Measurin

ABSTRACT. As part of a general study of arctic driftwood, 206 samples of driftw ood logs from the Mackenzie delta area were analyzed by dendrochronological methods (tree - ring studies). The aim was to detect the origin of the wood. Three forest stands in the delta were also sampled, and tree - ring chronologies were constructed. The Mackenzie drif twood can be divided into four groups: 1) driftwood originating from the upper Mackenzie delt a area with individual logs having up to 600 tree rings, 2) driftwood originating near the s outhern limit of the delta, 3) wood with relatively few tree rings with possible origin in the Li ard River drainage area, and 4) driftwood samples not datable with any available chronologies.

Three driftwood samples from the coast of Greenland could be correlated wit h tree - ring chronologies from the Mackenzie delta area and another three were correlated wit h chronologies from Alaska. American driftwood has not been detected in collections from Svalb ard and Iceland, although more than 200 samples have been analyzed from each area. This indicates that some American driftwood is transported from the Beaufort and Bering seas to the North Atlantic Ocean, probably via the western part of the East Greenland Current. This wood i s deposited on the coast of Greenland. American driftwood probably does not reach the islan ds in the central and eastern North Atlantic.

Key words: Mackenzie River, driftwood, dendrochronology, Arctic Ocean, East Gre enland Current, surface currents, drift ice, Canada, Alaska

RESUME. Dans le cadre d'une etude generale du bois flotte dans l'Arctique, on a analyse 206 echantillons de billes de bois flotte provenant du delta du Mac kenzie grace a des methodes dendrochronologiques (etudes des faisceaux de croissance de l'arbre). Le but etait de determiner l'origine du bois. On a aussi fait un echantillonnage de trois peuplements de forets dans le delta et on en a etabli l a chronologie des faisceaux de croissance. On peut diviser le bois flotte du Mack enzie en quatre groupes: 1) le bois provenant de la partie amont du delta du Mackenzie, a vec des billes possedant chacune jusqu'a 600 cernes concentriques, 2) le bois flotte provenant de la limite meridionale du delta, 3) le bois ayant relativement peu de cernes, proven ant peut - etre du bassin hydrographique de la riviere au Liard, et 4) des echantillons de bois flotte pour la datation desquels aucune chronologie n'etait disponible.

On a pu correler par dendrochronologie trois echantillons de bois flotte pr ovenant de la cote du Groenland avec des echantillons provenant de la region du delta du Mackenzie, et trois autres avec des echantillons provenant de l'Alaska. Dans de s collections provenant du Svalbard et d'lslande, on n'a pas detecte de bois flotte americain, bien qu'on ait analyse plus de 200 echantillons de chaque region. Cela revele qu'une petite quantite de bois flotte americain est transporte de la mer de Beaufort et de la mer de Bering jusqu'a l'ocean Atlantique nord, probablement par la partie occide ntale du courant est - groenlandais. Ce bois est depose sur la cote du Groenland. Le bois flotte americain n'atteint probablement pas les iles du centre et de l'est de l' Atlantique nord.

Mots cles: fleuve Mackenzie, bois flotte, dendrochronologie, ocean Arctique, cou rant est - groenlandais, courant de surface, glace de derive, Canada, Alaska

Traduit pour Arctic par Nesida Loyer.

INTRODUCTION

Driftwood occurs along many arctic shores. The amount of driftwood varies great ly, from only scattered logs to beaches completely covered with wood. Occasionally there are t hick piles of logs along stretches of the coast. In dry areas, driftwood can be preserved for thousands of years (Eurola, 1971; Blake, 1972; Haggblom, 1982; Bartholin and Hjort, 1987), in many cases with the tree - ring pattern preserved. By applying dendrochronological methods it should be possible to track the origin of the woo d and date it, if a circumpolar network of tree - ring master chronologies is available.

The driftwood originates in the circumpolar boreal forest regions that surr ound the Arctic. The rivers draining into the Arctic Ocean carry huge quantities of driftwood tha t derive either from living forests undercut by rivers or from logs that have come loose during timber floating. Much of this wood is caught in drifting ice transported by ocean currents and is eventually deposited along barren shores. Driftwood has been used for studying the isostat ic rebound after the last glaciation and Holocene eustatic transgressions (Blake, 1972; Salvigsen , 1981; Haggblom, 1982).

In recent years, the possibility of applying dendrochronological methods to driftwood has been suggested by Parker et al. (1983), Bartholin and Hjort (1987),and Eggertss on (1992). The suggestion was inspired by the work of Giddings (1941, 1943, 1952), who proposed the mapping of Arctic Ocean currents by using the tree - ring dating method. However, his p rimary interest was to cross - date the various driftwood populations along the Bering Sea coast and the arctic coast of Alaska. At the same time he noted that the source of some driftwood ca n be determined. For example, driftwood logs found on the shores west of Point Hope on the north coast of Alaska had a tree - ring pattern restricted to the Yukon region in Alas ka. They must therefore have drifted down the Yukon River into the Bering Sea and then north t hrough the Bering Strait into the Arctic Ocean.

The Mackenzie River Driftwood Project

The Mackenzie River system (Fig. 1) carries huge quantities of driftwood in to the Arctic Ocean, mainly during ice breakup at the beginning of June (Mackay, 1963). The t rees originate from living forests that have been undercut by the rivers (Fig. 2). Some logs a re subsequently and temporarily deposited on the riverbanks, whereas others are carried to the A rctic Ocean. Ocean currents and sea ice transport the wood and deposit it along the arctic sh ores. The driftwood can resist decay for a long time in the dry conditions of the Arctic.

The aims of the present study are 1) to describe and identify the origin of the wood that presently drifts down the Mackenzie River into the Arctic Ocean and 2) to provid e a dendrochronological master series for that wood, which then could be used to ide ntify and date Mackenzie driftwood deposited on arctic shores elsewhere.

The Mackenzie River system was chosen as one reference river for an overall study of arctic driftwood because it is one of two main sources of American driftwood to the Arc tic, the other being the Yukon River in Alaska (Eggertsson, 1992), which drains into the Bering Sea.

Kindle (1921) studied Mackenzie River driftwood and noticed that the great bulk of it came from the Liard River, which is the main tributary of the Mackenzie. He also not iced that spruce (Picea) and poplar (Populus) constituted the bulk of the wood.

The dendrochronology of the Mackenzie driftwood has not been studied before , but the work of Giddings (1941, 1952) on Yukon driftwood has been used to identify arctic dri ftwood of Alaskan origin (Eggertsson, 1992).

The Mackenzie River

The Mackenzie River and its tributaries drain an area of 1.84 million km'Sy mbol not transcribed'2. It is the second largest river system in North America, after th e Mississippi. It is the longest river in Canada and one of the ten longest in the world, measurin g 4220 km and flowing from Great Slave Lake in the south to the Beaufort Sea in the north, pas sing through boreal forest on its way to the arctic tundra. The main tributary is the Liard River, which originates in the Rocky Mountains of northern British Colombia and Yukon (Fig. 1 ).

Vegetation of the Mackenzie Delta

The northern part of the Mackenzie delta lies within the tundra zone, the s outhern part in boreal forest. The boundary zone between tundra and forested areas may be eithe r sharp or transitional. In general the gradual sequence from south to north is: 1) contin uous woodland, 2) open woodland, 3) patches of woodland and tundra with abundant willow (Salix) and ground birch (Betula glandulosa), 4) willow scrub (Salix nigra) and ground birch, 5) tu ndra with patches of willow scrub and ground birch, and 6) tundra (Mackay, 1963). The northern li mit of the woodland - tundra (3) is placed at the inferred limit of spruce (Picea). Althou gh isolated trees grow far beyond that limit, spruce is the dominant tree in the woodland - tundra but scattered balsam poplar (Populus balsamifera), tamarack (Larix laricina), and tree birches (Betula papyrifera) also occur. Spruce occurs mainly in the valleys and on slopes. The northernmost spruce in the delta grows today at 69Degree39'N, but during the Holocene climati c optimum (8500 - 5500 BP) spruce grew on the Tuktoyaktuk Peninsula 100 km farther north ( Ritchie and Hare, 1971).

Open woodland (2) covers most of the area south of Inuvik. The predominant tree there is white spruce (Picea glauca), which grows in valleys and on the river terraces, o ften as isolated stands. Black spruce (Picea mariana) grows in some bogs. Tamarack grows on ope n rocky slopes and in depressions together with spruce. Poplars and tree birches are sc attered. The area of continuous forest cover is relatively small, because it only occurs as narrow ribbons along the river channels and lakes (Mackay, 1963).

METHODS

During the summer of 1990, sampleswere collected from both living trees an d driftwood. The living tree samples were taken with a Swedish increment corer. From the dri ftwood logs a disk was sawn about I m from the root or thick end using a chain saw, and the length of the logs and their distance from the river were measured.

Three stands of Picea glauca were sampled in the Mackenzie delta (Fig. 1). The first was from a well - drained area on an island in the modern delta about 6 km southwest of Inuvik (Fig. 1:S2). A total of 26 core samples were collected from 24 trees.

The second stand was on wetter ground about 8 km north of Inuvik, on the ri verbanks of the East Channel (Fig. 1:S1). Here 34 cores were taken from 17 trees.

The third sampling locality was on the elevated Pleistocene delta 4 km east of Fort McPherson (Fig. 1:S3). Here 16 cores were taken from 8 young trees.

Driftwood samples were collected from the banks of the Mackenzie River (Fig . 3) between the village of Arctic Red River and Separation Point, just upstream of the delta (130 samples; Fig. 1:D2) and from the Beaufort Sea coast at Peninsula Point (Fig. 4), 10 km we st of Tuktoyaktuk and at Tuktoyaktuk itself (160 samples; Fig. 1:D1).

Most of the 130 driftwood logs sampled at Separation Point still retained t he root system and some bark. The main taxa were spruce (Picea) (80%) and poplar (Populus) (20%), although birches (Betula), alder (Alnus) and one tamarack (Larix) were also present. A t otal of 89 white spruce (Picea glauca) samples were measured and analyzed.

The driftwood logs from the Tuktoyaktuk - Peninsula Point area were more er oded than those at Separation Point. Some still retained the root but many were broken an d had no bark left. Of the samples, roughly 86% were spruce and 14% poplar. Some birches, al der, and tamarack were present but not sampled. From this area, 117 white spruce (Picea glauca) samples were measured and analyzed.

The tree rings in the driftwood samples were made more visible by cutting i nto the wood with a razor blade. Tree rings were measured in 1/100 mm units on an Aniol tree - ring measuring machine connected to a PC and analyzed using the CATRAS program (Aniol , 1983).

Two to four radii were measured in the driftwood samples and an average cur ve made for each tree/log. These tree curves were cross - correlated with each other using the CATRAS program's statistical technique (Aniol, 1983).

The CATRAS program calculates the percentage of agreement coefficients (sig n test) for all positions of overlap between two chronologies, taking no account of the magnitud es of the year - to - year changes in ring width. When the program calculates the t - value, a simple standardization is carried out where each ring width is converted to a percentag e of the mean in order to remove any trends from the basic data (Baillie and Pilcher, 1973).

Those tree curves showing high correlation values (t - and sign test values ) were visually checked by comparing the graphical plots of the tree curves. The best fitted we re used to build up mean chronologies from the driftwood samples. All chronologies were quality controlled by the COFECHA program (Holmes et al., 1986).

Absolute chronologies were built up for the living tree sites, one chronolo gy for each locality (Table 1). The statistical parameters are explained in Fritts (1976).

Additionally, chronologies from Giddings (1947), Cropper and Fritts (1981), Jacoby and Cook (1981) and Parker et al. (1973) proved useful for dating the driftwood tree curves and chronologies (Table 2).

The Greenland driftwood tree curves were standardized with a negative expon ential function to smooth the growth trends in width values (Holmes et al., 1986).

RESULTS

The Picea samples of the Mackenzie driftwood are divided into four groups b ased on possible origin and correlation with master chronologies.

The first group consists of long - living trees from the Tuktoyaktuk - Peni nsula Point driftwood collection (Fig. 1:D1). They have grown slowly, with an average tree - ring width as low as 0.22 mm and with up to 600 tree rings. The average tree - ring width does not generally decrease with age in these samples (no growth curve produced), indicat ing that the most important parameter controlling the radial growth in the trees is the tempe rature. According to Fritts (1976), the growth curve flattens as the tree becomes more s tressed. For a tree on a limiting site near the tree line the effects of the growth curve wou ld be very small because environmental parameters other than temperature decrease in importance a nd the temperature becomes the predominant growth factor (Fritts, 1976).

These long - living trees could be internally correlated to build up a floa ting chronology based on seven equally long driftwood tree curves (Table 3). This chronology co uld be synchronized with the living tree chronologies from the delta, giving high corre lation values with the delta mean chronologies (Table 4), with its end year dated to 1986. Figure 5 shows the Tuktoyaktuk - Peninsula Point driftwood chronology plotted against the chronolog y from south of Inuvik. From the plot it can be seen that the average treering width is lowe r in the driftwood chronology, indicating that the trees constituting the Tuktoyaktuk - Peninsula P oint chronology most probably originated in the upper delta area, close to the northern limit of the woodland tundra.

The second group consists of driftwood samples from Separation Point (Fig. 1:D2), showing high internal correlation values. These samples were internally cross - correla ted and six trees were used to construct a floating chronology (Table 3). This chronology could b e synchronized with the chronologies from the delta and its surroundings. It did not give as h igh correlations with the delta chronologies as the Tuktoyaktuk - Peninsula Point curve did, but the Separation Point chronology is still easily datable, with its end year 1989 (Table 5). Fig ure 6 shows the Separation Point driftwood chronology plotted against the chronology from south of Inuvik.

Trees giving the driftwood chronology from Separation Point have their orig in in the area south of the delta, possibly close to the sampling site. In this chronology, 40 driftwood samples from Separation Point could be dated, which is 45% of the analyzed samples from that area. Figure 7 shows the year of death of the dated samples from Separation Point. Wi th the help of the Separation Point chronology, seven samples from the Tuktoyaktuk - Peninsula Point driftwood collection could also be dated, which is 6% of the analyzed samples fr om that area.

The third group of driftwood consists of relatively young trees from both o f the sampling sites. These trees produced a growth curve indicating growth under more favoura ble environmental conditions than those trees growing in the delta area. Wood densi ty is lower than in older trees and reaction wood is more common. Internal cross - correlation i s not as common as in the first two groups, but it was possible to build up four floating chrono logies, with 3 - 7 trees in each (Table 3). Two of these chronologies gave relatively high correla tion values with a mean chronology from the Nahanni Butte area (Fig. 1) (from Parker et al., 1973 ), and the driftwood might therefore have its origin in the drainage area of the Liard Rive r.

The fourth group consists of driftwood samples where not more than two tree s can be joined into floating mean curves and which are not datable via any available chronology . A common feature of these samples are relatively few tree rings, 55 - 110, and compressio n wood is common, caused by tilting. These samples are therefore not good for tree - ring studies. Tilting occurs in trees that grow on unstable creeping slopes close to rivers, into whic h they eventually fall. From the pattern in the outermost tree rings it can be seen in many cases that the tree underwent stress before it fell into the river.

Many ice jams build up during ice breakup in the river and subsequent flood ing causes erosion (Kindle, 1921). Many driftwood logs have scars that they probably recei ved during such events.

When local people were asked if they knew the main origin of the driftwood, they usually thought it to be the Liard River or British Columbia. This is probably correct, because in the Rocky Mountains erosion is greater than in areas farther downriver. The ecologi cal conditions in the Liard River drainage area seem to be different from those in the delta. The trees are much younger, have broad tree rings, and produce a growth curve, whereas the del ta trees are older, with narrow tree rings and little evidence of a growth curve (Fig. 8).

North American Driftwood in the Greenland Sea

Driftwood can be found on the coasts of most islands in the north Atlantic Ocean. Most of this wood comes with the East Greenland Current via the Arctic Ocean (Fig. 9), p robably to a large extent transported frozen into the polar pack ice (Haggblom, 1982). As th e pack ice melts, the wood begins to float in open water until washed ashore on Greenland, Jan May en, Iceland, and Svalbard. Arctic driftwood is sometimes deposited on the Faroe Islands.

Driftwood collections from Greenland, Svalbard, and Iceland have been analy zed by the dendrochronological method.

The tree - ring pattern of six Picea driftwood samples from three localitie s on Greenland show high correlation values with tree - ring chronologies from Alaska or Canada but none of the samples from the driftwood collections on Iceland and Svalbard, although mor e than 200 driftwood logs have been analyzed from each of these localities. A dendrochrono logical and wood anatomical analysis of the samples from Iceland and Svalbard indicate their origin to be in the boreal and subboreal forests of Russia (Bartholin and Hjort, 1987; Eggert sson, 1992). To a large extent, the Greenland driftwood is probably also of Russian origin.

The first driftwood locality on Greenland is on Prinsesse Ingeborg Peninsul a (Fig. 9A) in the far north (81Degree30'N; collected by Svend Funder, Geological Museum, Copen hagen, 1988). From this collection two spruce samples could be dated in American chron ologies. One had a treering pattern showing that it came from the Mackenzie delta (Fig. 10) a nd another one showing that it came from Alaska (Fig. 11). The outermost ring on each sample w as dated to 1880 and 1923 respectively. The correlation values are shown in Tables 6 and 7.

The second is from Constable Pynt in Scoresby Sund along the central part o f the coast (Fig. 9B) (70DegreeN; collected by Christian Hjort, University of Lund, 1990). One sp ruce sample from this area could be dated via chronologies from the Mackenzie delta (Fig. 12 ) and one via chronologies from Alaska (Fig. 13), giving the date of the outermost ring as 188 8 and 1905 respectively. Tables 6 and 7 show the correlation values of the two samples.

The third driftwood locality was in the Godthab area on the southwest coast of Greenland (Fig. 9C), collected by Holst in 1880 (Ortenblad, 1881). (This collection belon gs to the Historical Museum in Stockholm.) Two driftwood samples from this collection coul d be dated, one in chronologies from the Mackenzie delta (Fig. 14) and one in chronologies f rom Alaska (Fig. 15), giving the date of the outermost ring as 1854 and 1831 respectively. The correlation values are shown in Tables 6 and 7. Figure 16 shows the tree - ring chronology s ites in Alaska.

Because North American driftwood has only been identified in driftwood coll ections from Greenland and not in Iceland and Svalbard, the possibility exists that North Ame rican driftwood is mainly transported southwards along the western flank of the East Greenland C urrent, entering it from the northwest (Fig. 9).

SUMMARY AND DISCUSSION

In summary, 290 driftwood samples were collected from the Mackenzie area. Of them, 206 Picea samples were measured and analyzed, of which 60 (29%) could be dated in ch ronologies from the delta area. The remaining 146 samples (71%) could not be absolutely da ted, possibly because of a lack of master chronologies from the Liard River drainage area and because of relatively few tree rings in some samples.

Six driftwood samples from the coast of Greenland could be shown to have a North American origin, three dated via tree - ring chronologies from the Mackenzie del ta area and another three with chronologies from Alaska. None of the driftwood samples from the collections in Svalbard and Iceland has proved to be of American origin. This i ndicates that although some American driftwood is transported from the Beaufort and Bering sea s to the North Atlantic and deposited on the coast of Greenland, American driftwood probably do es not reach the islands in the central and eastern North Atlantic.

Dendrochronology on driftwood is more complicated than dealing with living trees. The researcher has only the wood sample in his hands but does not know the ecologica l conditions in the forest of origin. The known parameters are only the species, the samplin g locality and the tree - ring pattern.

The species may tell about the possible origin. Pinus does not occur in th e Mackenzie material and does not grow in the watershed of the northward - and westward - fl owing rivers in Alaska (Hustich, 1966). This means that Pinus driftwood logs found on the arc tic shores most probably do not have their origin in the boreal forests of North America. The sa me can generally be said about Larix. Only one sample out of 290 from the Mackenzie delta was La rix, and according to Giddings (1953), it is a minor constituent of the Yukon river drift wood -- but it is the dominant tree genus of eastern Siberia. The Picea species dominate in the A merican driftwood (Picea glauca) and are also common in western Siberia (Picea abovata) (Hustich, 1966). These two species cannot be differentiated by the anatomy of the wood.

The sampling site locality can often reveal the origin of the sample. Drift wood samples collected at Separation Point have their origin in the drainage area of the Mack enzie River, upstream from the delta. Samples from Peninsula Point came from all over the Ma ckenzie drainage area. When dealing with samples collected in Greenland, Svalbard, and Iceland, a priori we only know that they originated somewhere in the circumpolar boreal for est regions.

The tree - ring patterns can tell us much more. With the help of master ch ronologies, the origin of the driftwood and the year when the wood began its drift can often be identified. This can only be done on a more regular basis if a good network of master chronologie s from around the Polar Basin is available. The Mackenzie River chronologies presented here s hould be useful in identifying Mackenzie driftwood along the shores of the Arctic and Subarctic. Thus the chronologies will help in mapping the pattern and velocity of the ocean currents , the drift pattern of the polar pack ice and, as a possible extension, climatic fluctuations mirror ed by changes in ice conditions indicated by varying driftwood frequencies.

TABLE 1. Statistics of the mean tree - ring curves from living Picea glauc a in the Mackenzie delta (sites shown on Fig. 1; terminology according to Fritts, 1976)

Fort N of Inuvik S of Inuvik McPherson (S1) ( S2) (S3)

Time span 1700 - 1989 1540 - 1989 1894 - 1989 Average mean sensitivity 0.180 0.197 0.143 Series intercorrelation 0.5 80 0.598 0.420 Autocorrelation 0.766 0.720 0.768 Number of trees 17 21 8

TABLE 2. Tree - ring chronologies from the Mackenzie drainage area (sites shown on Fig. 1)

Site Location Time span

N of Inuvik (S1) 68Degree25'N - 133Degree48'W 1700 - 1989 S of Inuvik (S2) 68Degree18'N - 133Degree45'W 1540 - 1989 Fort McPherson (S3) 67Degree28' N - 134Degree47'W 1894 - 1989 Spruce Creek 68Degree38'N - 138Degree38 'W 1570 - 1977 Delta chronology 68Degree00'N - 135Degree00'W 1700 - 1941 TT - HH 65Degree20'N - 138Degree20'W 1459 - 1975 Nahanni Butte area

61Degree00'N - 124Degree00'W 1800 - 1971

Table continiued...

Reference

N of Inuvik (S1) This paper S of Inuvik (S2) This paper Fort McPherson (S3) This paper Spruce Creek Cropper and Fritts, 1981 Delta chronology Giddings, 1947 TT - HH Jacoby and Cook, 1981 Nahanni Butte area Parker et al., 1973

TABLE 3. Driftwood chronologies from this study

Number of Number of Sensitivity Internal Time Possible origin trees rings correlation span of the driftwood

7 387 0.188 0.505 1600 - 1986 The delta area 6 169 0.173 0.546 1821 - 1989 S of the delta 3 157 0.190 0.5 14 1831 - 1987? Liard River area 5 105 0.208 0.500 1883 - 1987? Liard River area 5 139 0.217 0.566 1 - 139 ? 7 120 0.2 14 0.550 1 - 120 ?

TABLE 4. Correlation values of the Tuktoyaktuk - Peninsula Point (D1) driftwood chronology with chronologies from the Mackenzie delta (Fig. 1)

Tuktoyaktuk - Peninsula

Point chronology Site t Sign test (%)

N of Inuvik (S1) 13.12 73.2 S of Inuvi k (S2) 17.91 76.6 Fort McPherson (S3) 2. 58 63.0 Spruce Creek 15.50 70.4 Delta chro nology 19.99 78.4 TT - HH 5 .77 62.9 Separation Point (D2) 4.93 65.5

TABLE 5. Correlation values of the Separation Point driftwood chronology with c hronologies from the Mackenzie delta and surroundings

Separation Point

driftwood chronology Site t Sign test (%)

N of Inuvik (S1) 8.38 73.2 S of In uvik (S2) 7.11 69.0 Fort McPherson (S3) 4.31 67.9 Spruce Creek 5.62 67.9 Delta chronology 6.66 70.4 TT - HH 3.30 62.0 Tuktoyaktuk - Peninsula Point driftwood chronology (D1) 4.93 65.5

TABLE 6. Correlation values of driftwood samples from Greenland via chronologie s from the Mackenzie delta area

Ingeborg Peninsula, Constable Pynt, Greenland (A) (Fig. 10) Greenland (B) (Fig. 12) Site (see Fig. 1) t Sign test ( %) t Sign test (%)

N of Inuvik (S1) 5.23 64.2 7.86 67.3 S of Inuvi k (S2) 5.93 70.9 5.73 65.5 Spruce Creek 4.98 63.8 7.17 63.2 Delta chronology 6.10 63.8 8.26 64.6 TT - HH 4.02 65.0 3.73 62.2 Tuktoyaktuk - Peninsula Point driftwood chronology (D1) 5.10 66.5 11.06 70.8

Table continued...

Godthab, Greenland (C) (Fig. 14) Site (see Fig. 1) t Sign test (%)

N of Inuvik (S1) 4.52 63.2 S of Inuvik (S 2) 5.26 62.1 Spruce Creek 6.45 65.8 Delta chronology 6.08 63.0 TT - HH 3.43 60.7 Tuktoyaktuk - Peninsula Point driftwood chronology (D1) 7.42 64.1

TABLE 7. Correlation values of driftwood samples from Greenland via chronologie s from Alaska 'Symbol not transcribed'1

Ingeborg Peninsula, Constable Pynt,

Greenland (A) (Fig. 11) Greenland (B) (Fig.13) Site (see Fig. 16) t Sign test (%) t Sign test (%)

Koyuk 6.02 62.9 Cape Darby

6.07 60.2 Nulato 4.42 66.3 Fort Yukon Stephens Village Yukon River 5.10 64.0 6.02 64.9 Alaska Ran ge 6.02 65.9 Nunivak Island driftwood 6.07 60.2 6.38 67.1 Kobuk - Ko tzebue sites 6.71 62.0

Table continued...

Godthab, Greenland (C) (Fig. 15) Site (see Fig. 16) t Sign test (%)

Koyuk Cape Darby Nulato Fort Yukon 4.90 64.4 Stephens Village 5.50 70.5 Yukon River 6.30 67.2 Alaska Range Nunivak Island driftwood Kobuk - Kotzebue sites

'Symbol not transcribed'1 Chronologies from Giddings (1948, 1953), Oswalt (1958) , Van Stone (1958), and Cropper and Fritts (1981).

REFERENCES

ANIOL, R. 1983. Tree - ring analysis using CATRAS. Dendrochronologia 1:45 - 53 .

BAILLIE, M.G.L., and PILCHER, J.R. 1973. A simple crossdating program for tree - ring research. Tree - Ring Bulletin 33:7 - 14.

BARTHOLIN, T., and HJORT, C. 1987. Dendrochronological studies of recent driftw ood on Svalbard. In: Kairukstis, L., Bednarz, Z., and Feliksik, E., eds. Methods of dendrochronology I. Warsaw: Polish Academy of Sciences, Systems Research Insti tute. 207 - 219.

BEARMAN, G. 1989. Ocean circulation. Oxford: The Open University, Pergamon Pr ess. 238 p.

BLAKE, W., Jr. 1972. Climatic implications of radiocarbon - dated driftwood in the Queen Elizabeth Islands, arctic Canada. In: Vasari, Y., Hyvan, H., and Hicks, S., ed s. Climatic changes in arctic areas during the last ten - thousand years. Acta Universitati s Ouluensis, Series A., No. 3., Geology No. 1:77 - 104.

CROPPER, J. P., and FRITTS, H.C. 1981. Tree - ring width chronologies from the North American Arctic. Arctic and Alpine Research 13(3):245 - 260.

EGGERTSSON, 'O, 1992. Driftwood in the Arctic, a dendrochronological study. In : Bartholin, T.S., Berglund, B.E., Eckstein, D., and Schweingruber, F.H., eds. Tre e rings and environment. Lundqua Report 34:94 - 97.

EUROLA, S., 1971. The driftwood of the Arctic Ocean. Report from the Kevo Suba rctic Research Station 7:74 - 80.

FRITTS, H.C. 1976. Tree - rings and climate. London: Academic Press. 567 p.

GIDDINGS; J.L. 1941. Dendrochronology in northern Alaska. Fairbanks: Universit y of Alaska Publication No. IV. 107 p.

1943. The oceans, a plan for mapping arctic sea currents. Geographical Review 33:326 - 327.

1947. Mackenzie River delta chronology. Tree - Ring Bulletin (13) 4:26 - 29.

1948. Chronology of the Kobuk - Kotzebue sites. Tree - Ring Bulletin 14(4):26 - 32.

1952. Driftwood and problems of arctic sea currents, Proceedings of the America n Philosophical Society 96(2):131 - 142.

1953. Yukon river spruce growth. Tree - Ring Bulletin 20(1):2 - 5.

HAGGBLOM, A. 1982. Driftwood in Svalbard as an indicator of sea ice conditions.

Geografiska Annaler 64A:81 - 94.

HOLMES, R.L., ADAMS, R.K., and FRITTS, H.C. 1986. Tree - ring chronologies of w estern North America: California, eastern Oregon and north Great Basin. Chronology Se ries Vl. Tucson: University of Arizona. 40 p.

HUSTICH, I. 1966. On the forest - tundra and the northern tree lines. Annales U niversitatis Turkuensis A II(36). Reports from the Kevo Subarctic Research Station 3:7 - 47.

JACOBY, G.L., and COOK, E.R. 1981. Past temperature variations inferred from a 400 - year tree - ring chronology from Yukon Territory, Canada. Arctic and Alpine Research 13:409 - 418.

KINDLE, E.M. 1921. Mackenzie River driftwood. Geographical Review 11:50 - 53.

MACKAY, J.R. 1963. The Mackenzie delta area, N.W.T. Ottawa: Geographical Branc h Memoir 8. 202 p.

ORTENBLAD, V.T. 1881. Om Sydgronlands drifved. Bihang till Kungliga Svenska Vetenskapsakademiens forhandlingar 6(10):3 - 34.

OSWALT, W. 1958: Tree - ring chronologies in south - central Alaska. Tree - Rin g Bulletin 22(1 - 4):16 - 22.

PARKER, M.L., BRAMHALL, P.A., and JOHNSON, S.G. 1983. Treering dating of driftwood from raised beaches on the Hudson Bay coast. In: Harington, C.R. Cli matic change in Canada 3. Ottawa: National Museums of Canada. Syllogeus 49:220 - 272.

PARKER, M.L., JOZSA, L.A., and BRUCE, R.D. 1973. Dendrochronological investigat ion along the Mackenzie, Liard and South Nahanni River, N.W.T. Part II. Technical R eport to Glaciology Division, Water Resource Branch. Vancouver: Faculty of Forestry, Uni versity of British Columbia. 74 p.

RITCHIE, J.C., and HARE, F.K. 1971. Late - Quaternary vegetation and climate ne ar the arctic tree - line of northwestern North America. Quaternary Research 1:331 - 34 1.

SALVIGSEN, O. 1981. Radiocarbon dated raised beaches in Kong Karls Land, Svalba rd, and their consequences for the glacial history of the Barents Sea area. Geografiska Annaler 63(A):283 - 291.

VAN STONE, J.W. 1958. The origin of driftwood on Nunivak Island, Alaska. Tree - Ring Bulletin 22(1 - 4): 12 - 15.

Copyright Arctic Institute of North America Jun 1994
Provided by ProQuest Information and Learning Company. All rights Reserved