This article was downloaded by: [SUNY State Univ of New York Geneseo]On: 31 October 2014, At: 05:56Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

New Zealand Journal of Geology and GeophysicsPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/tnzg20

A late Quaternary pollen record from marine coreP69, southeastern North Island, New ZealandMatthew S. McGlone aa Landcare Research , P.O. Box 69, Lincoln, New ZealandPublished online: 23 Mar 2010.

To cite this article: Matthew S. McGlone (2001) A late Quaternary pollen record from marine core P69, southeastern NorthIsland, New Zealand, New Zealand Journal of Geology and Geophysics, 44:1, 69-77, DOI: 10.1080/00288306.2001.9514923

To link to this article: http://dx.doi.org/10.1080/00288306.2001.9514923

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose ofthe Content. Any opinions and views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be reliedupon and should be independently verified with primary sources of information. Taylor and Francis shall not beliable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilitieswhatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out ofthe use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

New Zealand Journal of Geology & Geophysics, 2001, Vol. 44: 69-770028-8306/01/4401-0069 $7.00/0 © The Royal Society of New Zealand 2001

69

A late Quaternary pollen record from marine core P69, southeasternNorth Island, New Zealand

MATTHEW S. McGLONE

Landcare ResearchP.O. Box 69Lincoln, New Zealand

Abstract Marine core P69 (115 km off the southeasternNorth Island) has already yielded a 26 000 yr record ofcarbonate and silica influx, δ O 1 8 stratigraphy, foraminifera,and sea-surface temperatures. A pollen analysis of the coreis presented here. The full-glacial (25 000-15 000 yr BP)pollen assemblages reflect a southern North Island landscapelargely covered with scrub and grassland, but only limitedareas of cool-temperate forest. Abundant reworked Tertiarypollen types indicate increased erosion at this time. Rapidspread of podocarp-dominant forest occurred between15 000 and 11 500 yr BP, an event that relates only in ageneral way to increasing sea-surface temperatures, butcoincides exactly with a sharp reduction of wind-inducedupwelling and terrestrially sourced quartz. The abruptmovement southwards of the glacially expanded zone ofstrong westerlies at c. 15 000 yr BP, rather than warming,appears to be the main factor controlling postglacialreafforestation.

Keywords Last Glaciation; postglacial; late Quaternary;palynology; paleoceanography; vegetation history; climatechange; Hawke's Bay; North Island; New Zealand

INTRODUCTION

Pollen analytical studies of terrestrial sequences on the NewZealand mainland provide a history of plant vegetation coverand climate change extending back through the Quaternaryand into the Tertiary. However, the majority of sequencescover only a small proportion of a single glacial-interglacialcycle. The few longer sequences (e.g., McGlone & Topping1983; Newnham 1992; Mildenhall 1995; Elliot 1998) havehiatuses and poor age control. New Zealand, therefore, hasno equivalent of the long, continuous vegetation records thathave proven so useful in exploring Quaternary environ-mental change elsewhere.

Marine cores close to the New Zealand mainland presentan opportunity to compensate for the fragmented nature ofterrestrial pollen records. The major advantage marine corespresent is continuous sedimentation over a long time span,combined with the independent time-scale provided byoxygen isotopes. Analysis of carbon isotopes, sediment, and

G00029Received II April 2000; accepted 1 November 2000

faunal variations gives valuable additional information asto water mass status, sea-surface temperatures, and influxof terrestrially derived sediments (Stewart & Neall 1984;Nelson et al. 1985; Dudley & Nelson 1988; Weaver et al.1998). Several marine cores close to the New Zealandmainland have been pollen analysed: a 60 000 yr record hasbeen produced from two cores off northeast New Zealand(Wright et al. 1995); a 350 000 yr record spanning fourglacial-interglacial cycles off southeast New Zealand(Heusser & van der Geer 1994); and a short, late Holocenecore from coastal waters near Poverty Bay (Wilmshurst etal. 1999).

However, marine palynology, although presenting manyopportunities, is complicated by the long transport route mostpalynomorphs follow before their ultimate incorporation inocean-floor sediments. Palynomorphs in marine sedimentsare transported by rivers and marine currents as well as byair, and hence a large proportion pass through intermediatestages, being held for variable lengths of time in river bank,estuarine, and coastal shelf sediments. Corrosion anddifferential loss of susceptible palynomorphs thus occurs,as well as enrichment by reworked, much older pollen andspores. As palynomorphs are derived over a large stretch ofcoastline, often it is not clear what the terrestrial source areasare for any given core, how they might be differentiallyrepresented, and how they might change in the course of aninterglacial-glacial cycle.

Wilmshurst et al. (1999) began work on resolving theseproblems through a detailed comparison of the late Holocenepollen record of a terrestrial core and an inshore marine corefrom Poverty Bay, North Island, and demonstrated a closerelationship between the two records but also somesystematic differences. However, more marine core recordsand more comparative work are needed before marinepalynological sequences can be confidently interpreted. Inthis paper, I present a pollen analysis and paleoenvironmentalinterpretation of marine core P69 (Stewart & Neall 1984),off the southeastern coast of the North Island (Fig. 1). Tohelp clarify problems surrounding palynological inter-pretation of marine cores, comparisons are made with twopreviously published marine and terrestrial pollen cores.

METHODS

A total of twenty-eight 1 cm3 subsamples were taken fromthe P69 core: the first 1.80 m was sampled at 20 cm intervals;the interval 1.85-2.45 m at 5-10 cm; and the basal 2.45-6.45 m portion of the core at c. 40 cm intervals. Pollen andspores were extracted by standard techniques of weak acidtreatment to remove carbonates, hydrofluoric acid to removesilica, and acetolysis to remove cellulose (Moore et al. 1991 ).Exotic marker spores were added to permit calculation ofpollen concentrations. All radiocarbon ages presented in thetext are based on the old half-life of 5568 years.

Dow

nloa

ded

by [

SUN

Y S

tate

Uni

v of

New

Yor

k G

enes

eo]

at 0

5:56

31

Oct

ober

201

4

70 New Zealand Journal of Geology and Geophysics, 2001, Vol. 44

\ .\ S

(X

f Hauraki (V> Gulf \ ',

^vV 8 \ ''' ^ ^ \ Co

\

1 176CE

•

man

l in

\

f • [ Maratoto |

{ NORTH ISLANDV NEW ZEALAND

á

\i

1i/

Jjs

X/

C7^

P

fj/

J//

1S8031

Bay ofPlenty

/^~J{ Hawk

/

/

178JE

EAST CAPECURRENT

/ East }y Cape /

>

r

I

• P69|

i0

i , 1

50 100 K

36JS

38"S

40°S

Fig. 1 North Island of New Zealand, showing location ofterrestrial and marine cores (filled circles) and general trend ofthe East Cape Current.

Pollen and spore profiles from all three diagramsdiscussed were subjected to detrended correspondenceanalysis (DCA) using DECORANA (Hill & Gauch 1980;Tilia program version 1.12—program by Eric C. Grimm,Illinois State Museum, 1992). DCA is an eigenvectorordination technique based on reciprocal averaging, andrearranges data in low-dimensional space so that the mostsimilar entities are closest together and the most dissimilarare farthest apart (Gauch 1982). Rare taxa and taxa generallyat low values except for occasional high representation in afew samples were excluded from the dataset, following therecommendation of Hill & Gauch ( 1980) to remove outliersand rare species. Taxa with five or less occurrences, or thatfailed to register at a level of 1% or more, were excluded.DCA was performed on the recalculated percentages afterexclusions.

CORE P69: PALYNOLOGY AND INTER-COMPARSION WITH OTHER MARINE ANDTERRESTRIAL CORES

Core P69 (40°23'S, 177°59.8'E) was recovered by a doublelength piston core in 2195 m of water in the OmakereDepression 115 km to the east of the Wairarapa coast, NorthIsland (Fig. 1). The core stratigraphy is described in Stewart& Neall (1984) and Nelson et al. (2000). The core is 6.63 min length, and consists of weakly bioturbated hemipelagicsediment, predominantly clayey medium silt. No visiblelayering was evident, aside from tephra horizons. There islittle evidence of turbidite deposition. Tephra stratigraphy

of the core is given in Table 1, and tephra ages given inFroggatt & Lowe (1990) were used to calculate sedi-mentation rates and the age model.

Pollen preservation and concentrationPollen and spore preservation is moderate, but a substantialnumber of grains were either obscured by dark mineral debrisor distorted beyond recognition. Pollen and spore concen-tration is relatively high for a marine core (6000-40 000grains/g dry wt), and reflects its relative closeness to shore.Dry weight concentrations, unconnected for accumulationrate, are between 6000 and 16 000 grains/g from the baseup to 2.50 m; between 2.50 and 1.00 m they rise steeply tonearly 40 000 grains/g, and then fall back to levels of 6000-22 000 grains/g. When these concentrations are adjusted forsediment accumulation rate, the highest pollen and sporeaccumulation rates occur from the base to 2.50 m, with onlyhalf those levels from 2.50 m to the surface.

Pollen and spore stratigraphyThe subgeneric classification of Nothofagus has been revised(Hill & Read 1991 ) and the new subgenera now align exactlywith the long-used pollen types: Nothofagus fusca typeequates to the subgenus Fuscospora; Nothofagus brassii typewith subgenus Brassospora; and Nothofagus menziesii typewith subgenus Lophozonia. In this paper I will useFuscospora and Brassospora instead of N. fusca type andN. brassii type, but retain Nothofagus menziesii as it is theonly extant New Zealand representative in its subgenus.

Pollen and spore percentages (Fig. 2) are based on apollen sum (range 147-290, x 236) of all taxa excludingspores and extinct pollen types; percentages of excluded taxaare calculated according to the same sum. The pollendiagram is divided into seven zones which are summarisedin Table 2.

Comparison with Lake Maratoto and Bay of Plentymarine core S803Lake Maratoto pollen diagram (Fig. 3) (Green et al. 1984;McGlone 1988) has been chosen for a terrestrial comparisonbecause it is one of the few that covers the same time rangeand, having a central lowland location in the northern NorthIsland, best represents the generalised pollen sources foroffshore core P69. It is taken from the centre of a small lake(altitude 60 m) in the Hamilton Basin. Deep-sea core S803(Fig. 4) is one of two pollen-analysed cores in the Bay ofPlenty (Wright et al. 1995), and is the only marine core tothe north of P69 that spans the same time range. Core P69,core S803, and Lake Maratoto share the same suite of datedtephras (Table 1), and can therefore be readily compared.

Table 1 Tephra layers in: marine cores P69 (depth below coretop in parentheses) and S803; and terrestrial core, Lake Maratoto,Hamilton. Tephra ages after Froggatt & Lowe (1990).

Tephra Ag

TaupoWhakataneRotomaWaiohauRerewhakaaituOkarekaKawakawa

ge( l 4CyrBP)

1850+ 104830 + 208530 + 10

11 850 ± 6014 700 ± 110c. 18 00022 590 ±230

X

X

X

X

X

X

P69

—(0.69(1.36(2.07(2.27(3.80(5.40

m)m)m)m)m)m)

S803

X

X

X

X-

-

X

L. Maratoto

X-

X

X

X

-

X

Dow

nloa

ded

by [

SUN

Y S

tate

Uni

v of

New

Yor

k G

enes

eo]

at 0

5:56

31

Oct

ober

201

4

McGlone—L. Quat. pollen record, core P69 71

Fig. 2 Marine core P69, percentage pollen diagram. Selected taxa only. Pollen sum: all taxa excluding spores and extinct types.

Fig. 3 Lake Maratoto, percent-age pollen diagram (McGlone1988). Selected taxa only. Pollensum: all taxa excluding spores andwetland types.

At a broad level, there is a close correspondence in thenature and timing of the vegetation changes recorded onlandand at sea. In P69, zone 6 (c. 19 000-15 000 yr BP),grassland and cool-temperate scrub and forest typesdominate, and this is compatible with the interval 16 000-14 700 yr BP at Lake Maratoto when the regional forest

dominants were Fuscospora and Nothofagus menziesii,existing as patches in a matrix of grassland and scrub(Newnham et al. 1989). Both cores record a major transitionto podocarp-broadleaved forest, beginning at around the timeof the Rerewhakaaitu Tephra (14 700 yr BP), and a periodof high abundance ofAscarina lucida and tree ferns between

Table 2 Zonation and summary pollen and spore stratigraphy, core P69.

Zone Depth (m) Age (yr BP) Description

12

34

5

6

7

0.0-0.300.30-0.90

0.90-1.701.70-2.00

2.00-2.40

2.40-4.25

4.25-6.45

0-0.3030-0.90

5970-10 26010 260-11 510

It 510-14 960

14 960-19 290

19 290-25 600

Appearance and increase of Pteridium esculentum; Pinus in uppermost sample.Phyllocladus, Agathis, Fuscospora and Brassospora increase, Cyathea smithii type,

Dacrycarpus dacrydioides and Podocarpus decrease.Dacrydium cupressinum abundant; Ascarina lucida consistently present in this zone.Podocarp tree pollen types and tree-fern spores dominant; Prumnopitys taxifolia major

contributor. Nothofagus types, shrubs, Asteraceae and grasses are at low levels.Prumnopitys taxifolia and other podocarp tree types and tree ferns increase steeply.

Nothofagus, Halocarpus, Phyllocladus, shrub types, Asteraceae and grass fall to lowpercentages or are eliminated.

Podocarp tree pollen at lowest levels; Metrosideros, Halocarpus, and grasses at highestpercentages. Empodisma and Cyperaceae at peak abundance.

High levels of Nothofagus (including extinct Brassospora), Libocedrus, Phyllocladusand Halocarpus. Shrub types (Coprosma, Myrsine, and Leptospermum), grasses andAsteraceae well represented. Intermediate levels of podocarp tree types.

Dow

nloa

ded

by [

SUN

Y S

tate

Uni

v of

New

Yor

k G

enes

eo]

at 0

5:56

31

Oct

ober

201

4

72 New Zealand Journal of Geology and Geophysics, 2001, Vol. 44

Fig. 4 Marine core S803, percentage pollen diagram (Wright et al. 1995). Selected taxa only. Pollen sum: all taxa excluding sporesand extinct types.

10 000 and 7000 yr BP. However, the time course of themajor podocarp tree species is substantially different, withDacrydium cupressinum reaching peak abundance between11 000 and 12 000 yr BP at Maratoto, but after 8000 yr BPin P69. Phyllocladus and, to a lesser extent, Fuscospora,become more common at Maratoto after 7000 yr BP, a trendseen in zones 3 and 2 in P69 as well.

Bay of Plenty core S803 bears rather less similarity toP69 than does Lake Maratoto. S8O3 has substantial amountsof warm-temperate podocarp and tree-fern taxa representedthroughout, and, although during the latter part of the LastGlaciation (40 000-15 000 yr BP), S803 has a highproportion of Fuscospora, Halocarpus, and Phyllocladus,there is a much lower representation of shrub and grasslandpollen types than in P69. As a consequence, the late-glacialtransition (15 000-10 000 yr BP), which is such a markedfeature of the Lake Maratoto and P69 pollen sequences, isnot as well defined, being mainly reflected in the decline ofNothofagus and rise of tree ferns. The postglacial sequenceshave more in common, and the Ascarina lucida rise isrecorded in both. The main differences are the higherproportion of Dacrydium cupressinum, Phyllocladus, andAgathis, and the lack of a late rise in the Fuscospora curveincoreS803.

Pollen and spore representationAll pollen and spores which are ultimately incorporated indeep-sea sediments follow a two-stage transport path. First,they are transported either by wind or by rivers and streamsto the ocean. Whether wind or fluvial processes dominatethe transport of pollen and spores to a particular marine coredepends very much on its location and the prominence ofwind-pollinated taxa in the vegetation (Mudie 1982; Sun etal. 1999). Second, once in the surface waters of the ocean,pollen and spores are incorporated into larger organiccomplexes of faecal pellets and flocculated/agglomeratedparticles (Pickrill 1987), which settle towards the bottom atthe same time that coastal currents and surface waters aremoving them along the coast and offshore (Heusser 1990).Settling through the marine column selects certain sizes and

shapes of pollen grains; Mudie (1982) noted, for instance,an increase in the percentage of the less-dense bisaccatepollen with increasing distance offshore, and size sorting ofsome pollen taxa favouring smaller sizes.

Reworking of older pollen and spores, and theirincorporation into marine sediments, has the potential toobscure the relationship between pollen and sporeassemblages and the contemporary vegetation cover. Thereare two major sources of reworked palynomorphs:sedimentary rocks—especially highly erodible ones such asmuds tone and limestone—and inshore marine sediments.Marine palynological records can therefore differ markedlyfrom those on the adjacent landmass according to theproportion of pollen and spores that come via an aerial orfluvial transport route, and according to the amount ofreworked palynomorphs they contain.

Some of the differences in pollen assemblages betweenP69 and Lake Maratoto are a result of differing source areas.While Lake Maratoto sits in the middle of a basin oncedominated by Dacrydium cupressinum forest (Newnham etal. 1989), before deforestation the mainland adjacent to andimmediately north of core P69 had abundant Prumnopitystaxifolia, which typically makes up one-half to three-quartersof the podocarp pollen influx in terrestrial cores (McGloneet al. 1993; Wilmshurst et al. 1999). Lake Maratoto also hasa much higher representation of non-wind-dispersed pollentypes such as Metrosideros, Nestegis, and Quintinia, whichare poorly represented in the extra-local pollen rain, whilewind-dispersed pollen, in particular that of podocarps andNothofagus, is overwhelmingly dominant in the pollenassemblages in core P69.

Corroded pollen and spores are much more common inmarine cores than in most lake or peat cores. Most soilsrapidly degrade pollen through biological activity, andtypically a residue of corrosion-resistant spores, and inparticular tree-fern spores, is left. Wilmshurst et al. (1999)showed that tree-fern spores were 2-6 times over-representedin a nearshore marine core relative to a mainland site. Tree-fern spore percentages are c. 2.5 times greater in core P69than in Lake Maratoto, which is representative of many

Dow

nloa

ded

by [

SUN

Y S

tate

Uni

v of

New

Yor

k G

enes

eo]

at 0

5:56

31

Oct

ober

201

4

McGlone—L. Quat. pollen record, core P69 73

Fig. 5 Scatter plot of pollenpercentages of Brassosporaversus Fuscospora in core P69assemblages. A, Pre-12 000 yr BP(R2 = 0.766; 0.919 if outlierremoved). B, Post-12 000 yr BP(R2 = 0.057).

40

30

CO

I 20

10

20 r

20

10 15 20 0

Brassospora %

North Island terrestrial pollen profiles. Disproportionatesurvivorship, therefore, is the best explanation for the higherrepresentation of fern spores in P69 relative to terrestrialcores.

Wilmshurst et al. (1999) noted in their Poverty Bayterrestrial-marine comparison that Fuscospora was nearlythree times as abundant in the marine core, and theysuggested that reworking of pollen from older Pleistoceneor Tertiary deposits contributed to this over-representation.Brassospora forms approximately one-third of totalNothofagus in core P69 and, as Fuscospora is alsorepresented in Tertiary deposits, it seems likely that someof its pollen is derived from the same source. To test thishypothesis, the correlation between Brassospora andFuscospora (and for other typical cool-climate taxa) wascalculated for the pre-12 000 yr BP interval, and also forthe post-12 000 yr BP interval (Table 3; Fig. 5). As erosionwas greater during the full-glacial period, and the resultingreworked pollen formed a greater proportion of the wholebecause of lower contemporary pollen production, it shouldbe expected that at least some taxa typical of partly vegetatedlandscapes would show a positive and significant correlationwith Brassospora before 12 000 yr BP, and this proves to

Table 3 Pearson's correlation coefficient (r) and significance (P)between Nothofagus subgenus Brassospora and selected pollentaxa, for the intervals before 12 000 yr BP (n = 17) and after 12 000yr BP (n = 11) in core P69. Brassospora is taken as a proxy forreworking of Tertiary-aged material into the core.

Before 12 000 yr BP After 12 000 yr BP

Pollen taxon

FuscosporaPhyllocladusAsteraceaePoaceaeHalocarpusNothofagus menziesiiLibocedrusCoprosma

r

0.8750.7100.6480.5890.4960.054

-0.081-0.129

P (2-tail)

*********

NSNSNS

r

0.2390.341

-0.016-0.230-0.627-0.016-0.614-0.100

P (2-tail)

NSNSNSNS*

NS*

NS

, f<0 .05 , **,/>< 0.01,***, P < 0.001; NS = not significant at= 0.05.

be the case for Asteraceae, Poaceae, and Halocarpus, all ofwhich are unlikely to have been derived from Tertiarysources. However, there is a highly significant correlationfor Fuscospora, a somewhat less-strong relationship withPhyllocladus, but no significant correlation for Libocedrusand Nothofagus menziesii, two other cool-temperate trees.When compared with full-glacial pollen spectra from centraland southern North Island (McGlone et al. 1996, fig. 4.10),there is nearly three times as much Fuscospora in the marinecore, but comparable amounts of Phyllocladus, Libocedrus,and Nothofagus menziesii. This suggests that a largeproportion of the pre-12 000 yr BP Fuscospora pollen isreworked from Tertiary sediments. The late postglacialincrease in Fuscospora seen in core P69 most likelyrepresents real change rather than reworking, as it has anon-significant correlation over this interval withBrassospora (Fig. 5) and the timing of the increasecoincides well with inland North Island sites (McGloneetal. 1996).

Empodisma is present in zone 7, but common in zone 6,and this wetland plant may possibly indicate extensive areasof oligotrophic mires on the poorly drained soils of theexposed continental shelf. Similar Last Glaciationoccurrences of Empodisma are recorded both on the Bay ofPlenty coast (McGlone et al. 1984) and offshore cores DSDP594 and S803 (Heusser & van der Geer 1994; Wright et al.1995). However, Empodisma pollen is sometimes found inquantity in estuarine sediments in catchments with no extantsource of this pollen type (McGlone unpubl.), and it seemslikely in these cases that it is reworked from older Pleistocenesediments. In the pre-12 000 yr BP sediments of core P69,Empodisma has no significant correlation with Brassospora(r = 0.351 ; P = 0.168), or with the other abundant herbaceouswetland type, Cyperaceae (r = 0.066; P = 0.802). It does,however, have a significant correlation with Halocarpus (r= 0.787; P < 0.001 ), a shrub or small tree often abundant inoligotrophic wetlands. Therefore, if it is reworked, it mustbe derived from a different source to that of Brassospora—Pleistocene sediments exposed by very low sea levels is apossibility.

Comparison with S803 suggests that P69 assemblagesare strongly influenced by pollen and spores derived fromthe Bay of Plenty hinterland and northwards. Above the

Dow

nloa

ded

by [

SUN

Y S

tate

Uni

v of

New

Yor

k G

enes

eo]

at 0

5:56

31

Oct

ober

201

4

74 New Zealand Journal of Geology and Geophysics, 2001, Vol. 44

12 14 16 18

SST (°C)Oxygen Isotope

20 8 10 12

Winter &

14 16 18 20 22

Summer SST (°C)Foram

0 0.5 1 1.5 2

P69:Pollen DCAAxis 1

0 0.5 1 1.5 2 2.5

Maratoto:Pollen DCAAxis 1

3 3.5 0 0.5 1 1.5 2

S803:Pollen DCAAxis 1

Fig. 6 Comparison of SST estimates (8l8O and foram in i feral) from core P69 (from data presented in Nelson et al. 2000) with the firstaxis of a DCA (detrended correspondence analysis) of the pollen records from core P69, Lake Maratoto, and marine core S803 (eigenvalues0.62, 0.64, 0.23, respectively).

location of core P69, the East Cape Current dominates(Fig. 1), bringing surface waters from the Bay of Plenty andHauraki Gulf southwards. Some pollen types (Phyllocladus,Libocedrus, Agathis) present in abundance in the postglacialportion of core P69 and in core S803 are either not commonor absent from the mainland adjacent to P69. Agathisaustralis, Phyllocladus cf. trichomanoides, and Libocedruscf. plumosa are common constituents of pollen diagramsfrom the Bay of Plenty lowlands (McGlone 1989; Giles etal. 1999) but not farther south (e.g., Wilmshurst et al. 1997,1999), and this area is the only likely source for these taxa.Some of the Agathis grains may be reworked from Tertiarysediments, but the late postglacial increase has the righttiming and magnitude to match onshore pollen results fromnorth of latitude 38°S (McGlone 1988;Newnhametal. 1989)and core S803. Pollen, whether wind or water transportedfrom the adjacent mainland, is unlikely to settle at P69because of the southwards flow of the surface waters andthe time taken for the organic complexes to sink through>2000 m of water. The P69 pollen profile, therefore,probably represents onshore vegetation from Hawke's Baynorthwards. Most pollen is likely to have been derived fromEast Cape-Bay of Plenty sources, with diminishing butsignificant input from the Agathis-domm&ted regions ofthe Coromandel Peninsula northwards. The East CapeCurrent may have affected representation of glacialvegetation at core P69. Full-glacial sediments from coreS803 average 41% pollen of warm-temperate conifers{Agathis, Dacrydium, Podocarpus, Prumnopitys), and17% for the full-glacial interval in core P69. Agathiscannot have been derived south of Coromandel at thattime and, taking its percentages as a guide, as much as80% of the warm-temperate forest conifer pollen in coreP69 during the full-glacial may have been sourced fromCoromandel north.

The P69 pollen diagram, therefore, is probably bestregarded as a blended representation of onshore vegetationover some 400-500 km of coastline. During interglacials,when pollen production was high in the southern part of thisrange, it seems likely that southern sources dominated pollenrepresentation; during full-glacial periods, representationwas more heavily weighted towards the northern part of thisrange, as wind-pollinated trees were then largely absent inthe south.

POLLEN DCAs, SEA-SURFACE TEMPERATURES,AND TERRESTRIAL DUST INFLUX

A number of environmental indicators have been analysedfor core P69, including quartz and carbonate influx, 518O,8I3C; planktonic foraminiferal species; and estimates of sea-surface temperature (SST) have been derived from 8I8O andforaminiferal transfer functions (Stewart & Neall 1984;Nelson et al. 2000). The first axis of the DCA for the threepollen diagrams considered here are compared with the 8I8Oand foraminiferal SST estimates (Fig. 6), and the terrestriallyderived aerosolic dust component is compared with selectedpollen curves for the late-glacial period (Fig. 7).

The foraminiferal transfer function SST estimates (Fig.6) show lowest temperatures (av. 8-9°C winter, 14-15°Csummer) between 22 000 and 18 000 yr BP, and theanomalously high oxygen isotope temperature estimates forthe same period are explained by strong upwelling inducedby strong westerly winds, which affect the position of theanalysed foraminifera in the water column (Nelson et al.2000). After 15 000 yr BP, temperatures estimated fromoxygen isotopes drop sharply, which is suggested to be aconsequence of the weakened upwelling, and from 12 000yr BP are consistent with the winter SST foraminiferalestimates. From 18 500 to 14 700 yr BP, SSTs range from11 to 17°C, averaging c. 2-A°C below those of the Holocene.At c. 6000 yr BP, peak SSTs—c. 1.5°C higher than presentin winter, and 2.1°C in summer—are reached, and then fallsteadily towards the present.

The first axis of the DCA at P69 contrasts taxa of warm,mild, moist climates (high values) with cool-temperate taxa(low values), and is best regarded as a summary index ofconditions for maximum plant growth. It is not closelyaligned with either of the SST indices. A sharp rise in theDCA curve occurs just after 14 700 yr BP, well after theinitial rise of the foraminiferal SST estimates and beforethe steep rise of the oxygen isotope derived SST estimates.However, this major change in the DCA curve is linked veryclosely with the change in the upwelling indicators, namelythe steep decline in oxygen isotope SST estimates after14 700 yr BP and the greatly reduced influx of carbonateand aerosolic quartz (Fig. 7). The even steeper increase ofthe first axis of the DCA at Lake Maratoto coincides almostexactly with declining upwelling offshore, as do pollen-

Dow

nloa

ded

by [

SUN

Y S

tate

Uni

v of

New

Yor

k G

enes

eo]

at 0

5:56

31

Oct

ober

201

4

McGlone—L. Qiiat. pollen record, core P69 75

Fig. 7 Comparison for the late-glacial transition in marine coreP69 of selected pollen curves,with aerosolic silica (2-5 um sizeclass) and biogenic carbonateinflux (from data presented inStewart & Neall 1984).

J/ •

i ibooö 20000' i 2 3

based indicators of afforestation from the central NorthIsland (McGlone & Topping 1977; Newnham & Lowe2000).

INTERPRETATION OF ONSHOREENVIRONMENTS FROM MARINE CORE P69

Zone 7 (c. 26 000-19 000 yr BP) of P69 represents thecooling phase of the full-glacial that culminates in zone 6(c. 19 000-15 000 yr BP). Zone 7 assemblages have muchhigher percentages of warm-temperate conifers and loweramounts of shrubland-grassland pollen types than zone 6,suggesting a greater forest cover. This is consistent with midlast-glacial sites from the Bay of Plenty and Gisborne, whichvariously record shrubland-grassland-A^o^o/agwi1 forestmosaics and patches of podocarp-dominant forest insheltered sites (McGlone et al. 1984).

Interpretation of zone 6 assemblages is more problem-atical because erosion, and thus reworking of Tertiary pollenand spores, peaked, and dominance of southwards-transported pollen and spore types can be expected to havebeen greatest. When these allowances are made, only a smallamount of warm-temperate conifer pollen (c. 1-2%) andtree-fern spores (c. 2-3%) can have been derived from thesouthern part of the source area, and possibly no more than10-15% Fuscospora. Inland full-glacial sites (McGlone1988; Newnham et al. 1989; Pillans et al. 1993) have verylow to trace amounts of podocarp pollen, confirming thatno large areas of podocarp forest can have been present. Ifthe pollen record of P69 is interpreted correctly here, forestrefugia along the coast must have been small and scattered.Relatively high percentages of Halocarpus, Phyllocladus,scrub types and grasses, and high influx of terrestrial dust(Stewart & Neall 1984), indicate an incompletely vegetatedlandscape largely covered with scrub and grassland.

Zones 7 and 6 coincide with an increase in the influx ofaerosolic quartz to the site, signifying accelerated rates ofregolith and rock erosion, fluvial transport of debris tooutwash river plains, and aerial transport by strong westerlywinds out to sea (Fig. 7) (Stewart & Neall 1984). Influx ofbiogenic carbonate and silica in the core is greatest in these

two zones, indicating greater upwelling as a consequenceof stronger winds. Zone 6 also has the highest percentagelevels of reworked extinct Brassospora, which can only havecome from erosion of Tertiary rocks and, in particular, fromthe east coast limestones. There is thus a good correlationbetween the sediment indicators of onshore erosion andwind, and the pollen evidence for partly vegetated landscapesand erosion.

Zone 5 represents an abrupt (c. 15 000-11 500 yr BP)transition from full-glacial open vegetation to podocarpforest dominance. The prominence of Dacrydium cupress-inum at the very beginning of the transition is likely torepresent the influence of vegetation in the Bay of Plentyand Coromandel region, as Dacrydium cupressinum did notplay an important role in the late-glacial south of the WaikatoBasin-Bay of Plenty regions (McGlone & Topping 1977;Lees 1986; Newnham & Lowe 2000). This zone coincideswith the abrupt diminution at c. 14 700 yr BP of aerosolicquartz influx, and a sharp decline in productivity at the site,both of which are linked with a southwards contraction ofthe strong glacial westerly windflow (Fig. 7) (Stewart &Neall 1984).

It has long been suggested that lower average temper-atures themselves could not have eliminated forest from thelowland North Island during the full-glacial, and that otherfactors such as windiness, drought, and frost must have beencrucial (e.g., McGlone 1988). The extreme full-glacialclimate which dominated south of the Northland Peninsulawas sustained by a steep meridional temperature gradientand associated strong westerly winds across central andsouthern New Zealand (Wright et al. 1995). When thisgradient abruptly contracted south, shortly after 15 000 yrBP, forest rapidly expanded across lowland and lowermontane areas (McGlone & Topping 1977; Newnham et al.1989; McGlone et al. 1993; Newnham & Lowe 2000),despite SST temperatures still being only midway to theirpostglacial peak (Fig. 6) ( Nelson et al. 2000).

Zone 4(11 500-10 000 yr BP) represents the latter halfof the late-glacial transition. Podocarp forest had fullyreoccupied the lowland to montane coastal regions by 11 500yr BP, and almost certainly all but the higher montane and

Dow

nloa

ded

by [

SUN

Y S

tate

Uni

v of

New

Yor

k G

enes

eo]

at 0

5:56

31

Oct

ober

201

4

76 New Zealand Journal of Geology and Geophysics, 2001, Vol. 44

subalpine sites inland (McGlone & Topping 1977; Newnham& Lowe 2000). Low representation of Dacrydium cupress-inum and tree ferns suggests a drier climate than that of earlyand mid postglacial. Reworked Brassospora pollen is at itslowest level during zone 4, and this could indicate lesserosion because of low rainfall under a fully forestedlandscape.

Zone 3 (c. 10 000-6000 yr BP) has a higher repre-sentation of Dacrydium cupressinum and tree ferns andabundance of the small tree Ascarina lucida, all indicatorsof moist, mild climates. SSTs are at a maximum in this zone,being some 1.5°C warmer than at present (Fig. 6) (Nelsonet al. 2000). In the upper half of zone 3, Fuscospora andPhyllocladus increase, reflecting a New Zealand-wide trendtowards more Nothofagus forest in the south, andPhyllocladus trichomanoides in the north (McGlone et al.1996). Zone 2 (c. 6000-800 yr BP) represents similarvegetation communities to those of the previous two zones,but Ascarina lucida is now absent, and Dacrydiumcupressinum, Fuscospora, Phyllocladus, and Agathis areeither at peak levels or increasing. Peak postglacialpercentages of Brassospora pollen in this zone could indicateincreasing rainfall accelerating erosion.

Zone 1 covers the period of human impact, as evidencedby the rise of Pteridium esculentum and decline of all foresttrees as a consequence of Polynesian forest clearance, andthen the first trace of European settlement in the form ofexotic pine pollen. In the marine core from Poverty Bay(Wilmshurst et al. 1999), the grass pollen curve rises towardsthe surface as a consequence of clearance of forest andbracken and its replacement by pasture during the Europeanera. There is no increase of grass pollen in the P69 profile,and it would appear that the last 100-120 years is missingor poorly represented in the core.

CONCLUSIONS

Core P69 demonstrates that excellent, high-resolution pollenanalytical records can be obtained from marine cores closeto the New Zealand mainland. However, over-representationof corrosion-resistant spores, dominance of the assemblagesby wind-pollinated trees, and derivation of the pollen andspores from several hundred kilometres of coast, meansinterpretation of the cores in terms of onshore vegetation isimprecise. Reworking of earlier Pleistocene and Tertiarypollen and spores as a result of onshore erosion, and drift ofpollen southwards on the East Cape Current, makesassessment of Last Glaciation pollen assemblages problem-atical, and vegetation reconstructions must be regarded astentative. Postglacial pollen assemblages are much lessaffected by reworking, and are comparable to those fromterrestrial sites in the pollen source areas except for under-representation of taxa with non-wind-dispersed pollen andover-representation of tree-fern spores.

An unexpected finding of this work is the lack of a closecorrelation between full-glacial and late-glacial changes inSSTs and afforestation of the adjacent landmass as indicatedby marine pollen assemblages. Although it is well understoodthat mean annual temperatures and precipitation are onlytwo of several factors controlling forest distribution(McGlone 1988), nevertheless, the tendency has been tointerpret vegetation change in terms of temperature andprecipitation alone. The P69 record suggests that late-glacial

climate changes are better thought of as resulting from acomplete alteration to the climatic system as a result of thecontraction of the glacially expanded westerly winds. SSTshifts in the adjacent ocean are part of this change but, fromthe forest viewpoint, not necessarily at all times the mostimportant.

Before much further progress can be made in inter-pretation of terrestrial vegetation change, a systematiccomparison throughout the New Zealand region of marinecore-top pollen assemblages with a wide range of terrestrialsites is needed. There is, moreover, much more that can bedone with marine pollen records. Features of the offshorepollen assemblages such as corroded and reworkedpalynomorphs, and pollen of wetland plants, while of littleor no use in vegetation reconstructions, may prove highlyuseful as sensitive indicators of terrestrial erosion, rainfall,and sea-level. There is a great potential also for statisticalcomparison of long marine pollen records spanning severalglacial-interglacial cycles with other records from the samecores that depict global ice volume, SST, and upwelling,permitting a much more sophisticated understanding ofocean-land interactions.

ACKNOWLEDGMENTS

Bob Stewart suggested that I work on core P69, and Lionel Carterprovided access to the core material; both have also been of greatassistance through discussion of the results and provision ofinformation. I thank Janet Wilmshurst for reviewing the draftmanuscript; and Mike Elliot and an anonymous reviewer forhelpful comments in their referee reports for the journal. AlisonWatkins prepared the pollen slides, and Kirsty Cullen drafted thefigures. Finally, I would like to record my appreciation of PatSuggate's support over the years for Quaternary paleoecologicalresearch in New Zealand. Funds for this research were providedby the Foundation for Research, Science and Technology (contractC09624).

REFERENCES

Dudley, W.C.; Nelson, C. S. 1988: The δ I 3C content of calcareousnannofossils as an indicator of Quaternary paleo-productivity in the Southwest Pacific region. Note. NewZealand Journal of Geology and Geophysics 31: 111-116.

Elliot, M. B. 1998: Late Quaternary pollen records of vegetationand climate change from Kaitaia Bog, far northern NewZealand. Review of Palaeobotanv and Palvnologv 99:189-202.

Froggatt, P. C ; Lowe, D. J. 1990: A review of late Quaternarysilicic and some other tephra formations from NewZealand: their stratigraphy, nomenclature, distribution,volume and age. New Zealand Journal of Geology andGeophysics 33: 89-109.

Gauch, H. G. 1982: Multivariate analysis in community ecology.Cambridge, Cambridge University Press. 298 p.

Giles, T. M.; Newnham, R. M.; Lowe, D. J.; Munro, A. J. 1999:Impact of tephra fall and environmental change: a 1000year record from Matakana Island, Bay of Plenty, NorthIsland, New Zealand. In: Frith, C. R.; McGuire, W. J. ed.Volcanoes in the Quaternary. The Geological Society ofLondon Special Publications 161: 11-26.

Green, J. D.; McGlone, M.S.; Lowe, D. J.; Boubee, J. A. T. 1984:Aspects of the developmental history of Lake Maratoto,North Island, New Zealand. Verhandlungen Inter-nationalen Vereinigung fur Limnologie 22: 1416.

Dow

nloa

ded

by [

SUN

Y S

tate

Uni

v of

New

Yor

k G

enes

eo]

at 0

5:56

31

Oct

ober

201

4

McGlone—L. Quat. pollen record, core P69 77

Heusser, L. E. 1990: Northeast Asian pollen records for the last150,000 years from deep-sea cores V28-304 and RC14-99 taken off the Pacific coast of Japan. Review ofPalaeobotany and Palynology 65: 1-8.

Heusser, L. E.; van der Geer, G. 1994: Direct correlation ofterrestrial and marine palaeoclimatic records from fourglacial-interglacial cycles—DSDP site 594, southwestPacific. Quaternary Science Reviews 13: 257-272.

Hill, M. O.; Gauch, H. G. 1980: Detrended correspondenceanalysis: an improved ordination technique. Vegetatio 42:47-58.

Hill, R. S.; Read, J. 1991: A revised infrageneric classification ofNothofagus (Nothofagaceae). Botanical Journal of theLinnean Society 105: 37-72.

Lees, C. M. 1986: Late Quaternary palynology of the southernRuahine Range, North Island, New Zealand. New ZealandJournal of Botany 24: 315-329.

McGlone, M. S. 1988: New Zealand. In: Huntley, B.; Webb III, T.ed. Handbook of vegetation science, Vol. 7: Vegetationhistory. The Hague, Kluwer Academic Publishers. Pp.558-599.

McGlone, M. S. 1989: Application of pollen analysis toenvironmental monitoring. In: Craig, B. ed. Proceedingsof the Environmental Monitoring in New ZealandSymposium. Wellington, Department of Conservation. Pp.220-224.

McGlone, M. S.; Topping, W. W. 1977: Aranuian (post-glacial)pollen diagrams from the Tongariro region, North Island,New Zealand. New Zealand Journal of Botanv 21:749-760.

McGlone, M. S.; Topping, W. W. 1983: Late Quaternary vegetation,Tongariro region, central North Island, New Zealand. NewZealand Journal of Botany 21: 53-76.

McGlone, M. S.; Howorth, R.; Pullar, W. A. 1984: Late Pleistocenestratigraphy, vegetation and climate of the Bay of Plentyand Gisborne regions, New Zealand. New Zealand Journalof Geology and Geophysics 27: 327-350.

McGlone, M. S.; Salinger, M. J.; Moar, N. T. 1993: Palaeo-vegetation studies of New Zealand's climate since the LastGlacial Maximum. In: Wright, H. E.; Kutzbach, J. E; WebbIII, T.; Ruddiman, W. F.; Street-Perrrott, F. A.; Bartlein, P.J. ed. Global climates since the Last Glacial Maximum.Minneapolis, University of Minnnesota Press. Pp.294-317.

McGlone, M. S.; Mildenhall, D. C ; Pole, M. S. 1996: History andpaleoecology of New Zealand Nothofagus forests. In:Veblen, T. T.; Hill, R. S.; Read, J. ed. The ecology andbiogeography of Nothofagus forest. Yale, New Haven,University Press. Pp. 83-130.

Mildenhall, D. C. 1995: Pleistocene palynology of the PetoneandSeaview drillholes, Petone, Lower Hutt Valley, NorthIsland, New Zealand. Journal of the Roval Society of NewZealand 25: 207-262.

Moore, P. D.; Webb, J. A.; Collinson, M. E. 1991: Pollen analysis.London, Blackwell Scientific Publications. 216 p.

Mudie, P. J. 1982: Pollen distribution in recent marine sediments,eastern Canada. Canadian Journal of Earth Science 19:729-747.

Nelson, C. S.; Hendy, C. H.; Jarrett, G. R.; Cuthbertson, A. M.1985: Near-synchroneity of New Zealand alpine glaci-ations and Northern Hemisphere continental glaciationsduring the past 750 kyr. Nature 318: 361-363.

Nelson, C. S.; Hendy, 1. L., Neil, H. L.; Hendy, C. H.; Weaver, P.P. E. 2000: Last glacial jetting of cold waters through theSubtropical Convergence zone in the Southwest Pacificoff eastern New Zealand, and some geological impli-cations. Palaeogeography, Palaeoclimatologv, Palaeo-ecology 156: 103-121.

Newnham, R. M. 1992: A 30,000 year pollen, vegetation andclimate record from Otakairangi (Hikurangi), Northland,New Zealand. Journal of Biogeography 19: 541-554.

Newnham, R. M.; Lowe, D. J. 2000: Fine-resolution pollen recordof late-glacial climate reversal from New Zealand. Geology28: 759-762.

Newnham, R. M.; Lowe, D. J.; Green, J. D. 1989: Palynology,vegetation and climate of the Waikato lowlands, NorthIsland, New Zealand, since c. 18,000 years ago. Journalof the Royal Society of New Zealand 19: 127-150.

Pickrill, R. A. 1987: Circulation and sedimentation of suspendedparticulate matter in New Zealand fiords. Marine Geology74:21-39.

Pillans, B.; McGlone, M. S.; Palmer, A.; Mildenhall, D. C ;Alloway, B.; Berger, G. 1993: The Last Glacial Maximumin central and southern North Island, New Zealand: Apaleoenvironmental reconstruction using the KawakawaTephra Formation as a chronostratigraphic marker.Palaeogeographv, Palaeoclimatologv, Palaeoecology 101:283-307.

Stewart, R. B.; Neall, V. E. 1984: Chronology of palaeoclimaticchange at the end of the last glaciation. Nature 311: 47-48.

Sun, X. J.; Li, X.; Beug, H. J. 1999: Pollen distribution inhemipelagic surface sediments of the South China Sea andits relation to modern vegetation distribution. MarineGeology 156: 211-226.

Weaver, P. P. E.; Carter, L; Neil, H. L. 1998: Response of surfacewater masses and circulation to late Quaternary climatechange east of New Zealand. Paleoceanography 13: 70-83.

Wilmshurst, J. M.; McGlone, M. S.; Partridge, T. R. 1997: A lateHolocene history of natural disturbance in lowlandpodocarp/hardwoood forest, Hawke's Bay, New Zealand.New Zealand Journal of Botany 35: 79-96.

Wilmshurst, J. M.; Eden, D. N.; Froggatt, P. C. 1999: LateHolocene forest disturbance in Gisborne, New Zealand: acomparison of terrestrial and marine pollen records. NewZealand Journal of Botany 37: 523-540.

Wright, I. C ; McGlone, M. S.; Nelson, C. S.; Pillans, B. J.1995: An integrated Late Quaternary (Stage 3 to Present-day) paleoclimate and paleoceanographic record fromoffshore northern New Zealand. Quaternary Research44: 283-293.

Dow

nloa

ded

by [

SUN

Y S

tate

Uni

v of

New

Yor

k G

enes

eo]

at 0

5:56

31

Oct

ober

201

4