Dendroglaciological reconstruction of late Holocene glacier activity at Todd Glacier, Boundary Range, northwestern British Columbia Coast Mountains
Sarah Coulter Laxton B.A., University of Victoria, 2003 A Thesis Submitted in Partial Fulfillment of the
Requirement dbr the Degree of MASTER OF SCIENCE in the Department of Geography
Sarah Coulter Laxton, 2005 University of Victoria O
All rights reserved. This thesis may not be reproduced in whole or in part, by photocopy or other means, without the permission of the author.
Supervisor: Dr. D.J. Smith
Abstract
The purpose of this thesis was to apply a combined approach of dendroglaciollogy and stratigraphic analysis to reconstruct glacial fluctuations at a remote location in the Boundary Ranges of the northwestern Coast Mountains, British Columbia. The approach taken enables: 1) comparison of glacial fluctuations in Todd Valley with respect to others in the southern Boundary Ranges; 2) placement of glacier behavlour in the Boundary Ranges to be understood in the context of late Holocene climate oscillations and glacier fluctuations noted elsewhere in the northwestern Cordillera.
Subfossilised wood was sampled from eight sites in Todd Valley providing a thorough spatial representation of the area. Dendrochonologic dating resulted in the identification of three phases of late Holocene advance including the: Sage (BC 420- 200), Bridge (AD 405-640), and LIA (AD 1400- 1630, 1640- 11 950, 1250- 1420, 1200- 1390) Advances. This thesis provides a basis for more in-depth studies of localized glacier activity in the Boundary Ranges,
Table of Contents
Title Page
* 0
Abstract
. . . .
. . .
. .
.
. . . .
. . .
. . .
. . . . .
.
.
. . .
.
.
.
.
. . .
.
. . .
. .
. . . .
. . . .
. . .
.
l a* , e Table of Contents
...
.
.
.
.
.
.
...
.
.
.
.
...
laaList of Tables . . . ~ . ~ . ~ ~ . ~ ~ ~ o ~ ~ ~ o ~ o ~ ~ ~ . . ~ ~ ~ ~ . ~ . ~ . . ~ ~ o ~ ~ . ~ . o ~ ~ ~ ~ ~ ~ ~ . ~ ~ ~ ~ . ~ . ~ . o ~ ~ o . . . ~ ~ ~ . . ~ . ~ . . ~ ~ iv List of Figures
.
.
...
.
.
.
.
.
.
.
.
.
.
*
.
v Acknowledgements ...~...~e...o~..o...,~...o~~~.~e~s.~~.~~..~.~~~.~~~~~.~..,..~~... ix 1.0 Introduction... ...
.... . .
...
.
. .. . .
1 1.1 Research Context...
.
.
...~~~..~..~..s...o~~~~~m~.~.~~..~ 2 1.2 Format of Thesis...
...
....
..........e....e.eessa-...
..
e e o ~ o e . e o . s o s m e m .*" **.*...- 52.0 Bendroglaciological reconstruction of late Holocene glacier activity at Todd Glacier, Boundary Range, northwestern British Columbia Coast Mountains ...s.oo-. 6
2.1 Introduction
...
. . .
.
.
.
.
. . . .
..
.
. . .
.
. -.
a *..
".
.
a *. *..
*..
. -. -.
a *. , *..
. -.
".
+ +-
*6 2.2 Research Background...
..
...
...
.
.
...
..
....
.
2.3 Study Site...
11 2.4 Research Methods. . .
.
. .
.
.
.
.
.
.
. . .
.
. . . .
.
.
.
.
. . . .
.
. .. .
. .
.
.
.
.
.
.
14 2.5 Results ~o.~.a...o................................... 16 2.6 Discussion... ...
...
.
.
...
.
.
.
.
.
.
. e . e e o..
s * s . . s . ~ . . . ~ . . o o s e e o .as..a 26 2.7 Conclusion.... .
.
. . .
. .
. . .
.
. . .
. .
.
. . .
.
. . .
. ..
3 1 3.0 Summary....
... ...
. ~ . . ~ . ~ . . . . e . e s . . . ~ s . . o o ~ o e ...
*..-...
...
...
e...... ...
-,.32 Literature Cited. .
. . .
. . . . .. .
. . .
. . .
.
. .
..
. . .
. . . .
.
.
. . .
..
. .
. . . .. .
.
. . .
.. . .
. . .
. .
.
. .
. . .
,. . .
.3 5List of Tables
...
Table 1 . Annual retreat rates of glaciers and corresponding aerial photographs -69...
Table 2 . Radiocarbon ages BP, Todd Valley 70
Table 3 - Wood sampled at Site 1 during 2004 field season
...
71...
Table 4 - Wood sampled at Site 2 during 2004 field season 7%
...
Table 5 - Wood sampled at Site 3 during 2004 field season
.
.
...
2
...
Table 6 - Wood sampled at Site 4 during 2004 field season 72
Table 7 - Wood sampled at Site 5 during 2004 field season
.a.o...
73Table 8 - Wood sampled at Site 6 during 2004 field season ...,..e..e... 74
Table 9 - Wood sampled at Site 7 during 2004 field season
...
74...
List of Figures
...
...*...*.
Figure 1 - Map of Todd Valley area
....
44Figure 2 - Oblique aerial photograph of the Todd Glacier forefield. Todd Glacier located
in the background. Boulders are located at the head of the majority ~f the flutes (SAllen 2003). ...ee.e.
...
45 Figure 3 - Looking south towards Todd Glacier, In the background, the large maroonboulder on the left side of the photo is the starting point o f a
-
200 m long north-south..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
trending flute (S. Laxton 2004)
Figure 4 - The west face of the boulder at the head of a flute shown in Figure 3. Ice axe
and person for scale, The terminus of Todd Glacier and Site 2 are in the background (S. Laxton 2004)
...
....
...
...
...
46...
Figure 5 - Location of all eight study sites in Todd Valley (30BCC97 143 No.37) 46 Figure 6 - Sample TG04-878 displaying rings that have undergone postglacial
deformation. The bole has been compressed along the left and right sides of the
photograph resulting in rings that have been elongated and distorted
...
47 Figure 7 - Boundaries of the Site 1 study area marked by the yellow rectangle(30BCC97143 No.38)
...
47 Figure 8 - Looking westward towards Bug Glacier. The LIA trim line is visible on theupper right side of the photograph (S. Allen 2003)
...
48 Figure 9 - Photograph of the northeast facing aspect of Site 1 with Bug Glacier located in the background. The arrow is pointing to the sampling location of TG04-802 situated on an outermost LIA lateral moraine (S. Laxton 2004)...
58 Figure 10 - TG04-802, located underneath the white field notebook, after being sampled with a chainsaw. The stump was found wedged under a 1.5 m diameter boulder on the crest of the outermostBug Glacier LIA lateral moraine (Fig.9) (D. Smith 2004)
...
49 Figure 1 1 - Bole protruding from morainal debris located-
150 m from the north side ofthe terminus of Bug Glacier. The long axis of the bole is oriented parallel to the direction of glacier flow (S. Laxton 2004) ... 49 Figure 12
-
Boundaries of the Site 2 study area demarked by the yellow rectangleFigure 13 - Photograph taken from Todd Valley looking east towards Site 2, outlined by
...
the yellow box (S. Laxton 2004) 0 0
Figure 14 - Four upright boles protected from destruction by mass wasting and avalanche
by the adjacent bedrock outcrop. South facing samples TG04-864-867 were
photographed prior to chainsaw sampling. The boles were
-
100 yrs 018 prior to beingkilled during a LIA advance of Sage Glacier (19. Smith 2004) 1
Figure 15 - Samples in Figure 14 after being sampled with chain saw, Duct tape is
wrapped around the sample prior to the first cut to keep the wood intact during cutting. The second cut is made at the bottom border of the tape indicated by the yellow arrow (D. Smith 2004)
...
5 1 Figure 16 - Photograph of sample TGO4-871 at Site 2 prior to chainsaw sampling, Thebole was found protruding from the massive matrix supported sandy glacial diamict plastered to the east wall of Todd Valley
-
250 m north of the 2004 terminus of Sage...
Glacier (D. Smith) 52
Figure 17 - Boundaries of the Site 3 study area outlined by the yellow rectangle
(30BCC97143 No.38) ... 52
Figure 18 - The east-facing wall of Todd Valley and Site 8 are in the background of the photograph. Person using a chainsaw to sample a detrital bold found on the bottom of the glaciofluvial channel at Site 3. The cut banks are
-
9 m from the base of the incision to the top (S. Laxton 2004)...
53 Figure 19 - View towards the western wall of Todd Valley from the bottom of theglaciofluvial cut bank at Site 3. The southern side of the cut bank on the right side of the photograph is
-
9.5 m high. Field assistant is preparing sample TG04-838 for chainsaw...
sampling (S. Laxton 2004) .53
Figure 20 - Close up of needles and twigs found in the same horizon as TG04-838 (Fig.
...
19) (S. Laxton 2004) 54
Figure 21 -Boundaries of the Site 4 study are outlined by the yellow rectangle
(30BCC97143 No.38)
...
54 Figure 22 - View northwest towards the cut bank and exhumed paleoforest horizon foundat Site 4. The double-headed green arrow follows the paleoforest horizon. The yellow
...
arrow points towards the Bug Glacier LIA trim line at Site 1 (K. Penrose 2004) 55 Figure 23 - Photograph taken east of the Site 4 stratigraphic section looking towards the southwest. The section is located along the yellow measuring tape in the photograph. Wood fragments are visible around the back leg and hat of the person. The yellow arrow
...
vii
Figure 24 - View south towards the terminus of Todd Glacier taken from the bed of the glaciofluvial channel that runs through Site 4. The stratigraphic section is located 10 m northeast of where this photograph was taken. The black arrow in the upper right of the photograph points toward Site 7 (S. Laxton 2004)
... .
.
.
.
.
.
...
56 Figure 25 - View of the south facing stratigraphic section at Site 4, located on thenorthern bank of the glaciofluvial melt water channel draining Todd Glacier (Fig. 22). Unit 1 is indicated by the green double-headed arrow; unit 2, yellow; and unit 3, blue (K.
...
Penrose 2004) 57
Figure 26
-
Southeast exposure of glacial diamict located in Unit 1 of the stratigraphic section at Site 4 (S. Laxton 2004)....
.
.
...
58...
Figure 27
-
Close up of needle and twig macrofossils in Unit 2 (S, Laxton 2004) 58 Figure 28 - Close up of the diamict composing Unit 3 at the Site 4 stratigraphic section. Note that there is very little difference in facies between Units 1 (Fig.26) and 3 (S.Laxton 2004)
...
- 3 9 Figure 29-
Boundaries of the Site 5 study area demarcated by the yellow rectangle...
(30BCC97143 No.38) 59
Figure 30
-
Looking west towards Site 5 from the bottom of Todd Valley. The woody debris and boles are protected from avalanche and mass wasting events by the large bedrock outcrop (S. Allen 2003)...
-60 Figure 3 1-
View of detrital wood and boles protected from the bedrock outcrop. An arrow is pointing towards a bole (TG04-88 1) that was sampled (S. Laxton 2004)...
60 Figure 32-
Photograph of an in situ standing bole upslope of Site 5 on the northern wall of Two Glacial Valley (K. Penrose 2004) ... .6 1 Figure 33-
Boundaries of the Site 6 study area outlined by the yellow rectangle(30BCC97143 No.38)
...
62 Figure 34-
Small sand deposit located at the eastern limit of Site 6. The sand isdeposited by strong katabatic winds flowing off Todd Glacier (S. Allen 2003)
...
62 Figure 35-
Looking westward from the edge of the melt water channels that drain the Two Glaciers. Boles were discovered in this area of Site 6 protruding from the steep sided glacial outwash and melt out sediments. The yellow arrow points towards a bole protruding from the sediment (S. Laxton 2004)...
63Figure 36
-
Boundaries of the Site 7 study area outlined by the yellow rectangle...
...*...*...
(30BCC97143 No.38)
.
.
63Figure 37
-
View from a helicopter looking south towards Site 7. The arrow points towards a gully that has acted as a conduit for melt water draining off the southwest...
section of Todd Glacier (S. Laxton 2004)
...
...
64 Figure 38-
East face of the gully at Site 7. The arrow points to boles and detrital wood buried by an advance of Todd Glacier. Todd Glacier, Sage Glacier, and Site 2 are in the background (D. Smith 2004)...
64 Figure 39-
View towards the west-facing surface of the gully located at Site 7, The green arrow points to TG04-$73 and the yellow arrow indicates the trim line on the east wall of Todd Valley (D. Smith 2004)...
-65 Figure 40-
Photograph taken on the southern site of the bedrock promontory at Site 7looking south. Sample TG04-874 is located under the chainsaw in the foreground. The
...
source area of Todd Glacier is in the background (D. Smith 2004) 65 Figure 41-
Glacial advances recorded in the Boundary Ranges of the northwestern Coast Mountains (sources: Clague et al. 1992; Haspel et al. 2005; Jackson and Smith 2005;...
Lewis and Smith 2005)
...
.
.
66Figure 42 - Glacial advances recorded in the Boundary Ranges of the northwestern Coast
Mountains (sources: Clague et al. 1992; Haspel et al. 2005; Jackson and Smith 2005; Lewis and Smith 2005)
...
..67Acknowledgements
The process of writing this thesis has been a great adventure,
H
would like to thank my supervisor Dr. Dan Smith for providing me the opportunity to study amidst the spectacular mountains and glaciers of the Coast and Canadian Rocky Mountains. Thanks to Drs. Walker and Buffus for helping me meet deadlines at really short notice,Field and lab work was greatly supplemented by the presence of my UVTRL lab mates: Lex, Karen, Rochelle, Sandy, Kelly, Krista, Scott, Dave, and Muddy. Special thanks go out to: Rochelle for her sage advice and statistics prowess; Scott provided excuses to check the daily snow reports, Kelly who loved to practice her teenage runaway skills, shared in my passion for Nutella Surprise, and danced into the night at Kajo's bar in Stewart; Muddy, best field dog in the universe; and Sandy my dendroglaciological confident, and skiing pal.
JoAnne Nelson first introduced me to the mountains and glaciers in the Stewart area and the joys of field research. I am very appreciative of the time I spent learning from her in the mountains of northern B.C.
My family and grandparents kept me fed and happy. The Hill Family kindly provided accommodation in Terrace during the field season of '03. I would also like to recognize all of my friends who kept me sane, especially Keira, Megan, and Jen. Songs #10 and #6 provided the soundtrack for my two years in CorB235A.
1.0 Introduction
Changing weather patterns and their effects upon human infrastructure and natural resources have prompted a variety of climate change investigations (Evans and Clague 1994; Ftyder 1998; Bell 2003)- Many of these queries have focused on understanding climate change and the consequent effects on human populations (Ryder 1998; Isarin and Renssen 1999; Luckrnan 2000). To construct climate models that better predict future climate behaviour, high resolution proxy data from pre-twentieth century climate
conditions are essential (Bradley and Jones 1993; Bradley et al. 2003). Prior and ongoing research focused on elucidating long-term glacier-climate response relationships suggests that this avenue of research has the potential to provide significant insights into past climates (Evans and Clague 1994).
In the Coast Mountains of British Columbia, numerous researchers have
completed research at a variety of locations in an effort to reconstruct glacial oscillations during the Holocene (last 10,000 years) and late Pleistocene (1.8 ma.-10,000 years ago) (e.g., Luckman and Villalba 2001). The majority of this research has, however, been concentrated in the Coast Mountains south of Bella Coola and north of the Yukon-Alaska border. Far less effort has been directed to describing Holocene glacial activity in the northern Coast Mountains, despite the promise of significant insights demonstrated by Clague and Mathewes (1 996).
The purpose of this thesis is to provide a comprehensive reconstruction of late Holocene glacier activity at one site in the Boundary Ranges of the northwestern Coast Mountains. The findings of this research will reduce the spatial gap in glacial
reconstructions for the northern Coast Mountains, thus facilitating a broader understanding of late Holocene glacier behavior in Pacific North America.
1.1 Research Context
This section provides information pertaining to the general geography, bedrock geology, and Holocene glacial history of the study area. The geography section will cover the location of the Boundary Ranges, landscape, flora, and human settlement in the region. Limited by the availability of literature on the subject, the bedrock geology is briefly highlighted. The last subject covered in this section is the Holocene history of the study area, including major glacial fluctuations, and previous findings throughout the Coast Mountains.
Geography
Located at the northern reaches of the Coast Mountains, the Boundary Ranges extend northwestward from the Nass River to the British Columbia - Alaska border
(Fig.1). Extensive glaciation in the area during the Wisconsinan (- 25,000-10,000 yrs BP) resulted in the development of hanging valleys, steep-sided valley walls, and ice- carved trunk valleys (Hickin et al. 2001). The present day flora of the area from 400-
1000 m as1 belongs to the Mountain Hemlock biogeoclimatic zone (Meidinger and Pojar 1991).
The small adjacent towns of Stewart, British Columbia, and Hyder, Alaska, situated at the head of the Portland Canal are the main population centres of the Boundary Ranges. The population of both towns has fluctuated over the last hundred years from over 10,000 people in Stewart during boom times to the current populations of
-
700 in Stewart and-
200 in Hyder. Since 1898, mineral prospectors have been lured tothe area by the possibility of striking it rich. Logging, mineral exploration, and tourism are the main economic drivers of the two communities.
The geology of the southern area of the Boundary Ranges is primarily composed of the Mesozoic Bowser and Hazelton Assemblages. The Coast Crystalline Belt lies adjacent to and is included in part of the eastern edge of the Boundary Range (Grove 197 1). In the Stewart area, the bedrock consists primarily of Lower Jurassic Unuk River Formation rock, which is part of the Hazelton Group, a member of the Stewart Complex. The Unuk Formation consists of brick red to apple green (including grays, purples and maroons) epiclastic volcanic rocks and lithic tuffs, with closely associated pillow lavas, siltstones, and carbonate lenses (Grove 1986).
Holocene Glacial Activity
Four roughly synchronous primary phases of Holocene glacier expansion have been noted throughout the central and southern Coast Mountains. These include the Garibaldi (Ryder and Thomson 1986), Tiedemann (Ryder and Thomson 1986), Bridge (Allen and Smith 2003), and Little Ice Age (LIA) (e.g., Smith and Desloges 2000; Larocque and Smith 2003) Advances:
The Garibaldi phase occurred approximately BC 3920-4471 (Smith 2003).
Evidence of the Garibaldi phase has been identified at Fyles Glacier in the Central Coast Mountains (Laxton et al. 2003), in Garibaldi Provincial Park (Koch et al. 2004a), in Tchaikazan Valley (Smith 2003) and in the southern Coast Mountains (Ryder and Thomson 1986).
First identified at Tiedemann and Gilbert Glaciers, the Tiedemann Advance occurred sometime between 1489 BC-235 AD (Ryder and Thomson 1986).
Evidence for this mid-Neoglacial advance was identified by Desloges and Ryder (1990) in the Bella Coola area and at Lillooet Glacier (Reyes and Clague 2004). Evidence of glacial advances that correspond to Tiedemann activity includes those found at: Nabesna Glacier, Wrangell Mountain Range (AD 83) and Bernard Glacier (BC 1 15) (Denton and Karlen 1977; Wiles et al. 2002).
Identified at Bridge, Lillooet, and Tiedemann Glaciers, increasing evidence is being discovered in the central and southern Coast Mountains of a widespread period of advancing glaciers at
-
AD 645, now referred to as the Bridge Advance (Larocque and Smith 2003; Allen and Smith 2004; Reyes and Clague 2004). Radiocarbon dates of trees buried by a glacier advance-
AD 248-660 were obtained from wood located near growth position along a lateral moraine at Lillooet Glacier (Reyes and Clague 2004). At Bridge Glacier, radiocarbon dated snags from the glacier forefield indicate that Bridge Glacier advanced into a standing forest at-
AD 597-773 (Allen and Smith 2003,2004). Additional support for the Bridge Advance was previously reported by Ryder and Thomson (1986) who dated a log deposited in glaciofluvial sediments to-
AD 689 at Tiedemann Glacier.The confining dates of the LIA have been established at between AD 1200 and 1900 (Grove 1988). LIA glacial advances are recorded by moraine complexes throughout the Coast Mountains (Ryder and Thomson 1986). Sites containing evidence of LIA advances include that found at Tzeetsaysul Glacier (Smith and Desloges 2000), in the Waddington Range (Larocque and Smith 2003), at Lillooet Glacier (Reyes and Clague 2004), at Bridge Glacier (Allen and Smith 2004), in
the Bella Coola region (Ryder and Thomson 1986; Desloges and Ryder 1990), in Garibaldi Provincial Park ( K ~ c h et al. 2004b), at Wisqually Glacier (Sigafoos and Hendricks 1961), at glaciers located on Mount Rainier (Sigafoos and Hendricks 1972), and in Strathcona Provincial Park on Vancouver Island (Lewis and Smith 2004).
1.2 Format of Thesis
This thesis consists of three chapters. Chapter 1 provides the spatial and temporal context in which the research was conducted. Chapter 2 is a manuscript that summarizes the findings of my field investigations. Chapter 3 discusses my findings and compares them to investigations of glacier behaviour at other locations in the western Canadian Cordillera and makes recommendations for future research.
2.Q Dendroglaciological reconstruction of late Holocene glacier activity at Todd Glacier, Boundary Range, northwestern British Columbia Coast Mountains
2.1 Introduction
The climate of the western Canadian Cordillera is regulated by the proximity of the Pacific Ocean, which imparts a maritime influence on the adjacent mountains through a complex suite of regional climate-forcing mechanisms. Documented changes in
weather patterns in the northwest Pacific indicate that the climate of this region is not static and that shifts between climatic states have occurred not only repeatedly but often abruptly (Charles 1998). Broad-scale climate reconstructions and simulations have drawn attention to the impact of abrupt climate shifts at interdecadal (Gedalof and Smith 2001) to millennia1 scales (Harrison et al. 2003; Spooner et al. 2003). Reflecting still poorly understood changing air mass circulation and ocean dynamics; these climatic "events" exhibit temporal and geographic variations that exceed the bounds of climate variability in the observational record. Although these events may offer an explanation for the punctuated shifts in climates that led to apparently synchronous Holocene-age glacier advances in Pacific North America (Luclunan and Villalba 2001), relatively few studies have focused on providing the data sets necessary to verify the origin and consequence of these low frequency climatic shifts.
High-resolution records of past climate fluctuations are clearly required to detect and assess the long-term impact and significance of any climate-induced changes. However, quantifying natural climate variability in Pacific North America remains problematic due to the brevity of instrumental climate records (Easterling et al. 1996). Analysis of pre-twentieth century climate conditions necessarily relies upon developing
detailed proxy indicators, with glacier-climate response relationships potentially providing insights into long-term paleo-climate change.
Substantive insights into the long-term response of glaciers to climate fluctuations have previously been derived from the application of dendroglaciological research
methodologies (e.g., Luckman 1998; Smith and Desloges 2000; Wiles et al. 2002). Dendroglaciology describes the application of dendrochrsnologic techniques to reconstruct glacier fluctuations (Schweingruber 1989; Luckman 1995). In some applications radiocarbon dates established on in situ stumps and boles have been
dendrochronologically cross-dated to create floating tree-ring chronologies that identify approximately when advancing glaciers killed or damaged the trees (Luckman 1988; Wood and Smith 2004) In other situations, living tree-ring records have been used to assign calendar dates to subfossil wood fragments that were suspected of being killed by a glacier (Luckman 1995 ; Smith and Laroque 1996). Dendroglaciological reconstruction thus has the potential to provide insights into past climate fluctuations by describing intervals in time when glaciers were both advancing and absent from the landscape (Wiles et al. 1996). For instance, the dendroglaciologic dating of a forest killed by an advancing glacier provides evidence of a period of sustained positive glacier mass balance. In contrast, the documentation of ice-fiee conditions and tree growth on glacier forefields describes a sustained period of negative glacier mass balance conditions. Given these understandings, documentation of prehistorical glacial activity from subfossil tree material can provide annually-resolved insights into long-term climate changes that are often only poorly resolved by studies of pollen or by analyses of lake sedimentation rates (Barnes 2003; Menounos et al. 2004).
The purpose of this chapter is to present a dendroglaciological reconstruction of glacier activity in a remote mountain valley in the Boundary Ranges region of the British Columbia Coast Mountains (Fig, 1). The site was selected for detailed study folllowing the discovery of an exhumed forest horizon md exposed subfossil wood in near growth position in the summer of 2003. Preliminary dating of wood samples fssm two locations indicated the site had the potential to provide further information on the long-term behaviour of glaciers in this little studied region. Following a return visit to the site in 2004 to undertake a detailed site assessment, radiocarbon dating and tree-ring cross- dating of subfossil wood samples provided the data necessary to establish the late
Holocene glacial history of this valley. These insights are compared to previous findings and are used to assess the synchroneity of glacial fluctuations throughout the
northwestern Canadian Cordillera.
2.2 Research Background
The Boundary Ranges of the British Columbia Coast Mountains are part of a remote mountain chain that spans the British Columbia
-
Alaska border. Although little research has been undertaken to describe Holocene glacier activity in this region (Clague et al. 2004b), ongoing research efforts at recently deglaciated sites in the area suggest the record is largely synchronous with that described elsewhere within the contiguous North American Cordillera, where roughly four phases of Holocene glacier expansion have been noted (Clague et al. 2004a). Designated as the Garibaldi phase (Ryder and Thomson1986), the Tiedemann Advance (Ryder and Thomson 1986), the Bridge Advance (Allen and Smith 2003), and the Little Ice Age (LIA) Advances (Smith and Desloges 2000;
Larocque and Smith 2003), these periods of glacier activity have been recognized at numerous sites in Pacific North America.
In Alaska and the Yukon, dendroglaciologica~ evidence corresponding to the
Tiedemann Advance include dates obtained from a spruce tree rooted in growth position (AD 83) and transported wood found within glaciofluvial deposits (BC 1 15) at Nabesena Glacier, Wrangell Mountain Range, and organic mattes from an alluvial fan (BC 201) at Guerin Glacier in the St. Elks Mountains (Denton and Kmlen 1977, Wiles e6 al, 2002). Evidence for glacier advances dating to the Bridge Advance include in situ tree stumps (AD 688) found -1 km from the termini of Dinglestadt and Grewingk Glaciers in the Kenai Mountains of Alaska (Wiles and Calkin 1990). Dendroglaciological research methods employed by Wiles et al. (1999) at I9 locations within Prince William Sound led to the documentation of LIA advances ranging in age from AD 1270- 18 13. The
-
600 year spread in LIA fluctuations is common throughout Alaska, where glacial termini advanced during three distinct phases at: AD 1 190- 1248, 16 1 1-
17 15, and 1874- 1895 (Wiles et al. 1999).Within the central and southern British Columbia Coast Mountains, evidence for a glacial advance during the Garibaldi phase has been identified by radiocarbon dating woody debris found within prograded glaciofluvial delta deposits proximal to Fyles Glacier (BC 4766-3625) south of Bella Coola (Laxton et al. 2003) and in Garibaldi Provincial Park where Koch et al. (2004a) dated in situ stumps at Lava Glacier to BC 4460. Wood debris (BC 4044-3330) that had been flushed from a subglacial conduit at Tchaikazan Glacier in Ty'los Provincial Park provides additional dendroglaciological evidence in support of a Garibaldi phase (Smith 2003). Dendroglaciological evidence for
a regional episode of glacid expansion associated with the Tiedemann Advance comes from Tiedemann and Gilbert Glaciers where Ryder and Thompson (1986) discovered glacially deposited logs buried in moraine sediments (BC 1530), Radiocarbon dates of samples from in situ stumps and detrital logs at Bridge Glacier (AD 597-773; Allen and Smith 2004) and at nearby Lillooet Glacier (- AD 248-660; Reyes and Clague 2004) benchmarks the Bridge Advance in this region. Sites containing indications of LlA
advances in the south and central Coast Mountains include dendroglacislogieal evidence dating from between
-
AD 1200-1 900 at Tzeetsaysul Glacier (Smith and Desloges 2000); Lillooet Glacier (Reyes and Clague 2004); and Bridge Glacier (Allen and Smith 2004). Additional evidence is reported from various investigations within the Bella Coola region (Ryder and Thomson 11 986; Desloges and Ryder 1990); within the Waddington Range(Larocque and Smith 2003); within Garibaldi Provincial Park (Kocb et al, 2004b); and at sites in Strathcona Provincial Park on Vancouver Island (Lewis and Smith 2004).
Although there is little evidence supporting episodes of glacial expansion during either the Tiedemann or Bridge Advances in the Cascade and Olympic Mountains of Washington State, Crandell and Miller (1 964) have suggested that Winthrop Glacier on Mt. Rainier attained its maximum late Holocene extent between BC 1727-AD 87.
Numerous investigators have dated moraine deposits using botanical dating techniques to the LIA in this region (e.g., Sigafoos and Hendricks 1961; 1972).
2.3 Study Site
This research was undertaken in the headwaters of north-facing Todd Valley (unofficial name) (56"13'00 N Lat, 129 "46'00 W Long) in northwestern British
Columbia at
-
1000 m as1 (Fig. 1). Located 40 km northeast ofthe town of Stewart, the site is accessible only by helicopter. The Todd Valley study site has been glacially incised and is-
2.5 km long and-
Xkm.
wide. The present-day treeline within the valley is found at-
1500 m asl, at some distance above the LIA glacial trimline at-
1 188 rn asl.Five glaciers drain into Todd Creek at the southern end of the valley. Todd Glacier is the main trunk glacier and the primary source of north-flowing Todd Creek. Rapid glacial retreat and downwasting over the last decade has exposed subfossil detrital wood, as well as stumps and boles in near growth positions, believed to be associated with forests overrun and killed during different glacial advances.
Bedrock Geology
The bedrock geology of Todd Valley is composed of Unuk Formation rocks from the Lower Middle Jurassic Hazelton Group, consisting of brick red to apple green
epiclastic volcanic rocks and lithic tuffs, with closely associated pillow lavas, siltstones, and carbonate lenses (Grove 1986). Mineral exploration has been intermittent in the valley, with recent diamond-drilling programs concentrating on copper-gold deposits 2 km north of the 2004 Todd Glacier terminus.
Glacial History
Clague et al. (2004b) identified phases of Bridge, and LIA advances at Berendon and Frank Mackie Glaciers north of Stewart, British Columbia. Recent insights into late Holocene glacier activity in the Stewart-Meziadin highway corridor come from Barnes
(2003) who utilized tree ring chronologies from Bear Pass and varve chronologies from Meziadin Lake to compile a record of inter-decadal climate change that was significantly correlated to historical temperature and precipitation data from the t o w s of Stewart and Smithers, British Columbia.
Although there have been no previous investigations of glacier activity within Todd Valley, related research has been conducted in the immediate area. At Surprise Glacier, 1 1 km to the east of Todd Glacier, detrital boles and in situ woody debris found within distinct laterally continuous horizons in the proximal. slope of a prominent lateral moraine, provide dendroglaciological evidence for episodes of glacier expansion coeval to the Teidemann, Bridge, and LIA Advances (Jackson and Smith 2005). At nearby Bear Glacier, Haspel et al. (2005) and Spooner et aL (2005) have reported evidence of glacial oscillations that date to Tiedemann (BC 1 489- 1945) and LIA (- AD 1 162) Advances.
In the Stikine River valley,
-
120 km to the northwest of Todd Valley, Ryder (1987) found evidence of a LIA advance at Scud Glacier dating to AD 1525. More recently, Lewis and Smith (2005) have reported on advances of Forrest Kerr Glacier, an outlet of the Andrei Icefield, that date to BC 1 11 1 and AD 645. Miller and Anderson (1 974) suggested that in Alaska, based on radiocarbon dates of overridden stumps and post-glacial bogs, climatic cooling was well established in the northern Boundary Ranges by BC 486. Indications of Teidemann-age activity is provided by stumps sampled in Mendenhall Valley which yielded dates of AD 87 (Miller and Anderson 1974). An exhumed paleoforest of in situ stumps also reported by Miller and Anderson (1 974) indicates that Davidson Glacier was advancing down valley by AD 12 10 in response to changing LIA climates.Glacial Geomorphology
Until the 1950s, five glaciers in Todd Valley coalesced and advanced to a
terminal position
-
2.5h
from the present snout of Todd Glacier (Fig. I), Bug Glacier (unoflcial name) is a steeply sloping (1 1 ") east-facing cirque glacier and thenorthernmost glacier within the study area (Fig. 1). North-facing Todd Glacier is the southernmost glacier with the largest source area of all the glaciers studied (Fig. 1). East of Todd Glacier is west-facing Sage Glacier (unoflcial name) with a steep bedrock slope at its terminus (Fig. 1). East-facing Glacier One and Glacier Two (unoflcial names) are located within the headwaters of Two Glacier Valley (unoflcial name) and are the sources of the small creek that drains into a proglacial lake
-
700 m from the terminus of Todd Glacier in the centre of Todd Valley (Fig. 1). Within the last 25 years, all of these glaciers have retreated and downwasted, with terminal retreat rates varying from 9 to 76mlyr based upon analysis of aerial photographs from 1974- 1997 (Table I). SurJicial Geology
The recently exposed proglacial valley floor in front of Todd Glacier is characterized by massive units of till incised by migrating glaciofluvial meltwater
streams. Several stratigraphic units are visible in streamside exposures, including a 75 m long section where subfossil trees were found in near growth position interbedded
between two till units. This exhumed forest contains boles with root balls, twig, and needle macrofossils, and a thin litter layer.
The valley bottom proximal to the 2004 snout position of Todd Glacier is capped with small-scale flutes exposed between 1994 and 1997 (Fig. 2). Desloges and Ryder (1990) and Smith (personal communication) have noted the presence of similar fluting
features with small-superimposed transecting ridges at sites throughout the Coast Mountains. In this instance, the flutes range in length from 100-300 m, begin at large boulders, and continue d o m valley parallel with the flow direction of Todd Glacier (Figs 3 and 4). Small recessional moraines in the centre of the fluted area range in height from
15-50 cm and are oriented perpendicular to the flutes. The flutes, composed of sub- rounded to sub-angular clasts of local provenance, illustrate massive bedding with no preferred clast orientation, and we matrix-supported (50% fine gravel, 20% sand, and 30% silt) (Fig. 4). Fine sand and silt fmm abandoned stream channels and the proglacial lake proximal to the terminus of Todd Glacier are continually deposited onto the flutes by katabatic winds flowing down valley (Fig.34).
2.4 Research Methods
Fieldwork
Sample collection and field analyses were undertaken in Todd Valley during the summers of 2003 and 2004 (Fig. 5). A one-day reconnaissance trip in July of 2003 yielded twenty tree-disc samples taken from what is now Site 8, and identified sites where detailed studies were later undertaken. During ten days in July, 2004, an additional 1 10 disc samples were collected.
Detrital and in situ wood samples were selectively chosen for dendroglaciological analysis and conventional radiocarbon dating. As hundreds of potential samples were discovered throughout the study site, criteria were developed to restrict the number of samples selected in the field for later analysis. This was necessary due to the space and weight restrictions of the helicopter used to transport the samples from the field site. Sample site selection was determined based on the setting in which the sample was
situated. Preference was given to samples that were either in sitzk or in near growth position, as it was easier to determine the provenance of those samples. Preference was given to samples that: had at least 100 annual growth rings; illustrated year-to-year variability; had bark or illustrated with little or no perimeter rot; and, preferably, were located in situ or near-growth position.
Data collection focused on wood samples found at selected locations within the study area (Fig. 5). At each site, radial cross-sections were collected with a chainsaw by cutting across the radial axis of subfossil boles and stumps, Prior to cutting, each sample was wrapped with duct tape to maintain its integrity during cutting. After cutting, the samples were labeled and then taped again to ensure cohesion as they dried.
Laboratory Analysis
All samples were returned to the University of Victoria Tree-Ring Laboratory (UVTRL) to air-dry. Wood glue was applied to samples that fragmented during drying to keep them intact. Once the radial cross-sections were dry, one face of section was sanded with progressively finer sandpaper to 600 grit and scanned using an AGFA Duoscan optical scanner at a resolution greater than 800 ppi. Tree-ring widths were measured along a minimum of two radial axes to the nearest 0.01 mm using WinDENDROTM (Version 6.1 D, 1998), an image processing software designed for measuring tree-rings (Guay et al. 1992). Where the ring boundaries were difficult to identify, a 40X
microscope and Velmex-type stage measurement system were used to verify the ring boundary positions.
The samples were first internally cross-dated and then an attempt was made to cross-date them within and or between, the various study sites. Cross-dating was
accomplished with reference to common pointer years (narrow tree-rings)
(Schweingmber 1989) and the accuracy of the cross-dating checked using COFECHA (Grissino-Mayer 2001). Values greater than 0.328 1 for series intercorrelation (significant at the 99% confidence level) were accepted as providing statistically significant cross- dates (Holmes 31983). Whenever samples could not be cross-dated, or when the cross- dating required verification, attempts were made to cross-date the sample with
radiocarbon-controlled floating chronologies independently developed by Jackson and Smith (unpublished manuscript) at nearby Surprise Glacier.
Radiocarbon analyses were undertaken to anchor the floating chronologies in time (c.f. Luckman 1995)- Samples of perimeter wood were removed from selected samples, the number of annual growth rings counted, and the samples forwarded for analysis to the Beta Analytical Radiocarbon Dating Laboratory in Miami, Florida. The dates reported from the lab are in 14c yrs BP and were calibrated using the INTCAL 98 Radiocarbon Age Calibration (Stuiver and van der Plicht 31998).
2.5 Results
Detailed field investigations were completed at eight study sites to reconstruct the spatial-temporal pattern of glacier activity within the study area (Table 2 and Fig. 5). Five of the sites were located in north-south trending Todd Valley (Sites 1 to 4, and Site 8). The remaining three sites were located in east-west trending Two Glacier Valley (Sites 5 to 7). An examination of historical aerial photographs indicates that Sites 3,4, and 6 were buried by glacial ice in 1974 and that Sites l , 2 , 7, and 8 were all ice-free prior to 1974. Sites 3 and 6 were free of ice by 1997, and Site 4 was exposed between
subalpine fir (Abies lasiocarpa) remains, except for sample TG04-801 which was identified as mountain hemlock (Tsuga mertensiana),
Site 1
Site 1 is located near the 2004 terminus of Bug Glacier,
-
2 Ian north of the 2004 toe of Todd Glacier (Figs. 7 to % 1). The remnant of an interlobate moraine that separated the two glaciers is located immediately south of the Bug Glacier terminus and has contributed to the formation of a small proglacial lake. Examination of historical aerial photographs indicates that Bug Glacier retreated at a much slower rate from 1974 to present than the other glaciers within the study area (Table 1).Three samples were collected at Site 1. TG04-801 was a weathered bole fragment
-
10 cm in diameter (>237 rings) found wedged within a diamict-filled depression on a large, gossan stained, bedrock promontory below Bug Glacier. Dated to AD 1400- 1520, TG04-801 was apparently killed during an advance of Bug Glacier in the intervalbetween either AD 1400-1 520 or 1580-1630 (Table 2).
TG04-802 was located along the distal flank of the outermost lateral moraine of Bug Glacier. Found under a 1.5 m diameter boulder, TG04-802 appeared to be the remains of an in situ tree stump (63 rings) partially buried by distal moraine spillage in AD 1640-1950 (Figs. 9 and 10).
TG04-803 was a 1.5 m long bole fragment (15 cm diameter, 73 rings) found protruding from glacial sediments
-
100 m from the 2004 terminus of Bug Glacier (Fig.11). No date was established for this sample, as it did not cross-date with the other samples fiom Site 1 or with samples collected elsewhere in the study area.
Interpretation
The limited dating completed at Site 1 suggests that Bug Glacier advanced Into Todd Valley during the early- and late-LIA. The recent kill date established dong the distal slope of the northern lateral moraine describes the maximum LIA extent of Bug Glacier and may define the termination of LIA activity within Todd Valley (Table 3). On the other hand, since the response time of Bug Glacier to historical climatic fluctuations appears to lag behind that of the other glaciers in the valley (Table 26, this suggesti~n is made cautiously.
Site 2
Site 2 is located
-
400 m north of the 2004 terminus of Sage Glacier on the eastern wall of Todd Valley (Figs. 12 and 13). Characterized by large bedrock outcrops and unconsolidated lateral moraine sediments (sandy matrix-supported boulders and cobbles), this slope shows evidence of postglacial erosion and modification by snow avalanche activity. Numerous boles, detrital wood fragments, and stumps were found buried in the moraine sediments.Eight samples were collected from Site 2 in 2004. Four samples (TG04-868 to TG04-871) were found projecting from or lying on eroding morainal sediment at various locations 150-300 m from the 2004 terminus of Sage Glacier. Radiocarbon dated TG04- 871 [I16 rings; AD 1250-1420; Table 21 cross-dates (r = 0.424) with TG04-870 (80
rings) and suggests a contemporaneous origin for the buried wood at this location. Four rooted boles (TG04-864 to TG04-867) were located in growth position on the down-valley side of a large bedrock outcrop
-
400 m north of the 2004 terminus of Sage Glacier (Figs. 14 and 15). Two samples, TG04-866 (1 10 rings) and TG04-867 (94rings), cross-date with TG04-871 (r = 0.424), confirming that they were all killed during the same event (Table 4) (Figo 16).
Interpretation
Radiocarbon dating and cross-dating suggest that the trees at Site 2 were killled during an early LIA advance of either Sage andor T ~ d d Glaciers
-
AD 1250-1420, Given that a prominent medial moraine separated the two glaciers in 1974, it seems likely that Sage Glacier scoured the valley slope and killed the trees that were growing on the valley side,Site 3
Site 3 is an abandoned meltwater channel located -1 km north of the 2004 terminus position of Todd Glacier on the western side of the valley floor (Fig, 11 7). Meltwater flowing from Todd Glacier and Two Glaciers from 1994 to 1997 has incised through
-
9 m of valley bottom sediments at this location. Although stream avulsion and frontal retreat of Todd Glacier presently directs the meltwater channel down the eastern side of the valley into a proglacial lake, during times of high runoff and precipitation the normally dry channel bed at Site 3 still plays host to a small ephemeral stream (Fig. 18).The abandoned channel at Site 3 is
-
300m
long with vertical banks composed of glacial diamict of local lithology capped by a series of flutes transected by small end moraines. A single in situ bole (TG04-838; 89 rings) was found 6 m up the side of the bank within a 10 cm thick horizon of needles, small twigs, and litter overlain by 3 m of till (Figs. 19 and 20). Radiocarbon dated to AD 1260- 14 10 (Table 4), TG04-838 was killed and buried during an early LIA advance of Todd Glacier.Numerous other woody debris fragments, including some that were bullet-shaped and indicative of glacial transport, were found scattered across or within close proximity to the abandoned channel. TG04-835 (160 rings) was found protruding out of a small recessional moraine, 150 m south of the abandoned channel. At the easternmost extent of Site 3, TG04-836 (92 rings) and TG04-837 (1 13 rings) were found wedged among rocks
on the south side of the abandoned channel (Table 5). Although radiocarbon dates were not established for these samples, TG04-836 (r = 0.358) cross-dated into a radiocarbon-
anchored floating chronology constructed from samples from Sites 4 and 6, as well as from samples collected along the strand line of the downstream proglacial lake. Based on its placement within this floating chronology, it is presumed that TG04-836 was killed
-
AD 1260-1410.Interpretation
Radiocarbon dating and tree-ring cross-dating of two samples at Site 3 indicate that Todd Glacier advanced into an established forest at this location in
-
1 260- 14 1 0 AD (Table 5).Site 4
Site 4 is located
-
200 m from the 2004 terminus of Todd Glacier in the centre of Todd Valley (Fig. 21). Sequential aerial photographs from 1974 to 1997 indicate the site was ice covered until 1997, after which a meltwater stream draining Todd Glacier and Two Glacier Valley incised through-
5 m of valley bottom sediments (Figs. 22-24). Within the past few years the channel has aggraded and changed course, leaving behind scores of bole and detrital wood fragments scattered along the channel. Nine of thesesubfossils were sampled and incorporated into a local floating chronology anchored to TG04-838 (AD 1260-141 0) from Site 3 (Table 6).
The abandoned stream channel at Site 4 is distinguished by a
-
75 m long east- facing c u t b d containing three stratigraphic units (Fig, 25). The Iowemost unit (Unit 1) is composed of 4.5 m of a matrix-supported, massive lodgement till with no distinct bedding, organics, grading, or fabric present (Fig. 26). Clasts within the unit range from fine silt to large cobbles, are sub-angular to sub-rounded, m d are exclusively composed of local volcanics.Overlying this unit is a laterally extensive 30 cm thick organic horizon composed mainly of needle, twig, branch, and root macrofossils (Fig. 2%). Large boles, some with root balls intact, protrude from the exposure. The primary orientation of the boles' long axes is in an east-west direction, parallel to the flow direction of Sage Glacier and tangential to the flow of Todd Glacier.
Unit 3 overlays Unit 2 and is a 2.4 m thick matrix-supported, massive glacial
diamict capped by north-south trending flutes (Fig. 28). Clast sizes range from fine sand to large cobbles, with a-axes lengths of up to 20 cm. Angularity of the clasts range from sub-rounded to sub-angular. There are no organics present within the unit at the
exposure.
Dendroglaciologic investigations at Site 4 focused on ten wood samples (92 to 213 rings) collected from within Unit 2 (Table 6). A radiocarbon date for TG03-806 (101 rings) found in near growth position indicates that Unit 2 was buried
-
AD 420-200 (Fig. 27). The weight of the glacier and/or till overburden resulted in ring compression withinthe other nine samples collected from Unit 2 and rendered unsuccessful cross-dating attempts (Fig. 6).
Interpretation
Deposition of the lowermost stratigraphic unit at Site 4 predates AD 420-200 m d may be related to a Tiedemann-age or earlier advance of Todd Glacier, Unit 2 is
interpreted to be an exhumed forest horizon containing mature trees that were pushed and buried as Sage Glacier advanced into Todd Valley from the east in
-
AD 420-200. The common orientation of the boles and the presence of litter and a thin soil layer are convincing evidence that the wood was not deposited by snow avalanche activity. The deposition of Unit 3 dates to the period after AD 420-200. The presence of flutes on thesurface at Site 4, contiguous with those at Site 3, suggests a LIA modification of this unit.
Site 5
Site 5 consists of a large mixed assemblage of boles, stumps and woody debris located on the south-facing wall of east-west trending Two Glacier Valley adjacent to the head of Todd Valley (Figs, 29 and 30). Ranging from 2 m long boles with a radius of 20 crn to large branches and root fragments (Figs. 3 1 and 32), this detrital assemblage appears to have been pushed by glacial ice into a protected cavity on the lee side of a large bedrock outcrop. Radial samples fiom six boles (TG04-879 to TG04-885) indicate that these were mature trees (1 10-21 3 rings) when they were killed (Table 7). Although
four of these samples cross-date into a floating site chronology (r = 0.524; Table 7), the chronology did not cross-date with radiocarbon-dated samples from any of the other sites in either Todd or Surprise Valley.
Interpretation
Although no kill date was established for the trees at Site 5, their condition and placement suggest that they were deposited during an eastward advance of the Two Glaciers during the LIA, If the trees had been killed or deposited during an earlier
advance, the re-advancement of the Two Glaciers during the LIA would have presumably eroded them.
Site 6
Site 6 is located along the east-west trending Two Glacier valley floor, close to where it joins Todd Valley (Fig. 33). The recent and ongoing retreat of the Two Glaciers has resulted in the stream erosion of pre-existing glaciogenic valley-bottom sediments. This fluvial activity has resulted in the formation of isolated till islands surrounded by active and stabilized meltwater stream channels. One bullet-shaped stump (TG03-8 15;
165 rings) was found in situ within the basal till unit and yielded a date of AD 405-640, suggesting that it was killed during the Bridge Advance (Table 8).
Stream eroded and deposited wood debris was also found scattered on outwash deposits throughout the valley (Fig. 35). Twelve of these detrial boles and stumps (37 to 18 1 rings) were sampled and used to help construct two radiocarbon dated floating chronologies from Sites 4,7, and Surprise Glacier. Five samples collected from fluvial cut banks and till deposits cross-date into the 20 sample LIA floating chronology (r =
0.410) composed of samples from Sites 3 and 4. Six samples (TG04-846, TG04-856, TG04-U, TG04-S, TG04-23, TG04-34) were combined with TG03-815 and TG04-874 from Site 7 to create a floating chronology anchored to trees killed during glacier advances at AD 405-640.
interpretation
Dendroglaciological and radiocarbon evidence from Site 6 show that the Two Glaciers advanced into standing forests at least twice during the late Holocene. The earliest advance is coeval with the Bridge Advance at
-
AD 405-640; while LIA advances me represented by trees killed in-
AD 1 172- 128 1 (Table 8). Umfortunatelly, with only a single in situ detrital sample identified in Two Glacier Vdley, little detail can be provided as to the rate of glacial expansion along the valley floor.Site 7
Site 7 is located within a north-south trending bedrock gully found 400 m northwest of the terminus of Todd Glacier (Figs. 36 and 37), Although only recently deglaciated, and currently bare of trees, a buried horizon of woody debris was traced for
-
20 m within the till sediments draped down the valley wall. Exposed and eroded by stream incision during the retreat of Todd Glacier up and over the overlying bedrock ridge, woody debris also lies scattered on a large alluvial fan located below the gully on the floor of Two Glacier Valley.Five dendroglaciologic samples were collected at Site 7. Two samples (TG04- 872, 185 rings; TG04-873, 134 rings) were collected from within the exposed gully wall but did not cross-date with one another (Figs. 38 and 39). Two additional broken bole fragments (TG04-874, 106 rings; TG04-875,211 rings) were found lying on top of glacial debris on the south side of the Site 7 bedrock promontory (Fig. 40). TG04-874 yielded a Bridge Advance date of AD 480-520, and cross-dates into the eight-sample floating chronology referred to at Site 6 (Table 9). TG04-876 (233 rings) was taken from a broken bole fragment and could not be cross-dated.
Interpretation
Based on the single date established for Site 7, it seems probable that Todd Glacier advanced up over the bedrock ridge at
-
AD 480-520 during a period of glacier expansion associated with the Bridge Advance. At that time the mature forest existing on the ridge top was pushed down-slope by the advance of Todd Glacier and buried by subglacial sediments.Site 8
Site 8 is located within the remnants of a large lateral moraine plastered against the western wall of Todd Valley between Two Glacier Valjey and Bug Glacier.
Composed of over-consolidated matrix-supported cobbles and small boulders, the steep proximal slope of the moraine is unstable and actively eroding. Numerous wood
fragments occur throughout the exposure, but only TG04-805 (103 rings) could be sampled safely. Cross-dating with TG04-871 from Site 2 (Table lo), suggests that the wood fragments found within the lateral moraine may all date to
-
AD 11200-1390.Interpretation
Based on the single cross-date established at Site 8, it appears that the lateral moraine at this site was constructed during an early LIA advance of Todd Glacier northward down Todd Valley.
2.6 Discussion
The findings of these dendroglaciological investigations describe the timing of five discrete glacial advances in Todd Valley: an undated pse-Tiedemann or Tiedemam Advance; a "kite9' T i e d e m a Advance (herein referred to as the S a g e Advance'); the Bridge Advance; and at least three LIA Advances (Table 2) (Fig.41). The episodes of glacial advance in Todd Valley are synchronous with other proxy records of dimate fluctuations for the northeast Pacific region and Canadian Rocky Mountains during the late Holocene. These records include those derived from tree-rings, lake cores, lichen, and pollen proxies,
Bre-Sage
The earliest evidence of glacial activity in Todd Valley is the till unit ]located below the paleoforest horizon dated to BC 197 at Site 4. It is possible that the lower-most till unit at Site 4 was deposited during a period of expansion coeval with the BC 1 008 advance at Surprise Glacier or the -BC 1530-1393 advance at Bear Glacier (Haspel et al. 2005; Jackson and Smith 2005).
Sage Advance
The east-west a-axis orientation of the boles located in the exhumed forest
at Site 4 show that Sage Glacier advanced into the valley bottom at BC 197. This advance is synchronous with an advance at Berendon Glacier,
-
35 km northwest of Todd Valleywhere Clague and Mathewes (1 996) found that Berendon Glacier reached its maximum mid Holocene extents at BC 158 and 83 1. Using a combination of lacustrine and terrestrial evidence, Menounos et al. (2005) suggest that in
-
BC 794- 197 some glaciers in the southern Coast Mountains extended to their maximum downvalley positions.Corroborating evidence of an
-
AD 87 glacial expansion has been identified by Ryder and Thomson (1 986) in the Mt. Waddington area of the southern Coast Mountains, where they described the maximum Tiedemann Advance as having occurred at-
BC 197.Synchroneity with the Sage Advance at Todd Glacier is also evident by
palynological findings that demonstrate a cooling trend in the northern Boundary Ranges. Based on Tsuga heterophylla (western hemlock) pollen records in northwestern North America, Sgooner et al. (2003) suggested that there was an increase in precipitation from AD 87-BC 1 1 1 1. The Sage Advance is coeval with the cool-moist conditions prevailing from
-
BC 403-AD 436 (Pellat et al. 2000), and the initiation of %ate Holocene glacial activity in White Pass at-
AD 87 (Lamoureux and Cockburn 2005). Based on pollen reconstructions, Hebda (1 995) noted that cooling and moist conditions prevailed in the north of British Columbia during the interval between BC 3369-AD 5. Spooner (1 994) interprets the period from BC 2630-AD 5 as a time of cooling and increased moisture for northern British Columbia.Recognition of the period around
-
BC 197 as a time of cool temperatures, increased moisture, and glacier expansion in the northwest corner of British Columbia indicates that the Sage Advance discovered at Todd Glacier may be evidence of larger regional scale climate conditions conducive to positive mass balance glacier conditions. Bridge AdvanceEvidence of glacial activity coeval with the Bridge Advance was found at Sites 6 (AD 405-640) and 7 (AD 480-520). As samples from both Sites 6 and 7 cross-date, indicating contemporaneous kill dates, it is suggested that Todd Glacier advanced
northward over the intervening bedrock ridge and coalesced with the Two Glaciers advancing from the west.
This record of a first millennium AD glacier advance is supported by
dendroglaciologic investigations at a number of sites in Pacific North America (Reyes et al. 2005) and by independent proxy climate records that show regional cooling, between AD 400 and 700. At Fra& Mackie Glacier
-
35 km northwest of Todd Valley, plant macrofossils extracted from deltaic and glaciolacustrine sediments yielded radiocarbon dates indicating that the glacier was advancing by AD 340-570 (Clague et al. 2004b). Jackson and Smith (2005) were able to establish two first millennia1 dates from a stacked sequence of detrital wood buried in the proximal face of a large lateral moraine atSurprise Glacier,
-
1 l km east of Todd Valley, dating to AD 220- 540 and AD 530- 780. At Forrest Kerr Glacier in the Iskut region, Lewis and Smith (2005) radiocarbon dated sheared wood remains to AD 100-500, indicating that the glacier advanced into a forest at that time. A second subfossil wood radiocarbon date suggests that the glacier was continuing to advance downvalley [at AD 530-7801.In the southern Coast Mountains Allen and Smith (2004) radiocarbon dated an in
situ stump close to the contemporary terminus of Bridge Glacier. The sample dates to AD 430-650 indicating that during that time the area 200 m down valley from the 2002 terminus was covered by ice. Wiles et al. (1999) dated an in situ stump at Beare Glacier near the mouth of Icy Bay Alaska, and found that the glacier had almost reached its Holocene limit by AD 420-670. Hu et al. (2001) used geochemical proxies in lake sediments and found evidence for pronounced cooling in southern Alaska centered at AD 600.
Little Ice Age
The LIA has been recognized in the western Cordillera as an extended period of positive glacier mass balance conditions during the late Holocene (e.g., Ryder and Thomson 1986; Luckman 2000; Smith and Desloges 2000; karocque and Smith 2003). Since the LIA was the most extensive and latest glacial advance to occur within the study area, there was markedly more dendroglaciologicsal evidence recovered substantiating glacier advances during this time period (Figure 44). Dates obtained from Site 1 (AD 1400-1630, 1640-1950), Site 2 (AD 1250- 2420), Site 3 (AD 1260-1410), and Site 8 (AD 1200-1390) indicate that Sage, Todd, Bug and the Two Glaciers advanced at various times during the LIA.
Three phases of LIA advance were identified in the Todd Valley area: an early, mid, and late phase. Evidence for simultaneous LIA advances on the east and western walls of Todd Valley was indicated by the cross-dating of samples between Sites 2 md 8, providing supporting evidence for the early LIA advance of Sage Glacier (Table 2). The advance noted at Sites 2 and 8 is the earliest indication of LIA activity in Todd Valley. Other evidence of coeval mid-LIA advances at separate locations within the Todd Valley area includes cross-dated samples from Sites 3,4, and 6. The latest evidence of LIA glacier expansion was found at Site 1 and post dates all of the other LIA samples collected in Todd Valley by
-
200 years.The cross-dated samples from sites 2 and 8 provide strong indication that Sage and Todd Glaciers were advancing during the early LIA. Evidence of coeval glacial advances fiom the 1 2 ' ~ and 1 3 ' ~ centuries has been discovered in Alaska, throughout the Coast and Canadian Rocky Mountains. For instance, Wiles et al. (1999) identified a late
1 2th to early 1 3 ' ~ century advance in the Prince William Sound area of Alaska that was undenvay at AD 1 190 and continued until AD 1248. In the Kenai Mountains of Alaska, Wiles and Calkin (1 994) also found evidence of glacier expansion that began in AD 1300. Lichenometrically dated moraines near Waskey Lake Alaska indicated that glaciers there reached their maximum at
-
700 BP (Levy et ali. 2004). In the southern Coast Mountains Larocque and Smith (2003) used lichenometry to date the construction of the LIA moraines between AD 1203 and IL 226. In the Canadian Rocky Mountains Luckman (1 993) found glacially-overridden trees which show that glaciers there were advancing between AD 1 150 and 1350.Precipitation reconstructions based on tree-ring data sets suggests that the LIA expansions at Todd Valley were synchronous with reconstructed precipitation increases in the Coast Mountains, interior sf British Columbia, and Canadian Rocky Mountains (Table 2). Using a combination of tree-ring derived temperature and snowpack proxy indices from the area around Mt. Waddington in the Coast Mountains, Larocque and Smith (2005b) suggest that cool and wet conditions prevailed at AD 170 1
-
1.708, 17 16- 1727, 1735, 1742-1 765, 1787-1 792, 181 3-1 823, and at 1848-1 85 1. Tree-ring based reconstructions of summer temperature maximums in the interior of British Columbia using Englemann spruce (Picea engelmannii) chronologies point to two prolonged intervals-
AD 1620-1 71 0 and 1775-1 880 that were cooler than present (Wilson and Luckman 2003). Luckrnan et al. (1997) utilized Englemann spruce (Picea engelmannii) and subalpine fir (Abies lasiocarpa) tree-ring chronologies to create a densitometrically- based 91 0 year summer temperature reconstruction. The reconstruction identified major cold intervals at-
AD 1200-1 350, l69Os, and the 1800s that coincide with regional andlocal periods of glacier expansion in the Canadian Rocky Mountains. Periods of moraine building in the 1 3'h-14', early 18" and 1 gth centuries identified in the Canadian Rocky Mountains on the most part match patterns of decreased summer temperatures derived from tree-ring chronologies (Luckman 2000). The coolies intervals identified by Wilson and Luckman (2003) are coincident with the mid md late LIA pulses of glacial expansion at Todd Glacier (Table 2). Although these reconstructions are from the Canadian Rocky Mountains, the main cold periods in their record are coeval with the LIA glacier
expansions at Todd Valley (Luckman et al. 1997).
Radiocarbon dated sample TG04-802 indicates that Bug Glacier was expanding during the late LIA. This expansion could be linked to increases in precipitation that have been reconstructed from proxy data in the Coast and Canadian Rocky Mountains. Luckman (2000) notes that based upon annual precipitation reconstructions created for the Banff Powerhouse site, for the most part, the 1 7th and 1 8h centuries appear to be
warmer and drier than the preceding century. However, a moister and cooler period in the late 1 7 ' ~ and early 1 gth century was acknowledged. Using glaciolacustrine records from Hector Lake, Canadian Rocky Mountains, Leonard (1 986) identified increased sedimentation rates in the lake varves. The increase in sedimentation is contemporaneous with glacial advances in the 1 3th, early 14'~, 1 gth, and 1 9th centuries (Leonard 1986).
The temperature and precipitation variation inherent with these climate reconstructions have been used to develop proxy glacier mass balance records. In the Waddington area of the Coast Mountains, Larocque and Smith (2005a) were able to identify three periods of positive mass balance during the last 450 years at AD 1750s,
Canadian Rocky Mountains proxy glacier mass balance records identified two periods of increased positive glacier mass balance at AD 1696- 1720, and 11 8 1 0- 1825 (Watson md
Luckman 2004). Watson and Luckman (2004) interpreted these periods of increased glacial mass balance as being attributed to a combination of higher winter precipitation and reduced summer ablation.
2.7 Conclusion
Through the integration of stratigraphic and dendroglaciologic mall yses, three periods of glacial advance during the late Holocene have been identified in Todd Valley: the Sage (BC 420-200), Bridge (AD 405-640), and LIA (1 400-1 630, 11 640- 1950, 1250- 1420, 1200- 1390 AD) Advances. The episodes of positive glacier mass balance recorded by these advances are interpreted to be coincident with late Holocene climate proxies located elsewhere in the Coast Mountains and Canadian Rocky Mountains (Fig. 42). This synchronous relationship between climatic deterioration and glacier advancement in Todd Valley provides evidence that the glaciers in the study area can be effectively used as proxy indicators of regional climatic forcing events.
3.0 Summary
The construction of predictive climate models is essential to better understand the ways in which the climate may behave in the future. This is especially pertinent today, as changing climates are continuing to have negative impacts upon human settlements. The relationship between climate flux and glacier mass balance is very clear as glaciers retreat in partial response to current global temperature increases. Chronologies of past glacier activity have been used to strengthen high-resolution multiproxy data sets that record climate forcing events that have extended back thousands of years. This thesis identified
