular interest is an exposure of diamicton and overlying stratified silt, sand, gravel, and peat at about 2090 m elevation adjacent to Paradise Creek (site 1, Figs. 1 and 3). A pine log (diameter = 25 cm) was found within the stratified sediments, 3 m below the top of the section (Fig. 4A). and a fir log (length = 4.3 m; maximum diameter= 20cm) at the base of the section (Fig. 4B) was undoubtedly eroded from the same sediments. The presence of long straight logs at this sitc indicates that tall upright trees were growing in the lower part of the cirque. The nearest living pine and fir trees of comparable size in areas of similar aspect are, respectively, about 60 m and 130 m Iowcr in elevation. The stratified sediments at this site also contain abundant seeds, pollen, and spores which have provided additional information on early Holocene timberline positions in this area (see next section).


The fossil wood has yielded 13 radiocarbon ages ranging from 8100 + 90 yr B.P. to 9120 4- 120 yr B.P. (Table I). Eight of the dated samples are logs and branches found on the surface (GSC4364, -4962, and -4970; RIDDL-885 and -886) and in the uppermost 30 50cm of recent alluvium (GSC-4961; S-3200 and -3201). This wood clearly has been eroded from older colluvial and slopewash sediments. One of the dated samples (GSC4591) is from the diamicton unit, and another (RIDDL-975) is from the base of a wedge of stratified sediments overlying the diamicton. I-:inally, three samples (TO- 1549, - 1550, and - t 551 ) are from a peat bed within the stratified wedge at site 1. Inspection of the data in Table 1 shows that there is no relationship between radiocarbon age and sample elevation. Likewise, surface and subsurface samples yield similar ages. The data simply indicate that there was a period of higher timberline at Castle Peak from at least 9.1 ka to at least 8.1 ka. The seemingly fortuitous preservation of abundant 8 9 ka fossil wood is probably due to the fact that extensive colluvium and slopewash

TABLE I Radiocarbon ages of wood from the Castle Peak cirque 1.*(; age (yr B.P.) ~

Laboratory no. 2

6 t 3C

Sample no.

Site (Fig. 1)

Elevation (m)

Enclosing material 3

8100 8230 8340 8350 8580 8710 8760 8770 8810 8820 886(I 9010 9120

TO-1551 GSC-4364 TO-1549 TO-1550 RIDDL-885 RIDDL-886 S-320[ (.~SC-4962 S-3200 GSC-4961 GSC-4970 GSC-4591 Rll)DL-975


CP21 86TD-21A '~ CPI8 CP20 CPI s CPI0 a CPI6 CP6'* CPI4 CP5 s CP9 s CPT.* CP2'*

l 5 I I 1 6 3 3 3 3 5 3 1

209l 2192 2090 2090 2089 2237 2091 2091 2091 2091 2[92 2091 2090


j 90 +_ 110 _+ 80 _+ 90 _+ 90 _+ 130 -,- 110 _+ 90 _+ 130 + 9(1 -¢ 80 _- 90 -- 120

-23.8 -24.3 -24.6 • 22.4

Peat Peat

Gravel Gravel Gravel r)iamicton Graxcl

~RII)DI,, TO, and S ages have 1 a error terms: GSC ages have 2 o error terms. R I D D L and TO ages were determined by the accelerator mass spectrometry method: GSC and S ages arc conventional. 2Laboratories: GSC. Geological Survey of Canada; RIDDL, Radio-Isotope Direct Detection Laboratory (McMastcr University): S, Saskatchewan Research Council; TO. IsoTrace Laboratory (University of TorontoL 3Dashcs indicate that dated wood was at the surface. 4Pinto sp. (white pineL SAbles sp.


accumulated on the walls and floor of the cirque during this part of the Holocene. Pollen record

Methods Stratified sediments at site 1 (Fig. 1) were sampled in August 1989 for pollen analysis. A 90cm section was studied, ranging from sandy gravel at the base, through peat and silty peat, into silt, sand, and gravel at the top (Fig. 5). Samples of ca. 50-100cm 3 were cut from a cleaned face of the exposure at 2-5 cm intervals, placed in labelled plastic bags, and stored at 4°C until they were subsampled. Subsamples of 2 ml were measured by displacement in distilled water, sieved at 250 p,m to remove coarse debris, and processed using a standard chemical maceration with 10% HC1, 48% HF, 6% KOH, and acetolysis (Faegri and lversen, 1975). A tablet containing 16,180 + 1460 Eucalyptus pollen grains was added to each subsample before processing to calculate pollen concentrations. Clay-rich samples were microsieved with 7 lam Nitex mesh (Cwynar et al., 1979) after dispersion in sodium pyrophosphate solution. Samples from three modern moss polsters at wet sites in the present alpine/subalpine transition at Castle Peak were sampled and processed for pollen in the same way as the fossil material. Processed residues were stained with safranin and mounted on glass slides in both glycerine jelly and silicone oil. Pollen counts (237-389 per sample) were made in evenly spaced transects at 500 x magnification. Difficult and rare grains were checked with the modern pollen reference collection at Simon Fraser University. Sums for percentage calculation included all trees, shrubs, and herbs, but excluded aquatics and spores. Results The results of the pollen analysis are summarized in Fig. 5, along with the radiocarbon ages and stratigraphy of the sampled section. All abundant and/or climatically significant taxa are shown, but some rare types were excluded to simplify data presentation. "The gaps in the pollen diagram at


10-20cm and 80-90cm are due to low pollen concentrations, poor pollen preservation, or both. At all other levels, pollen recovery and preservation was good. More than 50% of all pine pollen could be separated into the yellow pine (Pinus contorta type) and white pine (Pinus albieaulis type) categories. In addition to pollen and spores, the sediments contain other plant remains such as Pinus albieaulis seeds, conifer wood tracheid fragments, and stomatal guard cells from conifer needles (Fig. 6A). Two local pollen assemblage zones, which reflect different environments of deposition, are recognized. The lower zone (CP-I) encompasses the sandy gravel below 55 cm, while the upper zone (CP-2) spans the peaty part of the sequence. Zone CP- 1 (Pinus contorta--Shepherdia, 80-55 cm, <9120-8350 yr B.P.) records pioneering vegetation with maximum percentages of Pinus contorta type (Iodgepole pine), Alnus (alder), Cupressaceae (probably Juniperus, juniper), and Shepherdia canadensis (soopolallie) pollen. The most diagnostic species is Shepherdia canadensis (Fig. 6B), an insect-pollinated shrub with low pollen productivity. Pollen frequencies of this species in zone CP-I are 2-10%, values which indicate that the plant was present locally. This inference is strengthened by the presence of large clumps of Shepherdia pollen (Fig. 6C), suggesting that flowers or detached anthers were deposited in the gravel. Elaeagnus commutata (silverberry), another pioneering shrub in the same family as Shepherdia, has its only occurrence in this zone; two grains (Fig. 6D, E) were recovered from the basal sample. Pinus albicaulis type pollen (Fig. 6F) is present at low frequencies (3-4%) at the base of the zone and increases at the top. Abies pollen (Fig. 6F) exhibits a similar pattern. Zone CP-2 ( Pinus albicaulis-Abies-Cyperaceaeherbs; 55-0 cm; ca. 8350-8100 yr B.P.) is characterized by consistently high values of tree pollen, particularly Pinus contorta type (up to 55%), Pinus albicaulis type (16-36%), and Abies (3-7%). Cyperaceae pollen (Fig. 6G) increases steadily from the base of the zone to the top, accompanied by increases in herb pollen, notably Salix (willow), Epilobium (fireweed type, Fig. 6G), Trientalis (starflower), Parnassia (grass-of-Parnassus), Caltha


A ' I C A S T L E P E A K . SOL:I H E R N C O A S T M O t . N T A I N S , B R I I I S H



o %o




%o •

% .o~



"-~% - - - ~ - - - - - ~



~__~..\ 0



<<....>_. ~ ~

0 o L-

, \



. . . . .


L o-









- ~



. . . . . . . . . . .

% "~>~

.- o

o.><> .'~,

% ,~. ~

•" ~


r -o




F 0






L_._ 0








M---'~ ..... ;




_ o


.->. ~.-





%"<>o._<~ ~ I-


°,1,,, ~





~ ' .





....7^,.......... ~.1





ri- °[email protected]


i,}~ t














e, ~'"~'"



" i I ~#c/ , , ~ l ~ ; ~ l : ~ - . - 0

.,>-° ">~.~





T m ~ .~ .....






i¸ ~i

Fig. 6. Selected pollen and other microfossils from the section at site I. Magnifications are 500X for a, f, g, and h, and 1000x for b, c, d, and e. a. Pinus pollen grain in association with conifer guard cells (GC) and a coniferous wood tracheid fragment (T). b. Two joined pollen grains of Shepherdia canadensis from zone CP-I. c. Clump of Shepherdia canadensis pollen grains, d, e. Pollen grains of Elaeagnus commutata from the base of zone CP-I. f. Large Abies grain and typical white pine pollen grain (Pinus albicaulis type), characterized by warty verrucac (V) on the distal membrane, g. Phase contrast photo of Cyperaceae pollen (C) and triporate Epilobium grain with preserved viscin threads (VT) from zone CP-2. h. Geranium cf. G. richardsonii pollen grain from zone CP-2.


type (marsh marigold), Heracleum (cow parsnip), and Geranium cf. G. richards'onii (Fig. 6H). Potamogeton (pondweeds) pollen and the algal spores Sigmopollis also attain significant values here. These data indicate a trend towards increasing local wetness relative to zone CP-I.

Isotope data The oxygen and hydrogen stable isotope composition of wood cellulose preserves a record of the environmental conditions under which a tree lived (see for example, Burk and Stuiver, 1981; Yapp and Epstein, 1982; Edwards et al., 1985; Luckman et al., 1985; Edwards and Fritz, 1986; Ramesh et al., 1986). Although knowledge of the physiological processes that are involved remains incomplete (DeNiro and Cooper, 1990; Edwards, 1990), studies of modern trees have shown that oxygen and hydrogen isotope ratios in wood cellulose are determined primarily by the isotopic composition of the groundwater used by the tree and secondarily by fractionations occurring during evapotranspiration and cellulose synthesis.

Methods" Edwards and Fritz (1986) used data from modern trees to formulate a semi-empirical model that provides numerical estimates of groundwater isotopic composition and daytime growth season relative humidity ("photosynthetic humidity") from cellulose oxygen and hydrogen isotope measurements. Here we apply a slightly modified version of this model (Buhay et al., 1991) to compare past and present conditions near timberline in the Castle Peak cirque. Eleven samples of fossil wood and six samples of wood from trees living near timberline in the Castle Peak area were analyzed for 160 and 2H. All samples represent at least 10 years of growth; no attempt was made to analyze separate rings because of the "'noise" of year-to-year climatic variability. Cellulose was separated from finely ground wood samples through solvent extraction, bleaching, and alkaline hydrolysis, as described by Green (1963). Oxygen isotope analysis was performed on


COz gas produced from purified cellulose using the nickel pyrolysis technique (Thompson and Gray, 1977). Isotopic analysis of non-exchangeable carbon-bound hydrogen in the cellulose was carried out on H2 gas obtained from zinc reduction of H20 produced by combustion of nitrated cellulose in an oxygen atmosphere (DeNiro, 1981; Coleman et al., 1982). Oxygen (180/160) and hydrogen (/H/1H) isotope ratios are expressed as 6180 and 6ZH values, representing deviations in per mil (%o) from the SMOW standard ((~sample = Rsample/Rs•ow-1) × 103, where R= 180/160 or 2H/1H; see Fritz and Fontes, 1981). Cellulose 6180 and 62H values cited here have analytical uncertainties of + 0.3%o and + 6%o, respectively. The model used to interpret the isotopic data couples independent empirical relationships between (1) cellulose 6180 and environmental water 6180, (2) cellulose 62H and environmental water 62H, (3) and photosynthetic humidity (h) and the global Meteoric Water Line (MWL), the well known linear equation relating 6180 and 62H in precipitation (Craig, 1961). The model assumes fixed fractionation factors for biochemical effects during cellulose synthesis, and for equilibrium exchange and kinetic effects during evapotranspiration (see Fig. 8 caption).

Results The isotopic data for fossil and modern wood samples from the Castle Peak cirque indicate that the range of conditions near present timberline is essentially the same as that in the Castle Peak cirque during the early Holocene (Table 2), in spite of the differences in timberline elevation. Measured cellulose 6180 and 62H values from fossil and modern samples plot within a common field (Fig. 7), as do inferred values for water 61BO and h (Fig. 8). Data from the six modern tree samples roughly define the ranges of modern variability in 6180 and h with elevation in the Castle Peak cirque (Figs. 9 and 10). Most of the fossil data points also fall within these ranges, with the notable exceptions of samples CP9 and CP23.



TABLE 2 Isotopic data for wood samples from the Castle Peak cirque Sample








6180 *

62H *

6*80 *



CP51 C P62 CP91 CPI5' CPI6 CP22 CP23 CP26 CP27 CP28 CP29

3 3 5 3 3 I 5 7 7 7 7

2091 2091 2192 2091 2091 2089 2192 2085 2085 2085 2085

23.0 25.1 21.1 22.9 25.7 25.4 23.0 21.0 20.7 25.7 21.5


164 162 176 155 153 169 150 165 159 162 156

- 23.3 - 23.9 -24.6 - 21.6 - 22.5 - 25.3 - 20.7 - 22.5 - 21.3 - 24.2 -21.1

0.33 0.24 0.35 0.39 0.27 0.20 0.42 0.42 0.48 0.21 0.45


CP31 CP4A 2 CP4B' CP81 CP241 CP251

8 2 2 4 4 9

2082 2058 2058 2137 2137 2087

23.0 25. I 25.2 21.9 22.8 21.4


172 162 164 151 160 158

- 24.8 - 23.9 - 24.3 - 20.4 -22.5 - 21.4

0.28 0.24 0.23 0.47 0.37 0.45

*°/oo S M O W . I Abies sp. 2Pinus sp. (white pine).






• modern 1 fossil -160

Q; o

• .-




-r e~ tN

- 170















5180cellulose per mil SMOW

Fig. 7. P l o t o f 6 1 8 0 vs. 6 2 H f o r cellulose f r o m fossil a n d m o d e r n w o o d s a m p l e s f r o m the C a s t l e P e a k c i r q u e .
















-7 -w 0.4,'--

r"w z 0 0 -r r~


i {9.2-!


uncertainty I














5180wate r per mil

SMOW Fig. 8. Plot of inferred water 61sO vs. h for the Castle Peak samples. Note that the strength of the positive correlation between 6180 and h is exaggerated somewhat by the skewcd analytical uncertainties. The model used to derive these estimates involves the solution of three equations: (1) 61sO,.= A 6~SO..... + 1000(A- 1) (2) 62H,. = B f i 2 H . .... + 1000(B- 1) (3) 6~SOw,,,= 8 62H~.... + 10: where A = ~, 0y 0q- :~,(:~~k- l)h for ~,= 1.0264, ~ = 1.0095, zq = 1.0210 and B = /~,/3,.,6'k- [~',(/3,/Jk- l)h for ft,= 0.9520, [¢~= 1.0797,/Tk= 1.0185. The terms a and fl represent isotopic fractionations associated with cellulose synthesis (~, and ~,), and with equilibrium (~, and/7~) and kinetic (~ and ~k) effects during evapotranspiration. See Edwards and Fritz (1986) and Buhay et al. (1991) for a detailed discussion of the model formulation and modifications.

Discussion T h e fossil w o o d , r a d i o c a r b o n ages, and pollen d a t a from Castle Peak show that timberline was higher than present from before 9.1 ka until at least 8.1 ka. This suggests t h a t climate at times d u r i n g the early H o l o c e n e was w a r m e r t h a n t o d a y , which is consistent with theoretical p a l e o c l i m a t i c r e c o n s t r u c t i o n s based on M i l a n k o v i t c h o r b i t a l m o d e l s (Berger, 1978; K u t z b a c h , 1981; K u t z b a c h and G u e t t e r , 1984). T h e a l t e r n a t i v e e x p l a n a t i o n , that the s t u d y a r e a has been uplifted since 8 ka, is unlikely because there is no evidence for m i d d l e o r late H o l o c e n e tectonic o r glacio-isostatic uplift o f the m a g n i t u d e required to a c c o u n t for the timberline shift at Castle Peak ( C l a g u e et al., 1982). T h e height o f the s o u t h e r n C o a s t M o u n t a i n s a b o v e sea level d u r i n g the early H o l o c e n e was actually greater than t o d a y , becausc relative sea level a l o n g

the Pacific coast, at that time, was lower (Clague et al., 1982). A m i n i m u m vertical shift in timberline o f 60 m is indicated by differences in elevations o f fossil a n d living fir (the c o m p a r a b l e value for w h i t e b a r k pine is 130 m). These values are m i n i m a because the fossil w o o d p r o b a b l y has m o v e d d o w n s l o p e from g r o w t h sites a n d because the sites themselves m a y have been below timberline. If we a s s u m e an a d i a b a t i c lapse rate o f 6 . S / k m ( B a r r y a n d Chorley, 1968) a n d a m i n i m u m shift in timberline o f 60 m, m e a n s u m m e r t e m p e r a t u r e between 9.1 ka a n d 8.1 ka must have been at least 0.4~C w a r m e r than at present; a timberline shift o f 130 m implies a t e m p e r a t u r e difference o f 0.85~C. T h e p a l y n o l o g i c a l d a t a p r o v i d e a d d i t i o n a l .¢idence for w a r m e r c o n d i t i o n s d u r i n g the ~ ;y Holocene. The presence o f fossil pollen o~ ;)hepherdia canadensis, Elaeagnus commutata, and Gera-



•modern •fossil

[email protected]



[email protected]

tad @


[email protected]@




[email protected] m analytical















[email protected]

6180water p e r mi 1 SNOW Fig. 9. Inferred water 61"O vs. elevation for the Castle Peak samples. The slope of the modern range is based on linear regression of the living tree data.






•modern • fossil

[email protected]@

~L [email protected] ~E


[email protected] 1-




l @.4






Fig. I0. l n ~ r ~ d photosynthetic humidity (h) vs. elevation ~ r the Castle Peak samples. The s l o ~ of the modern range is b~ed on linear reg~ssion of the living tree data.


nium cf. G. richardsonii is significant, since none of these species is a component of alpine plant communities. Drier conditions during zone CP-I are suggested by the abundance of Shepherdia, which grows today in the study region on dry south-facing subalpine slopes in association with other shrubs and grasses (Selby and Pitt, 1984). Ogilvie (1990) also emphasizes the preference of Shepherdia for dry habitats and its common association with Pinus albicaulis in upper subalpine environments. The co-occurrence of Shepherdia and Elaeagnus pollen in zone CP-I is not unexpected, since both species are pioneers on mineral soils, and both have been reported from late Pleistocene-early Holocene sediments in southern and central British Columbia (Souther et al., 1987; Mathewes and King, 1989). However, Elaeagnus, unlike Shepherdia, has not been reported as occurring in the study area today (Selby and Pitt, 1984) or, for that matter, in the subalpine zone anywhere in British Columbia (Taylor and MacBryde, 1977); its fossils thus may represent a significant elevational range extension during the time of deposition of zone CP-I (<9.1-8.4 ka). Geranium richardsonii is an herb of montane to subalpine meadows, but, like all Geranium species, is not known to occur in the alpine zone (Taylor and MacBryde, 1977). All three of these indicator species are poor pollen producers and are insectpollinated, strongly suggesting that the plants were present in the Castle Peak cirque when their pollen was deposited in the sediments. Further support for higher early Holocene timberline at Castle Peak is provided by a comparison of tree pollen abundances in the three modern moss samples and the fossil peat (Table 3). Of particular interest are the relative abundances of TABLE 3 Comparison of Abies and Pinus albicaulis type pollen percentages in Castle Peak moss polsters and fossil peat (20-55 cm)


Moss polsters ( n = 3) Fossil peat (n = 8)

Pinus albicaulis





2 5

0.5-4 2.5-7

14.5 21

12--17 16-36


Abies and Pinus albicaulis type pollen, since fir and white pine are the only conifers recorded in the fossil wood assemblage. Both mean and maximum percentages of fir and white pine pollen in the fossil peat are substantially higher than those in the modern moss samples, suggesting greater abundance or reproductive success in the early Holocene, These data collectively support the contention that the summer growing season at Castle Peak between 9.1 ka and 8.1 ka was significantly warmer and probably longer than today, with upslope migration of subalpine plants, and substantially better growth and reproduction of trees. The presence of Shepherdia canadensis and Elaeagnus commutata pollen in zone CP-I also suggests that relatively dry conditions prevailed during that period. This was followed by increased local wetness during zone CP-2, as indicated by a decline in Shepherdia canadensis pollen and increases in Cyperaceae and herb pollen. Whether this change is only locally significant or whether it signals a trend to regionally wetter and cooler conditions is a subject for further investigation. The weak positive correlation between inferred water 6180 and h (photosynthetic humidity), which is evident in Fig. 8, is attributed to a combination of altitudinal effects on relative humidity and seasonal variation in the isotopic composition of precipitation. Although local topography and aspect can induce considerable variability, trees living at higher elevations, and thus closer to average cloud-base, should generally experience higher humidity during the growing season than those at lower elevations. In addition, because groundwater residence times at higher elevations are likely to be much shorter than at lower elevations, high-elevation trees can be expected to obtain more of their moisture from isotopically heavy summer rains, compared to trees at lower elevations, where isotopically light snowmelt would tend to constitute a larger proportion of local groundwater during the growing season. As mentioned previously, two fossil wood samples (CP9 and CP23) lie above the range of modern values on plots of 6180 vs. elevation and h vs. elevation (Figs. 9 and 10). These samples came from trees that evidently received larger


proportions of isotopically light snowmelt and experienced considerably lower photosynthetic humidity than would be expected at the same elevation under modern conditions, which is consistent with their location above modern timberline. Dryness at lower elevations during the early Holocene may have been enhanced by a reduction in the amount of moisture intercepted by the cirque, owing to higher cloud-base. Extrapolation of the modern h vs. elevation range shown in Fig. 10 suggests that the present, mean, daytime cloud-base during the growing season lies slightly below the col (elevation 2285 m) forming the east rim of the cirque. Displacement of the h vs. elevation range upslope enough to allow the establishment of CP9 and CP23 at 2192 m would shift mean cloud-base well above 2285 m, considerably reducing the amount of summer precipitation in the cirque.

Conclusions The presence of 9.1-8. I ka wood above timberline in the Castle Peak cirque and palynological data from a site on the cirque floor indicate that climate was warmer and perhaps drier during the early Holocene than today. This conclusion is in agreement with early Holocene paleoclimatic inferences from other sites in the Canadian Cordillera and western United States. Oxygen and hydrogen isotope data provide additional information about microclimatic conditions in the Castle Peak cirque during the early Holocene. Reconstructed water ~taO and photosynthetic humidity values for fossil wood samples fall within the same ranges as living trees in the same area. Increasing photosynthetic humidity with altitude indicated by the data from living frees suggests that conditions in the cirque at times of elevated timberline were drier than those prevailing today.

Acknowledgements Radiocarbon ages were provided by R.P. Beukens (lsoTrace Laboratory, University of Toronto), R.N. McNeely (Geological Survey of


Canada), and E. Nelson (Simon Fraser University). The fossil wood was identified by J. Gonzalez (Forintek Canada Corp.) and H. J6tte (Geological Survey of Canada). Isotope analyses were performed by the staff of the Environmental Isotope Laboratory, University of Waterloo. The work was supported by the Geological Survey of Canada, the Natural Sciences and Engineering Research Council of Canada (operating grants to RWM and TWDE), and the Waterloo Centre for Groundwater Research. Critical reviews by J.V. Matthews Jr. and two anonymous journal referees improved the paper. This is Geological Survey of Canada Contribution no.47091.

References Anderson, T.W., Mathewcs, R.W. and Schweger, C.E., 1989. Holocene climatic trends in Canada with special reference to the Hypsithermal interval. In: R.J. Fulton (Editor), Quaternary Geology of Canada and Greenland (Geol. Can., 1). Geol. Surv. Can., pp. 520-528. Baker, R.G., 1983. Holocene vegetational history of the western United States. In: H.E. Wright Jr. (Editor), Late-Quaternary Environments of the United States. 2. The Holocene. Univ. Minnesota Press, Minneapolis, MN, pp. 109-127. Barnosky, C.W., Anderson, P.M. and Bartlein, P.J., 1987. The northwestern U.S. during deglaciation; vegetational history and paleoclimatic implications. In: W.F. Ruddiman and H.E. Wright, Jr. (Editors), North America and Adjacent Oceans during the Last Deglaciation (Geol. North America, K-3). Geol. Soc. Am., Boulder, CO, pp. 289-321. Barry, R.G. and Chorley, R.J., 1968. Atmosphere, Weather, and Climate. Methuen, London, 319 pp. Berger, A., 1978. Long-term variations of caloric insolation resulting from the earth's orbital elements. Quat. Res., 9: 139-167. Buhay, W.M., Edwards, T.W.D. and Aravena, R., 1991. Finetuning the cellulose model for dendroclimatological interpretation. In: Proc. Int. Symp. Use of Isotope Techniques in Water Resources Development. IAEA-SM-319/10P, pp. 122-123. Burk, R.L. and Stuiver, M., 1981. Oxygen isotope ratios in trees reflect mean annual temperature and humidity. Science, 211: 1417-1419. Carrara, P.A., Trimble, D.A. and Rubin, M., 1991. Holocene treeline fluctuations in the northern San Juan Mountains, Colorado, U.S.A., as indicated by radiocarbon-dated conifer wood. Arct. Alp. Res., 23: 233-246. Clague, J., Harper, J.R., Hebda, R.J. and HQwes, D.E., 1982. Late Quaternary sea le'vels and crustal movements, coastal British Columbia. Can. J. Earth Sci., i9: 597-618. Clague, J.J. and Mathewes, R.W., 1989. Early Holocene thermal maximum in western North America: new evidence from Castle Peak, British Columbia. Geology, 17: 277-280. Coleman, M.L., Shepherd, T.J., Durham, J.E., Rouse, J.E. and

EARLY HOLOCENECLIMATEAT CASTLEPEAK, SOUTIIERNCOASTMOUNI'AINS,BRITISHCOLt,IMBIA Moore, G.R., 1982. Reduction of water with zinc for hydrogen isotope analysis. Analyt. Chem., 54: 993-995. Craig, H., 1961. Isotopic variations in meteoric waters. Science, 133: 1833-1834. Cwynar, L.C., Burden, E. and McAndrews, J.tl., 1979. An inexpensive sieving method for concentrating pollen and spores from fine-grained sediments. Can. J. Earth Sci., 16: 1115-1120. DeNiro, M,J., 1981. The effects of different methods of preparing cellulose nitrate on the determination of the D/El ratios of non-exchangeable hydrogen of cellulose. Earth Planet. Sci. Lett., 54: 177-185. DeNiro, M.J. and Cooper, L.W., 1990. Water is lost from leaves and trunks by fundamentally different mechanisms, Geochim. Cosmochim. Acta, 54: 1845-1846. Edwards, T.W.D., 1990. New contributions to isotope dendroclimatology from studies of plants. Geochim. Cosmochim Acta, 54: 1843-1844. Edwards, T.W.D. and Fritz, P., 1986. Assessing metcoric watercomposition and relative humidity from asO and 211 in wood cellulose: Paleoclimatic implications for southern Ontario, Canada. Appl. Geochem., 1: 715-723. Edwards, T.W.D., Aravena, R.O., Fritz, P. and Morgan, A.V., 1985. Interpreting paleoclimatc from '~O and 2H in plant cellulose: Comparison with cvidencc from fossil insects and relict permafrost in southwestern Ontario. Can. J. Earth Sci., 22:1720-1726. Faegri, K. and Iversen, J., 1975. Textbook of Pollen Analysis. Munksgaard, Copenhagen, 295 pp. Fritz, P. and Fontes, J.C., 1981. Introduction. In: P. Fritz and Fontes, J.C. (Editors), Handbook of Environmental Isotope Geochemistry. Elsevier, Amsterdam, pp. 1-19. Green, J.W., 1963. Wood cellulose. In: R.L. Whistler (Editor), Methods in Carbohydrate Chemistry. Academic Press, New York, NY, 3, pp. 9-20. Kearney, M.S. and Luckman, B.H., 1983a. Holocene timberline fluctuations in Jasper National Park, Alberta. Science, 221: 261-263. Kearney, M.S. and Luckman, B.H., 1983b. Postglacial vegetational history of Tonquin Pass, British Columbia. Can. J. Earth Sci., 20: 776-786. Kutzbach, J.E., 1981. Monsoon climate of the early Holocene: Climate experiment with the earth's orbital parameters for 9000 years ago. Science, 214: 59-61. Kutzbach, J.E. and Guetter, P.J., 1984. The sensitivity of monsoon climates to orbital parameter changes for 9000 yrs


BP: Experiments with the NCAR general circulation model. In: J. Imbrie and A. Bergers (Editors), Milankovitch and Climate Change. Reidel, Dordrecht, pp. 801-820. Luckman, B.H. and Kearney, M.S., 1986. Reconstruction of Holocene changes in alpine vegetation and climate in the Maligne Range, Jasper National Park, Alberta. Quat. Res., 26: 244-261. Luckman, BH., Hamilton, J.P., Jozsa, L.A. and Gray, J., 1985. Proxy climatic data from tree rings at Lake I,ouise. Alberta: a preliminary report. Geogr. Phys. Quat., 39: 127-140. Mathev, es. R.W., 1985. Paleobotanical evidence for climatic change in southern British Columbia during late-glacial and Holocene time. In: C.R. Harington (Editor), Climatic Change in Canada 5, Critical Periods in the Quaternary Climatic History of Northern North America. Natl. Mus. Can., Natl. Mus. Nat. Sci. Syllogeus Ser., 55: 397-422. Mathewes, R.W. and King, M., 1989. Holocene vegetation. climate, and lake-level changes in the Interior Douglas-fir biogeoclimatic zone. British Columbia. Can. J. Earth Sci., 26:1811 1825. Ogilvie, R.T., 1990. Distribution and ecology of whitebark pine in western Canada. In: Proc. Symp. on Whitebark Pine Ecosystems: Ecology and Management of a Ftigh-Mountain Resource. U.S. Dep. Agric., Forest Serv., Intermountain Res. Stn.. Gen. Tech. Rep., INT-.270, pp. 54-60. Ramesh, R., Bhattacharya, S.K. and Gopalan, K., 1986. Climaticcorrelations in the stable isotope records of silver fir (Abies pimlrow) trees from Kashmir, India. Earth Planet. Sci. Lett., 79:66 74. Selby, C.J. and Pitt, M.D., 1984. Classification and distribution of alpine and subalpine vegctation in the Chilcotin Mountains of southern British Columbia. Syesis, 17:13 41. Souther, J.G., Clague, J.J. and Mathewes, R.W., 1987. Nazko Cone: a Quaternary volcano in the eastern Anahim Bclt. Can. J. Earth Sci., 24: 2477--2485. Taylor, R.L. and MacBryde, B., 1977. Vascular Plants of British Columbia, a Descriptive Resource Inventory. Univ. British Columbia Press, Vancouver, B.C., 754 pp. Thompson. P. and Gray, J., 1977. Determination of '"O,,'~'O ratios in compounds containing C, H, and O. Int. J. Appl. Radiation Isotopes, 28:411-415. Yapp, C.J. and Epstein, S., 1982. Climatic significance of the hydrogen isotope ratios in tree cellulose. Nature, 297: 636-639.