lable at ScienceDirect

Quaternary International 235 (2011) 13e25

Contents lists avai

Quaternary International

journal homepage: www.elsevier .com/locate/quaint

Dinoflagellate cyst-based reconstructions of mid to late Holocenewinter sea-surface temperature and productivity from an anoxicfjord in the NE Pacific Ocean

R. Timothy Patterson a,*, Graeme T. Swindles b, Helen M. Roe c, Arun Kumar d, Andreas Prokoph a,e

aOttawa-Carleton Geoscience Centre and Department of Earth Sciences, Carleton University, 1125 Colonel By Drive, Ottawa, ON, Canada K1S 5B6b School of Geography, University of Leeds, Leeds LS2 9JT, UKc School of Geography, Archaeology and Palaeoecology, Queen’s University Belfast, Belfast BT7 1NN, UKdCenter for Petroleum and Minerals Research Institute, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabiae SPEEDSTAT, 19 Langstrom Crescent, Ottawa, ON K1G 5J5, Canada

a r t i c l e i n f o

Article history:Available online 6 July 2010

* Corresponding author. Tel.: þ1 613 520 2600; faxE-mail address: tpatters@earthsci.carleton.ca (R.T.

1040-6182/$ e see front matter � 2010 Elsevier Ltd adoi:10.1016/j.quaint.2010.06.016

a b s t r a c t

Published contemporary dinoflagellate distributional data from the NE Pacific margin and estuarineenvironments (n¼ 136) were re-analyzed using Canonical Correspondence Analysis (CCA) and partialCanonical Correspondence Analysis (pCCA). These analyses illustrated the dominant controls of wintertemperature and productivity on the distribution of dinoflagellate cysts in this region. Dinoflagellatecyst-based predictive models for winter temperature and productivity were developed from thecontemporary distributional data using the modern analogue technique and applied to subfossil datafrom two mid to late Holocene (w5500 calendar years before presentepresent) cores; TUL99B03 andTUL99B11, collected from Effingham Inlet, a 15 km long anoxic fjord located on the southwest coast ofVancouver Island that directly opens to the Pacific Ocean through Barkley Sound. Sedimentation withinthese basins largely comprises annually deposited laminated couplets, each made up of a winterdeposited terrigenous layer and spring to fall deposited diatomaceous layer. The Effingham Inlet dino-flagellate cyst record provides evidence of a mid-Holocene gradual decline in winter SST, ending with theinitiation of neoglacial advances in the region byw3500 cal BP. A reconstructed Late Holocene increase inwinter SST was initiated by a weakening of the California Current, which would have resulted ina warmer central gyre and more El Niño-like conditions.

� 2010 Elsevier Ltd and INQUA. All rights reserved.

1. Introduction

The climate along the west coast of Vancouver Island, BritishColumbia, is strongly influenced by seasonal changes in the relativestrength of the North Pacific High (NPH) and Aleutian Low (AL),which results in predominantly upwelling conditions prevailingduring the spring through to fall and downwelling conditionsduring the winter months (Ware and Thomson, 2000). At sub-decadal and longer time scales upwelling and climate variability arefurthermodulated by solar influenced changes, which influence thetrack of the jet stream (Christoforou and Hameed, 1997), as wellperiodic occurrences of the equatorial El Niño/La Niña cycles(Schwing et al. 2002). Superimposed on this climate variability isthe decadal-scale Pacific Decadal Oscillation (Chavez et al., 2003;

: þ1 613 520 5613.Patterson).

nd INQUA. All rights reserved.

Patterson, et al., 2004; Ivanochko et al., 2008), which is in turninfluenced by the Gleissberg solar cycle (Patterson et al., 2004;Shen et al., 2006) as well as several poorly understood, regionallyinfluenced, decadal and centennial-scale cycles and trends thatarise from global-scale teleconnections (Bond et al., 2001;Patterson et al., 2004, 2005; Bridgman et al., 2006). As the instru-mental record for this region spans only the last century, theidentification and understanding of decadal to centennial-scaleclimate trends and cycles can only be achieved through analysis ofproxies preserved in the geologic record. Dinoflagellate cystsprovide such a proxy as they preserve well and are importantprimary producers, responding to a variety of oceanographic vari-ables including paleo-currents, winter and summer sea-surfacetemperatures (SST), upwelling and productivity (e.g. de Vernalet al., 2001; Kumar and Patterson, 2002; Radi and de Vernal,2004; Radi et al., 2007; Marret et al., 2008; Pospelova et al.,2008; Radi and de Vernal, 2008). Dinoflagellate relative abun-dance varies according to the season, and they are particularly

R.T. Patterson et al. / Quaternary International 235 (2011) 13e2514

prevalent/common in areas of upwelling due to the increasedavailability of nutrients (Tappan, 1980; Pospelova et al., 2008).

Recent research on dinoflagellate cyst assemblages from the NEPacific region has resulted in the development of predictive modelsthat have provided important data on temporal variation for severalparameters, principally primary productivity, SST and salinity (deVernal et al., 2005; Radi et al., 2007; Pospelova et al., 2008).

This paper details an analysis of dinoflagellate cyst protists fromtwo piston cores collected from Effingham Inlet, a fjord on thesouthwest coast of Vancouver Island (Fig. 1). There is a directresponseof the residentwaterswithinEffinghamInlet to the easternNorth Pacific as it is contiguous with the open ocean (Thomson,1981). The anoxic and dysoxic conditions that prevail at depthwithin this fjord thusprovide ideal conditions for thepreservationofannually deposited laminated sediments and their contained dino-flagellate cysts, which archive a continuous record of seasonal-scaleclimate change spanning much of the Holocene (Patterson et al.,2000, 2004, 2005). Utilizing existing dinoflagellate cyst distribu-tional data for the NE Pacific region (Radi and de Vernal, 2004; Radi

Fig. 1. (A) Location map of southern British Columbia showing geographic featuresdiscussed in this paper and location of Effingham Inlet on SW coast of VancouverIsland. (B) Details of Effingham Inlet showing location of piston cores TUL99B03 andTUL99B11.

et al., 2007), which encompass the environmental gradients in thisregion a modern analogue-based model will be developed to carryout a paleoceanographic reconstruction of conditions withinEffingham Inlet through the Late Holocene.

2. Background

2.1. Setting

Effingham Inlet, on the southwest coast of Vancouver Island, isa 15 km long fjord, directly linked with the NE Pacific Oceanthrough Barkley Sound (Fig. 1). The area around the inlet is steepand rocky and surrounded by coniferous forest covered highmountains (700e1200 m) comprised of Mesozoic volcanic rocks.The slopes of Effingham Inlet below the water line are also steepwith little sediment drape (Patterson et al., 2000). Isostatic andeustatic sea-level changes in response to deglaciation have influ-enced circulation patterns in coastal British Columbia (Patterson,1993; Guilbault et al., 1997; Patterson and Kumar, 2002). By thelate Holocene, the interval spanned in this study, the coastline hadreached it’s present outline, greatly reducing the impact of sea-level as a variable in core interpretation in Effingham Inlet(Dallimore et al., 2008; Manounos et al., 2009)

The inlet is divided into two sub-basins, a 120 m deep innerbasin and a 210 m deep outer basin, which are separated by a 40 mdeep ‘inner’ sill located within a narrow channel (Fig. 1). A second65 m deep ‘outer’ sill separates the outer basin from Barkley Soundand the open ocean (Fig. 1). The waters below the sill depths in theinlet are infrequently mixed resulting in deep water circulationbecoming progressively restricted toward the head of the inlet. Thedeep water beneath the sill level in the outer basin is dysoxic whilethe waters in the inner basin beneath the sill depth are anoxic,receiving only infrequent infusions of oxygen (Patterson et al. 2000,2004). Effingham River with a drainage basin of about 79 km2

flowsinto the head of the inlet, and a small stream flows into the outerbasin. During the winter rainy season several additional ephemeralsmaller creeks contribute a significant flow of water into the inlet(Chang et al., 2003).

2.2. Climate-oceanographic conditions

Holocene productivity changes in Effingham Inlet can bedirectly correlated to the relative thickness of seasonally depositeddiatomaceous ooze and terrigenous material, which in turn relateto the marked seasonal variability in NE Pacific atmosphere andocean circulation patterns (Fig. 2). The darker colored terrigenouslayers are deposited during the rainier wintermonths whenmarineproductivity is low, and are principally comprised of detrital grainsthat are washed into the inlet via Effingham River and the smallcreeks along the margins of the steep shoreline (Chang et al., 2003).These darker layers alternate with lighter colored laminations thatare deposited between May and August, which are composedprimarily of the skeletal remains of diatoms and other phyto-plankton such as dinoflagellates.

The darker winter deposited laminations are produced whenthe AL pressure system with its associated counter-clockwise windpattern dominates the region (Fig. 2). These winds promotedownwelling and result in the development of the northeast PacificCoastal Current, which flows northwestward off the coast(Thomson, 1981; Thomson and Gower, 1998).

The light-colored summer laminations are produced when thecounter-clockwise NPHwinds generate strong Ekman transport (upto 20e40 m3 s�1/km (NOAA)), which pushes surface water awayfrom the Vancouver Island continental margin resulting inupwelling and the establishment of the southward flowing Shelf-

PacificOcean

PacificOcean

1000

1005

1010

1015

1015

1015

1010

1005

1005

1025

1020

1020

1025

1020

1020

1020

1030

AL

NPH

140°E 180° 140°W 140°W

140°W

60°N

140°E 180° 140°W

40°N

20°N

60°N

40°N

20°N

60°N

40°N

20°N

JANUARY

Atmospheric pressure (mb)

JULY

Atmospheric pressure (mb)

H

H

H

L

60°N

40°N

20°N

SubarcticCurrent

AlaskaCurrent

CaliforniaCurrent

SubarcticCurrent

AlaskaCurrent

CaliforniaCurrent

Shelf-breakCurrent

NE Pacific Coastal Current

VICC

VICC

CoastalUndercurrent

CoastalUndercurrent

A

B

C

D

Effingham

Inlet

Effingham

Inlet

Fig. 2. Maps showing main oceaneatmosphere circulation features for the NE Pacific region in winter (A,C) and summer (B,D) months. In summer northerly North Pacific High(NPH) winds generate the southward Shelf-Break Current at the surface, and a consequent offshore Ekman transport inducing upwelling, thus bringing high-salinity deep watercloser to the surface. During winter southerly Aleutian Low (AL) winds generate a northward drift producing the NE Pacific Coastal Current and a consequent onshore Ekmantransport. This causes an accumulation of low density less saline water on the surface, thus restricting the upwelling of deep water. The Vancouver Island Coastal Current (VICC) andCoastal Under Current are permanent features although they vary in strength seasonally.

R.T. Patterson et al. / Quaternary International 235 (2011) 13e25 15

Break Current (Fig. 2). This upwelling along the shelf break advectsnutrient-rich but poorly oxygenated deep water on to the conti-nental shelf, where it is carried landward through a series ofsubmarine canyons (Thomson andWare, 1996;Ware and Thomson,2005). The Coastal Under Current, which flows north along thecontinental margin at a depth of about 300 m also contributespoorly oxygenated and nutrient-rich water to the southwest coastof Vancouver Island. Estuarine inflow through Juan de Fuca Canyontransports the nutrient-rich water originating from both wind-induced upwelling and under current advection into Juan de FucaStrait (Fig. 1). These nutrient-rich waters undergo strong verticalmixing within the strait before eventually becoming entrained ina nutrient-rich surface outflow toward the open ocean (Crawfordand Dewey, 1989). This nutrient-rich water, with additionalcontributions from upwelled continental shelf water and theCoastal Under Current then flow northward over the inner shelf asa component of the Vancouver Island Coastal Current. It is thiswater that subsequently enters Effingham Inlet and contributesnutrients essential to phytoplankton growth (Thomson, 1981;Mantua et al., 1997; Thomson and Gower, 1998; Chang et al., 2003).

Through time changing climatic conditions result in shifts inthe relative influence of the NPH and AL causing shifts in surfacewind stress, mixed layer depth and ocean stratification. Thesechanges in turn influence the temperature, nutrient content andsalinity of surface waters where phytoplankton growth is

concentrated (Brodeur and Ware, 1992). These water propertyvariations are the principle drivers of changes in primaryproduction, influencing assemblage composition as well as theintensity and timing of phytoplankton blooms (Ware andThomson, 1991; Radi et al., 2007).

Although diatoms are the principal phytoplankton in theregion, dinoflagellates contribute significant populations. Dinofla-gellate cyst abundance provides a direct reflection of local marineproductivity, while changes in the relative abundance of specifictaxa are linked to differences in SST, nutrient supply change andsalinity (Kumar and Patterson, 2002; Patterson et al., 2005; Radiet al., 2007).

3. Methods

3.1. Sample collection and laboratory treatment

The two cores analyzed were collected from Effingham Inlet,one each from the inner and outer basins, using the researchvessel CCG J.P. Tully in October 1999. Prior to coring a seismicsurvey of the inlet was carried out using a 3.5 kHz air gun toidentify suitable coring sites away from the basin slopes and torecognize areas where the top few meters of sediment wererelatively gas free. Piston core TUL99B03 (49�4.275, 125�09.359),collected from 120 m water depth in the inner basin, was 1130 cm

Fig. 3. Generalized stratigraphic column of core TUL99B03 showing generalized sedimentological variability (after Dallimore, 2001; Chang et al., 2003) and INTCAL04-calibratedradiocarbon ages (only bolded twigs/wood 14C ages are used in reconstruction) correlated to the Winter Sea-Surface Temperature reconstruction. Root mean squared errors (RMSE)were �0.91 �C and 38.37 g Cm�2 for winter temperature and productivity respectively. Error bars are derived from 1000 bootstrap cycles. Comparison is also made between climatestages and glacier advances (‘Tiedemann’, ‘Peyto’, ‘Robson’) on Canadian Pacific Coastline (after Pellat and Mathewes, 1997) and major global cooling events as also indicated bydashed horizontal lines (Bond et al., 2001).

R.T. Patterson et al. / Quaternary International 235 (2011) 13e2516

long (Fig. 3), while the 1016 cm long piston core TUL99B11(49�2.632, 125�09.230) was collected from 205 m water depth inthe outer basin (Fig. 4). Upon return to shore the cores were storedas 1.5 m long sections, split, slabbed and X-rayed using conven-tional mammography X-ray film and medical X-ray equipment(Dallimore, 2001). The cores were subsequently logged and sub-sampled to carry out a variety of sedimentological, geochemical,micropaleontological and palynological studies. Subsamplesutilized for dinoflagellate cyst research (Patterson et al., 2005)were collected at the same depth intervals as sediment samplesused for diatom analysis (Hay et al., 2007) and fish scale counts(Wright et al., 2005; Patterson et al., 2005).

A total of 97 subsamples were taken for dinoflagellate cystanalysis; 48 from core TUL99B03, and 49 from core TUL99B11. Thesampling interval was nearly equidistant in both cores, w24 cm.The 4 cc subsamples weremacerated and processed using standardpalynological preparation techniques (see Patterson et al., 2005).The processed residues weremounted between a slide and glyceringel with two slides being prepared for each sample. Each slide wasanalyzed under light microscopy for dinoflagellate cysts using anOlympus BX51 microscope, generally at 400�. Species identifica-tion was primarily based on Rochon et al. (1999) with additionaltaxa identified based on Zonneveld (1997; e.g. Echinidium) and

Chatton (1914; e.g. Polykrikos kofoidii). The abundance of variousdinoflagellate cyst taxa was determined as well as the ratio ofmarine to terrestrial palynomorphs and the ratio of gonyaulacoid toperidinioid (G:P ratio) dinoflagellate cysts (see appendices inPatterson et al., 2005).

3.2. Stratigraphy and chronology

3.2.1. Radiocarbon datingRadiocarbon dates presented are reported in calibrated radio-

carbon years before present (cal BP; Table 1; Figs. 3 and 4). Radio-carbon dates for shell and woodmaterial were provided by IsoTraceRadiocarbon Laboratory using the 14C calibration program (Stuiverand Reimer, 1993; Stuiver et al., 1998a,b) and calibrated using theINTCAL04 dendrochronological database for terrestrial material,and the MARINE04 database for marine material (Reimer et al.,2004; Stuiver et al., 2005). The calibrated marine shell dates werenot used in the determination of the sedimentation rate as themarine reservoir effect has not been definitively established forrestricted fjords such as Effingham Inlet on the west coast of Van-couver Island (R. Beukens, personal communication, 2002; Reimeret al., 2004; Marine Reservoir Correction Database e http://intcal.qub.ac.uk/marine/). Additional calibration difficulties arise with

Sedimentology

Non-laminated

Indistinct laminae

Well laminated

Disrupted texture (gravity flow)

Core TUL99 B11

0

100

200

300

400

500

600

700

800

900

1000

2460±210

2600±240

BC G

laci

ers

Age(cal yr BP)

Unadjusted Depth(cm)

inae

ed

Core TUL9

Glo

bal G

laci

alAd

vanc

esM

ajor

D

roug

ht

0

500

1000

1500

2000

2500

1740±260

1060±160

Rhizosolenia setigera, AD 1940Minidiscus chilensis, AD 1840Thalassionema nitzshioides, AD 1550

1

2

3

1

2

3

7 8 9 10°C

Winter temperature

Fig. 4. Generalized stratigraphic column of core TUL99B11 showing generalizedsedimentological variability (after Dallimore, 2001; Hay et al., 2007) and INTCAL04-calibrated radiocarbon ages (only bolded twigs/wood 14C ages are used in recon-struction) correlated to the Winter Sea-Surface Temperature reconstruction. Rootmean squared errors (RMSE) were �0.91 �C and 38.37 g Cm�2 for winter temperatureand productivity respectively. Error bars are derived from 1000 bootstrap cycles.Comparison is also made between climate stages and glacier advances (‘Tiedemann’,‘Peyto’, ‘Robson’) on Canadian Pacific Coastline (after Pellat and Mathewes, 1997) andmajor global cooling events as also indicated by dashed horizontal lines (Bond et al.,2001).

R.T. Patterson et al. / Quaternary International 235 (2011) 13e25 17

utilizing shell dates obtained from cores recovered from deepwaterrestricted circulation basins such as those characterizing EffinghamInlet. This is because the marine model ages utilized in the marinecalibration dataset are only valid for the surface mixed layer, whichgenerally excludes samples from waters deeper than 75 m(P.J. Reimer, personal communication, 2008). Sedimentation ratesin the inner and outer basins of Effingham Inlet varied considerably.The development of the age model for the two cores analyzed isthus discussed separately below.

Table 1Radiocarbon ages obtained for cores TUL99B03 and TUL99B11.

Core Laboratorynumber

Description Depth (cm) Adju

TUL99B03 TO-8671 Wood 97 97TUL99B03 TO-8672 Shell frag. 169 161TUL99B03 TO-8673 Twig 286 278TUL99B03 TO-8674 Twig 553 529TUL99B03 TO-8675 Shell frag. 822 757TUL99B03 TO-8676 Wood 937 842TUL99B11 TO-8683 Shell frag. 531 476TUL99B11 TO-8684 Shell frag. 898 797TUL99B11 TO-8685 Wood 939 837TUL99B11 TO-8686 Shell frag. 969 867

3.2.2. Inner basin core TUL99B03Most of core TUL99B03 was comprised of laminated sediments

andwas deposited under primarily anoxic conditions (Fig. 3). It wasindependently determined that these laminations were depositedannually based on 14C and detailed diatom analysis (Chang et al.,2003). Core recovery was nearly complete (w97%) with the sedi-ments being predominantly made up of a mixture of diatom oozeand unconsolidated brown mud. Approximately 80% of the corewas composed of laminated sediment couplets as described above,with w18% comprising massive units. An additional w2% of thecore was composed of distorted laminae and graded beds. Themajority of the laminated couplets were found between w145 cmand w920 cm, ranging in thickness from 0.15 to 0.8 cm. The inter-vals above w145 cm and below w920 cm were characterized bythick sequences (>20 cm)without laminations. Particularly notablewas a significant debris flow near the base of the core. Overallthough there were relatively few instantaneous debris flow units,which had to be subsequently subtracted out to develop an accu-rate age model (Dallimore, 2001; Prokoph and Patterson, 2004;Patterson et al., 2004). The X-ray images of the core were scan-ned at 256 grey-scale resolution to generate a 130,787 pixel longand three pixels wide line-scan of the core (116 pixels¼ 1 cm).Approximately 12% of the image data had to be linearly interpo-lated due to intervals of poor quality imagery and lost coresegments. A complete and detailed description of the methodologyused for extracting the grey-scale values and lamination thicknessdata from TUL99B03 is found in Patterson et al. (2004) and Prokophand Patterson (2004).

The agemodel developed for TUL99B03 is in part based on threeaccurate radiocarbonwood dates, each of whichwas independentlycorrelated to each other by computer analysis of laminae countsand unit thickness data derived from X-ray interpretation(Patterson et al., 2004; Prokoph and Patterson, 2004; Fig. 5).Detailed analysis of a large number of wood-based radiocarbondates from a 40.9 m (Calypso) piston core (MD02-2494) collectedfrom nearby in the inner basin indicates that plant materialprovides a reliable indication of sediment age in Effingham Inlet(Dallimore et al., 2008).

Thickness measurements of the laminae couplets comparedagainst the accurate wood/twig 14C dates indicate that the sedi-mentation rate was consistent atw0.25 cm/year through the lowerw9.5 m of the core and decreased to w0.15e0.20 cm/year throughthe upper 1.8 m of core. Superimposed on the overall thicknesstrend were variations in lamination thickness of between 0.15 cmand 0.8 cm over intervals of <5 cm, indicating that short-termfluctuations in sedimentary rate could vary by up to a factor of 5(Prokoph and Patterson, 2004; Patterson et al., 2004). Witha sample spacing of w24 cm the dinoflagellate cyst temporalresolution obtained for TUL99B03 was w96 years.

sted depth (cm) 14C age Calibrated age ranges (yr BP)

IntCal04 MARINE04

DR¼ 330� 40

160� 40 145� 851770� 60 990� 902050� 70 2010� 1502830� 60 2930� 1503890� 90 3450� 1004190� 80 4695� 1802460� 90 1740� 2602830� 60 2325� 3452570� 100 2600� 2401820� 60 1060� 160

0 1000 2000 3000 4000 5000 6000Age (cal yr BP)

0 1000 2000 3000Age (cal yr BP)

200

0 0

400

600

800

1000

1200

Dep

th (c

m)

200

400

600

800

1000

1200

Dep

th (c

m)

+95%-95% +95%-95%

TUL99B03 TUL99B11

Minidiscus chilensis

Minidiscus chilensis

Thalassionema nitzschioides

Rhizolenia setigera

Refined age modelDiscarded radiocarbon dateShell-based radiocarbon dateTwigs/wood radiocarbon dateDiatom biomarker

Thalassionema nitzschioides

Fig. 5. Ageedepth model (linear regression as bold straight line) of cores (A) TUL99B03 and (B) TUL99B11. Age modes are based on 14C ages from twigs/wood (shaded squares) andbiostratigraphic markers (circles; see Hay et al., 2007). Other ages (see Table 1) are passively plotted (unfilled and black squares). 95% confidence interval of regression line is basedon error in 14C ages (dotted lines). Fine solid line. Refinement of ageedepth model based on counts and thickness of varves determined from X-ray images (for detail, see text).

R.T. Patterson et al. / Quaternary International 235 (2011) 13e2518

In an analysis of diatoms from this core Hay et al. (2007) dis-carded the lowermost radiocarbon wood date (TO-8676; Table 1)suggesting that the sample had been reworked. However,comparison of this datum with lamination counts using the othertwo wood-based radiocarbon dates as tie-points indicates that TO-8676 falls well within the 95% confidence interval of the indepen-dently generated regression curve, indicating that this radiocarbondate is indeed reliable (Prokoph and Patterson, 2004; Fig. 5). Hayet al. (2007) also proposed that over-penetration of coreTUL99B03 below the sediment/water interface was less thanpreviously supposed based on a single occurrence of the diatomMinidiscus chilensis Rivera 1984 near the top of the core, which wasused as a biostratigraphic marker (Hay, 2005). Hay placed thisdatum at AD 1860 based on comparison with results from othercores in the inlet where the species is often quite abundant. Weaccept the integration ofM. chilensis as a biostratigraphic marker incore TUL99B03. We also accept the proposed use of the diatomThalassionema nitzschioides Grunow ex. Mereschkowsky 1902 asa biomarker in core TUL99BO3 (found above 87 cm, first appear-ance AD 1550) as this well-known and large species appears atmany horizons in the core (Hay, 2005).

It was hoped that independent confirmation of the age modelcould be established by direct comparison with the age modeldeveloped for nearby Core MD02-2494. Comparison of TUL99B03with various age models developed for MD02-2494 (e.g. Dallimoreet al., 2008; Ivanochko et al., 2008) has proven to be problematic asthe MD02-2494 core was unfortunately obtained from an area ofthe inner basin that was subject to a large number of slumpingevents through the late Holocene, which has thus far made corre-lationwith the higher quality laminated record of TUL99B03 nearlyimpossible (A. Chang, personal communication, 2009).

3.2.3. Outer basin core TUL99B11Although core TUL99B11 contained several extended intervals

of laminated sediments, the core was also punctuated by numerousmassive intervals as well as several units deposited as instanta-neous debris flows (Dallimore, 2001; Fig. 4). The potential todevelop a high-resolution stratigraphy based on this core is thusinferior to the record contained within TUL99B03. In order todevelop a continuous ageedepth record, sedimentary units

confirmed to be the result of instantaneous gravity flow eventswere subtracted from the total core length prior to development ofthe age model (Hay et al., 2007). There was unfortunately onlya single wood radiocarbon date (TO-8685 dated at 2600� 240calibrated years before present (cal BP) from adjusted core depth867 cm) available to utilize in the age model (Fig. 5). This problemwas partially offset as Hay et al. (2007) proposed that three diatombiomarkers identified from near the top of the core might be usedto develop a reasonable age model for the core (Fig. 5). Thesebiomarkers were chronologically constrained based on previousanalyses of freeze and kasten cores from Effingham Inlet andincluded Rhizosolenia setigera Brightwell 1858, which appeared caAD 1940 (34 cm core depth), M. chilensis Rivera 1984, whichappeared ca AD 1840 (106 cm core depth) and a morphotype of T.nitzschioides Grunow ex. Mereschkowsky 1902, which appeared caAD 1550 (208 cm core depth; Hay, 2005). Based on linear regressionresults of the adjusted core length the average sedimentation ratefor this core was w0.32 cm/year. As the average sample spacing inthis core is w24 cm the dinoflagellate cyst temporal resolutionobtained for TUL99B03 was w72 years.

3.3. Statistical analysis

Contemporary dinoflagellate and environmental datasets fromthe NE Pacific margin (n¼ 76; Radi and de Vernal 2004) andestuarine environments (n¼ 60; Radi et al., 2007) were combinedto produce a training set comprised of several parameters encom-passing wide environmental gradients across the region (n¼ 136).We acknowledge that this data has been analyzed previously usingprincipal components analysis (Radi et al., 2007), but decided thata re-analysis of the data following the suggestions of Telford (2006)was beneficial. Canonical Correspondence Analysis (CCA) wastherefore used to explore the relationships between dinoflagellatecyst taxa and environmental variables using CANOCO version 4.5and graphed using CANODRAW version 4.0 (ter Braak, 1987, 2002;�Smilauer, 1992; ter Braak and �Smilauer, 2002). The chi-squaremeasure used in CCA has been criticised as it can give rare speciesa large influence on the analysis (Faith et al., 1987; Legendre andLegendre, 1998). However, CCA was used in this case as thespecies dataset contained many null abundances and the chi-

R.T. Patterson et al. / Quaternary International 235 (2011) 13e25 19

square distance should provide a good approximation for specieswith unimodal distributions along an environmental gradient (terBraak, 1985; Legendre and Gallagher, 2001). The downweightingof rare species caused negligible changes to the output of the CCA,suggesting that this was not a major problem in this case.

Pearson correlation analysis was used to determine the inter-correlations between environmental variables and a series ofpartial Canonical Correspondence Analyses (pCCA) (Borcard et al.,1992) were used to assess which variable(s) have a significantindependent effect on assemblage composition (cf Telford, 2006).Monte-Carlo permutation tests (499 random permutations) wereused to determine the statistical significance of these analyses, (e.g.Dale and Dale, 2002).

Predictive models were developed using the modern analoguetechnique (MAT), following the procedures described in de Vernalet al. (2005) using the C2 software package (Juggins, 2003). MAThas been shown to be one of the most reliable methods for pale-oenvironmental reconstruction and allows the reconstruction ofseveral parameters (Guiot and de Vernal, 2007; McKay et al., 2008).Hydrographic estimates were calculated from a series of 5 modernanalogues selected from the reference database after logarithmictransformation of the relative abundances of taxa (de Vernal et al.,2005). Reconstructions of winter temperature (�C) and productivity(g Cm�2) were carried out on the dinoflagellate cyst data fromEffingham Inlet core TUL99B03 from the inner basin and from coreTUL99B03 from the outer basin.

4. Results

4.1. Dinoflagellate cyst down-core distribution

The overall dinoflagellate cyst assemblage composition wassimilar both within and between cores TUL99B03 and TUL99B11.Observed cyst assemblages were characterized by moderately highspecies diversity with most samples containing between 14 and 18taxa although a few samples yielded as few as 10 taxa and others asmany as 21 species.

All samples quantified contained a usually well preserved anda diverse assemblage of dinoflagellate cysts. Due to problems withpreservation in some samples it was not always possible to assignspecimens to known dinoflagellate cyst genera or species. Thesespecimens could therefore not be included in subsequent statisticalanalysis. In addition to dinoflagellate cysts, other groups of paly-nomorphs observed included terrestrially-derived pollen andspores and fungal remains along with various types of organicdebris. A large number of foraminifera, thecamoebians, tintinnidsand copepod eggs were also observed (Patterson et al., 2005).

The terrestrial to marine palynomorph ratio (T/M ratio) rangedfrom a low of 0.35 (TUL99B11, 40e41 cm) to a high of 1.96(TUL99B03, 992e993 cm). In most samples this ratio was <1.

Peridinioid cysts comprised a significant proportion of theassemblage, primarily due to the abundance of Brigantedinium spp.and round brown cysts (RBC). For example, the ratio betweengonyaulacoid and peridinioid dinoflagellate cysts generally rangedfrom a low of 0.007 (TUL99B03, 300e301 cm) to a high of 0.84(TUL99B03, 758e759 cm). In an anomalous sample from TUL99B11,40e41 cm, where the gonyaulacoid dinoflagellate cyst Oper-culodinium spp. formed w75% of the sample assemblage the ratiowas 2.89.

Among the gonyaulacoid dinoflagellate cysts, Operculodiniumcentrocarpum and related morphotypes were quite abundant,generally ranging between w2% and w20% of the assemblage, andcomprising as much as w45% of the dinoflagellates in TUL99B03,758e759 cm and up to w75% in TUL99B11, 40e41 cm. Among theobserved peridinioid dinoflagellate cysts the relative abundance of

several taxa, particularly Brigantedinium spp., Islandinium minutum/spiny round brown cysts and round brown cysts were important. I.minutum/spiny round brown cysts varied between w2 and w17% inmost samples and occasionally reached abundances as high asw36%in TUL99B11, 184e185 cm, and w50% in TUL99B03, 300e301 cm.

Brigantedinium spp. are another group of protoperidinioid cystsrelated to RBC. All such forms showing an opening were groupedtogether as Brigantedinium spp. because the exact structure ofarchaeopyle could not be observed in all specimens. That said mostspecimens could probably be attributed to Brigantedinium simplexWall, 1965 ex Lentin andWilliams, 1993 as they were characterizedby a 2P hexagonal, equidimensional archaeopyle, as well as Brig-antedinium cariacoense (Wall, 1967) Lentin and Williams, 1993 fora number of specimens having a 2P hexagonal, and broader thanhigh archaeopyle.

All spherical brown protoperidinioid cysts without any visibleopening were classified as generic round brown cysts (RBC). Theyare a heterogeneous group of inaperturate, spherical, brown formsof various sizes having a psilate, verrucate to finely granulateautophragm, which have variously been assigned to dinocysts aswell as a variety of unknown/undifferentiated palynomorphs (Radi,pers comm. 2009). In addition Zonneveld (1997) assigned many ofthese forms to the dinoflagellate genus Echinidinum. The RBC wereexcluded from inclusion in the paleoenvironmental reconstructionas they probably represent multiple taxa with different tolerancegradients, which would negatively impact the reliability of thepredictive model.

Other peridinioid cysts including Lejeunecysta spp., Quinque-cuspis concreta, Selenopemphix nephroides, Selenopemphix quanta,Votadinium calvum, and Votadinium spinosum usually occurred asminor constituents of the dinoflagellate cyst assemblages. Theygenerally comprised <5% of the assemblage, but in few samplesthey formed/represented as much as w10%. Cysts of the gymno-dinioid dinoflagellate Polykrikos spp. also consistently occurred asminor assemblage constituents, generally making up <5% of theassemblage.

Therewere several pentagonal cysts observedwith a peridinioidmorphology. Unfortunately, due to poor preservation, insufficientmorphological details were present to place them in known taxa.All such morphotypes are thus referred to as ‘protoperidinioids’.These undifferentiated protoperidinioids were also excluded fromthe paleoenvironmental reconstruction using the same criteria asapplied to the RBC.

4.2. Analysis of the modern training set

CCA axis one explains 15.9% of the species data variance and56.0% of the specieseenvironment relationship variance and axesone and two together explain 21.9% of the species data variance and77.0% of the specieseenvironment relation variance (Table 2). Asa whole the measured environmental variables together explain28.5% of the variation in the dinoflagellate data. Monte-Carlopermutation tests illustrate the significance of axis one and allcanonical axes (p< 0.002). All of the environmental variables werecorrelated with axis one and several with axis two at the p< 0.01level (Fig. 6; Table 3). The dinoflagellate taxa are spread alongseveral of the environmental gradients and the samples from theNE Pacific margin and estuarine environments form two distinctgroups in the ordination, reflecting differences in salinity, temper-ature and productivity (Fig. 6). However, the environmental vari-ables are highly intercorrelated at the p< 0.01 level (Table 4), whichis a problem as it leads to a significant degree of redundancy in thedataset (Birks, 1995).

A series of pCCAs revealed that the environmental variablesexplain the following proportions of variance: winter temperature

Table 2Canonical Correspondence Analysis (CCA) results table.

Axes 1 2 3 4 Totalinertia

Eigenvalues 0.296 0.111 0.076 0.037 1.858Specieseenvironment correlations 0.798 0.744 0.699 0.449Cumulative percentage varianceOf species data 15.9 21.9 26.0 28.0Of specieseenvironment relation 56.0 77.0 91.3 98.3

Sum of all eigenvalues 1.858Sum of all canonical eigenvalues 0.529

Table 3Pearson r correlation coefficients between the environmental variables and CCAaxes one and two.

Variable Axis 1 Axis 2

Winter T 0.533** �0.322**Winter S 0.719** 0.283**Summer T 0.432** �0.510**Summer S 0.697** 0.269**Productivity �0.624** 0.218*

**p< 0.01, *p< 0.05.

R.T. Patterson et al. / Quaternary International 235 (2011) 13e2520

(12%, p< 0.002) summer temperature (8%, p< 0.004) winter salinity(8%, p< 0.004) summer salinity (3%, p< 0.0940) and productivity(10%, p< 0.002). The remaining 59% is due to intercorrelationsbetween the environmental variables (Fig. 7). Focus was subse-quently on the development of MAT models for winter temperatureand productivity as pCCA showed them to be independent, signifi-cant environmental controls on the assemblage composition (cfTelford, 2006). The MAT models have root mean squared errors(RMSE) of �0.91 �C and 38.37 g Cm�2 for winter temperature andproductivity respectively. Reconstructions of winter temperatureand productivity by MAT are shown in Figs. 8 and 9.

Fig. 6. (A) CCA biplot showing taxa and environmental data (B) CCA biplot showingthe samples and environmental data ordination. Open diamonds are estuarine samples(Radi et al., 2007), filled squares are samples from the NE Pacific margin (Radi and deVernal, 2004). See Table 5 for definitions of dinoflagellate cyst taxon codes.

5. Discussion

5.1. Controls on dinoflagellate cyst distribution in NE Pacific

The eastern margin of the NE Pacific region is characterized bytemperate waters, except for a pocket of cold water in the Gulf ofAlaska. In this region winter SST ranges from 4 to 11.4 �C alonga NeS gradient. Summer SST varies from 12 to 19 �C along an EeWgradient with the coldest waters being found along the coast whereNPH influenced coastal upwelling conditions prevail (Thomson,1981). The analyses presented here show that winter temperatureand productivity are the most statistically significant controls onthe distribution of dinoflagellate cyst assemblages in this region.Telford (2006) suggested that winter SST is not an importantenvironmental control on dinoflagellates and that summer SST asa control is more ecologically meaningful. However, Radi and deVernal (2004) found that there was no clear relationship betweensummer SST and the dinoflagellate cyst assemblages in the NEPacific as upwelling in summer is stronger in the southern coastalregion, resulting in only a very limited temperature gradient overthe study area.

Radi et al. (2007), using principal components analysis, sug-gested that productivity associated with seasonal upwellingconstituted the primary factor controlling the distribution patternof dinocyst assemblages and the distribution pattern followeda latitudinal gradient reflecting the importance of winter SST in thenorth-eastern Pacific margin. However, no variance partitioninganalysis was carried out to quantify the proportion of varianceexplained by each variable in this case. The results of pCCA on thecombined datasets presented here unequivocally shows thatwinter temperature explains a higher proportion of varianceoverall, with productivity a close second.

5.2. Late Holocene Winter SST trends

There has been considerable temporal variation in winter SST inthe NE Pacific through the last w5500 years based on paleo-temperature reconstructions derived from cores TUL99B03(w5500e500 calibrated radiocarbon years before present (cal BP))(Fig. 3) and TUL99B11 (w3000 cal BP-present) (Fig. 4). For coresTUL99B03 and TUL99B11 the sampling resolution of w96 and w72years respectively was too low to detect decadal-scale climatecyclicity that might be correlated with the PDO or cyclic pulses in

Table 4Pearson r correlation coefficients between environmental variables.

Variable Winter T Winter S Summer T Summer S Productivity

Winter T 0.482** 0.545** 0.509** 0.358**Winter S 0.482** 0.127 0.968** **�0.608Summer T 0.545** 0.127 0.15 **�0.603Summer S 0.509** 0.968** 0.15 **�0.572Productivity 0.358** **�0.608 **�0.603 **�0.572

**p< 0.01, *p< 0.05.

Table 5Dinoflagellate cyst taxon codes.

Code Taxon

BSPP Brigantedinium spp.AMIN Islandinium minutumLEJO Lejeunecysta olivaOCEN Operculodinium centrocarpumPKOF Cyst of Polykrikos kofoidiiQCON Quinquecuspis concretaSELE Selenopemphix nephroidesSQUA Selenopemphix quantaVCAL Votadinium calvumVSPI Votadinium spinosumPERI Undifferentiated ProtoperidinioidsBTEP Bitectatodinium tepikienseEGRA Echinidinium granulatumIACU Impagidinium aculeatumIPAT Impagidinium patulumIPAR Impagidinium paradoxumISPH Impagidinium sphaericumISTR Impagidinium strilatumIPAL Impagidinium pallidumAMIC Islandinium cezarePAME Cyst of Protoperidinium americanumLSAB Lejeunecysta sabrinaNLAB Nematosphaeropsis labyrinthusPDAL Cyst of Pentapharsodinium daleiPRET Pyxidinopsis reticulataSRAM Spiniferites ramosusSELO Spiniferites elongatusSMIR Spiniferites mirabilisSSPP Spiniferites spp.TVAR Trinovantedinium spp.SBEN Spiniferites bentoriiSBUL Spiniferites bulloidesISSP Impagidinium spp.LSSP Lejeunecysta spp.TAPP Trinovantedinium applanatumTRVA Trinovantedinium variabileDSPP Dubridinium spp.PSHW Polykrikos schwartziiEACU Echinidinium aculeatumEDEL Echinidinium delacatumESPP Echinidinium spp.CTYP Cyst Type A

R.T. Patterson et al. / Quaternary International 235 (2011) 13e25 21

solar activity, both of which have been identified as stronglyinfluencing the climate system of the NE Pacific (e.g. Christoforouand Hameed, 1997; Mantua et al., 1997; Mantua and Hare, 2002;Patterson et al., 2004, 2007; Chang and Patterson, 2005;Vázquez-Riveiros and Patterson, 2009). This lack of samplingresolution is unfortunate as it would have been useful to determineif winter SST followed the decadal-scale climate cyclicity recog-nized in previous time series analysis of Effingham Inlet records.For example, Patterson et al. (2004) found evidence of both the

8%

Summer temperature

8%Summer salinity

3%Productivity

10%

Intercorrelations59%

Winter temperature

12%

Winter salinity

Fig. 7. pCCA results for all environmental variables.

15e25 and 50e70 year bands of the PDO as well as the w88 yearGleissberg solar cycle in core TUL99B03, while Ivanochko et al.(2008) detected evidence of PDO cyclicity in core MD02-2494.General trends in late Holocene winter SST are identifiable in therecords from TUL99B03 and TUL99B11 though, which can becorrelated with other paleoclimatological/paleoceanographicrecords in the region (e.g. Chang et al., 2003; Patterson et al., 2004,2005; Hay et al., 2007; Ivanochko et al., 2008).

Sediments deposited in Effingham Inlet core TUL99B03 fromw5500 to 3600 cal BP provide evidence that the climate waswarmer and drier, with the NE Pacific weather pattern beinggenerally dominated by the NPH (Chang et al., 2003; Pattersonet al., 2004), in line with results from elsewhere in southernBritish Columbia (Pellat et al., 2001; Galloway et al., 2007, 2009,2010). These sediments are characterized by well-developedlaminae couplets deposited under primarily anoxic conditions, aswould be expected during a NPH dominated upwelling regime(Fig. 3). Further evidence of upwelling is provided by the highproductivity indicator taxa characterizing the dinoflagellate cystpopulations and diatom floras, which would have required a highupwelling generated nutrient supply (Chang et al., 2003; Chang andPatterson, 2005; Patterson et al., 2005; Hay et al., 2007). The winterSST reconstruction from the dinoflagellate cyst data from TUL99B03indicate that conditions were not constant through this intervalthough and provide evidence of distinct phases of gradual winterSST cooling and warming (>1 �C variability).

Modern atmospheric-oceanic conditions became establishedalong the Pacific coast of North America between w4200 and3000 cal BP (e.g. Nederbragt and Thurow, 2001; Chang et al., 2003;Patterson et al., 2004; Anderson et al., 2005; Barron and Bukry,2007; Kennett et al., 2007; Fisher et al., 2008; Barron, 2008). Thistransitionwas characterized by a stronger ALwith a center of actionpositioned further eastward and a weaker NPH positioned furtherto the south (Barron, 2008; Galloway et al., 2010). Determination ofwhen this change exactly occurred is difficult to determine withprecision as the Holocene transition from one climate state toanother in this region lasted for up to 2000 years, with considerablegeographic variability (Mathewes and Heusser, 1981; Gallowayet al., 2010).

In Effingham Inlet core TUL99B03 evidence of this transition tocooler and wetter winter conditions began from the base of thecore at w5500 cal BP with temperature steadily dropping untilw3500 cal BP, which correlates locally with palynological andphysical evidence of neoglacial advances (Tiedmenann, Peyto,Robson) along the southern British Columbia Pacific coast (Pellatand Mathewes, 1997; Pellat et al., 2001; Chang et al., 2003),evidence of wetter winters in SE Alaska (Fisher et al., 2008), as wellas a series of global cooling events (Bond et al., 2001) (Fig. 3). Theincreased AL influence directly impacted Effingham Inlet asincreased precipitation resulted in greater stratification of theupper water layer. As a result nutrient-rich waters entered Effing-ham Inlet and were carried to greater depths where they becameunavailable for phytoplankton utilization. In addition the change inwind patterns resulted in more oxygenation events (Pattersonet al., 2004). The shift from NPH to AL dominance is manifestedsedimentologically in both cores TUL99B03 and TUL99B11 by anincrease in the number of non-laminated units, indicatinga departure from the predominantly anoxic bottom-water condi-tions that had previously characterized sedimentation (Figs. 3 and4). Depressed winter SST conditions characterized Effingham Inletduring the onset of neoglacial conditions in the region, terminatingby w2800 cal BP. The generally warmer winter SST in EffinghamInlet after the termination of the neoglacial advances, mostnoticeable in core TUL99B03 (Fig. 3), may have come about asa result of increased seasonality of the California Current following

Fig. 8. Fossil dinoflagellate record from core TUL99B03 with reconstructions of winter temperature and productivity by WMAT. Root mean squared errors (RMSE) were �0.91 �Cand 38.37 g Cm�2 for winter temperature and productivity respectively.

R.T. Patterson et al. / Quaternary International 235 (2011) 13e2522

the mid to late Holocene transition (Barron et al., 2003, 2004). Thischange to the California Current would have resulted in strongerspring upwelling in a narrower zone near the coast. Aweakening ofupwelling in the early fall would have allowed warmer waters ofthe Central Gyre to move shoreward through the region inSeptember. These overall changes to the California Current wouldresult inwarmerwinter SST by causing an earlier onset of the UnderCurrent (Barron et al., 2003, 2004, Pers. Comm. 2009).

Winter SST warms at w2300e2200 cal BP in both cores (Figs. 8and 9), may represent an intensification of the AL andmoremodernconditions. Barron et al. (2004) argue that an increase in the warmwater diatom Pseudoeunotia doliolus (now Fragilariopsis doliola)that was reported at w3400 cal BP at offshore NE Pacific ODP site

Fig. 9. Fossil dinoflagellate record from core TUL99B11 with reconstructions of winter tempe38.37 g Cm�2 for winter temperature and productivity respectively.

1019 began to appear inshore off Eureka, California byw2300 cal BP in core TN02 0550 (Barron et al., 2004). This inshoremigration of F. doliola suggests that there was a further narrowingof the California Current in the fall, permitting warmer waters ofthe Central Gyre to penetrate further shoreward.

Barron (2008) has suggested that the warmer SST of the lateHolocene, also recognized in the Effingham Inlet records (Figs. 3and 4) was also characterized by an enhanced expression of ElNiño/Southern Oscillation throughout the region. As comparedwith the mid-Holocene the late Holocene seems to be generallymore El Niño-like or in a positive PDO state. The postw2300 cal BPinshore movement of warm waters from the Central Gyre andincreased expression of warm ENSO/PDO cycles may thus be

rature and productivity byWMAT. Root mean squared errors (RMSE) were �0.91 �C and

R.T. Patterson et al. / Quaternary International 235 (2011) 13e25 23

related (Barron, 2009, Pers. Comm.). In Effingham Inlet particularlywarmwinter SST conditions prevailed during thew1100 cal BP (AD900)ew700 cal BP (AD 1300) interval (Figs. 3 and 4), particularly incore TUL99B11. This period roughly corresponds with the so-calledMedieval Warm Period (MWP), an interval characterized by a pro-longed and pronounced megadrought throughout western NorthAmerica (Cook et al., 2004; MacDonald and Case, 2005). Modula-tion of teleconnected marine influences such as El Niño and thePDO controls the amount of moisture transported to the westernpart of the North American continent, which was greatly reducedduring theMWP (Cook et al., 2004). This interval seems to have alsocorresponded to higher rates of NPH influenced upwelling resultingin cooler summer NE Pacific SSTs, and an associated increase inmarine productivity in Effingham Inlet and elsewhere in the region(e.g. Ingram, 1998; Finney et al., 2002; Kim et al., 2004).

Winter SSTs subsequently progressively declined in the Effing-ham Inlet records, an interval corresponding to generally positivePDO conditions in the NE Pacific region and a return to less aridconditions throughout western North America (Cook et al., 2004;MacDonald and Case, 2005).

6. Conclusions

Canonical Correspondence Analysis (CCA) and partial CanonicalCorrespondence Analysis (pCCA) were used to re-analyzecontemporary dinoflagellate distributional data from the NE Pacificmargin and estuarine environments and highlighted that wintertemperature and productivity are the most important controls onthe distribution of dinoflagellate cysts in this region. Dinoflagellatecyst MAT models for winter temperature and productivity weredeveloped from the contemporary distributional data and appliedto subfossil data from two mid to late Holocene (w5500 cal BP-present) cores from Effingham Inlet.

The late Holocene MAT-based winter SST productivity recon-structions derived from cores TUL99B03 and TUL99B11 inEffingham Inlet produced results consistent with known paleo-ceanographic conditions in the NE Pacific. Particularly noteworthyin the Effingham Inlet records is the decline in winter SST afterw5500 cal BP that is associated with a decrease in the influence ofthe NPH through the Holocene. This drop in winter SST culminatedin the initiation of neoglacial advances in the NE Pacific byw3500 cal BP. Following the termination of glacial conditions incoastal areas there was a general increase inwinter SST through theLate Holocene linked to a weakening of the California Current,which would have resulted in a greater influence of the warmcentral gyre and overall more El Niño-like conditions. Winter SSTtemperature spikes coeval with megadrought conditions recordedin continental records are found in both core TUL99B03 andTUL99B11.

This research confirms that upwelling derived productivity andwinter SST are the main parameters influencing dinocyst distribu-tion in the North-eastern Pacific. Thus, dinoflagellate derivedreconstructions of productivity and winter SST primary are robustand constitutes a step forward in using dinocysts as paleoenvir-onmental tracers. Future research utilizing higher resolution dino-flagellate cyst datasets should provide a valuable contribution tobetter understanding the relationship between oceaneatmospherecirculation and climate change in the region.

Acknowledgements

This research was supported by an NSERC Strategic projectgrant, an NSERC Discovery Grant and a Canadian Foundation forClimate and Atmospheric Sciences research grant to RTP. We arealso grateful to R. Thomson (DFO, Sidney, BC) and Vaughn Barrie

(PGC-GSC, Sidney, BC), for providing ship time aboard the CCGSTulley, as well as for logistic and technical support. We thankTaoufik Radi, Université du Québec à Montréal, for contributing thecontemporary NE Pacific dinoflagellate distributional data utilizedin development of the training set and for his constructivecomments on the manuscript. We also thank John Barron of theUSGS, Menlo Park, California, for discussions on Late Holoceneoceaneatmosphere circulation in the NE Pacific and for his helpfulcomments on the manuscript. Thanks are also expressed to GillAlexander (Queen’s University Belfast) for cartographic support.Arun Kumar would like to thank King Fahd University of Petroleumand Minerals for permission to publish this paper.

References

Anderson, L., Abbott, M.B., Finney, B.P., Burns, S.J., 2005. Regional atmosphericcirculation change in the North Pacific during the Holocene inferred fromlacustrine carbonate oxygen isotopes, Yukon Territory, Canada. QuaternaryResearch 64, 21e35.

Barron, J.A., Heusser, L.E., Herbert, T., Lyle, M., 2003. High resolution climaticevolution of coastal northern California during the past 16,000 years. Paleo-ceanography 18, 1020e1035.

Barron, J.A., Heusser, L.E., Clark, A., 2004. High resolution climate of the past 3,500years of coastal northernmost California. In: Starratt, Scott W., Blomquist, NikkiL. (Eds.), Proceedings of the Twentieth Annual Pacific Climate Workshop, PacificGrove, CA, April 6e9, 2003. Technical Report 72 of the Interagency EcologicalProgram for the San Francisco Estuary. State of California, Dept. of WaterResources, pp. 13e22. http://tenaya.ucsd.edu/wdettinge/proc_2003.html.

Barron, J.A., Bukry, D., 2007. Development of the California current during the past12,000 yr based on diatoms and silicoflagellates. Palaeogeography, Palae-oclimatology, Palaeoecology 248, 313e338.

Barron, J.A., 2008. Transitioning out of the middle Holocene e enhancement ofENSO cycles between 4 and 3 ka along the Pacific coast of North America.American Geophysical Union Fall Meeting 2008, abstract #PP41B-1456.

Birks, H.J.B., 1995. Quantitative palaeoenvironmental reconstructions. In: Maddy, D.,Brew, J.S. (Eds.), Statistical Modelling of Quaternary Science Data. TechnicalGuide 5. Quaternary Research Association, Cambridge.

Borcard, D., Legendre, P., Drapeau, P., 1992. Partialling out the spatial component ofecological variation. Ecology 73, 1045e1055.

Bond, G., Kromer, B., Beer, J., Muscheler, R., Evans, M.N., Showers, W., Hoffmann, S.,Lotti-Bond, R., Hajdas, I., Bonani, G., 2001. Persistent solar influence on NorthAtlantic climate during the Holocene. Science 294, 2130e2136.

Bridgman, H.A., Oliver, J.E., Glantz, M.H., 2006. The Global Climate System: Patterns,Processes, and Teleconnections. Cambridge University Press, 331 pp.

Brodeur, R.D., Ware, D.M., 1992. Long-term variability of zooplankton biomass in thesubarctic Pacific Ocean. Fisheries Oceanography 1, 32e38.

Chang, A.S., Patterson, R.T., McNeely, R., 2003. Seasonal sediment and diatom recordfrom Late Holocene laminated sediments, Effingham Inlet, British Columbia,Canada. Palaios 18, 477e494.

Chang, A.S., Patterson, R.T., 2005. Climate shift at 4400 years BP: evidence fromhigh-resolution diatom stratigraphy, Effingham Inlet, British Columbia, Canada.Palaeogeography, Palaeoclimatology, Palaeoecology 226, 72e92.

Chatton, É., 1914. Les cnidocystes du péridinien Polykrikos schwartzii Bütschli.Archives de Zoologie Expérimentale et Générale 54, 157e194.

Chavez, F.P.,Ryan, J.R., Lluch-Cota, S.E.,Niquen,C.M.C., 2003. Fromanchovies to sardinesand back: multidecadal change in the Pacific Ocean. Science 299, 217e221.

Christoforou, P., Hameed, S., 1997. Solar cycle and the Pacific ‘centers of action’.Geophysical Research Letters 24, 293e296.

Cook, E.R., Woodhouse, C.A., Eakin, C.M., Meko, D.M., Stahle, D.W., 2004. Long-termaridity changes in the Western United States. Science 306, 1015e1018.

Crawford, W.R., Dewey, R.K., 1989. Turbulence and mixing: sources of nutrients onthe Vancouver Island continental shelf. AtmosphereeOcean 27, 428e442.

Dale, B., Dale, A.L., 2002. Application of ecologically based statistical treatments tomicropalaeontology. In: Haslett, S.K. (Ed.), Quaternary Environmental Micro-palaeontology. Arnold, London.

Dallimore, A., 2001. Late Holocene Geologic, Oceanographic and Climate History ofan Anoxic Fjord: Effingham Inlet, West Coast, Vancouver Island. UnpublishedPh.D. dissertation, Carleton University, Ottawa, Ontario.

Dallimore, A., Enkin, R.J., Pienitz, R., Southon, J.R., Baker, J., Wright, C.A., Pedersen, T.F.,Calvert, S.E., Ivanochko, T., Thomson, R.E., 2008. Postglacial evolution of a Pacificcoastal fjord in British Columbia, Canada: interactions of sea-level change, crustalresponse, andenvironmentalfluctuationsdresults fromMONAcoreMD02-2494.Canadian Journal of Earth Sciences 45, 1345e1362.

Faith, D.P., Minchin, P.R., Belbin, L., 1987. Compositional dissimilarity as a robustmeasure of ecological distance. Vegetation 69, 57e68.

Finney, B.P., Gregory-Eaves, I., Douglas, M.S.V., Smol, J.P., 2002. Fisheries produc-tivity in the northeastern Pacific Ocean over the past 2,200 years. Nature 416,729e733.

Fisher, D., Osterberg, E., Dyke, A., Dahl-Jensen, D., Demuth, M., Zdanowicz, C.,Bourgeois, J., Koerner, R.M., Mayewski, P., Wake, C., Kreutz, K., Steig, E., Zheng, J.,

R.T. Patterson et al. / Quaternary International 235 (2011) 13e2524

Yalcin, K., Goto-Azuma, K., Luckman, B., Rupper, S., 2008. The Mt Logan Hol-ocenedlate Wisconsinan isotope record: tropical PacificdYukon connections.The Holocene 18, 667e677.

Galloway, J.M., Patterson, R.T., Doherty, C.T., Roe, H.M., 2007. Post-glacial vegetationand climate history of two frog Lake, Central Coastal Mainland, BritishColumbia. Journal of Paleolimnology 38, 569e588.

Galloway, J.M., Patterson, R.T., Doherty, C.T., Roe, H.M., 2009. Post-glacial vegetationand climate dynamics at Tiny Lake, central mainland coast of British Columbia.Journal of Quaternary Science 24, 322e335.

Galloway, J.M., Babalola, L.O., Patterson, R.T., Roe, H.M., 2010. A high resolutionmarine palynological record from the central mainland coast of BritishColumbia, Canada: evidence for a mid-late Holocene dry climate interval.Marine Micropaleontology 75, 62e78.

Guilbault, J.P., Patterson, R.T., Thomson, R.E., Barrie, J.V., Conway, K.W., 1997. LateQuaternary paleoeceanographic changes in dixon Entrance, Northwest BritishColumbia: evidence from the foraminiferal faunal succession. Journal of Fora-miniferal Research 27, 151e174.

Guiot, J., de Vernal, A., 2007. Transfer functions: methods for quantitative paleo-ceanography based on microfossils. In: Hillaire-Marcel, de Vernal (Eds.), Proxiesin Late Cenozoic Paleoceanography. Elsevier, pp. 523e563.

Hay, M.B., 2005. Reconstitution des conditions océanographiques et climatiques del’holocène supérieur à Effingham Inlet, Colombie Britannique. UnpublishedPh.D. dissertation, Laval University, Quebec. http://www.theses.ulaval.ca/2005/23241/23241.pdf

Hay, M.B., Dallimore, A., Thomson, R.E., Calvert, S.E., Pienitz, R., 2007. Diatom recordof Late Holocene oceanography and climate along the West Coast of VancouverIsland, British Columbia (Canada). Quaternary Research 67 (1), 33e49.

Ingram, B.L., 1998. Differences in radiocarbon age between shell and charcoal froma Holocene shellmound in northern California. Quaternary Research 49,102e110.

Ivanochko, T.S., Calvert, S.E., Thomson, R.E., Pederson, T.F., 2008. Geochemicalreconstruction of Pacific decadal variability from the eastern North Pacificduring the Holocene. Canadian Journal of Earth Sciences 45, 1317e1329.

Juggins, S., 2003. C2 User Guide. Software for Ecological and Palaeoecological DataAnalysis and Visualisation. University of Newcastle, Newcastle upon Tyne.

Kennett, D.J., Kennett, J.P., Erlandson, J.M., Cannariato, K.G., 2007. Human responsesto Middle Holocene climate change on California’s Channel Islands. QuaternaryScience Reviews 26, 351e367.

Kim, J., Rimbu, N., Lorenz, S.J., Lohmann, G., Nam, S., Schouten, S., Ruhlemann, C.,Schneider, R.R., 2004. North Pacific and North Atlantic sea-surface tempera-ture variability during the Holocene. Quaternary Science Reviews 23,2141e2154.

Kumar, A., Patterson, R.T., 2002. Distributional controls on dinoflagellate cystassemblages in Effingham Inlet, Vancouver Island, British Columbia, Canada.Palaeogeography, Palaeoclimatology, Palaeoecology 180, 187e206.

Legendre, P., Gallagher, E.D., 2001. Ecologically meaningful transformations forordination of species data. Oecologia 129, 271e280.

Legendre, P., Legendre, L. (Eds.), 1998. Numerical Ecology, second English edition.Elsevier, Amsterdam.

Lentin, J.K., Williams, G.L., 1993. Fossil dinoflagellates: index to genera and species,American Association of Stratigraphic Palynologists, Contribution Series 28,864.

MacDonald, G.M., Case, R.A., 2005. Variations in the Pacific decadal oscillation overthe past millennium. Geophysical Research Letters 32 (L08703). doi:10.1029/2005GL022478.

Manounos, B., Osborn, G., Clague, J.J., Luckman, B.H., 2009. Latest Pleistocene andHolocene glacier fluctuations in western Canada. Quaternary Science Reviews28, 2049e2074.

Mantua, N.J., Hare, S.R., 2002. The Pacific decadal oscillation. Journal of Oceanog-raphy 58, 35e44.

Mantua, N.J., Hare, S.R., Zhang, Y., Wallace, J.M., Francis, R.C., 1997. A Pacific inter-decadal climate oscillation with impacts on salmon production. Bulletin of theAmerican Meteorological Society 78, 1069e1079.

Marret, F., Scourse, J., Kennedy, H., Ufkes, E., Jansen, J.H.F., 2008. Marine productionin the Congo-influenced SE Atlantic over the past 30,000 years: a novel dino-flagellate-cyst based transfer function approach. Marine Micropaleontology 68,198e222.

Mathewes, R.W., Heusser, L.E., 1981. A 12,000 year palynological record oftemperature and precipitation trends in southwestern British Columbia.Canadian Journal of Botany 59, 707e710.

McKay, J.L., de Vernal, A., Hillaire-Marcel, C., Not, C., Polyak, L., Darby, D., 2008.Holocene fluctuations in Arctic sea-ice cover: Dinocyst-based reconstructionsfor the eastern Chukchi Sea, Canadian Journal of Earth Sciences 45, 1377e1397.

Nederbragt, A.J., Thurow, J., 2001. A 6000 year varve record of Holocene sedimentsin Saanich Inlet, British Columbia, from digital sediment colour analysis of ODPLeg 169S cores. Marine Geology 174, 95e110.

Patterson, R.T., 1993. Late Quaternary benthic foraminiferal biofacies and paleo-ceanography of Queen Charlotte sound and southern Hecate Strait, BritishColumbia. Journal of Foraminiferal Research 23, 1e18.

Patterson, R.T., Guilbault, J.P., Thomson, R.E., 2000. Oxygen level control on fora-miniferal assemblage distribution in Effingham Inlet, Vancouver Island, BritishColumbia. Journal of Foraminiferal Research 30, 321e335.

Patterson, R.T., Kumar, A., 2002. Post-Glacial paleoceanographic history of SaanichInlet British Columbia based on foraminiferal proxy data. Journal of Forami-niferal Research 32, 110e125.

Patterson, R.T., Prokoph, A., Chang, A.S., 2004. Late Holocene sedimentary responseto solar and cosmic ray activity influenced climate variability in the NE Pacific.Sedimentary Geology 172, 67e84.

Patterson, R.T., Prokoph, A., Kumar, A., Chang, A.S., Roe, H.M., 2005. Late Holocenevariability in pelagic fish and phytoplankton productivity along the west coastof Vancouver Island, NE Pacific Ocean. Marine Micropaleontology 55, 183e204.

Patterson, R.T., Prokoph, A., Reinhardt, A., Roe, H.M., 2007. Climate cyclicity in lateHolocene anoxic marine sediments from the Seymour-Belize Inlet Complex,British Columbia. Marine Geology 242, 123e140.

Pellat, M.G., Mathewes, R.W., 1997. Holocene tree line and climate change on theQueen Charlotte Islands, Canada. Quaternary Research 48, 88e99.

Pellat, M.G., Hebda, R.J., Mathewes, R.W., 2001. High-resolution Holocene vegeta-tion history and climate from Hole 1034B, ODP leg 169S, Saanich Inlet, Canada.Marine Geology 174, 211e222.

Pospelova, V., de Vernal, A., Pedersen, T.F., 2008. Distribution of dinoflagellate cystsin surface sediments from the northeastern Pacific Ocean (43e25�N) in relationto sea-surface temperature, productivity and coastal upwelling. MarineMicropaleontology 68, 21e48.

Prokoph, A., Patterson, R.T., 2004. From depth-scale to time-scale: transforming ofsediment image color data into high-resolution time-series. In: Francus, P. (Ed.),Image Analysis, Sediments and Paleoenvironments. Developments in Paleo-environmental Research Series 7, Chapter 8. Springer, Dordrecht, pp. 143e164.

Radi, T., de Vernal, A., 2004. Dinocyst distribution in surface sediments from thenortheastern Pacific margin (40e60 degrees N) in relation to hydrographicconditions, productivity and upwelling. Review of Palaeobotany and Palynology128, 169e193.

Radi, T., Pospelova, V., de Vernal, A., Barrie, J.V., 2007. Dinoflagellate cysts as indi-cators of water quality and productivity in British Columbia estuarine envi-ronments. Marine Micropaleontology 62, 269e297.

Radi, T., de Vernal, A., 2008. Dinocysts as proxy of primary productivity in mid-highlatitudes of the Northern Hemisphere. Marine Micropaleontology 68, 84e114.

Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Bertrand, C.J.H.,Blackwell, P.G., Buck, C.E., Burr, G.S., Cutler, K.B., Damon, P.E., Edwards, R.L.,Fairbanks, R.G., Friedrich, M., Guilderson, T.P., Hogg, A.G., Hughen, K.A.,Kromer, B., McCormac, F.G., Manning, S.W., Ramsey, C.B., Reimer, R.W.,Remmele, S., Southon, J.R., Stuiver, M., Talamo, S., Taylor, F.W., van der Plicht, J.,Weyhenmeyer, C.E., 2004. IntCal04 terrestrial radiocarbon age calibration,26e0 ka BP. Radiocarbon 46, 1029e1058.

Rochon, A., de Vernal, A., Turon, J.L., Matthiessen, J., Head, M.J., 1999. Distribution ofdinoflagellate cysts in surface sediments from the North Atlantic Ocean and adja-cent basins and quantitative reconstruction of sea-surface parameters. AmericanAssociation of Stratigraphic Palynologists Contribution Series 35, 1e152.

Schwing, F.B., Murphree, T., Green, P.M., 2002. The Northern Oscillation Index (NOI):a new climate index for the northeast Pacific. Progress in Oceanography 53,115e139.

Shen, C., Wang, W., Gong, W., Hao, Z., 2006. A Pacific Decadal Oscillation recordsince 1470 AD reconstructed from proxy data of summer rainfall over easternChina. Geophysical Research Letters 33 (L03702). doi:10.1029/2005GL024804.

�Smilauer, P., 1992. Canodraw User’s Guide v.30. Microcomputer Power, Ithaca,New York.

Stuiver, M., Reimer, P.J., 1993. Extended 14C database and revised CALIB radiocarboncalibration program. Radiocarbon 35, 215e230.

Stuiver, M., Reimer, P.J., Bard, E., Beck, J.W., Burr, G.S., Hughen, K.A., Kromer, B.,McCormac, F.G., Plicht, J., Spurk, M., 1998a. INTCAL 98 Radiocarbon age cali-bration 24,000e0 cal BP. Radiocarbon 40, 1041e1083.

Stuiver, M., Reimer, P.J., Braziunas, T.F., 1998b. High-precision radiocarbon agecalibration for terrestrial and marine samples. Radiocarbon 40, 1127e1151.

Stuiver, M., Reimer, P.J., and Reimer, R.W., 2005. CALIB 5.0. [WWW program anddocumentation].

Tappan, H., 1980. The Paleobiology of Plant Protists. Freeman and Company, SanFrancisco, 1028 pp.

Telford, R.J., 2006. Limitations of dinoflagellate cyst transfer functions. QuaternaryScience Reviews 25, 1375e1382.

ter Braak, C.J.F., 1985. Correspondence analysis of incidence and abundance data:properties in terms of a unimodal response model. Biometrics 41, 859e873.

ter Braak, C.J.F., 1987. CANOCO e a FORTRAN Program for Canonical CommunityOrdination by (Partial) (Detrended) (Canonical) Correspondence Analysis,Principal Components Analysis and Redundancy Analysis (Version 2.1). Tech-nical Report LWA-88-02. GLW, Wageningen, The Netherlands.

ter Braak, C.J.F., 2002. CANOCO for Windows Version 4.51. Biometris-Plant ResearchInternational, Wageningen, The Netherlands.

ter Braak, C.J.F., �Smilauer, P., 2002. CANOCO Reference Manual and CanoDraw forWindows User’s Guide: Software for Canonical Community Ordination (version4.5). Microcomputer Power, Ithaca, New York, 500 pp.

Thomson, R.E., 1981. Oceanography of the British Columbia coast. Canada SpecialPublication of Fisheries and Aquatic Sciences 56, 291.

Thomson, R.E., Gower, J.F.R., 1998. A basin-scale oceanic instability event in the Gulfof Alaska. Journal of Geophysical Research 103, 3033e3040.

Thomson, R.E., Ware, D.M., 1996. A current velocity index of ocean variability.Journal of Geophysical Research 101, 14297e14310.

Vázquez-Riveiros, N., Patterson, R.T., 2009. Paleoceanographic history of the Sey-mour-Belize Inlet Complex, British Columbia, Canada, through the last 1100years. Quaternary Science Reviews 28, 2833e2850.

de Vernal, A., Henry, M., Matthiesen, J., Mudie, P.J., Rochon, A., Boessenkoal, K.P.,Eynaud, F., Grosfjeld, K., Guiot, J., Hamel, D., Harland, R., Head, M.J., Kunz-

R.T. Patterson et al. / Quaternary International 235 (2011) 13e25 25

Pirrung, M., Levac, E., Loucheur, V., Peyron, O., Pospelova, V., Radi, T., Turon, J.-L.,Veronica, E., 2001. Dinoflagellate cyst assemblages as tracers of sea-surfaceconditions in the northern North Atlantic, Arctic and sub Arctic seas: the new‘n ¼ 677’ data base and its application for quantitative paleoceanographicreconstruction. Journal of Quaternary Science 16 (7), 681e698.

de Vernal, A., Eynaud, F., Henry, M., Hillaire-Marcel, C., Londeix, L., Mangin, S.,Matthiessen, J., Marret, F., Radi, T., Rochon, A., Solignac, S., Turon, J.L., 2005. Thesurface ocean at middle to high latitudes of the Northern Hemisphere duringthe Last Glacial Maximum (LGM) estimates from dinoflagellate cyst assem-blages. Quaternary Science Reviews 24, 897e924.

Wall, D., 1967. The resting cysts of modern marine dinoflagellates and theirpaleontological significance. Review of Palaeobotany and Palynology 2,349e354.

Ware, D.M., Thomson, R.E., 1991. Link between long-term variability in upwellingand fish production in the northeast Pacific Ocean. Canadian Journal of Fisheriesand Aquatic Sciences 48, 2296e2306.

Ware, D.M., Thomson, R.E., 2000. Interannual to multidecadal timescale climatevariations in the Northeast Pacific. Journal of Climate 13, 3209e3220.

Ware, D.M., Thomson, R.E., 2005. Bottom-up ecosystem trophic dynamics deter-mine fish production in the Northeast Pacific. Science 308, 1280e1284.

Wright, C.A., Dallimore, A., Thomson, R.E., Patterson, R.T., Ware, D., 2005. LateHolocene paleofish populations in Effingham Inlet, British Columbia, Canada.Palaeogeography, Palaeoclimatology, Palaeoecology 224, 367e384.

Zonneveld, K.A.F., 1997. New species of organic walled dinoflagellate cysts frommodern sediments of the Arabian Sea (Indian Ocean). Review of Palaeobotanyand Palynology 97, 319e337.