Lagoon subsidence and tsunami on the West Coast of New Zealand
S.L. Nichol a,⁎, J.R. Goff b,1, R.J.N. Devoy c,2, C. Chagué-Goff b,1,B. Hayward d, I. James e
a School of Geography, Geology and Environmental Science, The University of Auckland, Private Bag 92019, Auckland, New Zealandb National Institute of Water and Atmospheric Research Ltd., PO Box 8602, Christchurch, New Zealand
c Department of Geography, University College Cork, Cork, Irelandd Geomarine Research, 49 Swainston Road, St. John's, Auckland, New Zealand
Multiproxy analysis of coastal wetland sediments inthe Northern Hemisphere has produced important re-cords of past environmental changes, including extremeevents such as earthquakes and tsunamis (e.g., Atwaterand Moore, 1992; Shennan et al., 1996). Research hasgenerally focused on sites where a demonstrable linkage
can be made between tsunamis and their generatingmechanism (e.g., Pacific northwest, Atwater, 1987;Clague, 1997; Clague et al., 2000). This work and workon other palaeotsunami deposits elsewhere (Dawsonet al., 1988; Bryant et al., 1992; Minoura et al., 1996)have served to considerably improve our geologicalunderstanding of palaeoseismic-related signatures pre-served in the sedimentary record. In New Zealand, wehave the opportunity to test and develop this approachby utilising a largely undisturbed and unstudied coastalsediment record that has formed in a seismically activesetting (Goff et al., 2001).
New Zealand sits astride the boundary between twomajor tectonic plates, the Australian and the Pacific. Inhistorical times seismic activity has ranged from barelydetectable earthquakes and tsunamis that have caused no
249S.L. Nichol et al. / Sedimentary Geology 200 (2007) 248–262
damage, to large-scale earth and sea movements (Eiby,1982; Hull, 1986). Recent reviews of past tsunamis inNew Zealand estimate that large, locally generatedevents, possibly with a nationwide impact throughoutthe North and South Islands, occur about once every500 years (Goff and McFadgen, 2002). The most recentevents in New Zealand occurred around the mid 15thcentury AD and are associated with a cluster of largeearthquakes, one of which was an Alpine fault rupture(Goff and McFadgen, 2002).
The Alpine fault is over 400 km long, extendingalmost the entire length of the South Island (Berrymanet al., 1992; Bull, 1996; Norris and Cooper, 1997). This
Fig. 1. Location map showing Okarito Lagoon, W
single right-lateral, oblique slip fault raises the SouthernAlps at a rate of about 5–8 m per 1000 years (Bull,1996) (Fig. 1). The last three major ruptures (magni-tudes of ca. 8.0 Mw) are believed to have occurred in AD1717 (along a 375 km segment), ca. AD 1630 and ca.AD 1460 (involving a minimum 300 km segment, witha vertical offset of 2.15±0.4 m) (Yetton et al., 1998;Wells et al., 1999). A smaller rupture occurred in AD1826 (Wells et al., 2001; Cullen et al., 2003).
This study examines the recent sedimentary record ofOkarito Lagoon on the West Coast of the South Island,with the aim of testing for evidence of co-seismic sub-sidence of the lagoon floor and tsunami inundation,
Fig. 2. Aerial photograph mosaic (unrectified) for Okarito Lagoontaken January 1988, showing core and trench sites for this study andinterpreted former tidal inlets (marked as blind channels). Source:Land Information New Zealand, courtesy Department of Conserva-tion. Crown Copyright Reserved.
given the context of frequent Late Holocene seismicity.To achieve this we undertook a multiproxy analysis ofsediment cores and trenches from the lagoon. This paperis an extension to earlier work on the same sample setpublished by Goff et al. (2004).
2. Okarito Lagoon
Okarito Lagoon, Westland (43°11′S, 170°14′E) isone of the largest estuarine inlets on the West Coast ofthe South Island. It is 10 km long and 2–4 km wide withan open water area of ∼20 km2 and fringing salt marsharea of ∼10 km2 (Fig. 1). The lagoon is separated fromthe sea by a narrow sand barrier with a maximumelevation of ∼4 m. The width of the barrier decreasessouthward from 500 m to 100 m at Okarito village,where an ephemeral tidal inlet provides the only directconnection to the open coast. From vertical aerialphotographs it is evident that the lagoon has a complexhistory of barrier breaching, with four previous tidalinlets ending as blind channels against the landward sideof the barrier (Fig. 2). The form of these channels hasbeen maintained as surrounding marshes have devel-oped upon probable relict flood tidal delta surfaces. Ourinvestigation focused on the southernmost marsh islandaround which the active inlet channel bifurcates, and onthe tidal flat to the south of the main channel. Tidalrange on the coast varies from 1 m on neaps to 2.1 m onspring tides. Holocene sea-level history for the WestCoast is poorly understood in detail, so we assume asimilar history for elsewhere in New Zealand whichrecords stillstand since ca. 6.5 ka (Gibb, 1986).
Rainfall in the region exceeds 10,000 mm/year,resulting in high erosion rates (5–12 mm/year) of thefractured, easily eroded schists and glacially derivedmaterials found to the east of the Alpine fault (Berry-man et al., 1992; Norris and Cooper, 1997). To the westof the Alpine fault, the overprint of Late Quaternaryglacial activity is evident in the large moraines thatsurround Okarito Lagoon (Griffiths and McSaveney,1986). Freshwater flow into the lagoon is derived fromthree sub-catchments; Okarito Forest (9700 ha) whichsupplies six small creeks that drain the moraines andenter along the northeast shoreline, and; Lake Mapour-ika (ca. 9000 ha) and Waitangi-taona River (8300 ha)catchments which drain to Okarito River and DeepCreek at the southern end of the lagoon. These catch-ments extend further inland to drain terrain formed inschist and glacial outwash and are traversed by theAlpine fault in their upper reaches (Fig. 1). The majo-rity of the sediment load is captured in Lake Mapourikaand Lake Wahapo, however, and it is the moraines that
251S.L. Nichol et al. / Sedimentary Geology 200 (2007) 248–262
provide the main source of coarse material enteringOkarito Lagoon (MacPherson, 1981). At its mouth,Okarito River is constructing a small (∼1 km2) deltathat has infilled the southeast corner of the lagoon. Araised shoreline bench, estimated at 1 m above presentwater level, fringes the landward side of Okarito La-goon and is backed by shallow (b5 m) caves cut intocliffed sections of moraines.
An additional key component to landscape develop-ment in the West Coast region is seismicity. In parti-cular, ruptures on the Alpine fault have been shown tohave resulted in widespread forest destruction due toground shaking (Wells et al., 1998, 2001; Cullen et al.,2003), river aggradation (Yetton et al., 1998), and rapidcoastal dune building (Goff and McFadgen, 2002); thelatter two being direct consequences of increased se-diment yield from hillslopes following earthquake-induced landslides. It is in this context of strong seismicoverprinting on the coastal landscape that we recognizedthe potential for a record of seismic events to be re-corded in Okarito Lagoon.
3. Data sources and analytical techniques
Sediment cores and trenches were used to investigatethe shallow stratigraphy of southern Okarito Lagoon.Cores were collected by the vibracore technique at twosites from the lagoon–marsh interface and close to highwater mark (Fig. 2). Compaction was measured prior tocore recovery. In the laboratory, cores were splitlengthwise, logged and sub-sampled for grain size,organic content, micro- and macrofossils, geochemical,and radiocarbon analyses. Trenches (1–3) were dug byhand to confirm the lateral continuity and character ofsedimentary units observed in the cores. Additionalsediment samples were collected from trenches. Grainsize was measured on 23 core and 12 trench sub-samplesusing a laser particle sizer (Galai™) system thatdetermines particle size using the time-of-transitionprinciple (Molinaroli et al., 2000). Results are reportedfor particle volume measurement. For trench 1, wedetermined grain size on two samples by mechanicalsieving at half-phi intervals from −1 to 4 phi. Organiccontent of cores was determined via loss-on-ignition(LOI) treatment of 17 sub-samples by ashing at 500 °C for4 h. Results are reported on a dry weight basis. Elementalanalysis of seven sub-samples fromcore 6was undertakenusing ICP-AES, following sample preparation by a multi-acid ‘total’ digestion. Geochemical data have beennormalised for grain size following Loring (1991).
For diatom analysis, 15 sub-samples from core 6were prepared following standard techniques (e.g., Bat-
tarbee, 1986). Fossil diatoms were mounted in Naphraxand counted using light microscopy at a magnificationof ×1000. A minimum count sum of 600 diatom valveswas used. Identification of diatom taxa and their palaeo-environmental interpretation is based upon establishedfloras and other texts, including Hustedt (1927–1966,1930, 1957), Van der Werff and Huls (1957–1974),Cleve-Euler (1951–1955), Krammer and Lange-Berta-lot (1986–1991), Hendey (1964), Round et al. (1990),Simonsen (1967), De Wolf (1982), Admiraal (1984),Denys (1991a,b), Vos and de Wolf (1993a,b, 1994),Foged (1977, 1978, 1979). John (1983), Crosby andWood (1958, 1959), Wood et al. (1959), and Wood(1961, 1963). Results are presented as a percentagefrequency diagram, showing the key diatom specieswithin salinity (Halobion) groupings.
For foraminiferal analysis, eight 10-cm3 samples ofsediment were processed from cores 4 and 6. Sampleswere washed over a 63-μm sieve to remove mud andheavy liquid floatation used to concentrate foram tests insandy samples. Census counts of all benthic foramswere made for each sample. Estimates of the tidal ele-vation at which each foram species was deposited weremade using the Modern Analogue Technique (Haywardet al., 2004), based on relative abundance data for 250benthic foram fauna from modern estuaries and coastallagoons around New Zealand (data from Hayward et al.,1999a). Similar techniques to determine estimates oftidal elevation based upon foram assemblages have beenused elsewhere (Scott and Medioli, 1986; Van de Plas-sche et al., 1998). In this study, elevation estimates arepresented as tidal range derived from the five mostsimilar modern faunas in the analogue set. The reli-ability of these estimates depends on a number offactors, including the range of tidal levels; depths andenvironments represented by the analogue samples; andthe breadth of the tidal and depth ranges of the majortaxa. Previous studies (e.g., Hayward et al., 1999b) haveshown that the most precise tidal ranges can be obtainednear high tide level from marsh faunas, with far broaderranges observed in intertidal mud and sand flats and insubtidal environments. Some taxa with restricted hightidal ranges in near normal salinity situations (e.g.,Haplophragmoides wilberti, Trochamminita salsa, Mili-ammina fusca) are known to live in abundance throughthe entire tidal range and also subtidally in more brack-ish environments (e.g., Hayward et al., 1999a). Thus anassessment of the setting and probable salinity of thelagoon or estuary at the time is important in estimatingtidal level.
Foram data also provided for calculation of an arti-ficial salinity index (SI). This was done using detrended
and canonical correspondence analysis, based on therelative abundance of foram species following Haywardet al. (2004). SI estimates range between 0 (freshwater)and 10 (normal marine salinity).
4. Results
4.1. Stratigraphy and sediment texture
All five sample sites preserve a similar stratigra-phic record within the upper metre (Figs. 3–5). Thebulk of the sediment matrix comprises massive beds
Fig. 3. Graphic logs fo
of fine to medium sand and silt with local shell andwood fragments. Within this, each core and trenchrecovered a buried marsh soil at variable depths,ranging from 50 cm (trench 3) to 88 cm (trench 1)with depths in the cores falling within this range. Thesoil unit is olive-brown to grey-olive silt that enclosesfine root hairs and wood fragments. It is in sharpupper contact with overlying sediments that are dis-tinguished by a colour change to light grey and ab-sence of organic material, an observation supportedby LOI results which show a 1–2% decrease acrossthis contact (Fig. 6).
r cores 4 and 6.
Fig. 4. Graphic logs for trenches 1–3.
253S.L. Nichol et al. / Sedimentary Geology 200 (2007) 248–262
In trench 1, the buried soil is overlain by a 23-cm-thick bed of shell-bearing gravel and coarse sand. Gra-vels are rounded and shell remains include whole andarticulated bivalves near the base of the unit thatbecome more commonly broken toward the upper con-tact at 65 cm depth. Shell species are dominated by thecockle Austrovenus stutchburyi, with rare examples ofthe intertidal bivalve Mactra ovata and the gastropodAmphibola crenata. In addition, remains of macro
algae, most likely seaweed, were recovered. On thelandward side of the lagoon (core 4 and trench 2)sediment above the soil comprises 20–30-cm-thick bedsof very fine sand that are massive and carry shell imprintsand wood fragments. In trench 2, the contact is on anirregular surface and grey fine sands have filled verticalcracks that penetrate 10 cm into the buried soil; thesecracks may have originated as burrows or, more likely,via desiccation of the soil when exposed subaerially. In
Fig. 5. (A) Core 6 showing soil surface buried 77 cm below modern tidal flat and articulated bivalve (Austrovenus stutchburyi) dated to 900±60 yearsBP; (B) trench 1 showing buried soil overlain by rounded pebbles, gravel and shell, grading upward to very fine sand. Scale card is 10 cm long;(C) trench 2 showing buried soil overlain by very fine sand; arrow indicates infilled desiccation crack or burrow. Tape is 50 cm long; (D) trench 3showing buried soil overlain by fine sand.
trench 3, the buried soil is overlain by a 4-cm-thick bedof coarse sand and medium-sized gravel that gradesupward to a 40-cm-thick deposit of very fine sand.
4.2. Geochemistry
A suite of elements assayed from core 6 show a clearpeak in concentration within the silty sands that overliethe buried soil (Fig. 6). Thus, iron (Fe), sulphur (S),titanium (Ti), strontium (Sr), barium (Ba), calcium (Ca)and sodium (Na) all increase markedly at 77 cm depthand then decrease within the successive bed of fine-medium sand to 20 cm depth. Although S occurrence insediments is mostly related to post-depositional diage-netic processes, Fe-rich sediments are more likely to fixS than Fe-poor sediments. Thus S and Fe can be usedtogether as indicators of palaeosalinity in sediments(Thomas and Varekamp, 1991; Daoust et al., 1996;Chagué-Goff and Goff, 1999). Na has also previouslybeen used as a proxy for marine influence in marsh andlagoon systems, because of the relatively high Na con-tent in seawater compared with freshwater (López-Buendía et al., 1999; Chagué-Goff et al., 2002). Simi-larly, Sr, Br and Ca also occur in higher concentrationsin marsh and wetland sediments inundated by saltwater
(Minoura et al., 1994; Chen et al., 1997). Titanium isused here as an indicator of heavy mineral concentration(e.g., magnetite). Therefore, the peak Ti concentrationdirectly above the buried soil within core 6 is interpretedto record the preferential sorting and concentration ofheavy minerals. In a lagoon, this particular responsewould most likely occur during a high energy event suchas initiated by tsunami, storm surge or large river flood.Collectively, the measured elements all indicate a sud-den rise in the salinity of lagoon waters, followed by agradual return to brackish conditions.
4.3. Diatoms
Diatom analysis of core 6 provides further evidenceof palaeoenvironmental change for Okarito Lagoon(Fig. 7). Fossil diatoms are generally well preserved andabundant, providing a flora of N240 taxa. In the lowerpart of the core (87–100 cm depth) brackish water taxa(mesohalobion, M; oligohalobion, OI-indifferent andOH-halophile) comprise N50% of the diatom assem-blage. Melosira juergensii and Mastogloia elliptica do-minate the M taxa, with lower but significant values ofMastogloia pseudoexigua, Cyclotella operculata var.mesoleia and Rhopolodia gibberula (OH — indicative
Fig. 6. (A) Geochemical data for core 6, also showing trends in silt and organic content. (B) Trends in silt and organic content for core 4.
255S.L. Nichol et al. / Sedimentary Geology 200 (2007) 248–262
of sandy conditions). The OI group, rich in Cocconeisplacentula and Cyclotella stelligera, maintains frequen-cies N32% throughout the interval. There is a distinctbackground marine influence (polyhalobion, P), but atfrequencies b15%. Marine diatoms are representedby Diploneis suborbicularis, Raphoneis surirella,Raphoneis minutissima, Thalassiosira eccentrica andParalia sulcata. These are littoral species, as are manyof the M group; together representing shallow marine-brackish water of sandy foreshore, sand flat and mudflat environments (Denys, 1985). Open coast marinespecies are also recorded in the sediments, confirmingtidal exchange at the site. These marine diatoms aresparse and often broken, indicating their presence isprobably the result of progressive inwashing and re-working of more open water, sandy marine sediments.
The interval between 67 and 87 cm depth is dis-tinguished by alternating peaks of marine (P) andbrackish–freshwater (OH, OI) groups. Marine diatomshave peak concentration at 85 cm (20%) and 73 cm(27%), the former associated with the buried soil andthe latter with the overlying fine sand. Species typicalof sandy–mud flat environments are prominent inthese apparently marine-enriched sediments, with in-creased values of Raphoneis spp., Cocconeis scutellum,Nitzschia granulata, N. punctata, Striatella unipunctataand Licmophora spp. Many other marine speciesoccur in low numbers, including Cerataulus turgidus,Actinocyclus octonarius var. crassus, Grammatophoraoceanica, Triceratium spp. and Navicula lyra.
A decline in marine taxa (P) occurs above 85 cmdepth through the buried soil and high values of brackish
(OH) and freshwater taxa (OI) are recorded, with indi-vidual maxima for R. gibberula, C. placentula andNavicula tripunctata. Full brackish water species (M)increase above the soil with Mastogloia spp., Melosirajeurgensii reaching peak frequencies at 75–73 cm, withlower values of Synedra fasciculata and Nitzschiasigma. Many of these OI–M group diatoms are welladapted to intertidal–salt marsh conditions and aretolerant of drying-air exposure.
Between 67 and 30 cm depth, freshwater (OI) speciesdominate and frequencies of brackish (M) diatomsdecline from N30% to ∼20%. Above 67 cm the marine
signal collapses to b2% above 50 cm. The OI and OHdiatoms (together N60% at the base of this interval) arejoined by a range of other fresh–brackish water species,increasing to near 80% at 30 cm. These include manyAchnanthes spp., with Achnanthes hungarica replacingC. placentula as the dominant taxon at the top of thezone. The diatom assemblage suggests an overall prog-ressively shallowing, fresh–brackish water and riverflood-dominated environment.
Many of the fresh and brackish water diatom valves(OI/OH and M groups) are relatively unworn and wholevalves dominate the general diatom assemblages (Fig. 7).
257S.L. Nichol et al. / Sedimentary Geology 200 (2007) 248–262
This lack of breakage is unusual given the potential forhigh-energy conditions and the reworking of sedimentsat the site. Many of the OI–Mdiatoms however representlow to medium salinity environments (Cl b20 ppt) andare predominantly of local to in situ origin, with growthunder intertidal–subaerial conditions. Peaks in key fresh-water (OI) species between 87 and 30 cm are associatedwith terrestrial plant debris most probably introducedduring river flooding. Overall, the diatom record fromcore 6 shows deposition generally in brackish to brack-ish–freshwater environments of estuary–lagoonal–marsh settings. Toward the top of the sedimentarysuccession a marked decline in the percentages of marinespecies and the associated increase in freshwater speciesindicate an increasing isolation from the sea.
The higher resolution diatom analysis between 67and 87 cm shows at least one distinctive marine ‘event’in the sedimentary sequence. It is worth noting that, if alower sampling resolution had been maintained, then theextremes in diatom values (75–70 cm) would not havebeen apparent. Two peaks in marine diatoms are re-corded. The first, at 85 cm, coincides with the base ofthe buried soil. In the context of a consistent marinediatom signal this peak is most likely an artifact of thestratigraphy and sedimentary processes. Marine diatomstend to be more heavily silicified and resistant to de-struction. Local reworking of sediments to concentratethe marine diatoms present may be the explanation, assupported by the increase in broken valves at this level.The second marine peak (at 73 cm) is more distinctiveand associated with the sand bed overlying the buried
Table 1Foraminifera data for Okarito cores 4 and 6
Core 6 a
Depth (cm) 5–7 50–52 57–59 62Number of species 5 3 2 5Total foram count 104 201 3 25Relative abundance %
Elevation (m above MSL) 1.9–2.2 2.0–2.2 n/a c 2.2Salinity index 2.6 2.4 2.6 3.3a Core 6 from edge of Leptocarpus similis salt marsh at neap high tide levb Core 4 from Leptocarpus similis salt marsh at mean high tide level, ∼2c n/a indicates insufficient foram preservation to allow application of MAT
soil. This peak is bracketed by higher concentrations offresh–brackish water diatoms (e.g., 80 cm within thesoil, 71 cm in fine sand) that we interpret to represent theprevailing background shallow water, intertidal to saltmarsh environment at the site.
4.4. Foraminifera
Forams extracted from cores 4 and 6 are used here toreconstruct palaeo-elevation ranges for depositionalsurfaces (Table 1). A total of 10 taxa are reported, invarying abundance and diversity through the sedimentrecord. In core 6, a sample taken directly below thecontact between the buried soil and overlying sand (79–81 cm) is co-dominated by agglutinated faunas thatindicate weakly brackish conditions (Ammotium fragile,H.wilberti) on a lagoon ormarsh surface that was close tomean high tide level (range 2–2.2m aboveMSL).Withinthe overlying sand (71–73 cm), the fauna is stronglydominated by calcareous taxa (i.e., Ammonia aoteana)indicative of increased salinity and a broader elevationrange spanning subtidal (2.5 m below MSL) to mid-tide(1.75 m above MSL). Four samples taken from theinterval 64–5 cm in core 6 all indicate a return to lowersalinity, based on consistently high values forH. wilbertiand the strong presence of A. fragile and M. fusca.Estimated elevations range from mid-tide (1.9 m aboveMSL) to mean spring high tide level (2.2 m aboveMSL),most likely in a brackish marsh environment.
Foram concentrations from the two samples takenfrom core 4 are very low, allowing estimation of tidal
level for only one sample (76–78 cm, Table 1). The faunafrom this sample indicate a wide elevation range, fromsubtidal (2.5 m below MSL) to mean high water neap(1.75 m above MSL).
In summary, the forams identified from these coresprovide evidence for abrupt changes in salinity duringthe accumulation of the upper metre of Okarito Lagooninfill. We have confidence in this evidence because thepreservation state of tests is moderate to good, there isno evidence for reworking of older tests into the sam-pled sediments and open marine fauna were not re-corded; together indicating an intact and in situ faunalassemblage that reliably reflects palaeo-salinity condi-tions at the sample sites.
4.5. Radiocarbon ages
Radiocarbon results for four samples taken from cores4 and 6 are presented in Table 2. Samples of buried soiland root material from within yielded similar uncalibrat-ed ages of 1800±60 years BP (Wk8989; core 4) and2170±70 years BP (Wk8987, core 6). Two specimens ofarticulated bivalve, taken from the base of the buried soilin core 6, yielded contrasting ages of 10,580±60 yearsBP (Wk8618) and 900±60 BP (Wk8986). We view theolder age as unreliable on the basis that sea level wasabout 30m lower than present at∼10 ka and the area nowoccupied by Okarito Lagoon would have ice-marginalterrain. Contamination of sample Wk8618 by old carbonis therefore assumed and the age rejected. The trueMiddle Holocene age for the lower part of the lagooninfill is supported by a C14 age of 6380±60 years BP onarticulated cockle (A. stutchburyi) recovered from 326 cmdepth in core 6 as part of a previous study (Goff et al.,2004).
5. Interpretation and discussion
The preservation of a buried soil at similar depthsbelow the modern salt marsh and fringing tidal flat ofOkarito Lagoon is strong evidence for localized sub-
sidence. Results from diatom- and foram-based recon-struction of palaeo-salinity support an interpretation ofabrupt changes in local environmental conditions acrossthe contact between the buried soil and overlying de-posits. Thus, we recognize an increase in the salinity oflagoon waters associated with increased water depth ofapproximately 50 cm. Further evidence for increasedtidal incursion at core site 6 is given by geochemical datathat record a marked increase in a suite of elements toindicate marine-enriched lagoon waters (especiallyS, Na, Ca and Ba), associated with sediments directlyabove the buried soil. We note, however, that thesesediments do not contain any open marine foraminiferaspecies, so that evidence for transport of material fromthe open coast is lacking.
Sediments directly overlying the buried soil range ingrain size from coarse sand and gravel (trench 1 and 3)to silty fine sand (core 4 and 6, trench 2). Our samplingdoes not allow for detailed mapping of these deposits,but local variability of grain size such as between core 6and trench 1 (a distance of ∼100 m) suggests thatsediment texture may vary with proximity to tidal chan-nels, with channels being the likely source for localreworking of coarse-grained sediments. These depositsgrade-upward to silty fine sands that contain an increas-ingly dominant freshwater diatom assemblage, whichsuggests that the site became less open to tidal influencefollowing deposition of the coarser bed. Abrupt declinesin geochemical indicators of saline conditions point to asimilar transition. On this basis, we interpret the finersediments as tidal flat deposits, formed under regular,low-energy conditions.
Here we propose a scenario to account for the ob-served changes in the shallow stratigraphy of OkaritoLagoon. Given the proximity of Okarito Lagoon to theAlpine fault, the most likely mechanism for subsidenceof the lagoon floor is earthquake activity. Co-seismicsubsidence has not previously been recognized for theWest Coast, but is well established for similar tectonicsettings elsewhere; most notably the Pacific Northwestcoast of the USA (Atwater, 1987). Subsidence of coastal
259S.L. Nichol et al. / Sedimentary Geology 200 (2007) 248–262
lowlands has also been associated with tsunami inun-dation of these coasts, with sedimentary evidence pre-served as anomalous deposits (sand sheets) across thesubsided surface (Atwater and Moore, 1992).
This leads to the question — did co-seismic subsi-dence of 0.5 m in Okarito Lagoon trigger a local tsunamithat left a sedimentary record? Based on the resultspresented, we highlight the following as potential tsu-nami evidence: (1) the gravel and pebble deposits intrench 1, and to a lesser degree, the coarse sand and finegravel in trench 3 are a compelling indicator of high-energy flow (metres per second) across the subsided soilsurface; (2) the abrupt rise in lagoon salinity as recordedby geochemical data, diatoms and forams is consistentwith, though not exclusive to, sudden incursion ofmarine waters, and; (3) the re-establishment of reducedsalinity conditions in sediments above the coarse-grained deposits, as indicated by diatoms particularly,strengthens the interpretation that higher salinities arenot the norm for Okarito Lagoon.
In evaluating the case for a tsunami in Okarito La-goon (and adjacent coast), we must also consider otherscenarios as explanation for the observed sedimentaryrecord. For these alternatives, we draw upon geomor-phic and historic evidence. We noted earlier that thebarrier to Okarito Lagoon has a series of blind channelsthat terminate behind the modern foredune (Fig. 2).These channels are interpreted as former tidal inlets.Given that the current (micro) tidal prism of the lagoononly requires one inlet, we assume that the lagoon hasalways functioned as a single-inlet system, the locationof the inlet has shifted and that at times the lagoon hasremained closed. Support for this assumption is foundin the raised shoreline bench and shallow caves thatfringe the landward side of Okarito Lagoon. The sim-plest explanation for this raised shoreline is that thelagoon was up to 1 m deeper and closed to tidal ex-change. An alternative is that the bench formed duringtheMiddle Holocene sea-level highstand. However, thehighstand remains to be convincingly documented forthis coast. Moreover, recent dendrochronological workon the rimu (Dacrydium cupressinum) and kahikatea(Dacrycarpus dacrydioides) forest that has colonizedthe bench indicates that the oldest trees post-date AD1832 (Goff et al., 2004). Thus, it is likely that low-ering of the lagoon water level occurred shortly priorto ca. AD 1830 and was presumably associated withbarrier breaching to re-establish tidal exchange, not achange in relative sea level. Goff et al. (2004) proposetsunami as the mechanism for barrier breaching andidentify a known tsunami in AD 1826 as the likelyevent.
Further indirect evidence for subsidence and tsunamialong this coast is found in historical accounts from themid-nineteenth century and is summarized in Goff et al.(2004). Of these accounts, keys for this study are tworeports of a drowned Maori village at Poherua (akaSaltwater) Lagoon 11 km to the north of Okarito Lagoon.The first report was from the explorer Thomas Brunnerin 1848 (republished in Pascoe, 1952) and the second in1864 by Dobson, who later wrote, “At Poherua lagoonthere had been at one time a Maori village of con-siderable size, the stumps of the posts of the housesshowing plainly…The land had sunk…, as these tracescould be seen at dead low water, spring tides.” (Dobson,1930, p.72). Dobson also noted that local Maori wereunaware the village had ever existed, which suggests tous that its subsidence and abandonment occurred wellbefore the 19th century. This site was last verified in the1980's (New Zealand Forest Service, in press).
Returning to the point regarding lagoon closure, thiswill be achieved by accretion of the beach berm on thebarrier, leading to blocking of the inlet (a process thatoccurs about once a decade today). Clearly, this requiresa positive sediment budget in the littoral zone and werecognize the large gravel-bed rivers (Waiho River andCook River) to the south of Okarito Lagoon as obviouslong-term sediment sources for barrier construction.However, it is unlikely that this supply has been at aconsistent rate. There is good evidence from the Haastcoast, ∼100 km to the south of Okarito Lagoon, thatbarrier construction has been episodic and linked tofluctuations in sediment discharge from rivers drainingthe Southern Alps. Reconstruction of the chronology ofdune-ridge formation by Wells and Goff (2006) hasidentified a link to known earthquakes on the Alpinefault. It follows that these events have led to increasedsediment supply to rivers from earthquake-inducedlandslides, and in turn, to the coast. Given this event-based fluctuation in sediment supply to barriers, perio-dic closure of lagoons along the West Coast is an ex-pected outcome.
To summarize, fluctuations in the salinity and waterlevel of Okarito Lagoon can be readily accounted for byperiodic opening and closing of the tidal inlet entrance.However, this cannot be used to explain the placementof gravel and coarse sand at relatively distal locations inthe lagoon, nor the preservation of a buried salt marshsoil. The shallow depth of the gravel and sand depositsalso precludes transport as a lag along a tidal channelbecause flow velocities in a b1 m deep channel inside amicrotidal lagoon would not exceed threshold for graveltransport. We also discount river floods as a transportmechanism on the basis that the competence of the
Okarito River is limited to transport of coarse sand to thedelta front, which is on the opposite shore of the lagoon,∼1.4 km from the site of trench 1. Tsunami remains theonly plausible mechanism for transport of coarse sandand gravel to this relatively quiescent location. And theassociation with a subsided soil provides the importantlink to palaeo-seismicity.
6. Age of this event?
Establishing the age for a palaeo-tsunami event isproblematic, with the best results often obtained bydating material in sedimentary units above and below aninterpreted tsunami deposit, to “age bracket” the event.Radiocarbon ages are often equivocal because of thepossible introduction of older reworked carbon (woodand other macro-/microfossil debris) (Dawson, 1994;Goff et al., 1998). We have already discounted onereported age (10,580±60 years BP) on this basis. Ack-nowledging the further possibility that organic materialin the buried soil contains old carbon derived from thecatchment, the two radiocarbon ages of 1800±60 yearsBP and 2170±70 years BP are interpreted as maximumages for deposition of the enclosing sediments. Ourremaining radiocarbon age of 900±60 years BP is de-rived from an in situ bivalve preserved within the buriedsoil. The calibrated age of this sample is in the range630–455 years BP (1320–1495 AD) and represents thebest age estimate for the subsidence event and tsunami.
This age correlates well with the earliest in a series ofrecognized ruptures on the Alpine fault, dated to ca. AD1450 (Bull, 1996; Yetton et al., 1998). Other ruptures aredated to AD 1620, 1717 and 1826, with the latter gene-rating a tsunami. These events are also correlated withlarge-scale forest disturbance due to severe groundshaking and landslides (Wells et al., 1998, 1999, 2001;Cullen et al., 2003). It seems highly likely therefore thatthe tsunami reported here is linked directly to the mid15th century event, either as the direct result of sub-sidence associated with onshore fault rupture, subma-rine fault rupture (the southern segment of the Alpinefault is submarine) or of submarine landsliding gene-rated by an earthquake. These sources remain to beverified, however. It is also suggested that the mid 15thcentury event caused sufficient forest destruction totrigger increased sediment supply to the coast and, inturn, to closure of the tidal inlet to Okarito Lagoon.
7. Conclusion
Multiproxy analysis of a distinct sedimentary succes-sion recovered at five sites in Okarito Lagoon provided
for reconstruction of palaeoenvironments that record co-seismic subsidence of a salt marsh surface and inun-dation of that surface by tsunami. Key criteria include aburied soil, associated with major upward changes ingrain size, geochemistry, diatom and foraminiferal as-semblages. Limited radiocarbon dating of the buried soilsuggests that the subsidence occurred in the mid 15thcentury and was linked to a rupture on the nearby Alpinefault. Of importance to this case is the necessity toseparate other possible mechanisms from the subsi-dence-tsunami interpretation. Thus, geomorphic changesto the barrier inlet are also recognized as an importantinfluence on lagoon water level and salinity, but not as acausal mechanism for the deposits preserved in the la-goon fill. Ultimately, however, there is a link becausethose changes in barrier form can be potentially tracedback to major seismic events in the region. Further workto test this hypothesis will require extension of the recordpresented here to demonstrate repetition of the processover time and across new study sites.
Acknowledgements
Financial support was provided to this project by theDepartment of Conservation (Hokitika, Southern Re-gional Office), West Coast Regional Council, The Uni-versity of Auckland, the Australian Institute of NuclearScience and Engineering, the New Zealand Foundationfor Research, Science and Technology, and GeoEnvir-onmental Consultants. We thank Quentin Smith for fieldwork assistance and Anne Hall, Keri Hulme and DebbieMcLachlan from Okarito for support with logistics andlocal information. Bruce McFadgen is thanked for pro-viding calibrations of radiocarbon ages. J.-P. Guilbault,A.R. Nelson, and W.B. Bull provided useful commentson earlier versions of this paper. Many thanks to DaveTappin for a comprehensive final review that signifi-cantly improved the paper.
References
Admiraal, W., 1984. The ecology of estuarine sediment-inhabitingdiatoms. Progress in Phycological Research 3, 269–322.
Atwater, B.F., 1987. Evidence for great Holocene earthquakes alongthe outer coast of Washington State. Science 236, 942–944.
Atwater, B.F., Moore, A.L., 1992. A tsunami about 1000 years ago inPuget Sound. Washington. Science 258, 1614–1617.
Battarbee, R.W., 1986. Diatom analysis. In: Berglund, B.E. (Ed.),Handbook ofHolocene Palaeoecology andPalaeohydrology.Wiley,Chichester, pp. 527–570.
Berryman, K.R., Beanland, S., Cooper, A.F., Cutten, H.N., Norris, R.J.,Wood, P.R., 1992. The Alpine fault, New Zealand: variation inQuaternary structural style and geomorphic expression. AnnalesTectonicae VI, 126–163.
261S.L. Nichol et al. / Sedimentary Geology 200 (2007) 248–262
Bryant, E.A., Young, R.W., Price, D.M., 1992. Evidence of tsunamisedimentation on the southeastern coast of Australia. Journal ofGeology 100, 753–765.
Bull, W.B., 1996. Prehistoric earthquakes on the Alpine fault, NewZealand. Journal of Geophysical Research 101 (B3), 6037–6050.
Chagué-Goff, C., Goff, J.R., 1999. Geochemical and sedimentologicalsignature of catastrophic saltwater inundations (tsunami), NewZealand. Quaternary Australasia 17, 38–48.
Chagué-Goff, C., Dawson, S., Goff, J.R., Zachariasen, J., Berryman,K.R., Garnett, D.L., Waldron, H.M., Mildenhall, D.C., 2002. Atsunami (c. 6300 years BP) and other environmental changes,northern Hawke's Bay, New Zealand. Sedimentary Geology 150,89–102.
Chen, Z., Chen, Z., Zhang, W., 1997. Quaternary stratigraphy andtrace-element indices of the Yangtze Delta, eastern China, withspecial reference to marine transgressions. Quaternary Research47, 181–191.
Clague, J.J., 1997. Evidence for large earthquakes at the Cascadiasubduction zone. Reviews of Geophysics 35, 439–460.
Clague, J.J., Bobrowsky, P.T., Hutchinson, I., 2000. A review ofgeological records of large tsunamis at Vancouver Island, BritishColumbia. Quaternary Science Reviews 19, 849–863.
Cleve-Euler, A., 1951–1955. Die Diatomeen von Schweden undFinnland. Kungliga Svenska Vetenskaps-akademiens HandlingarSerie 4, 2 (1), 3–163 (1951); 3 (3), 3–153 (1952); 4 (1), 3–158(1953); 4 (3), 3–225 (1953); 5 (4), 3–231 (1955).
Crosby, L.H., Wood, E.J.F., 1958. Studies on Australian and NewZealand diatoms, I: planktonic and allied species. Transactions ofthe Royal Society of New Zealand 85, 483–530.
Crosby, L.H., Wood, E.J.F., 1959. Studies on Australian and NewZealand diatoms, II: normally epontic and benthic genera. Tran-sactions of the Royal Society of New Zealand 86, 1–58.
Cullen, L.E., Duncan, R.P., Wells, A., Stewart, G.H., 2003. Floodplainand regional scale variation in earthquake effects on forests,Westland, New Zealand. Journal of the Royal Society of NewZealand 33, 693–701.
Daoust, R.J., Moore, T.R., Chmura, G.L., Magenheimer, J.F., 1996.Chemical evidence of environmental changes and anthropogenicinfluences in a Bay of Fundy saltmarsh. Journal of Coastal Re-search 12, 520–533.
Dawson, A., 1994. Geomorphological effects of tsunami run-up andbackwash. Geomorphology 10, 83–94.
Dawson, A., Long, D., Smith, D.E., 1988. The Storegga slides: evi-dence from eastern Scotland for a possible tsunami. Marine Geo-logy 82, 271–276.
Denys, L., 1985. Diatom analysis of an Atlantic-SubBoreal core fromSlijpe (Western Belgian coastal plain). Review of Palaeobotanyand Palynology 46, 33–53.
Denys, L., 1991a. A checklist of the diatoms in the Holocene depositsof the western Belgian coastal plain with a survey of their appar-ent ecological requirements. 1. Introduction, ecological code andComplete list. Belgische Geologische Dienst, Professional Paper1991/2, 246.
Denys, L., 1991b. A checklist of the diatoms in the Holocene de-posits of the western Belgian coastal plain with a survey of theirapparent ecological requirements. II. Centrales. Belgische Geo-logische Dienst, Professional Paper 1991/3, 247.
De Wolf, H., 1982. Method of coding of ecological data from diatomsfor computer utilisation. Mededelingen Rijks Geologiische Dienst36, 95–98.
Dobson, A.D., 1930. Reminiscences of Arthur Dudley Dobson. Whit-combe and Tombs, Wellington.
Eiby, G.A., 1982. Two New Zealand tsunamis. Journal of the RoyalSociety of New Zealand 12, 337–351.
Foged, N., 1977. Freshwater diatoms in Ireland. Bibliotecha Phyco-logica 34.
Foged, N., 1978. Diatoms in Eastern Australia. Bibliotecha Phycolo-gica 41.
Foged, N., 1979. Diatoms in New Zealand, the North Island. Bib-liotecha Phycologica 47.
Gibb, J.G., 1986. A New Zealand regional Holocene eustatic sea-levelcurve and its application to determination of vertical tectonicmovements. Royal Society of New Zealand, Bulletin 24, 377–395.
Goff, J.R., McFadgen, B.G., 2002. Seismic driving of nationwidechanges in geomorphology and prehistoric settlement—a 15thcentury New Zealand example. Quaternary Science Reviews 21,2313–2320.
Goff, J.R., Crozier, M., Sutherland, V., Cochran, U., Shane, P., 1998.Possible tsunami deposits of the 1855 earthquake,North Island,NewZealand. In: Stewart, I.S., Vita-Finzi, C. (Eds.), Coastal Tectonics.Geological Society Special Publication, vol. 133, pp. 353–374.London.
Goff, J.R., Wells, A., Chagué-Goff, C., Nichol, S.L., Devoy, R.J.N.,2004. The elusive AD 1826 tsunami, South Westland, New Zea-land. New Zealand Geographer 60, 14–25.
Griffiths, G.A., McSaveney, M.J., 1986. Sedimentation and rivercontainment on Waitangitaona alluvial fan—South Westland, NewZealand. Zeitschrift fur Geomorphologie 30, 215–230.
Hayward, B.W., Grenfell, H.R., Reid, C.M., Hayward, K.A., 1999a.Recent New Zealand shallow-water benthic foraminifera: taxon-omy, ecologic distribution, biogeography, and use in paleoenvir-onmental assessment. Institute of Geological and Nuclear SciencesMonograph 21.
Hayward, B.W., Grenfell, H.R., Scott, D.B., 1999b. Tidal range ofmarsh foraminifera for determining former sea-level heights inNew Zealand. New Zealand Journal of Geology and Geophysics42, 395–413.
Hayward, B.W., Scott, G.H., Grenfell, H.R., Carter, R., Lipps, J.H.,2004. Estimation of tidal elevation and salinity histories of shel-tered harbours and estuaries using benthic foraminifera. The Ho-locene 14, 218–232.
Hendey, N.I., 1964. An introductory account of the smaller algae ofBritish coastal waters. Part V: Bacillariophyceae (Diatoms). Fish-eries Investigation Series I. HMSO, London.
Hull, A.G., 1986. Pre-A.D. 1931 tectonic subsidence of Ahuriri La-goon, Napier, Hawke's Bay, New Zealand. New Zealand Journalof Geology and Geophysics 29, 75–82.
Hustedt, F., 1927–1966. Die Kieselalagen Deutschlands, Osterreichsund der Schweiz (I–III). In: Rabenhorst, L. (Ed.), Kryptogamen-Flora von Deutschland, Osterreich under der Schweiz. 7. Akade-mische Verlagsgesellshaft, Leipzig.
Hustedt, F., 1930. Bacillariophyta (Diatomaceae). In: Pascher, A.(Ed.), Die Susswasser-flora Mitteleuropas. Heft 10. Gustav FischerVerlag, Jena.
Hustedt, F., 1957. Die Diatomeenflora des Fluss-Systems derWeser imGebiet der Hansestadt Bremen. Abhandlungen Naturwissenschaf-tlicher Verein, Bremen 34.
John, J., 1983. The diatom flora of the Swan River estuary WesternAustralia. Bibliotecha Phycologica 64.
Fragilariaceae, Eunotiaceae (1991), pp. 1–576; 4. Teil: Achnantha-ceae, Kritische Erganzungen zu Navicula (Lineolatae) und Gom-phonemaGesamtliteraturverzeichnis (1991), pp. 1–437. In: Pascher,A. (ed) Süswasserflora von Mitteleuropa, 2 (1–4). Gustav FischerVerlag, Stuttgart.
López-Buendía, A.M., Bastida, J., Querol, X., Whateley, M.K.G.,1999. Geochemical data as indicators of paleosalinity in coastalorganic-rich sediments. Chemical Geology 157, 235–254.
Loring, D.H., 1991. Normalization of heavy-metal data from estu-arine and coastal sediments. ICES Journal of Marine Science 48,101–115.
MacPherson, J.M., 1981. Hydrology of Okarito Lagoon and the in-ferred effects of selective logging in Okarito Forest. New ZealandJournal of Marine and Freshwater Research 15, 25–39.
McFadgen, B.G., Manning, M.R., 1990. Calibrating New Zealandradiocarbon dates of marine shells. Radiocarbon 32, 229–232.
Minoura, K., Nakaya, S., Uchida, M., 1994. Tsunami deposits in alacustrine sequence of the Sanriku coast, Northeast Japan. Sedi-mentary Geology 89, 25–31.
Minoura, K., Gusiakov, V.G., Kuratov, A., Takeuti, S., Svendsen, J.I.,Bondevik, S., Oda, T., 1996. Tsunami sedimentation associatedwith the 1923 Kamchatka earthquake. Sedimentary Geology 106,145–154.
Molinaroli, E., De Falco, G., Rabitti, S., Portaro, R.A., 2000. Stream-scanning laser system, electric sensing counter and settling grainanalysis: a comparison using reference materials and marine sedi-ments. Sedimentary Geology 130, 269–281.
New Zealand Forest Service, 1984. Survey of the saltwater historicfeatures. Unpublished New Zealand Forest Service Report, Hoki-tika. 54 pp.
Norris, R.J., Cooper, A.F., 1997. Erosional control on the structuralevolution of range front faulting on the Alpine fault. GeologicalSociety of New Zealand Miscellaneous Publication 95A.
Pascoe, J.D. (Ed.), 1952. The Great Journey: An Expedition to Explorethe Interior of the Middle Island, New Zealand, 1846-1848.Pegasus Press, Christchurch. 111 pp.
Round, F.E., Crawford, R.M., Mann, D.G., 1990. The Diatoms: Bio-logy and Morphology of the Genera. Cambridge University Press,Cambridge.
Scott, D.B., Medioli, F.S., 1986. Foraminifera as sea-level. In: van dePlassche, O. (Ed.), Sea-level Research: A Manual for the Collec-tion and Evaluation of Data. Geo Books, Norwich.
Shennan, I., Long, A.J., Rutherford, M.M., Green, F.M., Innes, J.B.,Lloyd, J.M., Zong, Y., Walker, K.J., 1996. Tidal marsh strati-graphy, sea-level change and large earthquakes, I: a 5000 yearrecord in Washington, U.S.A. Quaternary Science Reviews 15,1023–1059.
Simonsen, R., 1967. Diatoms as indicators in the estuarine environ-ment. Veroffentl. Inst. Meersforsch, Bremerhaven 11, 287–291.
Stuiver, M., Reimer, P.J., Bard, E., Beck, J.W., Burr, G.S., Hughen,K.A., Kromer, B., McCormac, F.G., Plicht, J.V.D., Spurk, M.,
1998. INTCAL98 Radiocarbon Age Calibration, 24,000-0 cal BP.Radiocarbon 40, 1041–1083.
Thomas, E., Varekamp, J.C., 1991. Paleo-environmental analyses ofmarsh sequences (Clinton, Connecticut): evidence for punctuatedrise in relative sealevel during the latest Holocene. Journal ofCoastal Research 11, 125–158.
Van de Plassche, O., van der Borg, K., de Jong, A.F.M., 1998. Sealevel climate correlation during the past 1400 yr. Geology 26,319–322.
Van der Werff, A., Huls, H., 1957–1974. Diatomeënflora van Neder-land, vol. 1–10. Otto Koeltz Science Publishers, Koenigstein.Reprinted 1976.
Vos, P.C., de Wolf, H., 1993a. Diatoms as a tool for reconstructingsedimentary environments in coastal wetlands; methodologicalaspects. Hydrobiologia 269/270, 285–296.
Vos, P.C., de Wolf, H., 1993b. Reconstruction of sedimentary environ-ments in Holocene coastal deposits of the southwest Netherlands;the Poortvliet boring, as a case study of palaeoenvironmentaldiatom research. Hydrobiologia 269/270, 297–306.
Vos, P.C., deWolf, H., 1994. Palaeoenvironmental research on diatomsin Early and Middle Holocene deposits in central and north Hol-land (The Netherlands). Netherlands Journal of Aquatic Ecology28, 97–115.
Wells, A., Goff, J., 2006. Coastal dune ridge systems as chronologicalmarkers of palaeoseismic activity; a 650-yr record from southwestNew Zealand. The Holocene 16, 543–550.
Wells, A.M., Stewart, G.H., Duncan, R.P., 1998. Evidence of wide-spread, synchronous, disturbance-initiated forest establishment inWestland, New Zealand. Journal of the Royal Society of NewZealand 28, 333–345.
Wells, A.M., Yetton, M., Duncan, R.P., Stewart, G.H., 1999. Pre-historic dates of the most recent Alpine fault earthquakes, NewZealand. Geology 27, 995–998.
Wells, A.M., Duncan, R.P., Stewart, G.H., 2001. Forest dynamics inWestland, New Zealand: the importance of large, infrequent earth-quake-induced disturbance. Journal of Ecology 89, 1006–1018.
Wood, E.J.F., 1961. Studies on Australian and New Zealand diatoms,IV and V. Transactions of the Royal Society of New Zealand 88,669–712.
Wood, E.J.F., 1963. Studies on Australian and New Zealand diatoms,VI: tropical and subtropical species. Transactions of the RoyalSociety of New Zealand Botany 2, 189–218.
Wood, E.J.F, Crosby, L.H., Cassie, V., 1959. Studies on Australian andNew Zealand diatoms, III: descriptions of further discoid species.Transactions of the Royal Society of New Zealand 87, 211–219.
Yetton, M., Wells, A., Traylen, N.J., 1998. The probability and conse-quences of the next Alpine fault earthquake. EQC ResearchReport, vol. 95/193. Wellington.
