geomorphology, stratigraphy, and radiocarbon chronology of ...

schumacher.paul35

geomorphology, stratigraphy, and radiocarbon chronology of ...

G.H. DENTON ET AL.Fig. 1. Index map of the southern Lake District, Seno Reloncav’, and Isla Grande de ChiloŽ, Chile. The boxes show the locations ofthe glacial geomorphologic maps in Andersen et al. (1999) and in Figs 2Ð5. Valle Central is a large north-south trending depressionin the center of the map occupied by the lakes and gulfs. The numbered sites refer to Table 1.168GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTde los Andes between 40¡30' and 42¡25'S (Fig. 1).This sector of the Andes has a widespread low portion,with average summit elevations of 1500Ð2000m, composed chießy of crystalline rocks. Largevolcanoes with summit elevations of 1900Ð3200 mform the highest peaks. The volcanoes above 2000m elevation have extensive mantles of glacier ice,but only small glaciers are scattered on the lowercrystalline peaks. The modern snowline in an eastÐwest transect across the Andes at the latitude ofLago Llanquihue (41¡15'S), estimated by Porter(1981) from the median elevation of glaciers onhigh volcanic peaks, rises inland from about 1900m elevation at Volc‡n Calbuco alongside LagoLlanquihue to 2250 m elevation in the ArgentineAndes.The Cordillera de la Costa forms the westernmargin of Valle Central. The summits are generally350Ð800 m in elevation, but in places are as muchas 950 m. Today the Cordillera de la Costa is freeof glacier ice.Valle Central is largely above sea level and isabout 80 km wide in the northern half of the Þeldarea. The dominant topographic features are large,deep lakes in basins occupied at the LGM by piedmontice lobes emanating from the Andes. Theselakes include Lago Puyehue (185 m elevation),Lago Rupanco (115 m), and Lago Llanquihue (51m). Lago Llanquihue is commonly more than 200m deep, with a maximum depth of 320 m. Morainebelts and outwash plains dating to the LGM fringethe western margins of these lakes. Separating thelakes are outwash plains and moraine belts, alsolargely dating to the LGM. Farther west, Valle Centralfeatures older moraine belts and associatedoutwash plains. The lakes drain westward throughdeep spillways. In the case of Lago Rupanco andLago Llanquihue, the elevations of the outlets, andhence of the lake levels, are controlled by indurated,waterlain till deposits. Basalt bedrock controlsoutßow from Lago Puyehue. R’o PilmaiquŽn is theoutlet for Lago Puyehue, and R’o Rahue for LagoRupanco. R’o Maull’n now drains Lago Llanquihuesouthwestward through moraine belts and outwashplains to the PaciÞc Ocean. A former easternoutlet for Lago Llanquihue extends into R’o PetrohuŽ,which ßows into Estero Reloncav’.In the southern half of the Þeld area, Valle Centralis about 120 km wide and dominated by marineembayments. Seno Reloncav’ is commonly wellover 250 m deep, with maximum depths of morethan 450 m near Estero Reloncav’. At the LGM thisembayment was Þlled by a piedmont glacier fedthrough Estero Reloncav’. Scallop-shaped baysmark the western shore. Islands separated from themainland by sinuous channels occur near the westernshore. Islands divided by short, straight channelsnearly close off the southern margin of SenoReloncav’.South of Seno Reloncav’ there is an abrupt increasein the size of marine gulfs that containedpiedmont glaciers at the LGM. Seno Reloncav’ isabout the size of Lago Llanquihue, which in turn islarger than either Lago Rupanco or Lago Puyehue.But Golfo de Ancud, south of Seno Reloncav’, ismuch bigger than any of the water bodies farthernorth. It is commonly more than 250 m deep, andin places exceeds 300 m in depth. Just as with SenoReloncav’, southern Golfo de Ancud is fringed bydissected islands of glacial deposits.Northern Golfo Corcovado, with Isla Grande deChiloŽ at its western edge, is relatively shallowwith numerous islands of glacial deposits. Andeanice ßowed across the gulf, around the southern endof Cordillera de la Costa, and over southern IslaGrande de ChiloŽ onto the PaciÞc continentalshelf.The southern part of the Þeld area has characteristicsof a drowned terrain. Canal de Chacao is likelythe spillway of a former glacial lake in Golfo deAncud. The canals and scallop-shaped bays alongthe western margins of Seno Reloncav’ and Golfode Ancud were probably eroded by former icemarginalrivers that converged on Canal de Chacao.Taken together, Paso Tautil and Canal Calbucorepresent the former outlet for a glacial lake inSeno Reloncav’. Marine channels between dissectedislands along the southern margin of Seno Reloncav’may also have been spillways of a formerglacial lake in Seno Reloncav’. In similar fashion,the canals and marine passes that separate numerousislands of glacial drift south of Golfo de Ancudmay have originally been cut by rivers ßowing betweenformer lakes in present-day Golfo de Ancudand Golfo Corcovado. The drowned landscapemust be partially due to the sea-level rise of the lastdeglaciation. But many submerged features are toodeep to be explained entirely in this manner. Theimplication is of tectonic subsidence, consistentwith submerged alerce (Fitzroya cupressoides)stumps on the north shore of Seno Reloncav’ (Porter1981).The natural vegetation in the southern Lake Districtis illustrated along a transect from the southernLake District inland across the Andes (see Heusseret al. 1999; Moreno et al. 1999). Here Lowland De-GeograÞska Annaler á 81 A (1999) á 2169


G.H. DENTON ET AL.ciduous Forest (up to 200 m elevation) is succeededby Valdivian Evergreen Forest (200Ð600 m), NorthPatagonian Evergreen Forest (600Ð1000 m), andSubantarctic Deciduous Beech Forest (1000Ð1250m) (Heusser 1974, 1981). The shrub-like beechNothofagus antarctica occurs at tree line and inisolated locations to 1370 m. Above tree line isparkland and tundra consisting principally ofgrasses and ericaceous shrubs and herbs. Patchesof Magellanic Moorland exist above 700 m in thecoast range west of the southern Lake District. Onthe east slope of the coast range on Isla Grande deChiloŽ, Valdivian Evergreen Forest extends to 250m elevation and the North Patagonian EvergreenForest extends from 250 to 450 m (Villagr‡n 1988;Heusser et al. 1996). Above 450 m elevation thevegetation becomes more open and North Patagoniantree species occur as enclaves in a MagellanicMoorland complex.The evergreen forests and the high-elevationMagellanic Moorland complex reßect a cool-temperatewet climate that results from the cold, offshoreHumboldt Current combined with the SouthernHemisphere westerlies. Cyclonic depressionsare frequent throughout the year, resulting in highprecipitation and extensive cloudiness. Close tosea level at Puerto Montt on the northern shore ofSeno Reloncav’, mean annual temperature is11.2¡C, with a range from a winter mean of 7.7¡Cto a summer mean of 15.1¡C; mean annual precipitationis 2341 mm (Hajek and di Castri 1975).Previous workIn early studies BrŸggen (1950), Weischet (1958,1964), and Olivares (1967) recognized multiplemoraine belts of different ages west of Lago Puyehue,Lago Rupanco, and Lago Llanquihue. Mercer(1972, 1976), and Laugenie and Mercer (1973) differentiatedfour moraine belts west of Lago Llanquihue,the younger three of which had been previouslyrecognized by BrŸggen (1950) and Olivares(1967). The drifts associated with these morainebelts were named, from youngest to oldest,the Llanquihue (Heusser 1974), Casma, Colegual,and R’o Fr’o (Mercer 1976). Relative age characteristicssuggest that these four drifts represent atleast three, and perhaps four, glaciations. A radiocarbonchronology shows that only Llanquihuedrift represents the last glaciation. Llanquihue driftis little weathered, with near-surface volcanicclasts exhibiting rinds of 0.5 mm or less. The permeableportions of the Casma and Colegual driftsboth contain volcanic clasts with weathering rindsof about 2 cm and therefore may represent a singleglaciation (Mercer 1976). Volcanic clasts in permeableportions of R’o Fr’o drift have weatheringrinds of 4 cm. Mercer (1976) suggested that theoutermost group of Llanquihue moraines and outwashplains dates to >40,000 14 C yr BP and is probablyearly Llanquihue in age. The late Llanquihueice limit was placed well behind early Llanquihuemoraines near Puerto Varas, at the outer edge ofLlanquihue moraines near Frutillar Alto, and on thepeninsula at the mouth of Bah’a Mu–oz Gamero(Mercer 1976). A reworked peat clast in outwashgraded to the outermost late Llanquihue morainenear Frutillar Alto yielded an age of 20,100±55014 C yr BP (RL-116) (Mercer 1976). Basal peat fromthe Canal de Chanch‡n (named by Laugenie 1982)spillway channel yielded a date of 17,370±670 14 Cyr BP (RL-120) (Mercer 1976); a subsequent dateof basal peat of 18,170±650 14 C yr BP (GX-5274)is a minimum value for recession of ice lobe fromthe moraine at the mouth of Bah’a Mu–oz Gamero(J.H. Mercer, pers. comm. in Porter 1981). Thus thelate Llanquihue advance of the Lago Llanquihuepiedmont glacier was bracketed between 20,100and 18,170 14 C yr BP. For the adjacent Lago Rupancopiedmont lobe, dates of 19,450±350 14 C yr BP(I-5679) and 21,360±840 14 C yr BP (I-5678) (Mercer1976; Heusser 1974) apply to ashy peat belowtill on the outer late Llanquihue moraine.The subsequent history of the Lago Llanquihueice lobe was reconstructed from lake-level ßuctuationsassociated with the two main outlets(BrŸggen 1950). One of these is the present-dayR’o Maull’n outlet, anchored at about 51 m elevationby indurated waterlain till on the western lakeshore. The former eastern outlet is now plugged bya lahar fan from Volc‡n Calbuco, but is still lessthan 10 m above present-day lake level (Mercer1972, 1976). In a prominent terrace near LagoLlanquihue at Puerto Varas, organic beds are exposedin lakeside cliffs, in a road cut beside CalleSanta Rosa, and beneath the railroad bridge overRoute V-50. The thin laminated lacustrine silt andclay on these organic beds are overlain Þrst by thickdeposits of unsorted pebbles and cobbles in a sandymatrix and then by ash and pumice washed onto theterrace by the lake (Mercer 1976). This sequencewas interpreted as indicating ice recession from theeastern outlet of Lago Llanquihue in order for thelake level to drop enough for accumulation of organicsediments on the exposed terrace surface(Mercer 1976). Accumulation of organic sedi-170GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTments ended when a readvancing piedmont glacierlobe again blocked the eastern outlet, causing a risein lake level and deposition of lacustrine sedimentson the Puerto Varas terrace. This readvance extendedto undated moraine ridges located on the southernmargin of the lake 25 km from the westernshore. The interval of deposition of the organic sediment,termed the Varas interstade by Mercer(1972, 1976), was originally estimated to have begunbefore 16,270±360 14 C yr BP (RL-113) (woodfragment at base of organic silt at the railroadbridge site) and to have ended about 14,820±23014 C yr BP (I-5033) (top 5 mm of the organic bed atthe Calle Santa Rosa site) (Mercer 1972). The endof organic accumulation, and hence lake-level rise,was subsequently revised to about 13,300 14 C yr BPon the basis of three dates of peat and wood fromthe prominent organic bed at the railroad bridgesite (13,300±550 14 C yr BP, GX-2947; 13,200±320 14 C yr BP, GX-4169; and 13,750±295 14 C yrBP, GX-4170) (Heusser 1974, Mercer 1976).In a delineation of moraine belts and drift sheetsdifferent from that of Mercer (1976), Porter (1981)recognized four drift sheets, each dominated by till,deposited by the Lago Llanquihue piedmont icelobe; from youngest to oldest these were named theLlanquihue, Santa Mar’a, R’o Llico, and Carocol.A type section was assigned for each drift sheet.Only the little-weathered Llanquihue drift was recognizedas having distinctive morainal topography.The three older drift sheets were distinguished byrelative weathering criteria. A wide belt of SantaMar’a drift was mapped across most of the lowlandwest of the outer edge of Llanquihue drift. SantaMar’a drift is more weathered than Llanquihuedrift, in that it has a yellowish-brown color from limonitethat coats clasts and penetrates Þssures.Rind thicknesses on volcanic clasts near the surfaceof Santa Mar’a drift have a mean value of 2 mm.The broad crests within the Santa Mar’a drift sheetwere described as moraine ridges or dissected remnantsof a drift plain (Porter 1981). In any case, theSanta Mar’a drift includes the R’o Fr’o, Casma, andColegual drifts of Mercer (1976). Lying west of theSanta Mar’a drift, and exposed in a 10-km-widesurface belt near the Cordillera de la Costa, is theR’o Llico drift. Till of this drift is weathered to clay,and volcanic clasts have mean weathering rinds of10 mm. Carocol drift is largely buried by youngerdrifts and is highly weathered, with rinds on volcanicclasts having a mean thickness of 17 mm.Porter (1981) subdivided Llanquihue drift intothree units on the basis of morphostratigraphic andsedimentologic criteria. Type or reference sectionswere deÞned for Llanquihue I (the oldest and outermost),II, and III drifts. Each drift unit wasmapped near Lago Llanquihue and Seno Reloncav’.Llanquihue I drift, which in most placescomprises the outermost Llanquihue moraine belt,was certainly deposited more than 19,000 14 C yr BPand probably much more than 40,000 14 C yr BP(Porter 1981). The highest Llanquihue outwashplains are graded to the outer moraines of LlanquihueI drift. Llanquihue II drift comprises well-preservedmoraines which, between Puerto Octay andLlanquihue, are fronted by outwash that passesaround the Llanquihue I moraines. Near PuertoVaras, the Llanquihue II moraines are nested behindthe Llanquihue I moraines. The Llanquihue IImoraines were dated by Porter (1981) between20,100±550 14 C yr BP (RL-116) (the age of the reworkedpeat clast in outwash at Frutillar Alto; Mercer1972) and 18,900±370 14 C yr BP (UW-418)(the age of basal peat from Canal de la Puntilla).The Llanquihue III sediments in the Lago Llanquihuebasin are conÞned to the lakeside terrace describedby Mercer (1976), but in the Seno Reloncav’basin the Llanquihue III drift forms morainessituated only a few kilometers behind thoseof Llanquihue II age. On the northern shore of SenoReloncav’, Llanquihue III till overlies laminatedglaciolacustrine sediments; near the base of thesesediments at Punta Penas is a 20-cm-thick peat bedthat yielded dates of 15,220±160 14 C yr BP (UW-422), 15,400±400 14 C yr BP (W-948), and 14,200±135 14 C yr BP (UW-421) (Ives et al. 1964; Porter1981; Heusser 1981).The organic beds in the Puerto Varas terrace arecovered by a thin unit of glaciolacustrine sedimentsand/or by volcanic breccias that Porter(1981) interpreted as lahars. The volcanic brecciascan be traced on the lake-margin terrace fromPuerto Varas to near Volc‡n Calbuco, as well as betweenPuerto Varas and the R’o Maull’n outlet atthe town of Llanquihue. Porter (1981) argued thatthe Lago Llanquihue piedmont ice lobe must havereached the western margin of the lake and stoodagainst the terrace when the volcanic breccias weredeposited by lahars that originated on Volc‡n Calbuco.Otherwise, the lahars would have dischargeddirectly into the lake, rather than following the terracesurface along the lake edge to the R’o Maull’noutlet.By obtaining new dates of 13,965±235 14 C yr BP(UW-481) (Bella Vista Bluff site) and 13,145±23514 C yr BP (UW-480) (Northwest Bluff site) from or-GeograÞska Annaler á 81 A (1999) á 2171


G.H. DENTON ET AL.ganic beds in the Puerto Varas terrace, Porter(1981) postulated a more complex history of icemarginßuctuations during Llanquihue III timethan did Mercer (1976). From elevations and agesof the organic beds, Porter (1981) estimated that anice advance (and associated rise of lake level) beganprior to 15,700 14 C yr BP and culminated at15,000Ð14,000 14 C yr BP. Following ice recessionand lake-level drop, a second advance (and lakelevelrise) was underway by 13,965 14 C yr BP andculminated after 13,145 14 C yr BP at the westernmargin of Lago Llanquihue when laharic brecciaswere deposited on the terrace.Laugenie (1982) mapped moraine belts and outwashplains in the entire Chilean Lake District at ascale of 1:250,000. Near Lago Llanquihue, theLlanquihue moraine system of Laugenie (1982)corresponds with the Llanquihue drift of Porter(1981) and the Llanquihue moraine system of Mercer(1976). Laugenie (1982) divided the Llanquihueglaciation into three principal parts, each withsubsidiary stades and interstades: Eo-Llanquihue(>40,000 14 C yr BP), Meso-Llanquihue (25,000 to>40,000 14 C yr BP), and Neo-Llanquihue (25,000 to12,200 14 C yr BP). The outermost Trapen morainebelt near Seno Reloncav’ is assigned to the Eo-Llanquihue, along with the main outwash plaingraded to this moraine. Most outer moraine beltsfringing Lago Llanquihue (including the PuertoVaras moraine and the outer moraines near Bah’aMu–zo Gamero) are also Eo-Llanquihue. Largelynested within these outer moraines are those of theNeo-Llanquihue, including the San Antonio morainenear Seno Reloncav’ and the Frutillar morainenear Lago Llanquihue. The Neo-Llanquihuelimit near Bah’a Mu–oz Gamero is placed not at thePuerto Chico moraine, as suggested by Mercer(1976), but at the moraines at the head of the twoimportant spillway channels discussed earlier.Only near Bah’a Frutillar does the Neo-Llanquihuemoraine belt represent the outermost limit of theLlanquihue-age piedmont ice lobe, here dated byMercer (1972) to shortly after 20,100 14 C yr BP.Following the Puerto Varas interstade at 16,270Ð13,200 14 C yr BP, the Ensenada readvance, the lastof Neo-Llanquihue age, reached the R’o Pescadomoraine belt south of Lago Llanquihue (Laugenie1982).Meso-Llanquihue time was marked by restrictedAndean glaciers (Laugenie 1982), and the environmentis deÞned by a pollen diagram from nearLaguna Bonita alongside Lago Rupanco (Heusser1974). The older Moraine de Casma and Morainede Colegual, each fronted by an extensive outwashplain (Laugenie 1982), correspond with the Casmaand Colegual moraines and drift of Mercer (1976).Laugenie (1982) suggested that both moraines dateto the Colegual glaciation. The Moraine de Fresiaof Laugenie (1982), deposited during the FresiaGlaciation, corresponds with the R’o Fr’o morainebelt of Mercer (1976).On Isla Grande de ChiloŽ, Heusser and Flint(1977) mapped three major drifts and associatedmoraine belts; from oldest to youngest these areFuerte San Antonio drift, Intermediate drift, andLlanquihue drift. A mire situated on the outermostLlanquihue moraine at Taiquem— afforded a basalage of 42,400±1000 14 C yr BP (QL-1012) (Heusser1990). Dates of 14,355±700 14 C yr BP (GX-8686),14,970±210 14 C yr BP (I-12996), and 15,600±56014 C yr BP (GX-9978) come from a wood layer onthe surface of an organic deposit buried by Llanquihuedrift at Dalcahue (Mercer 1984). Samplesfrom deeper in the organic deposit are as old as17,750±560 14 C yr BP (Beta-33820) and 19,630±180 14 C yr BP (QL-4324) (Heusser 1990), affordinga minimum age for underlying Llanquihue till. Atthe Teguaco section, interdrift organic silt yieldeddates of 25,500±500 14 C yr BP (QL-1017) to30,000±300 14 C yr BP (QL-1019) (Heusser 1990).Finally, dates of basal organic samples from mireson the Llanquihue moraine belt on Isla Grande deChiloŽ indicate deglaciation prior to 13,000 14 C yrBP (Heusser 1990).MethodsFive steps were involved in establishing a radiocarbonchronology of Llanquihue drift and landforms.The Þrst was to produce glacial geomorphologicmaps of Llanquihue drift west of Lago Puyehue,Lago Rupanco, Lago Llanquihue, Seno Reloncav’,Golfo de Ancud, and northern Golfo Corcovado.Extensive aerial photographic analysis resulted inthe production of preliminary maps, which wereextensively Þeld checked over six years. After additionalaerial photographic analysis, the resultswere integrated into the maps in Andersen et al.(1999) and in Figs 2Ð6.The second step was to interpret the sedimentsexposed in borrow pits, road cuts, and natural sectionswithin each morphologic unit distinguishedon the maps. The third step was to describe the keystratigraphic sections. The fourth step was to datemultiple radiocarbon samples from stratigraphicsections, particularly where glacial sediments rest172GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTalso cannot be matched with the outer moraine atFrutillar Alto. Rather, the outer Llanquihue morainenear the Puntilla meltwater channel is youngerthan the Frutillar moraine element, and hence themoraines at the head of the Puntilla and Chanch‡nchannels are younger still.Such cross-cutting relationships carry the implicationthat the main Llanquihue outwash terrace isgraded to a complex of moraines of slightly differingages, and therefore itself cannot be strictly ofone age. A similar example comes from the morainebelts west of Seno Reloncav’, which arespread farther apart than those west of Lago Llanquihue.As a consequence the intermoraine areasare Þlled with subsidiary outwash terraces that convergeon drainage gaps through the outer morainebelt and are incised into the main Llanquihue outwashplain, where they form tributaries to the majorR’o Maull’n drainage channel from Lago Llanquihue.But the outer moraine belt itself is composedof numerous individual elements that are overlappingor occur one behind the other. The detailedmap pattern of these elements in Fig. 4 suggests thatthe outer moraine belt may represent several ßuctuationsof the Seno Reloncav’ piedmont glacier.The conÞguration of the Llanquihue morainebelts, along with the trend of surface till ßutes,shows the intersection of two former piedmont glaciersnorthwest of Calbuco (Seno Reloncav’ lobe inthe north and Golfo de Ancud lobe in the south). InFig. 4 the outermost moraine in the north appearsto be cut by younger moraines at this intersection.The implication is that the younger advance wasstronger for the southern than for the northern lobe,with ice from the southern lobe riding far up ontothe mainland. These overlapping moraines againpoint to the probability that the main Llanquihueoutwash plain is composite. North of the intersectionzone, the main outwash plain is graded to anouter prominent moraine belt west of Seno Reloncav’.South of the intersection zone, this outerbelt is dissected and nearly buried by outwash fromthe inner moraine belt, which here is close to themaximum Llanquihue position. Outwash graded tothis inner belt surrounds and passes through theremnants of the outer moraine belt.Mapping of moraine morphology along the outerLlanquihue drift limit west of Lago Llanquihueis complicated by outwash folded into ridges parallelto the former ice margin (Fig. 3). These foldedridges, described as push moraines (SchlŸchter etal. 1999), occur in four patches northeast of FrutillarAlto. The southern two patches have single ridges,whereas the northern two patches have multipleridges. Individual ridges are as much as 15 m in reliefand up to 1 km in length.Kame terracesThe third major morphologic feature depicted inFigs 2Ð5 involves kame terraces banked againsthigh ice-contact slopes alongside Lago Llanquihueand Bah’a Puerto Montt. A prominent terrace, inplaces up to 1 km broad, fringes much of the westernand southern shores of Lago Llanquihue. Onpromontories north of the R’o Maull’n outlet, theposition of coarse outwash in the terrace requires abuttressing ice lobe. In embayments the terrace iscomposed largely of lacustrine sediments, againrequiring an adjacent ice lobe. All the lakeside terracesare depicted as kame terraces in Fig. 3, withsome ice-contact slopes modiÞed by wave action.Kame terraces alongside Bah’a Mu–oz Gameroare composed of lacustrine sediments and have surfaceelevations of as much as 130Ð145 m. Most arebanked against an ice-contact slope about 30Ð40 melevation below the intakes for the Chanch‡n andPuntilla channels. A large terrace, again with a surfaceelevation of about 130 m, has an ice-contactslope that merges into the long moraine ridge projectingpart of the way across the mouth of Bah’aMu–oz Gamero as Punta Centinela (Fig. 3). As depictedin Fig. 3, Punta Centinela is on line with aprominent moraine ridge that trends eastÐwestabout 1 km inland from the northern lake shore.Laugenie (1982) referred to this ridge as the PuertoChico moraine. Terraces occur on both the proximaland the distal side of the Puerto Chico moraine. Theinner lacustrine terraces probably formed beside anice lobe that still Þlled much of Bah’a Mu–oz Gameroshortly after recession from the bayside ice-contactslope. As mentioned above, this high ice-marginallake in Bah’a Mu–oz Gamero probablydrained westward through the Chanch‡n and Puntillachannels. The moraine and terrace near PuntaCentinela most likely reßect a separate ice advance.Fig. 3 shows that the shores of Lago Llanquihue eastand southwest of Bah’a Mu–oz Gamero lack significantlake-marginal terraces. Instead, kame terracesalong the northern lake shore east of Bah’a Mu–ozGamero have moraines on their proximal sides andreßect an ice lobe that extended out of the main lakebasin. Likewise, the ice-marginal deposits alongsideBah’a Rincones, and as far south as the northshore of Bah’a Frutillar, do not feature kame terraces.In fact, extensive terraces extend only from theGeograÞska Annaler á 81 A (1999) á 2175


G.H. DENTON ET AL.north shore of Bah’a Frutillar to the R’o Maull’noutlet at the town of Llanquihue. It is also evidentfrom Fig. 3 that the lakeside terraces at Bah’a Frutillarare not continuous with those at Bah’a Mu–ozGamero. Therefore, kame terraces cannot be connectedas a continuous system from the R’o Maull’noutlet northward to Bah’a Mu–oz Gamero.In Fig. 3 we mapped a continuous kame terracealongside southern and western Lago Llanquihuefor 65 km from Punta Cabras to Bah’a Frutillar.From both north and south, the surface of this terracedeclines to a common elevation of about 70 mat the R’o Maull’n outlet. It seems probable thatthis terrace represents a coherent ice-marginal positionof the former Lago Llanquihue piedmontglacier, a situation compatible with the presence ofdebris-ßow volcanic breccia on the terrace surfacesouth of the R’o Maull’n outlet. There are at leastthree, and perhaps four, such debris-ßow deposits.They are generally sandy, although some havecoarse pumice and angular andesite clasts. Mostexhibit crude vertical sorting. Some Þll erosionalchannels as much as 2 m deep. Laminated glaciolacustrinesediments occur between some of thedebris-ßow units. Porter (1981) attributed thesesediment ßows to lahars from Volc‡n Calbuco,whereas H. Moreno (pers. comm. 1996) attributesthem to debris ßows associated with periodic outburstsof a lake dammed on the northern ßank ofVolc‡n Calbuco by the former Lago Llanquihuepiedmont ice lobe. In either case the distribution ofthe debris-ßow sediments requires that the LagoLlanquihue ice lobe stood against the terrace. Thusour reconstruction is similar to that of Porter(1981), except that we are not able to trace a continuouskame terrace north of Bah’a Frutillar.Prominent kame terraces near Bah’a PuertoMontt occur on the distal side of a discontinuousLlanquihue moraine belt along the upper lip of anextensive bayside ice-contact slope (Fig. 4). Thismoraine belt consists of small ridge fragments thatmark the marginal position of the former Seno Reloncav’piedmont glacier when the prominent kameterraces were formed. Other fragmentary ice-contactdeposits are on the proximal side of smaller andlower terraces beside Bah’a Puerto Montt. Overall,the geometry of the terraces and associated icemarginalfeatures shows that Ri— Coihu’n, whichnow ßows from Volc‡n Calbuco directly into SenoReloncav’, was deßected along the kame terracesby a blocking Seno Reloncav’ piedmont glacier.The resulting ice-marginal river built the prominentkame terraces near Bah’a Puerto Montt, and thenßowed northeast to R’o Maull’n, depositing subsidiaryoutwash terraces. Farther south extensive subsidiaryterraces separate Llanquihue moraine beltsof the Seno Reloncav’ piedmont glacier. The proximalends of these terraces are graded to the innermoraine belt adjacent to Bah’a Puerto Montt. Thedistal ends converge into narrow channels. One ofthese tributary channels exits southward from thesubsidiary plain and joins the channel of R’oGomez, which cuts through Llanquihue morainebelts and the main Llanquihue outwash plain. TheR’o Gomez channel also formed the main outßowfor a similar subsidiary terrace that heads at the innermoraine system northwest of Calbuco.Overall, Fig. 4 shows that high kame terracesnear Bah’a Puerto Montt merge into an extensivesystem of subsidiary outwash terraces, all of whichdrain to R’o Maull’n in narrow channels that cutthrough both the outermost Llanquihue morainebelts and the main outwash plain. These kame terracesall head at an inner moraine belt at the top ofan ice-contact slope beside Seno Reloncav’. NearBah’a Puerto Montt is a complex of small kame terracesbanked against the lower portions of this icecontactslope. These terraces, many of which havesmall moraine remnants on their ice-contact edges,were deposited by ice-marginal streams as theSeno Reloncav’ piedmont glacier lobe downwastedfrom the top of the gulfside ice-contact slope.Glacial sedimentsFive major types of glacial deposits make up Llanquihuedrift; gravelly diamicton, lodgement till,waterlain till, outwash, and laminated glaciolacustrinesediments. In addition, in-situ and reworkedpyroclastic ßows from Andean volcanoes mantlemoraine belts and outwash plains; remnants of ßowsediments also occur between glacial units instratigraphic sections. Fig. 7 shows a schematic diagramof how these Þve major types of deposits Þtinto a glaciogenic facies model related to morphologicalunits depicted in Figs 2Ð5.Gravelly diamicton and lodgement tillComplex glaciogenic sequences dominated bygravelly diamicton form moraine cores. These sequencesare as much as 10 m thick and feature massiveto faintly bedded gravelly diamicton separatedby lenses of sand and sorted gravelly outwash. Attheir distal ends, they are interlayered with outwashor lacustrine sediments. The enclosed clasts are176GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTExplanation for all stratigraphic sectionsFig. 7. Facies model of glaciogenic deposits in the southern Lake District. The cross-sections focus on the moraine belts along the westernmargin of lakes in Valle Central. The relic land surface prior to the LGM is accented to show the facies left during one glacial cycle(above) separated from the complex of facies left during earlier glacial cycles (below). Details for each of the major units are shownin the corresponding diagrams (AÐF). (A) Waterlain facies. Sediment supply is from super-, en- and subglacial positions, but in all casesthe depositional processes occur primarily in the proglacial lake. Subglacial till is absent, but local low-amplitude folding, faulting, andinjection of clastic dikes indicate large-scale fracturing of sediments. Deformation is limited to soft-sediment units that were depositedrapidly. See Turbek and Lowell (1999) for additional details. (B) A kame terrace. Such a terrace may be composite from more thanone glacial advance as outlined by the darker lines. Deposition in the lower portions is in a subaqueous environment, but a transitionto subaerial deposition commonly occurs within the stratigraphic sequence that makes up the terrace. Debris-flow sediments of volcanicbreccia occur on the terrace surface (Porter 1981). (C) An ice-contact slope. Glacial overrunning serves to smooth such slopes by infillingtopographic lows and by eroding topographic highs. Injections, clastic dikes, and transported blocks are common in the lowerparts of such slopes, and lodgement till is common at the tops of such slopes. See Turbek and Lowell (1999) for additional details.(D) Stacked moraine ridges that represent ice-marginal positions and may occur along the top of ice-contact slopes. The sediments inthese ridges vary from lodgement till to gravelly sediment flows, with most material being stratified. Deformation and transport of inclusionsare limited to the proximal sides of such ridges. (E) Composite moraine ridges that represent ice margins reoccupying the sameplace several times. Ridge cores may be deeply weathered, with pyroclastic flow sediments separating internal deposits of differentages. Erosion on the proximal side may result in old sediments being at or near the present land surface. (F) Ridges composed of outwashfolded beyond the former ice margin (SchlŸchter et al. 1999). Such ridges occur in a few restricted areas. A detachment plane probablyoccurs within the outwash at the base of the folds.GeograÞska Annaler á 81 A (1999) á 2177


G.H. DENTON ET AL.Fig. 7. Continued.178GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTFig. 7. Continued.GeograÞska Annaler á 81 A (1999) á 2179


G.H. DENTON ET AL.volcanic and crystalline rocks from the Andes. Weinterpret the faintly bedded diamictons as gravellysediment ßows from former ice margins at theproximal edges of moraine ridges and hillocks. Thebanded appearance comes from the stacking of severalindividual sediment ßows within each morainecore. Fig. 7D shows a schematic example of such amoraine core. Because most clasts in the gravellydiamictons are round to subround, the ßow materialwas probably derived from outwash terraces andformer moraine cores overrun by advancing ice.Compact diamicton mantles many moraines andice-contact slopes (Fig. 7). This diamicton commonlyachieves a thickness of 0.5Ð1.5 m, althoughin places it is as much as 3 m thick. It is light-to-mediumgray in color, compact, stony, and non-sorted.Clasts are volcanic and crystalline rocks from theAndes. Most clasts are round to subround and up tocobble size. Some are striated. The diamicton commonlyexhibits well-developed Þssility, featureslenses and boudins, and contains isolated large, orientedboulders. Fissility is associated with severeshear deformation that in places slices large boulders.We interpret this massive, compact diamictonas lodgement till. A separate facies, located beneaththe lodgement till, involves faintly beddeddiamicton layers formed by gravity ßows from adjacentadvancing ice.Both the gravelly glaciogenic sequences and thetill units have been deformed in many places, withconspicuous shearing and stacking of lenses. Welldevelopedshear planes have slickensides and oriented,pushed, and rolled boulders. Compressionjoints are common.Waterlain tillWaterlain till is widespread near the present-daywater level along the western shore of Lago Rupanco,Lago Llanquihue, and Seno Reloncav’, as wellas along the eastern coast of northern Isla Grandede ChiloŽ. Because it is so indurated, this waterlaintill controls the outlet elevations of Lago Rupancoand Lago Llanquihue.Figure 7A illustrates our view of the origin of waterlaintill. It typically consists of laminated glaciolacustrinesediments with lenses and discontinuousbeds of coarse material, including cobbles of volcanicand crystalline rocks from the Andes (Figs 8,9). These coarse sediments represent a ßow regimenfeaturing turbidites at all scales, with highly developedgrading distal from the former ice margin.Convolutions and interÞngering of Þne and crudestratiÞcation form irregular contact surfaces. Thereis occasional compaction faulting. Abrupt facieschanges are common. Most cobbles in the turbiditesare subangular to subround, and some are striated.Diapirs derived from underlying beds are commonlyinjected into these coarse lenses and beds. Flameand diapir structures are also present within the lacustrinebeds themselves. In addition, the lacustrinesediments show slump structures with consequentfolding of individual beds, including one case witha bed rolled around itself. Other structures includeclastic dikes of diamicton injected along jointplanes within lacustrine sediments. The waterlainunits are folded near promontories. Fine bedding inlacustrine beds, turbidites, and loading structures ofwaterlain till are preserved in detail. Basal lodgementtill does not occur near the lake edge below themodern high-water level. Above the high-watermark, the waterlain till passes beneath an organicbed within a kame terrace near the town of Llanquihue,and grades to surface lodgement till along thenorth shore of Bah’a Puerto Montt.OutwashOutwash of the main and subsidiary Llanquihueterraces shows coarsening toward the moraines thatmark coeval ice margins (Fig. 10), where subround-to-roundclasts of Andean origin are as muchas 0.3 m in diameter. In places, the outwash containsreworked organic clasts derived from peat andsoils developed on pyroclastic ßow deposits. Thestandard glacioßuvial outwash facies consists ofhorizontally stratiÞed, medium-to-bouldery gravel,with all transitions from sandy diamicton toopenwork gravel. Laminar cross-bedding is present,but not dominant. The outwash has extensivesand units. In positions near the former ice margins,it also contains massive, bouldery-to-sandy grainßows that commonly disrupt pre-existing bedding.In places, the advancing ice subjected the outwashto mechanical reworking, faulting, folding (withoverturned folds), boudinage, and elongations.Figs 11 and 12 give typical examples of outwashdeformed by advancing ice, whereas Fig. 13 showsFigs 8Ð10, see p. 189.180GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTFig. 11. Site 31 on east side of Route5 south of Puerto Varas. Location isgiven in Fig. 15. See legend in Fig. 7(p. 177) for lithologic details. Thelithological section records the featuresproduced by an advancing icelobe that highly deformed its own icecontactoutwash sediments. The easternmostsector of the outcrop displaysmechanically reworked stacksof gravel with rolled-up gravel-lensstructures. The glacier margin advancedto the top of the ridge, withcontinuing deformation of the outwash.Toward the top of the sectionthere was increasing accumulation offlow till. The inset shows stacking ofgravelly outwash with intense deformation(rolled-up lenses) and till bedscut by steep shear planes. The radiocarbonsamples listed in Table 1 forsite 31 do not come from this section,but rather from another nearby borrowpit on the west side of Route 5.Fig. 12. Road cut showing stratigraphic section through the core of a Llanquihue moraine ridge beside Route 5 at site 33 in Fig. 15.See legend in Fig. 7 (p. 177) for lithologic details. The section shows highly deformed glaciofluvial gravel and till thrust onto a coreof older fine-grained diamicton (open hatched signature) and outwash. The main ridge of the deformed sediment (southern sector) canbe interpreted as two compressed and thrust folds. The morphological expression of this deformation is a pronounced moraine ridge.The radiocarbon samples listed in Table 1 for the core of the moraine at site 33 are from a borrow pit 150 m west of this road cut.GeograÞska Annaler á 81 A (1999) á 2181


LLANQUIHUE DRIFTPyroclastic depositsFine-to-coarse pyroclastic sediment, ranging from0.5 m to several meters in thickness, covers morainebelts, outwash plains, and kame terraces. Italso makes up interdrift sediments exposed instratigraphic sections. The glass component of thismaterial is weathered, but feldspar crystals, pumicelithics, and small pieces of charcoal are commonlywell preserved. The deposits thicken in depressionsand thin on moraine ridges and hills. Thisvolcanic material likely represents the aggregateaccumulation from successive pyroclastic ßowsfrom Andean volcanoes. Some individual ßowscrossing the Lake District were hot enough to reducewood to charcoal, still enclosed in the pyroclasticsediments. It is difÞcult to determine thethickness of individual ßows, because in the Þeldarea they are not usually separated by airfall tephras.However, the overall Þne grain size suggeststhat only the thin distal facies of the ßows arepresent west of Lago Llanquihue and Seno Reloncav’,as well as on Isla Grande de ChiloŽ.Glaciogenic features and moraine preservationTwo features in Fig. 7 deserve further comment.The Þrst is that some moraines and kame terracesare sheared and stacked. The other is that many icecontactslopes commonly exhibit complex glacialdeposits from several ice advances. The implicationis that the individual ice-contact slopes in Figs2Ð5 do not necessarily depict a single ice-marginalposition, but may have been occupied repeatedly.Clay minerals produced by rapid weathering ofinterstitial volcanic glass cement drift in the southernLake District. The result is a striking preservationof glacial landforms, not only of Llanquihuebut also of pre-Llanquihue age. We made use of thispreservation of moraine belts and outwash plains,along with the relative weathering differences intill in the moraine belts, to separate Llanquihuefrom older drifts. Individual moraine belts, commonlyseparated by outwash plains, can be delineatedas in Fig. 3. For example, the main Llanquihueoutwash plain depicted in Fig. 3 separates aprominent set of moraines fringing Lago Llanquihuefrom older moraines and outwash plains fartherto the west. Weathering of near-surface till inthese successive moraine belts shows consistentlysharp differences that correspond with morphologicboundaries. Till in moraine belts near the lakesis virtually unweathered. Porter (1981) reportedweathering rinds of as much as 0.5 mm on near-surfacevolcanic clasts in this till. Oxidation of the matrixis slight and conÞned to the upper meter. Exceptwhere it is very compact, the till in the olderend-moraine systems is colored yellowish-brownthrough its observed thickness from limonite stainingalong Þssures and in clast sockets. On the basisof this weathering difference, then, the morainebelt fringing Lago Llanquihue is composed ofLlanquihue drift and the older moraines are composedof Casma/Colegual drift. The two well-preservedmoraine belts (each fronted by an outwashplain) that we mapped west of the Llanquihue morainebelts (Fig. 3) are equivalent to the Casma andColegual moraines of Mercer (1976) and Laugenie(1982).Radiocarbon dates of Llanquihue driftTable 1 gives more than 450 new radiocarbon datesof Llanquihue drift deposited by the Lago Rupanco,Lago Llanquihue, Seno Reloncav’, Golfo de Ancud,and northern part of the Golfo Corcovado piedmontglaciers. Table 2 gives weighted mean dates of selectedindividual organic beds within drift sequences.Site numbers for radiocarbon samples are listedin Table l and plotted in Figs 14Ð17 and in Fig 1. Additionaldates for pollen diagrams are given inHeusser et al. (1999) and Moreno et al. (1999). Wenow discuss the most pertinent of the dates from theregions of Lago Rupanco, Lago Llanquihue, SenoReloncav’, and northern Isla Grande de ChiloŽ, inFigs 14Ð17 (see PLATES 6Ð9 in separate cover)Fig. 14. Location of radiocarbon sample sites and stratigraphic sections cited in Table 1 and in text for Lago Puyehue and Lago Rupancomap sheet of Fig. 2. Plate 6.Fig. 15. Location of radiocarbon sample sites and stratigraphic sections cited in Table 1 and in text for Lago Llanquihue map sheet ofFig. 3. Plate 7.Fig. 16. Location of radiocarbon sample sites and stratigraphic sections cited in Table 1 and in text for Seno Reloncav’ map sheet ofFig. 4. Plate 8.Fig. 17. Location of radiocarbon sample sites and stratigraphic sections cited in Table 1 and in text for Isla Grande de ChiloŽ map sheetof Fig. 5. Plate 9.GeograÞska Annaler á 81 A (1999) á 2183


G.H. DENTON ET AL.each case starting with the youngest and progressingto the oldest dates. See Bentley (1997) for additionalnew radiocarbon dates from near LagoPuyehue.Lago Rupanco RegionLimiting maximum radiocarbon dates for the outermostlate Llanquihue moraine near Lago Rupancocome from the borrow pit at site 1. Mercer(1976) reported an age of 19,450±350 (I-5679) forashy peat just below till at this site. The speciÞcsample locality has been destroyed as the borrowpit has been expanded. The pit now features a capof lodgement till over a moraine core of mechanicallydislocated glacioßuvial and organic beds. Inonly one locality, in the southeast corner of the pit,is there an undisturbed transformation from peatsealed with gray silt giving way upward to sand andtill. A sample from the peatÐsilt contact yielded anage of 22,460±135 14 C yr BP (A-9516) (Table 1),which we take to be a close maximum for ice advancebecause the youngest dates of additionalsamples from translocated organic beds beneaththe till cap are 22,560±495 14 C yr BP (T-9660A)and 22,745±295 14 C yr BP (T-10309A) (Table 1).Lago Llanquihue RegionLakeside kame terraces. The prominent kame terracebeside Lago Llanquihue is dated at several localities.An important section occurs on the lakesideedge of the terrace near the town of Llanquihueabout 2 km south of the R’o Maull’n outlet (site23) (Figs 18 and 19). Situated 3.0 m above thepresent-day lake level, an organic silt bed near thebase of this section contains fossil beetles andwood fragments. Eight dates of wood from near thetop of this bed yielded an error-weighted mean of14,869±38 14 C yr BP (Tables 1 and 2, Fig. 18). Asmall wood piece from near the base of the organicbed yielded an AMS date of 15,600±120 (AA-21939) (Fig. 18). Thus the organic bed was depositedbetween 14,869 and 15,600 14 C yr BP.The Llanquihue organic bed represents an intervalwhen lake level was near or even below itspresent-day value. Below the organic bed is a compactunit of waterlain till that represents glacier advanceacross the site. Above the organic bed areglacioßuvial units that make up the kame terrace.Nearby, the terrace is capped by debris-ßow sedimentcomposed of silt-to-coarse-sand-sized pyroclasticmaterial. The Lago Llanquihue piedmontglacier must have been at or very close to the currentouter edge of the terrace to allow deposition ofthe glacioßuvial units in this topographic situationon the organic bed. It still must have been in essentiallythe same position when the debris-ßow volcanicsediments were deposited on the terrace surface.Therefore, the Llanquihue site affords evidencefor an advance of the Lago Llanquihue icemargin to a position alongside the prominent kameterrace on the western lake shore at 14,869 14 C yrBP, with the ice persisting until deposition of the debris-ßowvolcanic sediments on the terrace surface.An organic bed in the kame terrace at PuertoPhillippi (site 19) is sealed by glaciolacustrine sediments.The top of this bed dates to 14,650±75 14 Cyr BP (A-9423).We obtained new dates from three stratigraphicsections in the kame terrace at Puerto Varas. At BellaVista Bluff (site 30) a recent slip revealed a freshexposure of the organic bed and lacustrine sediments(Figs 20 and 21) previously reported by Porter(1981). The organic bed, located near the centerof the section, is 23 cm thick. Two wood pieces(each about 20 ´ 8 cm) occur within glaciolacustrinesediments 8 cm above this bed. These pieces,both probably washed into the lake, afforded agesof 14,430±140 14 C yr BP (A-6322) and 14,655±7514 C yr BP (A-8546) (Table 1). In-situ organic litter(with twigs) sealed with glaciolacustrine sedimentsoccurs on the top surface of the bed at 60.97m elevation (the present lake level varies but isclose to 51 m elevation). Samples from this litteryielded ages of 14,715±85 14 C yr BP (A-8549R),14,560±95 14 +80C yr BP (T-9656A), 14,595 14 C yrBP (A-9187), 14,585±80 14 -75C yr BP (A-9191),14,600±75 14 +80C yr BP (A-9185), and 14,135 -7514 Cyr BP (A-9189) (Table 1). The weighted mean is14,540±29 14 C yr BP (Table 2). Most basal samplesfrom the organic bed date to 15,200Ð15,300 14 C yrBP, but the oldest are 15,635±95 14 C yr BP (A-8550)and 15,730±160 14 C yr BP (A-9190). Therefore,the Bella Vista Bluff organic bed was deposited betweenabout 15,700 and 14,532 14 C yr BP, veryFig. 19, see p. 192.Fig. 21, see p. 192.184GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTFig. 18. Composite lithostratigraphy oflakeside cliff section near the town ofLlanquihue at site 23 in Fig. 15. Seelegend in Fig. 7 (p. 177) for lithologicdetails. Radiocarbon dates are describedin Table 1. Unit A is waterlain till atthe base of the section, indicative ofglacier advance. It consists of stronglydeformed silty sand and sand, interbeddedwith laminated clayey silt and withflows of massive-to-bedded diamictonthat includes boulders and cobbles.Unit A is overlain by a nonglacial complexwith numerous radiocarbon samples.It consists of a fining-upward sequence,with basal boulders and silty tosandy gravel grading upward into massiveclayey sandy silt with beetle andwood fragments, which is overlain firstby clayey silt and clayey peat withwood fragments and then by clayey siltwith wood fragments. The non-glacialbeds are only preserved at this locality;they are missing a few meters to thesouth and approximately 10 m to thenorth. Unit B is a coarsening-upwardsequence grading from bedded-tocross-bedded,medium-to-coarse sand,with pumice clasts and rip-up fragmentsfrom lower units, into sandygravel and sand with well-developedclimbing ripples and then to massiveand cross-bedded gravel. A highamount of volcaniclastic sand occursthroughout the upper part of the section.This sequence represents the iceadvance responsible for the kame terrace.Unit B is overlain by topsoil developedin pyroclastic flow sediments.close to the age range of the organic bed near thetown of Llanquihue. The Bella Vista bed is a lateralcontinuation of the one that previously yielded anage of 13,965±235 14 C yr BP (UW-481) (Porter1981); the reason for the difference is not known(Table 1, Fig. 20).Below the Bella Vista organic bed are gray, horizontallybedded glaciolacustrine sediments, withoutdropstones, that give way upward to beachsand. Above the organic bed is laminated glaciolacustrinesilt, overlain sharply along a beddingplane by at least three units of debris-ßow volcanicbreccia (3Ð4 m thick). Glaciolacustrine sedimentsseparate the lowest two volcanic ßow deposits. Thehighly erosive ßows that deposited the lowest brecciaunit peeled away lacustrine sediments along abedding plane and overturned them into a fold. Theorientation of this fold indicates that the debris ßowpassed from south to north across the top of the terrace.At Northwest Bluff (site 28) an organic bed at 63m elevation rests on a coarse gravel deposit in thelakeside terrace (Porter 1981). The top of the organicbed is sealed by glaciolacustrine silt and Þnesand, which in turn is overlain by several units ofthe same debris-ßow volcanic breccia found at BellaVista Bluff. Two samples from the sealed top ofthe organic bed yielded ages of 14,825±110 14 C yrBP (A-9523) and 14,925±95 14 C yr BP (A-9524)(Table 1); the weighted mean is 14,882±72 14 C yrBP (Table 2). This is the same organic bed that previouslyyielded an age of 13,145±235 14 C yr BP(UW-480) (Porter 1981), an unexplained discrepancy.GeograÞska Annaler á 81 A (1999) á 2185


G.H. DENTON ET AL.Fig. 20. Schematic composite lithostratigraphyof stratigraphic section atBella Vista Bluff (site 30 in Fig. 15),Puerto Varas. See legend in Fig. 7(p. 177)for lithologic details. Unit A islaminated glaciolacustrine clay and siltcoarsening upward to silt at the top ofthe unit. Unit B is an organic bed (up to26 cm thick) capped by lacustrine siltsand clays that make up unit C; woodfragments (up to 15 cm long and 7 cmin diameter) occur in the organic bedand again in lacustrine silts and clays 3to 10 cm above the organic bed. Unit Dis composed of debris-flow deposits ofvolcanic breccia that fine upward fromgranules and gravel at the base to coarsesand at the top; folded beds of lacustrinesediment occur within unit D, either asrip-up clasts or as rotated lacustrinebeds partially entrained in debris-flowsediments. Unit E is a fining-upwardsequence of lacustrine sand and silt.The coarse fraction (medium-grainedsand) is composed of volcanic pumice;numerous climbing ripples and diapirsoccur near the base of unit E. Unit F isa debris-flow deposit that fines upwardand includes some organic rip-upclasts. This unit shows a sharp planarcontact with the underlying lacustrinesediments. Unit G is the uppermostdebris-flow unit of volcanic brecciathat forms the terrace surface. The topmost40 cm feature large (1Ð2 cm diameter)pumice fragments; the base ofthis unit is cemented with iron oxides,forming an indurated layer 5 cm thickon top of the intermediate debris-flowunit. Insets give details of organic bedand of radiocarbon sample localities.We obtained new dates from the railroad bridgesite (26) of Heusser (1974), Mercer (1976), Porter(1981), and Hoganson and Ashworth (1992) inwestern Puerto Varas. Heusser (1974) previouslydescribed the section as exposing, from top to bottom,laharic sediments as reinterpreted by Porter(1981) (470 cm), peat (10 cm), gyttja (70 cm),peat (5 cm), gyttja (165 cm), and gravel (75 cm).A piece of wood from 15 cm above the base of thelower gyttja dated to 16,270±360 14 C yr BP (RL-113) (Mercer 1972, 1976; Heusser 1974). Thebase of the upper, 10-cm-thick peat bed yieldedan age of 14,250±400 14 C yr BP (GX-2948), andthe top of this peat bed gave an age of 13,300±550 14 C yr BP (GX-2947) (Heusser 1974). Additionalages given by Mercer (1976) of this upperorganic bed are 13,200±320 14 C yr BP (GX-4169)for the uppermost 0.5 cm of peat and 13,750±295186GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFT14 C yr BP (GX-4170) for wood at 2 cm depth.Hoganson and Ashworth (1992) reported an ageof 14,060±450 14 C yr BP (GX - 5507) for the topof the peat bed.It proved difÞcult for us to piece together thestratigraphic section, because a concrete retainingwall and two abutments have been built since thesite was Þrst described. We therefore received permissionfrom town ofÞcials to re-excavate key portionsof the section including the part now behindthe retaining wall. Fig. 22 gives our interpretationof the section on the south side of Route V-50 nearthe railroad bridge, based on the new hand-dug excavations,as well as on notes from S.C. Porter andA. Ashworth (written comm. 1996). The lowermostunit consists of horizontally laminatedglaciolacustrine silt with dropstones, 48 cm thick,exposed 14 m east of the railroad abutment. Poorlysorted, coarse gravel with a sandy matrix, 92 cmthick, rests on the glaciolacustrine unit. The uppersurface of the gravel unit dips downward towardthe west. The crystalline and volcanic erratics withinthis deposit were derived from the Andes. Onegranite boulder is 50 cm long. About 9 m east of therailroad abutment, organic silt, 26 cm thick with afaint gyttja layer, occurs between the gravel unitbelow and the debris ßow of volcanic brecciaabove. The organic silt is exposed laterally for atleast 5 m. It comfortably overlies the gravel unit,and has a highly eroded contact with the overlyingvolcanic debris-ßow sediments. Several diapirs oforganic silt rise into the overlying debris-ßow sediments.In the re-excavated portion of the section behindthe retaining wall described by Heusser (1974) andHoganson and Ashworth (1992), organic silt, 1.68m thick with wood pieces (one in vertical growthposition) and diffuse brown bands, rests on sandand gravel. This silt is capped by a 26-cm-thick organiclayer with peat and wood. The top of the organiclayer, in turn, is sealed by a 29-cm-thick massive,coarsening-upward volcanic sand unit. Thisvolcanic sand is overlain, in turn, by a silt unit withtwo 1-cm-thick peat beds that include small woodfragments. The upper 4.2 m of the exposed sectionfeature several compact, massive units of volcanicbreccia emplaced by debris ßows across the site,just as at Bella Vista Bluff and Northwest Bluff. Pyroclasticßow deposits cap the sequence where it isundisturbed.We think that the gravel unit exposed 14 m eastof the abutment corresponds with the gravel unitdescribed by Heusser (1974) and Mercer (1976) atthe base of the section behind the retaining wall.This gravel unit, in turn, probably corresponds withgravel exposed on the laminated glaciolacustrinesilt unit west of the concrete abutment for the roadoverpass. The organic silt resting on the gravel eastof the railroad bridge abutment almost surelymatches the thick lower organic silt unit betweenthe abutments. The same debris-ßow volcanic brecciasextend across the entire section. The upperpeat beds are missing east of the railroad abutment,probably due to erosion by debris ßows. The suddenloading from the emplacement of the debrisßowvolcanic sediments produced diapirs of organicsilt. Our correlations suggest that both the coarsegravel unit and the laminated lacustrine sedimentwith dropstones, both exposed 7Ð14 m east of therailroad abutment, project stratigraphically belowthe organic silt and peat beds between the abutments.We undertook extensive dating of the two thinpeat layers encased in the silt that overlies the 29-cm-thick massive volcanic sand unit above theprominent organic bed. Our Þrst attempt yieldedunsatisfactory results, because we had not yet receivedpermission to reopen the excavation andhence were forced to sample in an area with potentialrootlet and groundwater contamination. Whensubjected to standard acidÐbaseÐacid pretreatment,Þbrous peat from the upper of the two thinlayers yielded an age of 12,840±90 14 C yr BP (AA-7459); a redating of the same sample after strongacidÐbaseÐacid pretreatment afforded an age of13,940±85 14 C yr BP (AA-7459B). This discordancesuggests contamination by humates. An ageof 14,175±110 14 C yr BP (AA-7460) was obtainedfor a Þbrous peat sample from the lower of the twothin layers. But this date may also be too young, asa small wood piece from the same horizon gave anage of 14,600±110 14 C yr BP (AA-7465) with standardacidÐbaseÐacid pretreatment and of 14,350± 9014 C yr BP (AA-7465C) with acidÐbaseÐacid pretreatmentand then extraction of cellulose. Anothersmall wood sample from the lower of the two thinpeat layers yielded ages of 14,290±100 14 C yr BPand 14,460±85 14 C yr BP (ETH-13528) for thesame graphite target. But a complete redating withseparate pretreatment yielded an anomalouslyFig. 22, see p. 190.GeograÞska Annaler á 81 A (1999) á 2187


G.H. DENTON ET AL.young age of 12,950±100 14 C yr BP (ETH-13528),again suggesting contamination. Therefore weconsider this initial set of dates from the two thinpeat beds to be unreliable.Following re-excavation, we collected four uncontaminatedsamples from the lower of the twothin peat layers that yielded concordant ages of14,600±100 14 C yr BP (AA-21011), 14,360±10014 C yr BP (AA-21012), 14,570±150 14 C yr BP (AA-21013), and 14,680±100 14 C yr BP (AA-21014)(Table 1); the weighted mean is 14,550±54 14 C yrBP (Table 2).The prominent 26-cm-thick organic bed haspockets of wood in the upper 7 cm. We subdividedthree individual wood samples in order to obtainreplicate ages for each. Replicates of one samplewere 14,415±70 14 C yr BP (A-9193) and+7514,475 -7014 C yr BP (A-8551). Replicates of anotherwood sample were 14,620±90 14 C yr BP (A-9194) and 14,735±75 14 C yr BP (A-8553). The+80third replicate set gave ages of 14,57014 C yr BP(A-8554) and 14,52014 -75+80-75 C yr BP (A-9195). Individualdates of other wood samples from this samehorizon are 14,625±75 14 C yr BP (A-8552), 14,930±80 14 +110C yr BP (A-8175), and 14,285 -10514 C yr BP(A-8775) (Table 1). Porter (1981; written comm.,1996) reported a date of 15,050±100 14 C yr BP(QL-1335) for wood from this horizon (Table 1).Overall, the weighted mean age for wood samplesfrom the upper 7 cm of the organic bed is 14,613±25 14 C yr BP (Table 2). It was samples from this 26-cm-thick organic layer that previously yielded agesof 13,200 to 14,200 14 C yr BP (Heusser 1974; Mercer1976; Hoganson and Ashworth 1992). Again,the reason for the discrepancies of these earlierages with our new dates is not known.We obtained 11 new dates of wood pieces fromvarying depths in the organic silt and sand unit thatunderlies the 26-cm-thick organic bed (Table 1,Fig. 22). These dates range from 17,320±140 14 Cyr BP (AA-23726) to 17,880±140 14 C yr BP (AA-23732), implying rapid deposition of the silt andsand unit. Three dates of a small tree in growth position116Ð126 cm below the top of the silt and sandunit are: 17,505±150 14 C yr BP (A-9526.l), 17,520±140 14 C yr BP (AA-23729), and 17,790±140 14 Cyr BP (AA-23731). All the new dates are considerablyolder than the age of 16,270±360 14 C yr BP reportedearlier for wood from near the base of thesilt unit (Mercer 1972, 1976; Heusser 1974).Again, the reason for the discrepancy is not known.At the Calle Santa Rosa site (site 27) in PuertoVaras, situated about halfway between the railroadbridge and Bella Vista Bluff, Mercer (1976) describedan organic bed between units of horizontallylaminated glaciolacustrine sediment. Here thekame terrace is also capped by the debris ßows ofvolcanic breccia present at Bella Vista Bluff,Northwest Bluff and the railroad bridge site. Organicmatter sealed at the surface of the organic bedat 62 m elevation (Porter 1981) at the Calle SantaRosa site yielded an age of 14,820±230 14 C yr BP(I-5033) (Mercer 1976).From this revised chronology, we subdivide thesediments in the kame terrace at Puerto Varas intoa threefold sequence; the units above the organicbeds, the organic beds themselves, and the units belowthe organic beds. Exposed below the organicbeds is horizontally laminated lacustrine silt withoutdropstones at the Bella Vista Bluff (Fig. 20),Bella Vista Park (Porter 1981), and Calle SantaRosa (Mercer 1972, 1976) sites. At the NorthwestBluff site, gravel underlies the organic bed (Porter1981). At the railroad bridge site, a thin gravel unitand laminated glaciolacustrine silt with dropstonesunderlie the organic beds. Above the organic bedsare glaciolacustrine sediments at all sites except therailroad bridge; at least three units of debris-ßowsediments of volcanic breccia cover the terrace atall sites.The organic beds in the Puerto Varas terrace liewithin an elevation range of 60Ð67 m. We surveyedthe top of the newly exposed Bella Vista Bluff organicbed at 60.97 m elevation, and Porter (1981)placed the top of this bed at 60.8 m in an adjacentexposure now covered. We surveyed the top of theprominent 26-cm-thick organic layer at the railroadbridge site at 67.12 m elevation, comparedwith PorterÕs (1981) value of 66.2 m. Porter (1981)placed the top of the organic beds at 60.1 m elevationat Bella Vista Park, at 62.0 m at Calle Santa Rosa,and at 63.0 m at Northwest Bluff.The dates of peat and wood in these organic layersoriginally had a signiÞcant scatter between15,715 14 C yr BP and 13,200 14 C yr BP (Mercer1972; Heusser 1974; Porter 1981; Hoganson andAshworth 1992). But the top of the Bella Vista Bluffbed is now dated to 14,540 14 C yr BP (Table 2), andthe base of the organic layer is 15,700 14 C yr BP (A-9190) (Table 1). These results are in reasonableagreement with the single age of 15,715±440 14 Cyr BP (GX-5275) from within the organic bed at thenearby Bella Vista Park site (Porter 1981). The topof the Northwest Bluff organic bed now dates to14,882 14 C yr BP (Table 2) rather than 13,145 14 C yrBP (Porter 1981). The upper portion of the 26-cm-188GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTFig. 8. Waterlain till on a promontoryalongside Lago Llanquihuenear Bella Vista Bluff (site 30) inPuerto Varas.Fig. 9. Waterlain till on a promontoryalongside Lago Llanquihuenear Bella Vista Bluff (site 30) inPuerto Varas.Fig. 10. Outwash at site 11 at FrutillarAlto. The surface of this outwashunit, shown in fig. 7 of Andersenet al. (1999), is graded to theoutermost Llanquihue moraine.The borrow pit at site 11 yielded numerousdates of reworked organicclasts (Table 1) that are maximumvalues for the outwash and hencefor the outer Llanquihue moraine towhich the outwash is graded.GeograÞska Annaler á 81 A (1999) á 2189


G.H. DENTON ET AL.thick organic layer at the railroad bridge site nowdates to 14,613 14 C yr BP (Table 2) rather than to13,200Ð14,060 14 C yr BP. The original age of theCalle Santa Rosa organic bed (14,820±230 14 C yrBP, I-5033; Mercer 1972) is compatible with thenew dates. The lower of the two 1-cm-thick peatlayers at the railroad bridge site is now placed at14,550 14 C yr BP (Table 2). Near-basal wood in theorganic silt and sand at the railroad bridge site datesto about 17,800 14 C yr BP (Table 1) rather than16,270 14 C yr BP (Mercer 1976). Thus the tops of theorganic beds within the Puerto Varas embaymentnow yield nearly identical dates. However, basal organicsediments at the inland railroad bridge site are190GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTabout 2000 14 C yr older than basal organic sedimentsat the lakeside Bella Vista Bluff site. In turn,the base of the organic layer at the Llanquihue siteis the same age as at Bella Vista Bluff.The Lago Llanquihue piedmont glacier stoodagainst the kame terrace during deposition of sedimentsoverlying the organic deposits. Depositionof glacioßuvial sediments on the organic bed in thekame terrace near the town of Llanquihue (site 23)requires an adjacent buttressing ice margin. Depositionof the debris-ßow sediments of volcanic materialon the terrace also requires that ice stood atthe terrace edge between the R’o Maull’n outlet andPunta Cabras, so that ßows of volcanic sedimentwere directed along the terrace surface from R’oPescado to R’o Maull’n. That deposition of the glacioßuvialsediments and the volcanic ßow sedimentsin the terrace near the town of Llanquihuewas nearly simultaneous is shown by dates at thetop of the organic beds at the Llanquihue and PuertoVaras stratigraphic sites.The position of glaciolacustrine sediments onthe top of the Bella Vista Bluff organic unit is bestexplained by a buttressing ice lobe that dammed alocal, ice-marginal lake in the Puerto Varas embayment.This conclusion is supported by the similarityin ages of the top of the Bella Vista Bluff and Llanquihueorganic beds. This local ice-dammed lakedid not ßood the railroad bridge site, where volcanicßow breccias afford the only evidence of a buttressingice margin younger than the organic beds.At the Northwest Bluff site the position of thepoorly sorted gravel below the organic bed probablyrequires an adjacent ice margin. The glaciolacustrineunits beneath the organic beds at the railroadbridge and Bella Vista Bluff sites both probablyreßect an ice lobe in the Puerto Varas embayment.The simplest explanation is that theglaciolacustrine units beneath organic beds at therailroad bridge and Bella Vista Bluff sites are thesame age and hence reßect the same adjacent icemargin. But there are two serious problems withthis explanation. First, numerous dropstones arepresent in the glaciolacustrine units below the organicbeds at the railroad bridge but not at the BellaVista Bluff site. Also, the close minimum ages forthe glaciolacustrine units are substantially different,about 17,800 14 C yr BP at the railroad bridgesite and 15,730 14 C yr BP at Bella Vista Bluff. Thusit is more likely that the two glaciolacustrine unitsrepresent separate events. The glaciolacustrine unitat the railroad bridge site probably represents a buttressingice margin deep in the Puerto Varas embaymentat shortly before 17,800 14 C yr BP. Thelower glaciolacustrine unit at Bella Vista Bluffprobably represents an adjacent ice margin shortlybefore 15,730 14 C yr BP. Glaciolacustrine sedimentsdeposited in the youngest of these ice-marginallakes would make up the core of what is nowthe eastern part of the overall Puerto Varas terrace,but would not have been deposited on the organicbeds at the railroad bridge site.This interpretation of the terrace sequence atPuerto Varas is in accord with our reconstruction oflake-level ßuctuations. The rim of indurated waterlaintill exposed along the western shore of LagoLlanquihue now controls the R’o Maull’n outlet.We postulate that this outlet would always quicklyFig. 22. Stratigraphic section at the railroad bridge site (26), Puerto Varas. Location given in Fig. 15. See legend in Fig. 7 (p. 177) forlithologic details. The upper panel gives the position of hand-dug and road-cut sections at the railroad bridge site. The lower panel givesour interpretation of the stratigraphic data presented in (a); the location of the road surface, railroad abutment and road abutment areall indicated on the diagrams. Top panel shows Unit A, the lowermost unit exposed at the railroad bridge site, west of the railroad abutment.Unit A is a glaciolacustrine deposit formed of silt and clay with numerous dropstones. This glaciolacustrine deposit is unconformablyoverlain by coarse-grained gravel (unit B) with individual crystalline clasts (granite) up to 50 cm in diameter. At the easternend of the railroad bridge site, the gravel is, in turn, overlain unconformably by medium-grained and coarse-grained debris-flow brecciaunits (I 1 and I) that show fining-upward sequences. Just east of the retaining wall, the gravel is overlain by massive organic silts (unitF). The organic silts form well-defined diapirs and flame structures that penetrate the overlying debris-flow sediments. Bottom panelgives the location and stratigraphy of two hand-dug sections between the railroad and road abutments. Coarse-grained debris-flow sedimentsagain form the uppermost unit (unit I). The western hand-dug section has two thin peat layers separated by fine- to mediumgrainedvolcanic sands (unit H) from a lower, prominent organic bed. The organic layer (unit G) in this western section is as much as26 cm thick, and in the upper 7 cm contains numerous pieces of wood 5Ð8 cm in length. The inset on the left of the upper panel showsthe detailed stratigraphy revealed in the eastern hand-dug reexcavation behind the retaining wall. This re-excavation shows the followingstratigraphy. Unit I is composed of debris-flow volcanic sediments. Unit H and the upper two thin organic beds (both of which occurin the excavation 1.5 m to the east) are not exposed at this site (emplacement of the overlying debris-flow sediments probably removedthese deposits). Unit G is a thick (up to 26 cm) organic bed with numerous wood fragments. Unit F is organic silt (120 cm thick) withtwo diffuse horizons of increased organic content. Wood fragments, many in growth position, occur through this unit. Unit E is finegrainedorganic sand and silt with large fragments of wood. Unit D is an admixture of fine- and medium-grained organic sand, withabundant wood fragments, many in growth position. Unit C is gray, coarse-grained sand with small pebbles.GeograÞska Annaler á 81 A (1999) á 2191


G.H. DENTON ET AL.Fig. 13. Stratigraphic sectionthrough a moraine ridge located akilometer south of site 33 on westside of Route 5. Location given inFig.15. The photograph shows amoraine ridge resting on an intactoutwash unit. The two units havean erosional contact. Weatheredpyroclastic deposits rest on the morainesurface. Nearby morainesoverlie the same outwash unit,without any glacial deposits on theoutwash in areas between the moraines.Fig. 19. Stratigraphic section at site 23 near the town of Llanquihue.See Fig. 18 for details. The organic beds are near the bottomof the photograph. Overlying the organic bed are the glaciofluvialsediments that make up the kame terrace at this locality. Underlyingthe organic bed is waterlain till that crops out along the shoreof Lago Llanquihue.Fig. 21. Overview of the Bella Vista Bluff section (site 30 in Fig.15). The organic bed is in the center of the photograph at the topof the pickax handle. See Fig. 20 for description of units.192GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTabFig. 24. Photographs of interdriftpyroclastic flow sediments at thecrossroad section B at site 25 inFig. 15. The location of these pyroclasticflow sediments, as well asthe character of the drift unitsabove and below, is shown in Fig.23b. (a) Taken from the center ofthe section, looking south. (b) Takenfrom the same place, lookingnorth.cut through terrace sediments and moraines, bothcomposed largely of unconsolidated glacioßuvialgravels or gravelly sediment ßows, and stabilize atthe level of the indurated compact waterlain till.The highest outcrop of waterlain till yet located isat 53 m elevation (about 2 m above the present-daylake level) at the Llanquihue site 23, about 2 kmsouth of the R’o Maull’n outlet. Thus we think thatice advance over the eastern R’o PetrohuŽ outletwould not by itself cause a lake-level rise, becauseindurated waterlain till would still control the R’oMaull’n outlet. If this is correct, then higher-thanpresentlake levels recorded in the Puerto Varas terracesections represent local glacial lakes impoundedby an adjacent piedmont ice lobe. Conversely,a fall in lake level signiÞes recession of apiedmont lobe from the Puerto Varas embayment.Thus we infer that lake-level lowerings leading todeposition of organic silt and sand at the railroadbridge site by 17,800 14 C yr BP and at the Bella VistaBluff site by 15,730 14 C yr BP (A-9190) werecaused by ice recession from the Puerto Varas embayment.Flooding of the organic bed at Bella VistaBluff at 14,540 14 C yr BP (Table 2) was caused bya readvance of the piedmont glacier to the proximaledge of the kame terrace.An important remaining question concerns theextent of ice recession when the organic beds weredeposited in the Puerto Varas embayment. Mercer(1976) suggested deglaciation of the eastern R’oPetrohuŽ outlet. Overßow through R’o PetrohuŽalso would have required glacier recession fromGeograÞska Annaler á 81 A (1999) á 2193


G.H. DENTON ET AL.both Seno Reloncav’ and Estero Reloncav’. By ourreconstruction, however, a lowered lake levelmerely requires that the ice margin pull back fromthe western lake shore. Near the town of Llanquihue(site 23), the base of the organic-rich silt bed isonly 3 m above present-day lake level. Hence lakelevel dropped to or below 54.0 m elevation duringdeposition of the organic beds at both the PuertoVaras and Llanquihue sites, because of their nearlyidentical ages. Therefore, it is just possible that theLlanquihue (and hence the Bella Vista Bluff) organicbed formed while Lago Llanquihue drainedthrough the R’o Maull’n outlet. But it is equallypossible that Lago Llanquihue drained through theR’o PetrohuŽ outlet when the organic beds were depositedat Llanquihue and in the Puerto Varas embayment.We merely point out that such easterlydrainage is not a necessity.Llanquihue moraine belt. The Llanquihue morainesbetween the lakeside kame terraces and themain Llanquihue outwash plain farther west havebracketing dates from several key localities. Minimumlimiting ages come from organic material onor within the moraines. New basal dates from miresare 20,160±180 14 C yr BP (AA-9296) for Canal dela Puntilla (site 6), along with 20,580±170 14 C yrBP (AA-9303) and 20,380±170 14 C yr BP (AA-9298) for Canal de Chanch‡n (site 5). The basal agefor Canal de la Puntilla is the end member of a consistentsequence of dates through the thickness ofthe mire (Moreno et al. 1999). These basal ages areminimum values for this entire sector of the Llanquihuemoraine belt, which is traversed by the twomeltwater channels. Moreover, the channels havenot been scoured by meltwater or lake spilloversince organic deposition began on the channelßoors implying that readvances of the Lago Llanquihuepiedmont glacier have not been extensiveenough to dam a lake that spilled over into thechannels, let alone reach the bayside ice-contactslope. Such a situation is consistent with that atPuerto Varas, where the railroad bridge site has notbeen overrun by ice since at least 17,800 14 C yr BP.Similar basal ages come from sediment cores indepressions elsewhere on the Llanquihue morainebelt. At Fundo Li–a Pantanosa (site 8), situated justwithin the two outermost Llanquihue moraines, a3.50-m-deep mire yielded near-basal ages of19,768±397 14 C yr BP (AA-14774) and 19,993±257 14 C yr BP (AA-15900) (Table 1) (Heusser et al.1999). A 7.07-m-deep mire just within the outermostLlanquihue moraine at Fundo Llanquihue(site 18) afforded ages within the basal meter of20,645±220 14 C yr BP (UGA-6907), 20,890±18514 C yr BP (UGA-6908), 20,455±180 14 C yr BP(UGA-6909), 20,650±175 14 C yr BP (UGA-6910),20,680±175 (UGA-6912), and 20,585±170 14 C yrBP (UGA-6913) (Table 1) (Heusser et al. 1999). A+250basal date of 19,760 -24014 C yr BP (A-8534R)comes from a 2.7-m-thick mire in a small depressionon the proximal side of the two outer Llanquihuemoraines near Frutillar Alto (site 14) (Table 1).Maximum limiting ages for the Llanquihue morainebelt are derived from organic clasts reworkedinto outwash of the main Llanquihue plain. Theyoungest ages are 23,020±280 14 C yr BP (QL-4539) from north of Bah’a Mu–oz Gamero (site 3);20,840±400 14 C yr BP (QL-4527) from near Canal+180de la Puntilla (site 4); 22,70014 C yr BP (A-9514), 22,790±410 14 -175C yr BP (A-6197), 21,120±180 14 C yr BP (AA-25374) and 21,840±700 14 C yrBP (QL-4548) from near Frutillar Alto (site 11); and22,250±220 14 C yr BP (Wk-2536) and 22,985±23514 C yr BP (TUa-470A) from southwest of PuertoVaras (site 32). As mentioned above, Mercer(1972) reported an age of 20,100±500 14 C yr BP(RL-116) for a peat clast reworked into Llanquihueoutwash at Frutillar Alto (site 13) about 1.5 kmsouth of site 11 where we obtained many samples.Also, a clast dated to 22,870±310 14 C yr BP (A-6196) is reworked into ice-contact drift south ofBah’a Frutillar at site 17.The map pattern of Llanquihue moraines, alongwith the geometric relations of Canal de Chanch‡nand Canal de la Puntilla to these moraines, are keyto interpretation of these dates. The minimum agesof 20,580±170 14 C yr BP (AA-9303) for Canal deChanch‡n (site 5) and of 20,160±180 14 C yr BP(AA-9296) for Canal de la Puntilla (site 6) apply toall of the Llanquihue moraines distal to the lakesideterraces that fringe Bah’a Mu–oz Gamero. Theyare consistent with ages for basal organic materialat Fundo Llanquihue of as much as 20,890±18514 C yr BP (UGA-6908). Hence the outer Llanquihuemoraines at these sites antedate 20,580Ð20,89014 C yr BP. From moraine geometry it is difÞcult toargue that the outermost Llanquihue moraine atFrutillar Alto is younger than these ages. Indeed,Fig. 3 implies that the Frutillar Alto frontal moraineis older, not younger, than the frontal Llanquihuemoraine cut by Canal de Chanch‡n and Canal de laPuntilla. Also, the outermost Llanquihue morainenear Frutillar Alto is probably equivalent to the outermostmoraine near the Fundo Llanquihue site.The overall implication is that the maximum age of194GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFT20,100±500 14 C yr BP (RL-116) for the outermostLlanquihue moraine at Frutillar Alto (Mercer1972) is too young by more than 790 14 C yr. A furtherimplication is that the age of 20,840±400 14 Cyr BP (QL-4527) for an organic clast in outwash(site 4) near Canal de la Puntilla may be too young,because basal dates from the Puntilla and Chanch‡nchannels suggest that this outwash antedates20,120Ð20,580 14 C yr BP. There is strong paleoenvironmentaland chronologic evidence from Canalde la Puntilla (Moreno et al. 1999) and Canal deChanch‡n (Table 1) that the Lago Llanquihue icelobe did not advance into this moraine system after20,580 14 C yr BP. Hence, most new chronologicdata show that the entire Llanquihue moraine beltwest of Lago Llanquihue has a minimum age of20,580Ð20,890 14 C yr BP and a maximum age of23,020Ð22,250 14 C yr BP (see below for alternateinterpretation).The revised chronology indicates that the Llanquihuemoraine belt west of Lago Llanquihue cannotbe subdivided into two units that relate to twoepisodes of glacier expansion. The outermost moraineswere formerly called Llanquihue I and takento antedate the LGM (Porter 1981). Likewise, mostof the main Llanquihue outwash terrace graded tothese moraines was also thought to antedate theLGM (Mercer 1976; Porter 1981). Moraines ofLlanquihue II age, with a mapped outer limit wellwithin the Llanquihue moraine belt except nearFrutillar Alto, were believed to represent the LGM,and were placed between 18,900 and 20,100 14 C yrBP (Mercer 1976; Porter 1981). Part of the reasonfor this original subdivision involved old dates ofintertill wood samples from exposures within themoraine belt west of Puerto Varas (Mercer 1976),which we now discuss.The morainal landforms near the crossroad site(25) at the intersection of Routes 5 and V-50 westof Puerto Varas reßect old drift overrun by ice thatdeposited the outermost Llanquihue drift fartherwest at the LGM. The section beside Route V-50near the southwest corner of the crossroad site wasoriginally described by Mercer (1972). Calledcrossroad south in Fig. 23, this section shows intensemechanical deformation of the upper glacialunit that makes up the moraine surface. The crossroadnorth section in Fig. 23 also shows complexshear deformation of the uppermost glacial unit, aswell as of the weathered interdrift pyroclastic ßowunits (Fig. 24).At the crossroad south section in Fig. 23, woodhas been sheared off the surface of the interdrift pyroclasticßow sediments into the overlying glacialunit. The location of dated wood and charcoal samplesis shown in Fig. 23. Because they fall at theouter age limit of reliable radiocarbon dating, theseages are all regarded as minimum values. Fromwood and peat in the interdrift pyroclastic ßow depositsof the crossroad south section, Mercer(1976) previously reported dates of > 39,900 14 C yrBP (I-4170) and > 39,900 14 C yr BP (I-5032) and+2300Porter (1981) gave a date of 57,800 -320014 C yr BP(QL-1336). Again, these ages are minimum values.The mechanical deformation of drift above thepyroclastic ßow sediments could have been impartedduring glacier overriding of the crossroadsections. Hence, the stratigraphic sequences at thecrossroad site cannot be used to assign an age to theouter Llanquihue moraines farther west. Instead,the pertinent dates come from a stratigraphic exposure(site 24) on the north side of Route V-50 whereit passes through the outer Llanquihue moraines500 m west of the crossroad site. The exposure revealsorganic pyroclastic ßow sediments (remobilizedby glacier overriding) that separate a lower,little-weathered outwash (>30 m thick) from upperthin till with morainal topography of moderate relief.Dates of 25,020±290 14 C yr BP (UGA-6939),26,150±220 14 +325C yr BP (UGA-6821), 29,350 14 Cyr BP (A-7712), and 29,36014 -310+215-210 C yr BP (A-7715)pertain to the remobilized organic pyroclastic sedimentsbeneath the till. The outermost Llanquihuemoraine adjacent to the Fundo Llanquihue site 18,with its minimum dates of 20,680Ð20,890 14 C yrBP, can be traced directly to the outermost Llanquihuemoraine both west and south of Puerto Varas(Fig. 3). In turn, outwash graded to this same morainesouth of Puerto Varas is younger than 22,250±220 14 C yr BP (Wk-2536) (site 32), as describedabove. Hence, we conclude that the outermostLlanquihue moraine was deposited during theLGM, despite the presence of only older depositsat the crossroad site.An organic silt clast reworked into the outwashbeneath the remobilized pyroclastic sediments atsite 24 yields a date of >44,400 14 C yr BP (UGA-6938). The outwash has a ßattish terrace surfaceFig. 23, see pp. 197Ð198.Fig. 24, see p. 193.GeograÞska Annaler á 81 A (1999) á 2195


LLANQUIHUE DRIFTFig. 23. Stratigraphic sections at the crossroadsite (25) at the intersection of Routes 5 and V-50 in the Llanquihue moraine belt west of PuertoVaras. See legend in Fig. 7 (p. 177) for lithologicdetails. Panel a gives location of individual sections.Sections A and B are selected for detaileddescriptions because of new road constructionin 1996/97. Panel b shows section A along thewest side of Route 5 north of the intersection.This exposure shows highly deformed Llanquihuedrift that rests conformably on older driftand gravel units (southern half of section). MSPis a master shear plane with well-developedslickensides at the base of the Llanquihue sedimentcomplex. Insets (a), (b), and (c) give detailsof the degree of syndepositional and postdepositionalshear deformation. Panel c showssection B along Route V-50, west of the intersectionwith Route 5. Here highly deformedLlanquihue drift is stacked on beds of weatheredpyroclastic flow sediments. The base of the outcrop(eastern and central part of section) showsolder drift and gravel. Inset (a) of Panel c showsdetails of weathered pyroclastic flows, withthree distinct depositional units resting on induratedolder drift. The top unit has abundant woodand charcoal fragments. Inset (b) shows an exampleof mechanical stacking of weathered pyroclasticflows at the base of Llanquihue drift,with a complex cross-cutting pattern of shearplanes. Panel d displays a schematic compositelithostratigraphic section for the crossroad site.Included are the composite lithostratigraphywith average thickness, which varies considerablyfrom site to site, radiocarbon dates on woodand charcoal (Table 1), lithologic descriptions,and stratigraphic interpretations.GeograÞska Annaler á 81 A (1999) á 2197


G.H. DENTON ET AL.Fig. 23. Continued.198GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTthat, by projection, forms the main Llanquihueplain. The terrace surface can be traced eastwardfor 300 m beneath the late Llanquihue morainal topographyof moderate relief, where it grades to alarge buried moraine ridge. Hence the ice front associatedwith the terrace was situated only a fewhundred meters behind the position reached by thelate Llanquihue piedmont lobe. The outwash of theterrace is far less weathered than any deposits exposedin the crossroad sections below the pyroclasticßow sediments and lying beyond the range of radiocarbondating. Hence, it is probably of middleor early Llanquihue age.Additional evidence for middle or early Llanquihueglacial drift comes from two important exposuresin the Llanquihue moraine belt at the top ofthe ice-contact slope above the lakeside kame terraces.One is alongside Bah’a Frutillar Bajo (site16) and the other is alongside Bah’a Mu–oz Gameronear Puerto Octay (site 7).Figure 25 depicts the Bah’a Frutillar Bajo section(site 16), where a lens of organic pyroclasticßow sediments separates two till units (Fig. 26),both of which are compact, little weathered, medium-grayin color, and characterized by boudinagestructures and numerous shear planes. Both aretaken to be lodgement tills. Where they are not separatedby organic pyroclastic sediments, the twotills could be mistaken for a single unit. The intertillorganic pyroclastic sediments are as much as 95 cmthick and have an exposed lateral extent back intothe face of about 15 m. A shear plane constitutes thecontact between the organic bed and the upper tillfor 7.5 m of lateral extent on the lakeside portion ofthe exposed section. In the inland portion of the exposure,a wedge-shaped unit of silt and gravellydiamicton with crude layering separates the organicbed from the overlying till. The basal shear planeof the upper till passes inland from the top of the organicbed to the top of this diamicton wedge. Beneaththe shear plane, the top of the lakeside end ofthe organic bed has been removed. But beneath theinland diamicton wedge, the vegetation litter of theold land surface of the organic bed is preserved intact.We attribute the sealing diamicton to sedimentßows off an advancing ice margin, and the subsequentshearing to subglacially imparted stressesbeneath the glacier ice that deposited the upper till.Hence dates of the sealed land surface pinpoint anadvance of the Lago Llanquihue piedmont lobeacross the lakeside ice-contact slope at the inneredge of the Llanquihue-age moraine belt. Seventeensuch dates of organic pyroclastic sediment,wood, and Þbrous organic matter from different localitiesalong the exposed sealed land surfaceyielded a weighted mean of 26,797±65 14 C yr BP(Tables 1 and 2).Dates from the base of the Frutillar Bajo organicbed at three localities are 34,765±840 14 C yr BP(UGA-6945), 34,985±440 14 C yr BP (UGA-6919),and 36,960±550 14 C yr BP (UGA-6724) (Table 1,Fig. 25). Other dates of the organic bed are givenin Table 1. Thus the ice-free interval separatingdeposition of the two till units lasted at least 10,00014 C yr, from prior to 36,960 14 C yr BP (Table 1) to26,797 14 C yr BP (Table 2). The lack of signiÞcantweathering of the till below the organic bed suggestsa Llanquihue age. Such a conclusion is in accordwith the pollen record from the organic bed ofcontinued Subantarctic Parkland cold and wet conditions(Heusser et al. 1999). There is not any suggestionof an interglacial evergreen forest or a lowlanddeciduous forest. The implication of this pollenrecord is that the till below the organic silt bedwas deposited subsequent to the last interglaciationand hence is middle or early Llanquihue in age.It is worth pointing out that an exposure 200 mto the east, but in the same topographic position onthe ice-contact slope as the Frutillar Bajo site, revealscoarse stratiÞed drift with an enclosed organicclast dated to 22,870±310 14 C yr BP (A-6196)(site 17). Evidently this same ice-contact slope wasoccupied at least twice, once during the advance at26,797 14 C yr BP and again during the young advancethat reached the outer Llanquihue moraineswest of Frutillar Bajo.A road cut through the ice-contact slope abovethe lakeside terraces west of Bah’a Mu–oz Gameroat the Puerto Octay site (7) exposes two superimposedoutwash units separated by a 90-cm-thicklayer of organic pyroclastic ßow sediment (Fig.27). Dates of the uppermost organic sediment fromÞve separate localities along the lateral extent (150m) of the organic bed (Table 1) give a weightedmean of 29,363±178 14 C yr BP (Table 2). Both outwashunits head at the ice-contact slope, wherepoorly sorted stratiÞed drift with enclosed bouldersis exposed. Thus both were deposited in an iceproximalposition when a former piedmont glacierstood at the top of the slope, which is within 2.2 kmFigs 26Ð27, see p. 196.GeograÞska Annaler á 81 A (1999) á 2199


G.H. DENTON ET AL.Fig. 25. Schematic composite lithostratigraphy of the Frutillar Bajo section at site 16 in Fig. 15. See legend in Fig. 7 (p. 177) for lithologicdetails. Unit A is gravelly sand to sandy gravel, with well-rounded components, a polymictic gravel fraction, and abundant volcaniclasticsediment. The upper 5 m of unit A exhibit lenses of volcaniclastic clayey silt in a coarsening-upward sequence, which is overlain bymassive basal till. Unit B is matrix-supported sandy-silty gravel, with some boulders and with mechanically emplaced clasts of beddedsilty sand; the unit is sheared and is probably a lodgement till. Unit C is a matrix-supported, sandy-silty, bouldery gravel that is highlycemented, with yellowish-whitish fissility infillings; it is probably a basal till. Unit D is a bedded sandy gravel, interpreted to be glaciofluvialoutwash. ISB consists of non-glacial beds; the lower unit is massive clayey silt with charcoal and with a basal orange-browngravel layer composed dominantly of pyroclastic flow sediments. The upper unit is massive weathered sandy silt with charcoal, interpretedto be a tuffaceous pyroclastic flow sediment. Top (inset) shows a preserved grass layer covered by glacial silt (inset shows locationof radiocarbon samples in Table 1). The lower part of unit E is matrix-supported sandy-silty gravel, with subrounded clasts, and faintbedding, interpreted to be till flows. The upper part of unit E is matrix-supported silty-sandy gravel with some boulders and mechanicallyemplaced clasts of silty-sandy rhythmites; this upper part is lightly sheared and probably represents lodgement till. TS is the topsoil,which is largely composed of pyroclastic flow sediments. Units AÐD are interpreted as sediments of a single glacier advance, with unitB and unit C being formed during oscillations of the ice margin. ISB represents a non-glacial interval. Unit E represents a glacier advanceduring which sediment flows from an ice margin sealed intact part of the top of unit B.200GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTof the outer Llanquihue moraine belt. The uppersurface of the organic bed between the outwashunits lacks a preserved land surface and hence maynot be original. In fact, in a locality not sampled atthe head of the ice-contact slope, the upper surfaceof the organic bed has been sheared by overridingice. However, the consistency of the dates suggests,but does not prove, that little material has beeneroded off the top of the organic bed. The argumentfor this inference is that the ages given in Table 1increase markedly with depth within the organicbed, and hence that the upper surface would haveyielded highly disparate ages if even a few centimetershad been eroded. Therefore, we suggest thatthe age of about 29,363 14 C yr BP probably affordsa close limiting value for deposition of the upperoutwash and hence for glacier advance to the top ofthe ice-contact slope. However, given the lack of apreserved land surface, we cannot be sure that theadvance was not younger than 29,363 14 C yr BP.Four basal samples from the inter-outwash organicpyroclastic bed at differing localities alongthe exposure yield limiting minimum dates for thelower outwash unit of 33,900Ð39,340 14 C yr BP (Table1). Overall, the Puerto Octay exposure revealsan organic bed composed largely of pyroclasticßow sediments that range from ³ 39,340 to 29,36314 C yr BP in age and that separate two ice-proximaloutwash units. The lower outwash unit shows littleweathering and, in the absence of the interveningorganic bed, could not have been distinguishedfrom the upper outwash unit. The pollen recordfrom the organic bed shows the continuous presenceof a Subantarctic Parkland, with cold and wetconditions, similar to the situation at the FrutillarBajo site (Heusser et al. 1999). Notably missing isan indication of the numerous thermophilic treespecies of an interglacial evergreen forest or deciduouslowland forest. Just as at the Frutillar Bajosite, the implication is that the lower outwash unitwas deposited subsequent to the penultimate interglaciationand is therefore probably of middle orearly Llanquihue age.Seno Reloncav’ RegionAn extensive system of Llanquihue morainesfringes Seno Reloncav’ (Fig. 4). The main Llanquihueoutwash plain is graded to the outermost ofthese moraines. Subsidiary Llanquihue outwashplains are nested within the outer moraine systemand are graded to inner moraine belts. A high icecontactslope occurs alongside Bah’a PuertoMontt, where prominent kame terraces, togetherwith small moraine fragments, occur on the proximalside of this ice-contact slope. We recognizeseveral glacier advances into these moraine beltsand the gulfside ice-contact slope. We now discussthe evidence for these advances, starting with theyoungest.Dates at three sites document a glacier advancethat culminated shortly after 15,000 14 C yr BP. AtPunta Penas on the north shore of Seno Reloncav’(site 47), lodgement till unconformably overlies aglaciolacustrine unit (Fig. 28). Any lake in this positionwould have been proglacial. The thickness ofthe glaciolacustrine unit is variable, reßecting theamount of lacustrine material removed by overridingice. Near the base of the glaciolacustrine unit isa 20-cm-thick organic layer about 7.2 m abovehigh-tide mark. Dates of this organic bed are15,940±315 14 C yr BP (T-10296A) at the top,16,275±440 14 C yr BP (T-10297A) in the middle,and 16,000±275 14 C yr BP (T-10298A) at the base.This 20-cm-thick organic layer is overlain Þrst bylacustrine sediments with several thin organic layers,and then by glaciolacustrine sediments withoutorganic layers. Two of the thin organic layers areparticularly prominent, with organic macrofossilsand wood on the upper surfaces. The dates of surfaceorganic matter from the lowest of these twothin layers are 15,675±90 14 C yr BP (A-9019),+11015,665 14 C yr BP (A-9020), and 15,200±12014 -105C yr BP (AA-21009) (Table 1), giving a weightedmean of 15,554±60 14 C yr BP (Table 2). The datesof surface organic litter from the highest of thesetwo thin beds are 15,020±120 14 C yr BP (AA-21010), 15,000±75 14 C yr BP (A-9017), 14,735+80-7514 C yr BP (A-9018), and 14,835±80 14 C yr BP(A-9506) (Table 1), giving a weighted mean of14,879±42 14 C yr BP (Table 2).These organic beds point to a time of low lakelevel that requires the former piedmont ice lobe tobe pulled back from the western shore of the SenoReloncav’ embayment enough to allow drainagethrough Canal de Chacao at a time when sea levelwas more than 100 m below the present-day value(Fairbanks 1989). Subsequent ßooding of the peatbeds by a proglacial lake, accompanied by rapidaccumulation of glaciolacustrine sediments beginningabout 14,879 14 C yr BP, reßects an advance ofthe Seno Reloncav’ piedmont lobe that culminatedinland of the Punta Penas site when the surface tillwas deposited.A small, subsidiary Llanquihue outwash plaingraded to the ice-contact slope west of Bah’a Puer-GeograÞska Annaler á 81 A (1999) á 2201


G.H. DENTON ET AL.to Montt (sites 44, 45) yielded reworked organicclasts with dates of 16,900±120 14 C yr BP (A-6491), 15,640±100 14 C yr BP (A-6492), 16,060±120 14 C yr BP (UGA-6942), and 15,040±100 14 C yrBP (A-6493) (Table 1). Each date affords a maximumlimiting value for advance of the Seno Reloncav’piedmont ice lobe to the upper edge of theice-contact slope at this locality.At site 73 on Isla Maillen in Seno Reloncav’, acompact peat mat is sealed by laminated glaciolacustrinesediments that are overlain by lodgementtill. The date of 15,220±80 14 C yr BP (A-9507) (Table1) for the top of the peat registers an advance ofthe Seno Reloncav’ piedmont glacier over IslaMaillen.Farther south near Calbuco (site 94), lodgementtill 1.5 m thick covers a delta built into a proglaciallake within the limits of a Llanquihue moraine beltabout 6 km west of Seno Reloncav’. Organic clastsreworked into the foreset beds of the delta yielded+150dates of 15,285 14 C yr BP (A-7702), 15,500±8514 -145C yr BP (A-7698), 15,535±85 14 C yr BP (A-7697),16,160±80 14 C yr BP (A-8776), 15,730±70 14 C yrBP (A-8777), 15,365±85 14 C yr BP (A-8778),+8515,390 -8014 C yr BP (A-8779), 15,110±100 14 C yrBP (A-9427), 15,130±125 14 C yr BP (A-9425),15,250±175 14 C yr BP (AA-25373), and 14,900+160-15514 C yr BP (A-9426) (Table 1). These dates affordmaximum limiting values for the advance ofthe Golfo de Ancud piedmont ice lobe registered bythe lodgement till on the delta.The data from sites 44, 45, 47, and 94 thus documenta major advance that culminated shortly after15,000 14 C yr BP. Fig. 4 shows that the northernthree sites reßect advance of the Seno Reloncav’lobe only to the ice-contact slope near Bah’a PuertoMontt; in one place (sites 44, 45) the lobe reachedthe top of this slope, allowing meltwater to deposita small tongue of outwash within the Llanquihuemoraine system. Farther south, however, the cor-202GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTrelative advance of the Golfo de Ancud lobe passedinland over the Calbuco site 95. To conÞrm the realityof this difference, we undertook a coring programto determine how far the last advance extendedinto the Llanquihue moraines west of Seno Reloncav’.We noted earlier that the complicated morainebelts fringing Seno Reloncav’ could besubdivided into two, and perhaps three, groups onthe basis of morphologic relationships. The resultingbasal ages of mires in depressions on theseLlanquihue moraine belts are listed in Table 1.Each of these dates affords a minimum age for theunderlying morainal topography. It is not possibleto determine how close these minimum ages are tothe time of deposition of the moraine, unless closelybracketing maximum ages are also available.There are now 21 basal dates from mires on theinner and intermediate moraine belts fringing SenoReloncav’ (Table 1). Eight dates fall between13,000 and 14,000 14 C yr BP, three fall between14,000 and 15,000 14 C yr BP, and one falls between15,000 and 16,000 14 C yr BP. The key site is theHuelmo mire (site 81), on the proximal side of acontinuous moraine belt that trends northÐsouthjust inland of the ice-contact slope beside Seno Reloncav’(Moreno 1998). The basal date for theHuelmo mire is 16,410±110 14 C yr BP (AA-23244), a value consistent with dates of samplesfrom higher in the core. The implication is that thereadvance of the Seno Reloncav’ piedmont lobeabout 14,900 14 C yr BP did not reach these moraines,but must have terminated instead on the icecontactslope alongside Seno Reloncav’. This is entirelyconsistent with the situation farther northnear Bah’a Puerto Montt where the piedmont icelobe barely reached the top of this ice-contactslope. Thus the relatively limited extent of the SenoReloncav’ piedmont lobe during this advance issimilar in magnitude to the extent of the Lago Llanquihuelobe.To check the conclusion from the coring programthat the inner prominent moraine belt besideSeno Reloncav’ antedated the advance at shortlyafter 15,000 14 C yr BP, we also dated organic clastsreworked into the associated subsidiary outwashterrace (Table 1). With one exception, all the datesare 25,300±300 14 C yr BP (Wk-2541) (site 71) orolder. The single exception is a reworked organicclast dated to 19,820±185 14 C yr BP (TUa-468A)(site 53). These dates show that the inner moraineFig. 28. Composite stratigraphic section at Punta Penas at site 47 in Fig. 16. See legend in Fig. 7 (p. 177) for lithologic details. Thesite consists of two exposures; the first is a sea cliff in the tidal zone and the second is a road cut just inland and northwest of the seacliff. Datum for this site is the high-tide level. There are five basic stratigraphic units at Punta Penas. Unit A includes interbedded horizontalsand, ash, and gravel layers exposed within the present tidal zone. There are some reworked organic clasts. Unit A also containsorganic silt beds with wood and large tree remains rooted in place. Porter (1981) reported an age of 42,400±500 (QL-1337) from anin situ stump. The beds of unit A have been eroded and infilled several times, with channels cutting the organic beds and, in turn, infilledwith partly consolidated pebble gravels. The horizontally laminated stratified deposits of unit A represent former littoral zones, withaccompanying changes in base level resulting from submergence during tectonic events. Unit B is outwash made up of horizontallybedded gravel with sand layers. The maximum thickness of this outwash unit is 2.7 m. Locally this gravel is matrix-supported and beddingis not well defined. All enclosed clasts of pebbles and cobbles are well rounded. There are rare boulders up to 30 cm in diameter,as well as a reworked inclusion of organic silt. The most probable source for this inclusion is an organic bed in unit A, suggesting thaterosion and reworking occurred subsequent to deposition of unit A. The lower contact of the outwash unit was not observed. Unit Cis basal till with subglacial thrust planes. Unit C makes up most of the exposure and forms the core of the sea cliff. There are four lithologicsubunits. At the top, subunit C4 is a diamicton with a silty matrix with a clast concentration of less than 10%. Fracture lines arecommon in this subunit. Observed thickness is 2.8 m, but vegetation and reworking make this a minimum value. A pervasive shearplane marks the contact with subunit C3, which has a granular matrix and a clast concentration of over 20%. An irregular contact separatesthe diamicton of unit C3 from the underlying laminated silt unit C2. The silt laminations are 3Ð8 mm thick and are disturbed.There are high-angle faults through the thickness of the silt bed. A 2-m-long clastic dike penetrates the silt near its upper contact withthe overlying diamicton. Clastic dikes also occur in several other parts of the section. The silt bed has a thickness of 1 m, and is not inan original depositional orientation. It appears to have been a distal glaciolacustrine package prior to translocation. Subunit C1, the lowestpart of unit C, consists of translocated gravels approximately 1 m thick and rising 13¡ toward the northwest. Beds within this gravelunit have a vertical offset of at least 1 m. At the base of this unit is a sheared zone containing numerous angular clasts; at several locationsthe individual fragments of the clasts are adjacent to each other, indicating that the clasts were broken and transported less than 2 cm.This sedimentary package is interpreted as subglacial thrust slices at the base, with reworking of these sediments into subglacial till atthe top. Unit D is a proglacial lacustrine sequence, best exposed in a road cut 38 m northwest of the limit of the sea cliff. The base lies3.6 m above datum. Sediments in this sequence extend to at least 10 m above the datum, where they lie below the top of a terrace. Thebasal 2 m are composed of finely laminated silt with a few organic beds. At 5.2 m above the datum is a 20-cm-thick organic bed withlocalized disruptions of the bedding. Above this 20-cm-thick organic bed, there are individual thin (~1 cm thick) organic beds, of whichtwo are prominent. The dates reported in Table 1 come from these three organic horizons, as shown on the diagram. Bedded coarsesand lies 2Ð3 m above the 20-cm-thick organic silt bed. Here the entire sequence coarsens upward, with a small gravel unit near theupper contact. The thickness of unit D is variable because it is cut by unit E, a basal till with a silty matrix. Clast concentrations in thetill range from 10 to 20%. Fracture planes are parallel to the base of the till and reflect subglacial deposition.GeograÞska Annaler á 81 A (1999) á 2203


G.H. DENTON ET AL.belt is late Llanquihue in age. But none of the samplesis in the age range of the reworked organicclasts at the Calbuco site or from the young outwashplain near Bah’a Puerto Montt. Hence theseresults are consistent with the basal age of Huelmomire in indicating that the inner moraine belt alongsideSeno Reloncav’ was not deposited during theadvance shortly after 15,000 14 C yr BP.The mapping program, however, showed quite adifferent situation near the Calbuco site (94), as alreadysuggested by the chronologic results. Here aßuted surface drift suggests a strong advance fromthe southeast that overran the site and probablyreached close to the outer Llanquihue moraineslimit (Fig. 4). This advance was of a piedmont glacierlobe from Golfo de Ancud, not from Seno Reloncav’.Again, the implication is that the advanceshortly after 15,000 14 C yr BP was stronger for thesouthern than for the northern lobe. As discussedbelow, such a strong advance is also characteristicof the northern part of the Golfo Corcovado lobe,which achieved its maximum on Isla Grande deChiloŽ during an expansion that culminated closeto 14,805 14 C yr BP (Table 2). Thus there is a majorincrease in the strength of the advance shortly after15,000 14 C yr BP between the Lago Llanquihue andSeno Reloncav’ piedmont lobes to the north, andthe Golfo de Ancud and the northern Golfo Corcovadolobes to the south.Numerous new dates afford maximum limitingvalues for the outermost Llanquihue moraine beltdeposited by the Seno Reloncav’ piedmont glacier.This belt forms a continuous arc from the interlobateintersection north of the Calbuco site (94) tothe east of the town of Alerce (Fig. 5). Organicclasts reworked into the moraine core, or into outwashgraded to the moraine, afford the followingages: 26,230±340 14 C yr BP (A-6487); 28,970±410 14 +160C yr BP (A-6488); 28,060 14 C yr BP (A-8780); 27,405 14 -155+350C yr BP (A-8781); 27,510±53014 -335+325C yr BP (UGA-6984) and 29,410 -30014 C yr BP (A-8783) from the core of a moraine north of Bah’aPuerto Montt (site 33); 29,080±360 14 C yr BP (A-6777) from outwash that passes beneath the outer+390Llanquihue-age moraine (site 40); 32,78014 Cyr BP (A-8785) and 26,47514 -385+265-255 C yr BP (A-8787)+315from outwash at site 60; 32,23514 -300 C yr BP (A-+2408789) from outwash at site 61; 26,560 14 -235 C yr BP(A-8790) from outwash at site 62; and 29,605+285 14-275 C yr BP (A-8791) from outwash at site 63 (Table1).Fourteen samples of basal organic material frommires in topographic depressions afford minimumlimiting ages for the outer moraine belt (Table 1).It is certain that this outer moraine belt antedates18,000 14 C yr BP, as four basal dates fall between18,000 and 19,000 14 C yr BP. There is also a highlikelihood that the moraine is older than 20,480+325 14 C yr BP (A-8525), 21,230±169 14 C yr BP-315+415(AA-16244), and 22,680 - 39514 C yr BP (A-8526),the near-basal dates of mires at sites 56, 66, and 55,respectively.The pre-late Llanquihue stratigraphic record isrevealed in sections exposed during building constructionon the gulfside ice-contact slope nearBah’a Puerto Montt on the western bank of CanalTenglo (site 46) (Fig. 29), as well as on the gulfsideroad at Punta Pelluco (site 48) (Fig. 30). The lowerpart of the Canal Tenglo section reveals debris-ßowsediments and ßuvial units, both composed of pyroclasticmaterials, interbedded with lacustrinesediments and peat beds. None of these lower unitscontains crystalline clasts from the Andes. Numerousdates show that the lower peat beds range in agefrom about 29,385 14 C yr BP (Table 2) to 39,660+640- 59514 C yr BP (A-7699) (Table 1). These dates documenta long interval when the Seno Reloncav’piedmont glacier lobe had withdrawn from the icecontactslope near Puerto Montt alongside SenoReloncav’. As discussed in Heusser et al. (1999),pollen analysis of all the dated peat beds showsSubantarctic Parkland environmental conditions.During this ice-free interval Canal Tenglo wasprobably the channel of a river draining Volc‡n Calbuco.The ßuvial units were deposited rapidly in thechannel, as shown by climbing ripple structuresthat in some cases are deformed because of sedimentloading. The debris-ßow units of volcanicmaterial, one of which encloses a tree in growth po-+265sition dated to 31,905 -26014 C yr BP (A-7721), areprobably the distal facies of lahars that radiated outwardin drainage channels from Volc‡n Calbuco.Therefore, it is likely that each package of volcanicsediment between peat beds represents rapid depositionassociated with a discrete volcanic event.The upper part of the Canal Tenglo section isdistinctly different from the lower part, in that itcontains two thick ice-proximal outwash units,each with crystalline cobbles and boulders of Andeanorigin. Dates of associated peat beds affordclose maximum ages for these outwash units. Sandthat coarsens rapidly upward into coarse gravel ofthe lower outwash unit rests on the youngest of thelower peat beds mentioned above. The undisturbedtop of the peat bed yielded six consistent ages withan error-weighted mean of 29,385±95 14 C yr BP204GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTFig. 29. Schematic composite lithostratigraphy of a sequence of glaciofluvial, fluvial volcanic, and peat units stacked in an ice-contactslope that rises 130 m above Canal Tenglo. See legend in Fig. 7 (p. 177) for lithologic details. Unit A features fibrous, organic-rich siltsealed with fine-grained lacustrine sediments. Unit B consists of coarse-grained volcanic sand with large-scale, cross-bedded stratification;this unit fines upward from small granules to fine-grained silt and sand. Unit C is fibrous peat with wood (some pieces in growthposition) interbedded with coarse-grained volcanic debris-flow sediments; some reworked wood fragments occur within the volcanicsand. Unit D consists of thin beds (about 10 cm thick) of peat interbedded with coarse-grained volcanic sand. Units E and G featurelarge, cross-bedded volcanic sand that fines upward to fine-sand and silt. Unit F is fibrous peat with numerous wood fragments. Somewood fragments are sealed beneath laminated silt and clay. Unit H is peat with a low organic content. Unit I features cross-bedded volcanicsand; this unit fines upward from coarse-grained sand to fine sand and silt. Unit J is peat with a low organic content. Unit K isfibrous peat with numerous wood fragments (up to 10 cm in length) some sealed by silt and clay. Unit L is fibrous peat sealed by laminatedsilt and clay. Unit M is stratified silt and sand (with flame structures and well-developed diapirs). Unit N features well-definedhorizons of peat (with low organic content) interbedded with well-stratified medium- to fine-grained sand. Unit O is fibrous peat withwood fragments sealed by laminated silts. Unit P is a coarsening-upward sequence of massive and stratified sand, gravel, cobbles, andboulders of Andean origin. It is interpreted to be glaciofluvial in origin. A major shear plane occurs a third of the way up this unit fromthe base. Unit Q is fibrous peat capped with laminated silts. Unit R features coarse gravel and fine-grained cobbles of Andean originresting unconformably on unit Q; it is interpreted to be glaciofluvial in origin.GeograÞska Annaler á 81 A (1999) á 2205


G.H. DENTON ET AL.Fig. 30. Stratigraphic section at Punta Pelluco (site 48 in Fig. 15). Unit A is an intertidal delta sequence with horizontal beds of coarsevolcanic sand, locally cross-bedded. Thin silt beds are concentrated in zones that are 1Ð2 m apart. Unit A extends upward from thepresent intertidal zone, with a maximum observed thickness of 11 m. Clastic dikes penetrate the unit from below. See Table 1 for adescription of the radiocarbon sample from unit A. Unit B is a chaotic mixture of boulder gravel to coarse silt, deposited in a high-energyenvironment associated with proglacial drainage. Basal gravel and reworked organic clasts indicate a major unconformity. The unit isfiner-grained on the western side of the outcrop than on the eastern side. Unit B2 is a proglacial channel gravel deposited in a highenergyenvironment associated with proglacial drainage. Coarse gravel is eroded on Unit B. Unit C features basal glacial thrust planeswith imbricated slices of volcanic sand separated by gravel and diamicton. The lower contact is a major thrust slice. Unit D includesbasal glacial thrusts with included blocks of sand and gravel. The unit has an erosive basal contact, with reworked organic clasts. Allbeds are reoriented, and several thrust planes occur within individual beds. Unit E is basal till with a sandy silt matrix and with fissility.(Tables 1 and 2). A similar situation occurs higherin the section, where proximal outwash with Andeancrystalline clasts overlies another peat bed.The lowest peat in this bed yielded a date of 24,055±150 14 C yr BP (A-7912) (Table 1); the top of thepeat bed afforded two consistent dates with aweighted mean of 23,059±87 14 C yr BP (A-7626and T-11330A) (Table 2). This date is a close maximumvalue for the ice advance represented by theupper outwash unit.At site 86 about 3.8 km south of Huelmo in Fig.5, an organic bed 500 m long separates glacioßuvialunits exposed in a seaside cliff alongside SenoReloncav’. The upper glacioßuvial unit underliesLlanquihue moraines. The upper surface of the organicbed is sealed intact with a thin silt layer thatgives way to a coarsening-upward outwash sequence.Four dates of the upper surface of the peatbed yielded a weighted mean of 22,570±87 14 C yrBP (Table 2) for an advance of the Seno Reloncav’piedmont lobe into the Llanquihue moraine belt.Shells dated to 20,925±115 14 C yr BP (A-7627)are embedded in marine silt and mud at the presentshoreline of Seno Reloncav’ at site 95. The shellbearingsediments have been glacially displaced,and thus record ice-free conditions at or slightlyseaward of the site.Overall, the stratigraphic sections alongsideSeno Reloncav’ record a long ice-free interval inthe outer moraine belt that began prior to 39,66014 C yr BP and ended with the deposition of proximaloutwash about 29,385 14 C yr BP (Table 2) at CanalTenglo. Another major readvance is recorded at theCanal Tenglo section by renewed deposition ofproximal outwash and probable glacier overridingshortly after 23,059 14 C yr BP. The same advance ismost likely recorded south of Huelmo (site 86) at22,570 14 C yr BP (Table 2). We estimate that this advanceprobably culminated with the deposition ofthe outer Llanquihue moraine belt fringing SenoReloncav’, which has maximum dates of 26,230Ð26,475 14 C yr BP and minimum dates of 21,230Ð22,680 14 C yr BP. Recession was under way fromthe outer moraine belt by 21,230Ð22,680 14 C yr BP.The coast was ice-free at 20,925 14 C yr BP. Ayounger readvance culminated shortly after 15,000206GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFT14 C yr BP. To the north, the relatively small SenoReloncav’ piedmont lobe reached only to the westernedge of the marine embayment. To the south,the large Golfo de Ancud piedmont lobe advancedbeyond the coast well into the Llanquihue-age morainebelt. Although the intermediate moraine beltsof the Seno Reloncav’ ice lobes are undated, theymust represent additional advances.Northern Isla Grande de ChiloŽFigure 5 depicts Llanquihue moraine belts in thenorthern sector of eastern Isla Grande de ChiloŽ.The main Llanquihue outwash plain is graded tothe outermost Llanquihue moraine belt, and subsidiaryoutwash terraces are graded to morainesnested behind the outermost belt. We have not yetcarried out extensive Þeld work on Isla Grande deChiloŽ, but we report here new radiocarbon datesfrom several sites described previously.The Taiquem— mire (site 96) is in a depressionwithin the outermost Llanquihue moraine belt(Fig. 5). Thirty dates were obtained from a corethrough the 6.55 m of peat and lacustrine sedimentsin this mire (Heusser et al. 1999). These dates areprogressively older with increasing depth in thecore. The youngest is 10,355±75 14 C yr BP (AA-17974) at 90 cm depth, and the oldest is > 49,89214 C yr BP (AA-14770) at 600 cm depth (Heusser etal. 1999). The results show that the outermostLlanquihue moraine in the eastern sector of northernIsla Grande de ChiloŽ antedates the LGM. Thedates show continuous, or nearly continuous, accumulationof peat and lacustrine sediment. Thepollen record documents the existence of a SubantarcticParkland environment during cold stadials,from at least 15,000 14 C yr BP back to > 49,892 14 Cyr BP. In middle Llanquihue time, SubantarcticForest characterizes interstadials. Conspicuouslyabsent from the entire record prior to 15,000 14 C yrBP is an indication of the present-day interglacialrain forests of Isla Grande de ChiloŽ. The implicationof these data is that the outer Llanquihue moraineat Taiquem— is considerably older than49,892 14 C yr BP and yet postdates the penultimateinterglaciation. The most likely situation is that thisouter moraine is middle or early Llanquihue in age.There is evidence for at least two advances intothe Llanquihue moraine belt on Isla Grande de ChiloŽduring the LGM. The youngest of these advancesis documented at the Dalcahue site (98) (Fig. 31),previously described by Laugenie (1982) and Mercer(1984), where an organic bed of pyroclasticsand and silt accumulated in a small basin on compact,medium-gray lodgement till with Andean volcanicand crystalline erratics. In turn, the organicbed is overlain by a moraine with a core of glacioßuvialsand and gravel capped by thin (0.20Ð1.5m) lodgement till with several large (up to 1.5 mlong), striated crystalline erratics derived from theAndes. The basal part of the glacioßuvial sequencein the moraine core consists of Þne- to mediumbeddedsand that seals intact the upper surface ofthe organic bed. Thus the Dalcahue organic bed occursbetween two glacial drift units, both representingadvances of an Andean piedmont glacier lobeonto Isla Grande de ChiloŽ. A pollen proÞle constructedat centimeter intervals for the thickest portion(152 cm) of the organic bed (location B in Fig.31) shows the continuous existence of a SubantarcticParkland environment from shortly after 25,17614 C yr BP until about 14,720 14 C yr BP (Heusser etal. 1999). Dates of the uppermost organic sedimentat the location of the pollen proÞle are 14,720±10014 C yr BP (A-6189) and 14,770±110 14 C yr BP (A-6190). Dates for the base of the organic bed are28,558±366 14 C yr BP (AA-19438) at 151 cm+225depth; 30,070 -21514 C yr BP (A-7685), also at 151cm depth; and 27,910±315 14 C yr BP (AA-19439)at 152 cm depth (Table 1). Table l shows numerousadditional dates from within the organic bed at thepollen site. About 10 m south of the location of thepollen proÞle, the upper surface of the organic bed,here only 90 cm thick, reveals a wood layer buriedintact by the stratiÞed sand at the base of the glacioßuvialsequence. An additional 33 dates both ofwood and of organic silt and sand range from15,260±115 14 C yr BP (UGA-6983) to 14,480±18014 C yr BP (UGA-6918) (Table l). Overall, the totalof 35 dates of wood and organic sediment from theintact original surface of the organic bed at the twosampling localities yields a mean of 14,805±18 14 Cyr BP (Table 2), which should accurately record thelast ice advance over the Dalcahue site. The progressivelyincreasing ages with depth within theDalcahue organic bed indicate that this advancewas the most extensive in at least the last 30,070 14 Cyr BP, and therefore the most extensive of the LGM.Fig. 31, see p. 196.GeograÞska Annaler á 81 A (1999) á 2207


G.H. DENTON ET AL.The advance of the northern part of the GolfoCorcovado piedmont lobe that passed over the Dalcahueorganic bed reached at least to the Llanquihuemoraine at the upper edge of the high ice-contactslope (site 99) 1.5 km north of the Dalcahuesite. An exposure at the top of this slope shows icecontactgravels unconformably cut into till and outwash.The till occurs along much of the ice-contactslope of this moraine, and probably corresponds tothe basal till beneath the nearby Dalcahue organicbed. These relationships suggest that the slopemarked an ice-marginal position on at least two occasions.The Þrst probably corresponds with theadvance dated at more than 30,070 14 C yr BP at theDalcahue site. The second corresponds with the advanceover the Dalcahue organic bed, as shown byages of organic clasts reworked into the ice-contactgravel at the top of the slope. The youngest of theseclasts dates to 14,820±450 14 C yr BP (QL-4532)and affords a maximum limiting age for the last advanceto the head of the ice-contact slope. Thisclast must have been derived from the upper surfaceof the Dalcahue (or equivalent) organic bed.Other reworked clasts in the ice-contact gravelyielded ages of 20,030±160 14 C yr BP (UGA-6972)and 19,840±180 14 C yr B.P (UGA-6979). Theseclasts must have been derived from deeper in theDalcahue organic bed (or from an equivalent organicbed), because the advance over the Dalcahueorganic bed is the only one that could have been responsiblefor reworking such young organic clasts.We obtained preliminary limiting dates for glacierrecession by coring mires in morainal topographysouth of Dalcahue. Our mapping programshows that the outer Llanquihue moraines bendsharply southwestward near Dalcahue, and that theLlanquihue glacial limit trends toward the coastand out to sea near the middle of Isla Grande deChiloŽ. Thus the sites we cored south of Dalcahue,as well as those cored by Villagr‡n (1988) in southeasternIsla Grande de ChiloŽ, all lie well within thelimit reached by Andean ice during the youngestLlanquihue advance recorded at the Dalcahue site.The Mayol mire (site 100) yielded a sequence ofconsistently older dates with increasing depth thatculminates in a basal age of 14,941±97 14 C yr BP(AA-20370) (Heusser et al. 1999). Near-basal agesfrom the Estero Huitanque mire (site 101) are13,345±105 14 C yr BP (TUa-258A) and 14,350±240 14 C yr BP (Beta-62029) (Heusser et al. 1996).Farther south at site 102, basal gyttja from a mireon Llanquihue drift dates to 13,560±95 14 C yr BP(T-10307A). Villagr‡n (1988) reported near-basaldates of 13,040±210 14 C yr BP (Beta-10481) fromPuerto Carmen (site 103), and 13,100±260 14 C yrBP (Beta-10485) from Laguna Chaiguata (site 104).The Þnal site that we examined is at Teguaco (site97), previously described by Heusser (1990). Herea road cut exposes a 28-cm-thick organic silt andgyttja bed sealed by laminated, light-gray glaciolacustrinesediments that are overlain Þrst by sand andthen by ice-proximal glacioßuvial deposits. Twelvesamples of gyttja with macrofossils from the uppersurface of the organic bed and of wood pieces encasedwithin the glaciolacustrine sediments yieldedan error-weighted mean of 22,295±40 14 C yr BP(Table 2). Dates from near the base of the organic+190silt and gyttja bed are 27,34514 C yr BP (A-7693)and 26,630±175 14 -185C yr BP (A-7692) (Table l). Organicsand from the base of the underlying 30-cmthickorganic sand unit afforded a date of 29,240±600 14 C yr BP (QL-4546).The Teguaco section is located within a deep,southeast-trending river valley. We interpret thestratigraphic sequence as representing an incursionof an Andean piedmont glacier lobe onto IslaGrande de ChiloŽ, damming a glacial lake thatßooded the organic silt bed. The laminated glaciolacustrinesediments sealed the top of the organicbed. The wood in the glaciolacustrine silts was eitherßoated off the land surface, or else was washedinto the lake. The advancing piedmont ice lobe thendeposited the coarsening upward sequence thatculminated in ice-proximal outwash.As currently understood, the moraines andstratigraphic sections on Isla Grande de ChiloŽrecord three major advances of Andean piedmontglaciers into the Llanquihue moraine belt. The oldestadvance, represented by the outermost morainebelt and by the main Llanquihue outwash plain, occurredprior to 49,892 14 C yr BP. The youngest advanceculminated at close to 14,805 14 C yr BP at amoraine located 1.5 km northwest of the Dalcahuesite (98). In Fig. 5 this moraine, which representsthe maximum position achieved by Andean piedmontice during the LGM, can be traced northwardat least 30 km from the vicinity of the Dalcahuesite, where it lies within 4 km east of the maximumLlanquihue moraine dated to > 49,892 14 C yr BP atTaiquem— (site 96). Recession from the ice limitreached during the youngest advance was underwaysouth of the Dalcahue site at Mayol (site 100)by 14,94l 14 C yr BP. The overall chronology of theDalcahue site indicates that this youngest advancewas the most extensive of the LGM. At least oneadditional advance during the LGM, dated to208GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFT22,295 14 C yr BP at Teguaco (site 97), culminatedbehind the position reached during the youngestadvance that overrode the Dalcahue site.DiscussionIce-front dynamicsThe Llanquihue-age piedmont glacier lobes in thesouthern Lake District had ßattish surface slopes.For example, the moraine belts mapped in Fig. 3 indicatethat the surface slope of the Lago Llanquihuepiedmont lobe rose inland toward the Andes atabout 3.0 m/1000 m during the LGM. Similar valuesapply to the former Rupanco and Puyehue icelobes (Fig. 2). When it stood at the edge of the lakesidekame terrace at the culmination of the readvanceabout 14,550Ð14,869 14 C yr BP, the LagoLlanquihue lobe again had an upper surface thatrose inland at only about 2.2 m/1000 m. These ßattishlobes deposited only waterlain till, not meltoutor basal lodgement tills, at the edge of lake and marinebasins, in sharp contrast to the situation withinthe moraine belts, where only meltout, ßow, andlodgement tills are present. The waterlain tills atthe periphery of the basins were soft at the time ofglacial deposition and overriding, as shown by theclastic dikes of diamicton injected into the laminatedglaciolacustrine sediments and by the diapirstructures caused by loading of the lacustrine sedimentswith gravelly debris-ßow material from adjacentice. And yet the Þne laminations in the glaciolacustrinesediments that make up much of thewaterlain till do not show deformation by pervasiveshear from overriding glacier ice.We suggest that waterlain till is the characteristicdeposit recording the passage of piedmont icelobes through the lake and marine basins. The ßatnessof the ice lobes probably occurred because theglacier ice was supported by high pore-water pressurein the waterlain till. Most of the basal ice ßowprobably occurred at the iceÐsediment interfacerather than by pervasive shearing of the underlyingwaterlain till. Once the ice ßowed out of the basininto the fringing moraine system with its widespreadporous outwash and gravelly diamicton,water ßowed away from the glacier base, and thepredominant basal deposit was lodgement till.Thus we postulate that the major facies depositedwithin the lake basins involved settling from siltplumes associated with meltwater discharge, alongwith sediment ßows of gravelly diamicton from theice margin. Basal sliding took place at the sedimentÐiceinterface and till was not lodged on thelake ßoor. In contrast, the major facies depositedwhen ice lobes advanced out of the lake basins intothe moraine belts were outwash, gravelly sedimentßows, and lodgement till.Accuracy of radiocarbon chronologyThe validity of the reconstructions of glacier ßuctuationsdepends on the accuracy of the radiocarbonchronology. Overall, the dates are remarkablyconsistent, but there are some problems. Becausecritical organic horizons are commonly isolated instratigraphic sections, many individual results cannotbe judged within the context of a consistent sequenceof dates. In these cases, we obtained numerousdates of individual horizons, from variousorganic materials where possible, in order to establishthe age from weighted mean values. The resultsof such exercises at the Dalcahue (98) andFrutillar Bajo (16) sites yielded two additional insights.First, individual dates of wood and organicsediment were indistinguishable. Second, the datesof the upper surface of the Dalcahue (site 98) organicbed have a range of about 780 14 C yr(15,260±115 14 C yr BP, UGA-6983, to 14,480±18014 C yr BP, UGA-6918) that is not dependent on thetype of material analysed or on the dating laboratory(Table 1). Thus an individual date can be asmuch as 350Ð400 14 C years different from themean age of an organic horizon. Similar examplescome from individual dates within the series obtainedfor the upper surface of organic beds at thePuerto Varas railroad bridge (site 26) and Bella VistaBluff (site 29) sections (Table 1). Third, manydates reported by Mercer (1976) and Porter (1981)for organic beds in the Puerto Varas embayment aretoo young, compared with the new dates, by asmuch as 1600 14 C yr. Such discrepancies suggestcaution in interpreting the age of glacial events atother sites from individual dates. Hence, we emphasizereplicate dating wherever possible in orderto obtain the accuracy and precision necessary toestablish the chronology of glacier ßuctuations.The most reliable results occur where a sequenceof dates is integrated with multiple analyses of variousmaterials from a critical horizon. Examples arethe ages of the upper surface of the organic beds atthe Dalcahue (site 98) and Frutillar Bajo (site 16)sites. The next most reliable results come from individualorganic horizons with numerous analyses;an example is the organic bed at the Llanquihue site(23). Equally reliable are the results of the basaldates from cores and mires that have a sequence ofGeograÞska Annaler á 81 A (1999) á 2209


G.H. DENTON ET AL.consistent ages. Examples are the cores from FundoLlanquihue (site 18) (Heusser et al. 1996, 1999)and Canal de la Puntilla (site 6) (Moreno 1997;Moreno et al. 1999). The least reliable results aresingle dates of soil, gyttja, or peat clasts reworkedinto outwash. These clasts were eroded either froma pre-existing land surface or from organic beds exposedin stratigraphic sections. In such cases we attemptedto date multiple reworked clasts from individualoutwash units to achieve consistency in theresults. We note that different parts of large clastscan obviously have widely varying ages (Table 1).Single dates of organic material from the base ofcores through mires on Llanquihue moraine beltsalso may not be reliable. Again, where possible, weobtained radiocarbon dates from multiple cores toachieve consistency in the results.Preferred chronology for Llanquihue glacialadvancesA fundamental question is whether the low-slopingpiedmont glacier lobes in the southern Chilean LakeDistrict and Isla Grande de ChiloŽ reacted in unisonto climate forcing or whether they exhibited the outof-phasepulsing behavior suggested for low-slopingsouthern lobes of the Laurentide Ice Sheet andfor low-sloping ice streams of the West Antarctic IceSheet (Alley and Whillans 1991; Clark 1994). Todiscriminate between these two possibilities, we obtainednumerous dates of the youngest major advanceof the Lago Llanquihue, Seno Reloncav’,Golfo de Ancud, and northern Golfo Corcovadopiedmont lobes. The details of each site are givenabove and the dates are listed in Table l. The maximumextent of the Lago Llanquihue lobe during thisadvance is dated to 14,869 14 C yr BP at the Llanquihuesite (23; Table 2); 14,540 14 C yr BP at the BellaVista Bluff site (30) in Puerto Varas (Table 2);14,550Ð14,613 14 C yr BP at the railroad bridge site(26) in Puerto Varas; 14,820 14 C yr BP at the CalleSanta Rosa site (27) in Puerto Varas (Mercer 1976);14,882 14 C yr BP at the Northwest Bluff site (28) inPuerto Varas (Table 2); and 14,650 14 C yr BP at thePuerto Phillippi site (19) (Table 1). For the Seno Reloncav’lobe this advance is dated to 14,879 14 C yrBP at Punta Penas (47) (Table 2); 15,220 14 C yr BP atthe Isla Maillen site (73) (Table 1); and shortly after15,040 14 C yr BP at the top of the ice-contact slopeon the western side of Bah’a Puerto Montt (45) (Table1). For the Golfo de Ancud lobe, this readvanceis dated to shortly after 14,900 14 C yr BP west of Calbuco(94) (Table 1). Numerous dates from the Dalcahuesite (98) place the last maximum of the northernpart of the Golfo Corcovado lobe at close to14,805 14 C yr BP (Table 2); general recession wasunderway at the Mayol site (100) by 14,941 14 C yrBP (Table 1). Therefore, we conclude that these fourpiedmont glacier lobes ßuctuated in near synchrony(within the limits of radiocarbon dating) during theyoungest major pulse of the LGM.To test whether such synchronous behavior wascharacteristic of other ice-marginal ßuctuations,we examined the chronology of the outer Llanquihuemoraine belts of the northern three piedmontlobes, and of the early LGM advance onto IslaGrande de ChiloŽ. Here, interpretation of the situationis more complicated than with the youngestreadvance. Therefore we established a preferredand an alternate chronology. Our preferred chronologystarts with dates of in-situ samples in threestratigraphic sections. The Þrst is from site l at theouter limit of Llanquihue moraines fringing LagoRupanco. Here peat sealed by silt below the till capof this moraine is 22,460 14 C yr old, thus affordinga close maximum date for this moraine (Table 1).On the western coast of Seno Reloncav’ at site 86,a major readvance into the Llanquihue morainebelt is dated to 22,570 14 C yr BP (Table 2). Theavailable bracketing ages suggest that the outermostLlanquihue-age moraine belt west and northof Seno Reloncav’ was deposited during this advance.Farther south at Teguaco on Isla Grande deChiloŽ, a major glacial advance documented by numerousdates at 22,295 14 C yr BP (Table 2) terminatedbehind the position reached at 14,805 14 C yrBP. Finally, we consider the chronology of the outermostLlanquihue-age moraine belt fringing LagoLlanquihue. Most of the bracketing dates implythat this moraine belt is older than 20,890Ð20,580and younger than 23,020Ð22,250 14 C yr BP. It isthus possible that this moraine belt also was depositedduring the major readvance dated elsewhere toabout 22,295Ð22,570 14 C yr BP.Older advances are recorded for individuallobes. The most Þrmly established of these earlieradvances is recorded at the Frutillar Bajo section(site 16) (Fig. 25), where a landscape surface wasburied intact by glacial deposits at 26,797 14 C yr BP.A less certain situation (because of the lack of alandscape buried intact) involves the advance at29,363 14 C yr BP recorded at the Puerto Octay site(7) and 29,385 14 C yr BP Canal Tenglo site (46).From the sections at Canal Tenglo and PuertoOctay, it appears that late Llanquihue glacial conditionsbegan at about 29,400 14 C yr BP. From the210GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTchronology of the last major advance and from theminimum ages of deglaciation on Isla Grande deChiloŽ, we conclude that the LGM ended shortlyafter 14,550Ð14,869 14 C yr BP. The glacial recordgiven here and the pollen records presented inHeusser et al. (1999) and in Moreno et al. (1999)together suggest that full-glacial or near-full-glacialconditions persisted for most of this 15,000 14 Cyr interval. The only discrepancy between theserecords occurs near the beginning of this interval,where a glacial maximum apparently occurredabout 29,400 14 C yr BP, but Subantarctic Forest lingereduntil a few thousand years later at Dalcahueand Taiquem—. During this long interval, glacierpiedmont lobes probably achieved maxima at14,550Ð14,869 14 C yr BP; 22,295Ð22,570 14 C yr BP;26,797 14 C yr BP; and 29,363Ð29,385 14 C yr BP.There is a suggestion from the Puerto Varas embaymentof additional maxima shortly before17,800 14 C yr BP and again shortly before 15,70014 C yr BP. The amount of recession between thesemaxima cannot yet be determined from direct evidence.But the pollen records from Dalcahue (site98) and Taiquem— (site 96) on Isla Grande de ChiloŽ,along with those from Fundo Llanquihue, Canalde la Puntilla, and Alerce in the southern LakeDistrict, suggest persistent cold, wet conditionsthroughout late Llanquihue time, with mean summertemperature as much as 6Ð8¡C below thepresent-day value and precipitation about doublethat of today. Therefore, it is likely that glaciers remainedrelatively large throughout this long intervalof cold, wet climate. However, two times of minorglacier recession deserve mention. The Þrst isthe Varas interstade of Mercer (1976). From theLlanquihue (23) and Bella Vista Bluff (30) sites, wenow place this recession between 15,700 14 C yr BPand 14,869 14 C yr BP. From the pollen analysis ofthe Bella Vista Bluff and Llanquihue sites given inHeusser et al. (1996, 1999), as well as of the FundoLlanquihue (18) and Canal de la Puntilla (6) sites(Heusser et al. 1999; Moreno et al. 1999), we concludethat open Subantarctic Parkland with coldand wet conditions persisted through this interval.Pollen of thermophilic trees is rare or lacking. TheLago Llanquihue piedmont glacier lobe recededfrom the western lakeshore during this interval, butfrom the available lake-level history we cannot determineif the lobe evacuated the lake basin. Thesecond minor interval involved ice recession fromthe outermost Llanquihue moraine system west ofLago Llanquihue prior to 20,890 14 C yr BP at FundoLlanquihue (site 18); 20,100 14 C yr BP at Canal dela Puntilla (site 6); and 20,580 14 C yr BP at Canal deChanch‡n (site 5). Similar recession had occurredfrom the western edge of the Seno Reloncav’ marineembayment by 20,925 14 C yr BP. The pollenrecords from the Fundo Llanquihue and Canal dela Puntilla cores again show persistence of SubantarcticParkland under cold and wet conditions duringthis recession.Massive glacier recession marked the end offull-glacial climate immediately after the culminationof the last major glacier advance at 14,550Ð14,869 14 C yr BP, and ice had receded deep into theAndes, within 10 km of modern glaciers, by 12,310±360 14 C yr BP (RL-1892) (Heusser 1990) (site105) (Table 1). This last advance reached the westernshore of Lago Llanquihue and Seno Reloncav’.On Isla Grande de ChiloŽ it was the maximum advanceof the LGM. Pollen records from Canal de laPuntilla (Moreno et al. 1999) and Fundo Llanquihue(Heusser et al. 1999) show cold, wet environmentalconditions during this advance. The pollenrecord places emphasis on the importance of theclimate warming that accompanied this glacier recession.A marked rise of Nothofagus within a SubantarcticParkland environment is taken to be evidenceof the initial warming at 14,600 14 C yr BP(Moreno 1998; Moreno et al. 1999). The pollenrecords from Canal de la Puntilla (Moreno 1994,1997; Moreno et al. 1999), Fundo Llanquihue, andAlerce (Heusser et al. 1996, 1999) all highlight theinvasion of the Lago Llanquihue and Seno Reloncav’lowlands by thermophilic tree species ofthe North Patagonian Evergreen Forest beginningat close to 14,000 14 C yr BP. The Taiquem— pollenrecord (Heusser et al. 1999) shows this to be themost important vegetation change since prior to49,842 14 C yr BP. A closed-canopy North PatagonianEvergreen Forest then came into existence at12,700Ð13,000 14 C yr BP, and reached its fullest developmentat about 12,200Ð12,500 14 C yr BP.The paleoenvironment changes inferred frompollen are similar to those previously interpretedfrom the beetle record in the Chilean Lake District(Hoganson and Ashworth 1992; Ashworth andHoganson 1993). However, new dates (Table 1)from the Puerto Varas railroad bridge site (26) nowshow that 10 of the 12 beetle samples from the organicsilt unit refer only to 17,350Ð17,880 14 C yrBP, and not to a continuous record through the lastseveral thousand years prior to the end of the LGM.The remaining two beetle samples come from theprominent organic bed on the silt unit; it is unclearwhether these two samples should be assigned toGeograÞska Annaler á 81 A (1999) á 2211


G.H. DENTON ET AL.17,350Ð17,880 14 C yr BP or to 14,613 14 C yr BP. Inview of this revised chronology, the available fossilbeetle data show cold, wet conditions at the followingtimes within the LGM: 16,000Ð18,170 14 C yrBP (Canal de Chanch‡n, site 5); 17,350Ð17,880 14 Cyr BP (railroad bridge, site 26); and 15,715 14 C yrBP (Bella Vista Park, site 29).Dates of interdrift organic sediments at the PuertoOctay (7), Frutillar Bajo (16), and Canal Tenglo(46) sites indicate an interval of ice-free conditionsalong the western shores of Lago Llanquihue andSeno Reloncav’, beginning about 36,960Ð39,66014 C yr BP and ending about 26,940Ð29,400 14 C yrBP. We have not found evidence of Andean glacieradvance into the outer Llanquihue-age morainesystem during this interval. By necessity, then, theextensive outwash plains graded to Llanquihuemoraine belts must have been inactive throughoutthis middle Llanquihue non-glacial interval. Wecannot yet determine from direct evidence themagnitude of middle Llanquihue recession of Andeanpiedmont and mountain glaciers. But the pollenrecords from the Puerto Octay, Frutillar Bajo,and Canal Tenglo sites are consistent in showingthe persistence of a cool, wet Subantarctic Parklandenvironment through the later part of middleLlanquihue time in the Lago Llanquihue and SenoReloncav’ areas.Llanquihue drift units that are little weatheredoccur beneath dated middle Llanquihue interdriftorganic deposits near Lago Llanquihue. At thePuerto Octay site (7), outwash dated to >39,34014 C yr BP represents a glacier advance at least to thetop of the adjacent ice-contact slope. At FrutillarBajo (site 16) a coarsening-upward sequence ofglacioßuvial deposits is capped by till, all olderthan 36,960 14 C yr BP. The associated glacier advancewas beyond the top of the ice-contact slopeand into the Llanquihue moraine belt. West ofPuerto Varas at the outer edge of the Llanquihuemoraine belt at site 24, thin late Llanquihue driftoverlies thick outwash with an intervening organiclayer. The terrace morphology of the outwash canbe traced eastward beneath the thin late Llanquihuedrift to a massive overridden moraine located about500 m behind the outermost late Llanquihue icemarginalposition. At the nearby crossroad site 25,the upper portion of the weathered interdrift pyroclasticsediments was sheared into overlying driftduring a glacial advance over the site. The likelysituation is that this advance culminated at the massiveoverridden moraine to which the pre-late Llanquihueoutwash terrace is graded.The outermost moraine depicted in Fig. 5 forIsla Grande de ChiloŽ is also middle or early Llanquihuein age, as shown by the basal date of>49,892 14 C yr BP for the mire at Taiquem— (site96). For two reasons it seems probable that theLlanquihue moraine at Taiquem— was depositedduring marine isotope stage (MIS) 4. First, the moraineis beyond the range of accurate dating. Second,close-interval pollen analysis of the Taiquem—core shows Subantarctic Parkland and SubantarcticEvergreen Forest persisting back to the time ofdeposition of the moraine (Heusser et al. 1999). Ifthis Llanquihue moraine antedated the penultimateinterglaciation, then pollen of thermophilic trees ofthe North Patagonian and Valdivian Evergreen Forestswould be expected in the older part of theTaiquem— record (such trees invaded the lowlandsof the Lake District and Isla Grande de ChiloŽ earlyin the present interglaciation). Hence, the absenceof such pollen strongly suggests that the morainepostdates the last interglaciation. An old Llanquihueadvance that antedates 49,892 14 C yr BP andpostdates the penultimate interglaciation is mostlikely to be equivalent in age to MIS 4. Old Llanquihuedrift in the moraine belt near Lago Llanquihuealso is probably equivalent in age to MIS 4;however, it is more difÞcult to rule out a youngerage because, unlike the situation at Taiquem—, thechronology and pollen records of the interdrift organicdeposits extend back only to about 40,00014 C yr BP. In any case, the implication is that thesnowline in this sector of the Chilean Andes wasdepressed as much during the old Llanquihue advance(s)as during the LGM.Alternate late Llanquihue chronologyWe do not wish to leave the impression that our preferredchronology for glacial maxima during theLGM is without problems. For example, only limitingdates bracket the outer Llanquihue morainebelts west of Lago Llanquihue and Seno Reloncav’.Hence the proposed linkage of these outermoraines with the stratigraphic evidence of an advanceat 22,295Ð22,570 14 C yr BP could be questioned.In fact, there are now four dates that do notsupport this linkage. Three are of reworked organicclasts in outwash graded to the outer Llanquihuemoraines. The Þrst of these clasts, from site 4 westof Bah’a Mu–oz Gamero, dates to 20,840 14 C yr BP.As described above, the geomorphologic relationshipsdepicted in Fig. 2 suggest that this date maybe too young. The dated clast is reworked into out-212GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTwash graded to the outermost Llanquihue moraine,situated on the upper edge of a high ice-contactslope. Proximal to this outer ice-contact slope areseveral subsidiary Llanquihue outwash terracesthat, in turn, grade to a moraine belt and ice-contactslope alongside Bah’a Mu–oz Gamero. The meltwaterchannels of Canal de Chanch‡n and Canal dela Puntilla head at this lakeside ice-contact slope,and hence the minimum ages for these two channels(20,580±170 14 C yr BP, AA-9303; and 20,160±180 14 C yr BP, AA-9296) apply to all the Llanquihuemoraines and ice-contact slopes west of Bahi‡Mu–oz Gamero. Thus it is difÞcult, although notimpossible, to argue that the outermost Llanquihuemoraine and outwash plain are younger than20,840 14 C yr BP.The situation is more problematic with regard tothe date of 20,100 14 C yr BP from the second ofthese organic clasts, reworked into outwash thatpasses beneath the outer Llanquihue moraine atsite 13 near Frutillar Alto (Mercer 1976). As outlinedabove, the geomorphologic tracing of thisouter moraine at Frutillar Alto to the northeastshows that it is older than the outer (and hence oldest)Llanquihue moraine cut by Canal de Chanch‡nand Canal de la Puntilla. Therefore, the same argumentconcerning the age and geomorphologic positionof these channels implies that the date fromFrutillar Alto is too young. This conclusion is augmentedby the fact that the outer Llanquihue morainebelt at Fundo Llanquihue (site 18) can betraced northward to the outer Llanquihue moraineat Frutillar Alto. But at Fundo Llanquihue this outermoraine belt has minimum ages that extend backto 20,890 14 C yr BP (UGA-6908), again suggestingthat the maximum age of 20,100 14 C yr BP derivedfrom the reworked clast in outwash at Frutillar Altois too young.The third of these reworked clasts occurs in outwashgraded to the outermost Llanquihue moraineat site 11 near Frutillar Alto. This clast dates to21,120±180 14 C yr BP (AA-25374) (Table 1). Takenat face value, this result could be used to arguethat the age of the outer Llanquihue-age morainesystem west of Lago Llanquihue is close to 21,00014 C yr BP. Such an age comes from taking the availablemaximum (23,020 14 C yr BP at site 3; 22,79014 C yr BP, 22,700 14 C yr BP, and 21,120 14 C yr BPat site 11 at Frutillar Alto; and 22,250 14 C yr BP and22,985 14 C yr BP at site 32) and minimum (20,58014 C yr BP at Canal de Chanchˆn, site 5; 20,160 14 Cyr BP at Canal de la Puntilla, site 6; 20,890 14 C yrBP at Fundo Llanquihue, site 18) bracketing datesof the outer moraine belt at face value. But we preferto reserve judgment on this important issue untiladditional maximum limiting dates are available.A date of 19,450 14 C yr BP (I-5679) of ashy peatfrom below till at site 1 in the outer moraine nearLago Rupanco previously reported by Mercer(1976) also does not Þt our preferred chronologyfor the outer Llanquihue moraine system in thesouthern Lake District. We have been unable toreplicate this age, despite dating several samplesfrom a similar stratigraphic position in the sameborrow pit (Table l), and hence we have not includedit in our preferred chronology.These potential problems with the preferredchronology leave open the distinct possibility thatthere were several advances (rather than one advance)into the outer moraine belt between 20,000and 23,000 14 C yr BP. The earliest would have beenthat dated at 22,295Ð22,570 14 C yr BP from stratigraphicsections. It would have represented theLlanquihue maximum for the former Lago Rupancoand Seno Reloncav’ piedmont glacier lobes.Farther south on Isla Grande de ChiloŽ, as well asnear Lago Llanquihue, this advance would haveculminated short of the outer late Llanquihue moraines.A subsequent advance at close to 21,000 14 Cyr BP would have formed the outermost Llanquihue-agemoraines west of Lago Llanquihue, andelsewhere would have culminated short of the outermostLlanquihue-age moraines.Conclusions¥ The glacial chronology presented here suggeststhat the LGM in the region of the southern LakeDistrict, Seno Reloncav’, and Isla Grande deChiloŽ began at 29,400 14 C yr BP and endedshortly after 14,550Ð14,805 14 C yr BP. Withinthis interval of nearly 15,000 14 C yr, glacial advancesinto the outer Llanquihue moraine beltsculminated at about 29,400 14 C yr BP, 26,79714 C yr BP, 22,295Ð22,570 14 C yr BP, and 14,550Ð14,869 14 C yr BP. Additional glacial advancesinto the outer moraine belt may have occurred atclose to 21,000 14 C yr BP, shortly before 17,80014 C BP, and shortly before 15,730 14 C yr BP.Snowline depression during these advances wasabout 1000 m (Porter 1981; Hubbard 1997).Pollen records given in Heusser et al. (1999) andin Moreno et al. (1999) show that the LGM environmentin the southern Lake District remaineda Subantarctic Parkland under cold (6Ð8¡C depression of mean summer temperatureGeograÞska Annaler á 81 A (1999) á 2213


G.H. DENTON ET AL.below the current value) and wet conditions.However, at Dalcahue on Isla Grande de ChiloŽ,Subantarctic Forest was present until about25,176 14 C yr BP, then to be replaced by SubantarcticParkland (Heusser et al. 1999). At Taiquem—,Subantarctic Forest dominated untilabout 26,019 14 C yr BP (Heusser et al. 1999).These results may mean that the estimated beginningof the LGM from the glacier record isseveral thousand years too old.¥ Andean piedmont glacier lobes did not advanceinto the outer Llanquihue moraine belts duringthe portion of middle Llanquihue time between29,400 14 C yr BP and >39,660 14 C yr BP.¥ Early or middle Llanquihue glacier expansion tonear the outer limit of Llanquihue-age morainesoccurred more than about 40,000 14 C yr BP forthe Lago Llanquihue piedmont lobe. The Golfode Ancud and northern Golfo Corcovado piedmontlobes reached their maximum at the outermostLlanquihue moraine on Isla Grande deChiloŽ prior to 49,892 14 C yr BP. Pollen analysisof the Taiquem— mire suggests that this moraineprobably dates to the time of MIS 4. The implicationis that the Andean snowline was then depressedas much as during the LGM.AcknowledgementsThis research was supported by the OfÞce of ClimateDynamics of the National Science Foundation,the Lamont-Scripps Consortium for ClimateResearch of the National Oceanic and AtmosphericAdministration, the National Geographic Society,the Norwegian Research Council, and the SwissNational Science Fund. P.I. Moreno was supportedby the EPSCoR Program of the National ScienceFoundation. W. S. Broecker invited G. Denton, C.Heusser, and L. Heusser on the Þeld trip (funded byExxon) to southern South America in 1989 that initiatedthis work; we thank him for his continuedsupport and for numerous discussions about thesigniÞcance of our Southern Hemisphere paleoclimatedata. The Þeld work was carried out in co-operationwith the Servicio Nacional de Geolog’a yMiner’a, Chile. Arturo Hauser Y. and Jorge Mu–ozB. helped greatly with logistics and geological advice.Hugo Moreno R. afforded enormous aid inidentifying and interpreting the pyroclastic ßowsediments on all the glacial landforms, as well asthe debris-ßow deposits of volcanic breccia on thekame terrace near Puerto Varas. Richard Kellydrafted many of the maps and Þgures. Dr Seymour214typed the manuscript. M. Dubois, D. King, C.Latorre, A. Moreira, S. Moreira, and A. Silva assistedin the Þeld. C. Porter built Þeld equipmentand aided in the Þeld work. M. Boegel and Y.Hechenleitner were very helpful in making localarrangements. We thank W. Beck, G.Bonani, C.Eastoe, S. Gulliksen, I. Hajdas, A. Hogg, T. Jull, R.Kalin, R. Nydal, and M. Stuiver for providing radiocarbondates. We are grateful to A. Ashworthand S. Porter both for discussions and for informationabout stratigraphic sections in the southernLake District. We thank R. Rutford for energetichelp in the Þeld. This paper is funded in part by agrants/cooperative agreement from the NationalOceanic and Atmospheric Administration. Theviews expressed herein are those of the author(s)and do not necessarily reßect the views of NOAAor any of its sub-agencies.George H. Denton, Department of Geological Sciencesand Institute for Quaternary Studies, BryandGlobal Sciences Center, University of Maine, Orono,Maine 04469, USA.Calvin J. Heusser, 100 Clinton Road, Tuxedo, NewYork 10987, USA.Thomas V. Lowell, Department of Geology, Universityof Cincinnati, Cincinnati, Ohio 45221,USA.Christian SchlŸchter, Institute of Geology, Universityof Bern, 3012 Bern, Switzerland.Bj¿rn G. Andersen, Institute for Geology, Universityof Oslo, NO-0316, Blindern, Oslo 3, Norway.Linda E. Heusser, Lamont-Doherty Earth Observatory,Palisades, New York 10964, USA.Patricio I. Moreno, Institute for Quaternary Studies,Bryand Global Sciences Center, University ofMaine, Orono, Maine 04469, USA.David R. Marchant, Department of Geology, BostonUniversity, 675 Commonwealth Avenue, Boston,Massachusetts 02215, USA.ReferencesAlley, R.B. and Whillans, I.M., 1991: Changes in the West AntarcticIce Sheet. Science, 254:959Ð963.Andersen, B.G., Denton, G.H. and Lowell, T.V., 1999: Glacialgeomorphologic maps of Llanquihue drift in the area of thesouthern Lake District, Chile. Geografiska Annaler, 81A:155Ð166.GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTAshworth, A.C. and Hoganson, J.W., 1993: The magnitude and rapidityof the climatic change marking the end of the Pleistocenein the mid-latitudes of South America. Palaeogeography,Palaeoclimatology, Palaeoecology, 101:263Ð270.Bentley, M.J., 1997: Relative and radiocarbon chronology of twoformer glaciers in the Chilean Lake District. Jorunal of QuaternaryScience 12:25Ð33.Blunier, T., Chappellaz, J., Schwander, J., DŠllenbach, A.,Stauffer, B., Stocker, T.F., Raynaud, D., Jouzel, J., Clausen,H.B., Hammer, C.V. and Johnsen, S.J., 1998: Asynchrony ofAntarctic and Greenland climate change during the last glacialperiod. Nature, 394:739Ð743.Bond., G.C. and Lotti, R., 1995: Iceberg discharges into the NorthAtlantic on millennial time scales during the last glaciation.Science, 267:1005Ð1010.BrŸggen, J., 1950: Fundamentos de la geolog’a de Chile. InstitutoGeogr‡fico Militar. Santiago.Clark, P.U., 1994: Unstable behavior of the Laurentide Ice Sheetover deforming sediment and its implications for climaticchange. Quaternary Research, 41:19Ð25.Denton, G.H., Heusser, C.J., Lowell, T.V., Moreno, P.I., Andersen,B.G., Heusser, L.E., SchlŸchter, C. and Marchant,D.R., 1999 : Interhemispheric linkage of paleoclimate duringthe last glaciation. Geografiska Annaler, 81 A:107Ð153.Fairbanks, R.G., 1989: A 17,000 year glacio-eustatic sea levelrecord: Influence of glacial melting rates on the YoungerDryas event and deep-ocean circulation: Nature, 342:637Ð642.Hajek, E.R. and di Castri, F., 1975: Bioclimatografia de Chile.Imprenta editorial de la Universidad Cat—lica de Chile. Santiago.Heusser, C.J., 1974: Vegetation and climate of the southern ChileanLake District during and since the last interglaciation.Quaternary Research, 4:290Ð315.Ñ 1981: Palynology of the last interglacialÐglacial cycle in midlatitudesof southern Chile. Quaternary Research, 16:293Ð321.Ñ 1990: Chilotan piedmont glacier in the southern Andes duringthe last glacial maximum. Revista Geol—gica de Chile, 17:3Ð18.Heusser, C.J. and Flint, R.F., 1977: Quaternary glaciations andenvironments of northern Isla ChiloŽ. Geology, 5:305Ð308.Heusser, C.J., Heusser, L.E. and Lowell, T.V., 1999: Paleoecologyof the southern Chilean Lake DistrictÐIsla Grande de ChiloŽduring middle-late Llanquihue glaciation and deglaciation.Geografiska Annaler, 81 A:231Ð284.Heusser, C.J., Denton, G.H., Hauser, A., Andersen, B.G. andLowell, T.V., 1995: Quaternary pollen records from the ArchipiŽlagode ChiloŽ in the context of glaciation and climate.Revista Geol—gica de Chile, 22:25Ð46.Heusser, C.J., Lowell, T.V., Heusser, L.E., Hauser, A., Andersen,B.G. and Denton, G.H., 1996: Full-glacialÐlate-glacial paleoclimateof the southern Andes: Evidence from pollen, beetlesand glacial records. Journal of Quaternary Science, 11:173Ð184.Hoganson, J.W. and Ashworth, A.C., 1992: Fossil beetle evidencefor climatic change 18,000Ð10,000 years BP in south-centralChile. Quaternary Research, 37:101Ð116.Hubbard, A.L., 1997: Modelling climate, topography and paleoglacierfluctuations in the Chilean Andes. Earth Surface Processesand Landforms, 22:79Ð92.Ives, P.C., Levin, B., Robinson, R.D. and Rubin, M., 1964: U.S.Geological Survey radiocarbon dates VIII. Radiocarbon,6:37Ð76.Laugenie, C., 1982: La RŽgion des Lacs, Chili Meridional. UnpublishedPhD thesis. Universit de Bordeaux, Bordeaux,France. Volume 1, 1Ð332, Volume 2, 335Ð822.Laugenie, C.A. and Mercer, J.H., 1973: Southern Chile: A chronologyof the last glaciation. In: IX INQUA Congress Abstracts.Christchurch, N.Z. 202Ð203.Lowell, T.V., Heusser, C.J., Andersen, B.G., Moreno, P.I., Hauser,A., Heusser, L.E., SchlŸchter, C., Marchant, D.R. andDenton, G.H., 1995: Interhemispheric correlation of latePleistocene glacial events. Science, 269:1541Ð1549.Mercer, J.H., 1972: Chilean glacial chronology 20,000 to 11,000Carbon -14 years ago: Some global comparisons. Science,176:1118Ð1120.Ñ 1976: Glacial history of southernmost South America. QuaternaryResearch, 6:125Ð166.Ñ 1984: Late Cainozoic glacial variations in South Americasouth of the equator. In: Vogel, J.C.D. (ed.): Late CainozoicPaleoclimate of the Southern Hemisphere. A.A. Balkema.Rotterdam. 45Ð58.Moreno, P.I., 1994: Vegetation and climate near Lago Llanquihuein the Chilean Lake District between 20,200 and 9500 14 C yrBP. MS thesis. University of Maine. Orono, Maine. 50 p.Ñ 1997: Vegetation and climate near Lago Llanquihue in theChilean Lake District between 20,200 and 9500 14 C yr BP.Journal of Quaternary Science, 12:485Ð500.Ñ 1998: Termination of the last ice age in mid-latitudes of SouthAmerica. PhD thesis. University of Maine. Orono, Maine. 187 p.Moreno, P.I., Lowell, T.V., Jacobson, G.L. Jr and Denton, G.H.,1999: Abrupt vegetation and climate changes during the lastglacial maximum and last termination in the Chilean LakeDistrict: A case study from Canal de la Puntilla (41¡S). GeografiskaAnnaler, 81 A:285Ð311.Olivares, R.B., 1967: Los glaciones cuaternarias al oeste de LagoLlanquihue en el sur de Chile. Revista Geogr‡fica, 67:100Ð108.Porter, S.C., 1981: Pleistocene glaciation in the southern LakeDistrict of Chile. Quaternary Research,16:263Ð291.SchlŸchter, C., Gander, P., Lowell, T.V. and Denton, G.H., 1999:Glacially folded outwash near Lago Llanquihue, southernLake District, Chile. Geografiska Annaler, 81 A:347Ð358.Sowers, T. and Bender, M., 1995: Climate records covering thelast deglaciation. Science, 269:210Ð214.Steig, E.J., Brook, E.J., White, J.W.C., Sucher, C.M., Bender,M.L., Lehman, S.J., Waddington, E.D., Morse, D.L. and Clow,G.D., 1998: Synchronous climate changes in Antarctica andthe North Atlantic. Science 282:92Ð95.Turbek, S.E. and Lowell, T.V., 1999: Glacial deposition along anice-contact slope: An example from the southern Lake District,Chile. Geografiska Annaler, 81 A:325Ð346.Villagr‡n, C., 1988: Late Quaternary vegetation of southern IslaGrande de ChiloŽ, Chile. Quaternary Research, 29:294Ð306.Ward, G.K. and Wilson, S.R., 1978: Procedures for comparing andcombining radiocarbon age determinations: A critique. Archaeometry,20:19Ð31.Weischet, W., 1958: Studien Ÿber den glazial bedingten Formenchatzder sŸdchilenischen LŠngssenke im West-Ost Profilbeiderseits Osorno. Petermanns Geographische Mitteilungen,102:161Ð172.Ñ 1964: Geomorfolog’a glacial de la regi—n de los lagos. Communicacionesde la Escuela de Geolog’a No. 4. Universidadde Chile, Santiago. 36 p.Manuscript received Febr. 1998, revised and accepted Nov. 1998.GeograÞska Annaler á 81 A (1999) á 2215


G.H. DENTON ET AL.Table 1. Radiocarbon dates associated with glacial deposits in the region of the southern Lake District, Seno Reloncav’ and Isla Grandede ChiloŽ from the University of Arizona Laboratory of Isotope Geochemistry (A), the NSF-Arizona Accelerator Mass Spectrometry(AMS) Facility (AA), the University of Georgia Radiocarbon Laboratory (UGA), the Trondheim Laboratoriet for Radiologisk Datering(T and TUa), the University of Washington Quaternary Isotope Laboratory (QL), the University of Waikato Radiocarbon Laboratory(Wk), ETH-Hšnggerberg AMS Facility (ETH), and Beta Analytic (Beta). See original references for dating laboratories of dates reportedby previous investigators. The locations of the sample sites are given in Figures 1 and 7-10. Asterisks indicate samples used in Table 2.SiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description $ ‘ 5XSDQFR ERUURZ SLW 6DPSOH IURP WRS FP RI ILEURXV SHDW 7RS RI SHDWLV VHDOHG E\ D WKLQ OD\HU RI VLOW RYHUODLQ E\ VDQG FP DQG WKHQ WLOO FP 'DWH LV FORVH PD[LPXP IRU WKH JODFLHU DGYDQFH UHSUHVHQWHG E\ WLOO$ ‘ 6DPSOH IURP WRS FP RI ILEURXV SHDW EHVLGH VDPSOH $7 $ ‘ $VK\ SHDW OD\HU PHFKDQLFDOO\ GLVSODFHG LQWR PRUDLQH FRUH 0D[LPXPDJH IRU PRUDLQH7 $ ‘ $VK\ SHDW OD\HU PHFKDQLFDOO\ GLVSODFHG LQWR PRUDLQH FRUH 0D[LPXPDJH IRU PRUDLQH7 $ ‘ 7RS RI SHDW OD\HU IROGHG DQG PHFKDQLFDOO\ GLVSODFHG LQ PRUDLQH FRUH0D[LPXP DJH IRU PRUDLQH7 $ ‘ %DVH RI SHDW OD\HU VDPH OD\HU DV VDPSOH 7$ IROGHG DQGPHFKDQLFDOO\ GLVSODFHG LQ PRUDLQH FRUH 0D[LPXP DJH IRU PRUDLQH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH7 $ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ :RRG IURP EHG PHFKDQLFDOO\ GLVSODFHG LQWR PRUDLQH FRUH 0D[LPXPDJH IRU PRUDLQH$ $VK\ SHDW PHFKDQLFDOO\ GLVSODFHG LQWR PRUDLQH FRUH EHQHDWK WLOO0D[LPXP DJH IRU PRUDLQH$$ ‘ 3HDW PHFKDQLFDOO\ GLVSODFHG LQWR PRUDLQH FRUH EHQHDWK WLOO 0D[LPXPDJH IRU PRUDLQH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ :RRG VKHDUHG LQWR PRUDLQH FRUH EHQHDWK WLOO 0D[LPXP DJH IRU PRUDLQH8*$ ! :RRG LQ EHG PHFKDQLFDOO\ GLVSODFHG LQWR FRUH RI PRUDLQH 0D[LPXPDJH IRU PRUDLQH7 'LWWR H[FHSW GLIIHUHQW VDPSOH216 $ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK 0D[LPXP DJH IRU RXWZDVK 4/ ‘ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK EHQHDWK WLOO DW RXWHU PDUJLQ RIPRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH4/ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH4/ ! 'LWWR H[FHSW GLIIHUHQW VDPSOH4/ ! 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ! 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ! 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ! 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ 'LWWR H[FHSW GLIIHUHQW VDPSOH 4/ ‘ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK JUDGHG WR RXWHUPRVW /ODQTXLKXHPRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH4/ 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ! 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ /DUJH RUJDQLF FODVW ZLWK VWUDWLILHG RUJDQLF OD\HUV 6DPSOH IURP WRS RIFODVW &ODVW UHZRUNHG LQWR RXWZDVK 0D[LPXP DJH IRU RXWZDVK8*$ ! 6DPH DV 8*$ EXW IURP FHQWHU RI FODVW $$ ‘ &DQDO GH &KDQFKiQ :RRG IURP P GHSWK LQ P FRUH WR EDVH RIPLUH LQ &DQDO GH &KDQFKiQ &RUH ORFDWHG RQ FKDQQHO ERWWRP P ZHVWRI 5RXWH 98GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTSiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description$$ ‘ :RRG IURP P GHSWK LQ P FRUH WR EDVH RI PLUH LQ &DQDO GH&KDQFKiQ &RUH VLWH ZLWKLQ P RI $$$$ ‘ :RRG IURP P GHSWK LQ P FRUH WR EDVH RI PLUH LQ &DQDO GH&KDQFKiQ &RUH VLWH ZLWKLQ P RI $$$$ ‘ %DVDO J\WWMD IURP D P FRUH IURP D PLUH LQ &DQDO GH &KDQFKiQ &RUHVLWH ZLWKLQ P RI $$$$ ‘ %DVDO RUJDQLF PDWWHU IURP P FRUH ORFDWHG P QRUWK RI $$$$ ‘ %DVDO RUJDQLF PDWWHU IURP P FRUH ORFDWHG P QRUWK RI $$$$ ‘ %DVDO RUJDQLF PDWWHU IURP P FRUH ORFDWHG P QRUWK RI $$$$ ‘ %DVDO RUJDQLF PDWWHU IURP P FRUH ORFDWHG P QRUWK RI $$$$ ‘ %DVDO RUJDQLF PDWWHU IURP P FRUH LQ PLUH RQ IORRU RI &DQDO GH&KDQFKiQ P HDVW RI 5RXWH 98 $$ ‘ &DQDO GH OD 3XQWLOOD :RRG IURP P GHSWK LQ P FRUH WR EDVH RIPLUH LQ &DQDO GH OD 3XQWLOOD 6HH 0RUHQR HW DO IRU ORFDWLRQ DQGDGGLWLRQDO GDWHV%HWD ‘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‘ ¥ 2UJDQLF VLOW DQG ILQH VDQG IURP WRS RI EHG RI ZHDWKHUHG RUJDQLFS\URFODVWLF IORZ VHGLPHQWV EHWZHHQ WZR RXWZDVK XQLWV RQ WKH VRXWK VLGHRI 5RXWH 98 6DPSOH VLWH RQ HDVWHUQ HQG RI H[SRVXUH QHDU LFHFRQWDFW VORSH 3RUWHU P HDVW RI VDPSOH 4/4/ ‘ 2UJDQLF VLOW DQG ILQH VDQG IURP QHDU EDVH RI VDPH RUJDQLF OD\HU DQG VLWHDV 4/ 3RUWHU 4/ ‘ 2UJDQLF VLOW DQG ILQH VDQG IURP XSSHU FP RI D FPWKLFN RUJDQLF EHGRI ZHDWKHUHG S\URFODVWLF IORZ VHGLPHQWV EHWZHHQ WZR RXWZDVK XQLWV RQVRXWK VLGH RI 5RXWH 98 6DPSOH IURP QHDU ZHVWHUQ HQG RI VHFWLRQ8*$ 2UJDQLF VLOW DQG ILQH VDQG IURP XSSHU FP RI D FPWKLFN EHG RIRUJDQLF S\URFODVWLF IORZ VHGLPHQWV EHWZHHQ WZR RXWZDVK XQLWV RQ VRXWKVLGH RI 5RXWH 98 6DPSOH FROOHFWHG P HDVW RI VDPSOH 4/8*$ 'LWWR H[FHSW VDPSOH IURP FP GHSWK8*$ 'LWWR H[FHSW VDPSOH IURP FP GHSWK LQ RUJDQLF EHG8*$ 2UJDQLF VLOW DQG ILQG VDQG IURP XSSHU FP RI D FPWKLFN EHG RIRUJDQLF S\URFODVWLF IORZ VHGLPHQWV EHWZHHQ WZR RXWZDVK XQLWV RQ VRXWKVLGH RI 5RXWH 98 6DPSOH FROOHFWHG P HDVW RI VDPSOH 4/8*$ ‘ 'LWWR H[FHSW VDPSOH FROOHFWHG DW FP GHSWK LQ RUJDQLF EHG8*$ 'LWWR H[FHSW VDPSOH FROOHFWHG DW FP GHSWK LQ RUJDQLF EHG4/ ‘ 'LWWR H[FHSW VDPSOH FROOHFWHG QHDU EDVH RI RUJDQLF XQLW $$ ‘ )XQGR /LxD 3DQWDQRVD *\WWMD IURP FP GHSWK LQ D P FRUH LQPLUH LQ RXWHU /ODQTXLKXH PRUDLQH EHOW 0LPLPXP DJH IRU PRUDLQH 6HH+HXVVHU HW DO $$ ‘ 'LWWR H[FHSW VDPSOH FROOHFWHG DW P GHSWK$ 5HZRUNHG RUJDQLF FODVW LQ IROGHG RXWZDVK GLVWDO WR WKH RXWHUPRVW SRVLWLRQ UHDFKHG E\ ODWH /ODQTXLKXH SLHGPRQW LFH 0D[LPXP DJH IRURXWZDVK DQG IROGLQJGeograÞska Annaler á 81 A (1999) á 2217


G.H. DENTON ET AL.SiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description :N ‘ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK 0D[LPXP DJH IRU RXWZDVK :N ‘ )UXWLOODU $OWR 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK JUDGHG WR PD[LPXP/ODQTXLKXH PRUDLQH 6DPSOH VSOLW DQG GDWHG DW WZR ODERUDWRULHV0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH$ ‘ 6SOLW RI VDPH VDPSOH DV :N$ ‘ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK JUDGHG WR PD[LPXP /ODQTXLKXHPRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH4/ ‘ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK JUDGHG WR PD[LPXP /ODQTXLKXH‘ PRUDLQH 6DPSOH FRXQWHG WZLFH 0D[LPXP DJH IRU RXWZDVK DQGPRUDLQH4/ ‘ /DUJH RUJDQLF FODVW UHZRUNHG LQWR RXWZDVK JUDGHG WR PD[LPXP/ODQTXLKXH PRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH 6DPSOHWDNHQ IURP WRS RI FODVW4/ ‘ 6DPH FODVW DV 4/ VDPSOH IURP PLGGOH RI FODVW$ ‘ 6DPH FODVW DV 4/ VDPSOH IURP EDVH RI FODVW4/ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK JUDGHG WR PD[LPXP /ODQTXLKXHPRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ ‘ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK JUDGHG WR PD[LPXP /ODQTXLKXHPRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH 5/ ‘ ¥ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK EHQHDWK WLOO RI RXWHU /ODQTXLKXHPRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH )URP 0HUFHU 8: ‘ ¥ 'LWWR H[FHSW VDPSOH IURP 3RUWHU $5 2UJDQLF VDQG IURP P LQ P FRUH WKURXJK PLUH LQ FKDQQHOEHWZHHQ /ODQTXLKXH PRUDLQH ULGJHV 0LQLPXP DJH IRU PRUDLQHV $ ! 2UJDQLF VLOW OD\HU LQ ODFXVWULQH VHGLPHQWV LQ ODNHVLGH NDPH WHUUDFH $ )UXWLOODU %DMR 6PDOO SLHFHV RI ZRRG DQG ILEURXV RUJDQLF PDWWHU IURP ROGODQG VXUIDFH VHDOHG LQWDFW RQ WRS RI RUJDQLF EHG 6HH )LJV DQG IRUORFDWLRQ RI VDPSOHV$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH218GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTSiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description8*$ 2UJDQLF VLOW DQG ILQH VDQG IURP GHSWK RI FP LQ )UXWLOODU %DMR RUJDQLFEHG )LJV DQG 6DPSOH IURP VLWH RI SROOHQ SURILOH LQ +HXVVHU HWDO 8*$ 'LWWR H[FHSW VDPSOH IURP GHSWK RI FP8*$ ‘ 'LWWR H[FHSW VDPSOH IURP GHSWK RI FP8*$ ‘ 'LWWR H[FHSW VDPSOH IURP GHSWK RI FP DW EDVH RI RUJDQLF EHG8*$ ‘ 2UJDQLF VLOW DQG ILQH VDQG IURP GHSWK RI FP DW VLWH RI SROOHQ SURILOH LQ+HXVVHU HW DO 8*$ ‘ 'LWWR H[FHSW IURP GHSWK RI FP DW EDVH RI RUJDQLF EHG8*$ ‘ 2UJDQLF VLOW DQG ILQH VDQG IURP GHSWK RI FP DW EDVH RI RUJDQLF EHG P QRUWK RI SROOHQ VHFWLRQ LQ +HXVVHU HW DO $ ‘ 2UJDQLF FODVW IURP LFHFRQWDFW VWUDWLILHG GULIW DW WRS RI LFHFRQWDFW VORSH0D[LPXP DJH IRU GULIW 8*$ ‘ )XQGR /ODQTXLKXH *\WWMD IURP P GHSWK LQ P FRUH WKURXJKPLUH MXVW LQVLGH RXWHUPRVW /ODQTXLKXH PRUDLQH 6HH +HXVVHU HW DO8*$ ‘ 'LWWR H[FHSW VDPSOH IURP P GHSWK8*$ ‘ 'LWWR H[FHSW VDPSOH IURP P GHSWK8*$ ‘ 'LWWR H[FHSW VDPSOH IURP P GHSWK8*$ ‘ 'LWWR H[FHSW VDPSOH IURP P GHSWK8*$ ‘ 'LWWR H[FHSW VDPSOH ZDV FOD\H\ J\WWMD IURP P GHSWK $ ‘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‘ 7RZQ RI /ODQTXLKXH 6PDOO ZRRG SLHFH LQ RUJDQLF VLOW EHG 6HH )LJ IRU ORFDWLRQ RI VDPSOH(7+ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH(7+ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH(7+ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH(7+ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH(7+ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOHGeograÞska Annaler á 81 A (1999) á 2219


G.H. DENTON ET AL.SiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description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‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ! 2UJDQLF FODVW UHZRUNHG LQWR VOLJKWO\ ZHDWKHUHG RXWZDVK WKDW XQGHUOLHV WKHUHPRELOL]HG S\URFODVWLF IORZ VHGLPHQWV (7+ !‘ &URVVURDG VLWH 6PDOO ZRRG SLHFH VKHDUHG IURP VXUIDFH RI ZHDWKHUHGS\URFODVWLF IORZ VHGLPHQWV LQ 6HFWLRQ % RI )LJ (7+ !‘ 6PDOO ZRRG SLHFH LQ ZHDWKHUHG LQWHUGULIW S\URFODVWLF IORZ XQLWV LQ6HFWLRQ % RI )LJ (7+ !‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH(7+ !‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH(7+ !‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ! HVW 6PDOO ZRRG SLHFH IURP RUJDQLF KRUL]RQ LQ WRS FP RI LQWHUGULIWZHDWKHUHG S\URFODVWLF GULIW XQLWV RYHUODLQ E\ WLOO DW VHFWLRQ & LQ )LJXUH RQ HDVW VLGH RI 5RXWH $$ ‘ HVW 5DLOURDG EULGJH 3XHUWR 9DUDV )LEURXV SHDW IURP XSSHU RI WZR WKLQ SHDWOD\HUV WKDW RYHUOLH WKH FPWKLFN PDLQ RUJDQLF EHG LQ WKH UDLOURDGEULGJH VHFWLRQ 6WDQGDUG DFLGEDVHDFLG SUHWUHDWPHQW 6HH )LJ IRUORFDWLRQ$$% ‘ HVW 5HSHDW RI $$ ZLWK VWURQJ DFLGEDVHDFLG SUHWUHDWPHQW$$ ‘ HVW )LEURXV SHDW IURP ORZHU RI WZR WKLQ SHDW OD\HUV DERYH WKH PDLQ RUJDQLFEHG LQ WKH UDLOURDG EULGJH VHFWLRQ )LJ $$ ‘ HVW 6PDOO ZRRG SLHFH IURP ORZHU RI WZR WKLQ SHDW OD\HUV DERYH WKH PDLQRUJDQLF EHG LQ WKH UDLOURDG EULGJH VHFWLRQ 6WDQGDUG DFLGEDVHDFLGSUHWUHDWPHQW )LJ $$& ‘ HVW $ VSOLW RI VDPH VDPSOH DV $$ ZLWK DFLGEDVHDFLG SUHWUHDWPHQWDQG WKHQ H[WUDFWLRQ RI FHOOXORVH(7+ ‘ 6PDOO ZRRG SLHFH IURP ORZHU RI WZR WKLQ SHDW OD\HUV DERYH WKH PDLQRUJDQLF EHG LQ WKH UDLOURDG EULGJH VHFWLRQ )LJ (7+ ‘ 6DPH JUDSKLWH WDUJHW DV (7+ )LJ (7+ ‘ 6DPH VDPSOH DV (7+ DQG (7+ 5HGDWHG ZLWK VHSDUDWHSUHWUHDWPHQWV )LJ $$ ‘ 3HDW IURP ORZHU RI WZR WKLQ SHDW OD\HUV DERYH WKH PDLQ RUJDQLF EHG LQ WKHUDLOURDG EULGJH VHFWLRQ )LJ $$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ :RRG IURP XSSHU FP RI FPWKLFN PDLQ RUJDQLF EHG LQ WKH UDLOURDGEULGJH VHFWLRQ )LJ $ $ VSOLW RI VDPH VDPSOH DV $220GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTSiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description$ ‘ :RRG IURP XSSHU FP RI FPWKLFN PDLQ RUJDQLF EHG LQ WKH UDLOURDGEULGJH VHFWLRQ )LJ $ ‘ $ VSOLW RI VDPH VDPSOH DV $$ :RRG IURP XSSHU FP RI FPWKLFN PDLQ RUJDQLF EHG LQ WKH UDLOURDGEULGJH VHFWLRQ )LJ $ $ VSOLW RI VDPH VDPSOH DV $$ ‘ :RRG IURP XSSHU FP RI FPWKLFN PDLQ RUJDQLF EHG LQ WKH UDLOURDGEULGJH VHFWLRQ )LJ $ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH4/ ‘ ¥ :RRG IURP FPWKLFN PDLQ RUJDQLF EHG LQ WKH UDLOURDG EULGJH VHFWLRQ)URP 3RUWHU ZULWWHQ FRPPXQLFDWLRQ $$ ‘ 6PDOO ZRRG SLHFH FP EHORZ WRS RI RUJDQLF VLOW XQLW WKDW OLHV EHQHDWKWKH PWKLFN PDLQ RUJDQLF OD\HU )LJ $$ ‘ 'LWWR H[FHSW FP EHORZ WRS RI RUJDQLF VLOW XQLW$$ ‘ 'LWWR H[FHSW FP EHORZ WRS RI RUJDQLF VLOW XQLW$$ ‘ 'LWWR H[FHSW FP EHORZ WRS RI RUJDQLF VLOW XQLW$$ ‘ :RRG LQ YHUWLFDO SRVLWLRQ DW FP EHORZ WRS RI RUJDQLF VLOW XQLW 7KLVZRRG LV SDUW RI WKH UHPQDQWV RI D VPDOO WUHH LQ JURZWK SRVLWLRQ$$ ‘ :RRG IURP FP GHSWK IURP VDPH WUHH DV $$$ ‘ :RRG IURP FP GHSWK IURP VDPH WUHH DV VDPSOHV $$ DQG $$$ $ VSOLW RI VDPSOH $$$ ‘ 6PDOO ZRRG SLHFH FP EHORZ WRS RI RUJDQLF VLOW XQLW )LJ $$ ‘ 'LWWR H[FHSW FP EHORZ WRS RI RUJDQLF VLOW EHG$$ ‘ 'LWWR H[FHSW FP EHORZ WRS RI RUJDQLF VLOW EHG , ‘ ¥ &DOOH 6DQWD 5RVD 3XHUWR 9DUDV 8SSHUPRVW SDUW RI RUJDQLF EHG DW PHOHYDWLRQ VHDOHG LQWDFW E\ JODFLRODFXVWULQH VLOW XQLW 0HUFHU %\RXU LQWHUSUHWDWLRQ GDWHV SRQGLQJ RI ORFDO ODNH E\ SLHGPRQW JODFLHU LQ3XHUWR 9DUDV HPED\PHQW $ ‘ 1RUWKZHVW %OXII 3XHUWR 9DUDV 8SSHU FP RI D FPWKLFN EHG RIJ\WWMD VHDOHG DW WKH WRS E\ ODPLQDWHG JODFLRODFXVWULQH VHGLPHQWV DW PHOHYDWLRQ$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH *; ‘ ¥ %HOOD 9LVWD 3DUN 3XHUWR 9DUDV 3HDW IURP ZLWKLQ RUJDQLF EHG DW P HOHYDWLRQ 7RS RI RUJDQLF EHG VHDOHG E\ JODFLRODFXVWULQHVHGLPHQWV 3RUWHU $ ‘ %HOOD 9LVWD %OXII 3XHUWR 9DUDV :RRG LQ JODFLRODFXVWULQH VHGLPHQWV FP DERYH WKH WRS RI D SURPLQHQW FPWKLFN RUJDQLF EHG 7KH WRSVXUIDFH RI WKH RUJDQLF EHG LV DW P HOHYDWLRQ 6HH )LJ IRUVWUDWLJUDSK\ DQG VDPSOH ORFDWLRQ$ ‘ :RRG LQ JODFLRODFXVWULQH VHGLPHQWV FP DERYH WKH WRS RI D SURPLQHQWFPWKLFN RUJDQLF EHG 7KH WRS VXUIDFH RI WKH RUJDQLF EHG LV DW P HOHYDWLRQ )LJ 7$ ‘ 2UJDQLF VLOW IURP XSSHU FP RI WKH SURPLQHQW FPWKLFN RUJDQLF EHGDW P HOHYDWLRQ )LJ $5 ‘ 2UJDQLF WUDVK OD\HU RI VPDOO ZRRG SLHFHV DQG PDFURIRVVLOV VHDOHG E\JODFLRODFXVWULQH VHGLPHQWV RQ WKH VXUIDFH RI WKH SURPLQHQW FPWKLFNRUJDQLF EHG DW P HOHYDWLRQ )LJ $ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$5 ‘ 2UJDQLF VLOW IURP ZLWKLQ FPWKLFN RUJDQLF EHG )LJ GeograÞska Annaler á 81 A (1999) á 2221


G.H. DENTON ET AL.SiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description$5 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK WKDW SDVVHV EHQHDWK WLOO RI RXWHUPRVW/ODQTXLKXH PRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH:N ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH :N ‘ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK 0D[LPXP DJH IRU RXWZDVK:N ! 'LWWR H[FHSW GLIIHUHQW VDPSOH78D$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ ‘ 2UJDQLF FODVW UHZRUNHG LQWR JUDYHOO\ IORZ WLOO LQ GHIRUPHG FRUH RIPRUDLQH ULGJH 0D[LPXP DJH IRU PRUDLQH ULGJH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $5 ‘ 2UJDQLF VLOW IURP P GHSWK LQ P FRUH WKURXJK PLUH RQ/ODQTXLKXH PRUDLQH $5 2UJDQLF VLOW\ VDQG IURP P GHSWK LQ P FRUH WKURXJK PLUHLQ VPDOO EDVLQ RQ /ODQTXLKXH PRUDLQH $5 2UJDQLF VDQG IURP P GHSWK LQ P FRUH WKURXJK PLUH LQGHSUHVVLRQ EHWZHHQ /ODQTXLKXH PRUDLQH ULGJHV$5 2UJDQLF VDQG IURP P GHSWK LQ P FRUH WKURXJK PLUH LQGHSUHVVLRQ EHWZHHQ /ODQTXLKXH PRUDLQH ULGJHV$5 ‘ 2UJDQLF VDQG IURP P GHSWK LQ P FRUH WKURXJK PLUH LQGHSUHVVLRQ EHWZHHQ /ODQTXLKXH PRUDLQH ULGJHV $ ‘ *\WWMD IURP P GHSWK LQ P FRUH WKURXJK PLUH RQ/ODQTXLKXH PRUDLQH $5 ‘ 2UJDQLF ILEHUV IURP P GHSWK IURP P FRUH WKURXJK PLUH RQ/ODQTXLKXH PRUDLQH $ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK 0D[LPXP DJH IRU RXWZDVK4/ ! ¥ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ ! 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK EHQHDWK RXWHU HGJH RI /ODQTXLKXHPRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH222GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTSiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ 'LWWR H[FHSW GLIIHUHQW VDPSOH :N ‘ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK 0D[LPXP DJH IRU RXWZDVK:N ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH :N ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH4/ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ &DQDO 7HQJOR )LEURXV SHDW DQG ZRRG IURP WRS VXUIDFH RI SHDW OD\HU6HH )LJ IRU VDPSOH ORFDWLRQ DQG VLWH VWUDWLJUDSK\7$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 3HDW IURP EDVH RI SHDW OD\HU 6HH )LJ IRU VDPSOH ORFDWLRQ DQG VLWHVWUDWLJUDSK\$ )LEURXV SHDW DQG ZRRG IURP WRS VXUIDFH RI SHDW OD\HU 6HH )LJ IRUVDPSOH ORFDWLRQ DQG VLWH VWUDWLJUDSK\$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 5HPDLQV RI WUHH LQ JURZWK SRVLWLRQ RQ WRS VXUIDFH RI SHDW OD\HU )LJ $ )LEURXV SHDW DQG ZRRG IURP WRS VXUIDFH RI SHDW OD\HU )LJ 7$ ‘ 3XQWD 3HQDV *\WWMD IURP WRS RI FPWKLFN RUJDQLF EHG 6HH )LJ IRU VDPSOH ORFDWLRQ DQG VLWH VWUDWLJUDSK\7$ ‘ 2UJDQLF VLOW IURP PLGGOH RI FPWKLFN RUJDQLF EHG )LJ 7$ ‘ 2UJDQLF VLOW IURP EDVH RI FPWKLFN RUJDQLF EHG )LJ $ ‘ 6XUIDFH RUJDQLF WUDVK OD\HU LQFOXGLQJ OHDYHV DQG ZRRG IURP WKH ORZHU RIWKH WZR SURPLQHQW EXW WKLQ RUJDQLF OD\HUV KLJKHU LQ WKH VHFWLRQ WKDQ WKHFPWKLFN RUJDQLF OD\HU )LJ $ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOHGeograÞska Annaler á 81 A (1999) á 2223


G.H. DENTON ET AL.SiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description$$ ‘ 6XUIDFH RUJDQLF WUDVK OD\HU LQFOXGLQJ OHDYHV DQG ZRRG IURP WKH KLJKHVWRI WKH WZR SURPLQHQW EXW WKLQ RUJDQLF OD\HUV KLJKHU LQ WKH VHFWLRQ WKDQ WKHFPWKLFN RUJDQLF OD\HU )LJ $ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ 3XQWD 3HOOXFR *\WWMD IURP RUJDQLF OD\HU LQ VHFWLRQ DW 3XQWD 3HOOXFR6HH )LJ IRU VDPSOH ORFDWLRQ DQG VLWH VWUDWLJUDSK\$ *\WWMD IURP RUJDQLF OD\HU D VDPH VWUDWLJUDSKLF SRVLWLRQ DV VDPSOH $ EXW LQ JXOIVLGH FOLII LQ HPED\PHQW P ZHVW RI 3XQWD 3HOOXFR$ 'LWWR H[FHSW GLIIHUHQW VDPSOH $$ ‘ )XQGR 6DQWD (OHQD *\WWMD IURP P GHSWK LQ P FRUH WKURXJKPLUH RQ /ODQTXLKXH PRUDLQH EHOW :N ‘ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK 0D[LPXP DJH IRU RXWZDVK$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ! 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH78D$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ! 'LWWR H[FHSW GLIIHUHQW VDPSOH :N ! 'LWWR H[FHSW GLIIHUHQW VDPSOH:N ! 'LWWR H[FHSW GLIIHUHQW VDPSOH7$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ 'LWWR H[FHSW GLIIHUHQW VDPSOH $$ ‘ 2UJDQLF FODVW UHZRUNHG LQWR JUDYHOO\ RXWZDVK EHQHDWK RXWHU HGJH RI/ODQTXLKXH PRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH78D$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK 0D[LPXP DJH IRU RXWZDVK7$ ‘ 5HGDWLQJ RI VDPH FODVW DV $ $ 2UJDQLF VLOW IURP P GHSWK LQ P FRUH WKURXJK PLUH RQ/ODQTXLKXH PRUDLQH EHOW$$ ‘ 5HSOLFDWH RI $ $ 2UJDQLF VLOW IURP P GHSWK LQ P FRUH WKURXJK PLUH RQ/ODQTXLKXH PRUDLQH EHOW$$ ‘ 5HSOLFDWH RI $ $ *\WWMD IURP P GHSWK LQ P FRUH WKURXJK PLUH RQ/ODQTXLKXH PRUDLQHV $ 2UJDQLF VLOW IURP P GHSWK LQ P FRUH WKURXJK PLUH RQ/ODQTXLKXH PRUDLQHV $$ ‘ 2UJDQLF VLOW IURP P GHSWK LQ P FRUH WKURXJK PLUH LQ/ODQTXLKXH PRUDLQHV $ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK 0D[LPXP DJH IRU RXWZDVK$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH224GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTSiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description$ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ ‘ 2UJDQLF ILEHUV IURP P GHSWK LQ P FRUH WR EDVH RI PLUH LQFKDQQHOV RQ RXWHU /ODQTXLKXH PRUDLQH EHOW $ 5 2UJDQLF PDWUL[ RI JUDYHO IURP P GHSWK LQ P FRUH WR EDVHRI PLUH RQ RXWHU /ODQTXLKXH PRUDLQH EHOW $$ ‘ %DVDO J\WWMD IURP P GHSWK LQ FRUH WR EDVH RI PLUH RQ RXWHU/ODQTXLKXH PRUDLQH EHOW $$ ‘ /D &DPSDxD *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQRXWHUPRVW /ODQTXLKXH PRUDLQH EHOW 6HH +HXVVHU HW DO $$ ‘ *\WWMD IURP P GHSWK LQ VDPH FRUH DV $$$ 2UJDQLF VLOW IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ RXWHUPRVW /ODQTXLKXH PRUDLQH EHOW $$ ‘ *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ /ODQTXLKXHPRUDLQH EHOW $$ ‘ *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ /ODQTXLKXHPRUDLQH EHOW :N ‘ 2UJDQLF FODVW UHZRUNHG LQWR RXWZDVK EHQHDWK RXWHU HGJH RI /ODQTXLKXHPRUDLQH 0D[LPXP DJH IRU RXWZDVK DQG PRUDLQH :N ! 5HZRUNHG RUJDQLF FODVW LQ RXWZDVK 0D[LPXP DJH IRU RXWZDVK $ ‘ ,VOD 0DLOOHQ *\WWMD IURP WRS RI RUJDQLF OD\HU VHDOHG E\ JODFLRODFXVWULQHVHGLPHQWV WKDW DUH RYHUODLQ E\ WLOO &ORVH PD[LPXP DJH IRU LFH DGYDQFHRYHU VLWH $$ ‘ *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ /ODQTXLKXHPRUDLQH $$ ‘ *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ /ODQTXLKXHPRUDLQH $$ ‘ *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ /ODQTXLKXHPRUDLQH $ 2UJDQLF VLOW IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ/ODQTXLKXH PRUDLQH $ *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ/ODQTXLKXH PRUDLQH$$ ‘ %DVDO J\WWMD IURP UHFRULQJ RI VDPH PLUH P IURP VLWH RI $GeograÞska Annaler á 81 A (1999) á 2225


G.H. DENTON ET AL.SiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description $ ‘ )LEURXV RUJDQLF PDWHULDO LQ RUJDQLF VDQG DW P GHSWK LQ PFRUH WR EDVH RI PLUH RQ /ODQTXLKXH PRUDLQH $ ‘ )LEURXV RUJDQLF PDWHULDO DW P GHSWK LQ P FRUH WR EDVH RIPLUH RQ /ODQTXLKXH PRUDLQH $ +XHOPR *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUHLQ WKH FRDVWDO /ODQTXLKXH PRUDLQH EHOW &RUH WDNHQ QHDU HGJH RI PLUH$$ ‘ *\WWMD IURP EDVH RI P FRUH WDNHQ EHVLGH FRUH RI VDPSOH $$$ ‘ *\WWMD IURP P GHSWK LQ P FRUH LQ WKH FHQWHU RI WKH +XHOPRPLUH &RUH WR EDVH RI PLUH 7KLV GDWH LV SDUW RI D FRQVLVWHQW VHTXHQFH RIDJHV LQ WKLV FRUH 0RUHQR $$ ‘ *\WWMD IURP P GHSWK LQ WKH VDPH P FRUH DV $$ 7KLVGDWH LV SDUW RI D FRQVLVWHQW VHTXHQFH RI DJHV LQ WKLV FRUH 0RUHQR $ ‘ *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUH LQ FRDVWDO/ODQTXLKXH PRUDLQH EHOW $ *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUH LQ/ODQTXLKXH PRUDLQH EHOW $ ‘ /DJR &RQGRULWR *\WWMD IURP P GHSWK LQ P FRUH WR EDVH RIODFXVWULQH VHGLPHQWV LQ /DJR &RQGRULWR $ ‘ )LEURXV PDWHULDO LQ RUJDQLF VDQG IURP P GHSWK LQ P FRUHWR EDVH RI PLUH RQ /ODQTXLKXH PRUDLQH EHOW $ :RRG DQG ILEURXV SHDW IURP XSSHU VXUIDFH RI RUJDQLF EHG VHDOHG E\RXWZDVK WKDW LV RYHUODLQ E\ /ODQTXLKXH PRUDLQH EHOW 7KH RUJDQLF EHG LVDERXW P ORQJ DQG H[SRVHG LQ VHD FOLII NP VRXWK RI +XHOPR 'DWHLV FORVH PD[LPXP IRU JODFLHU DGYDQFH RYHU VLWH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ ‘ *\WWMD IURP P GHSWK LQ P FRUH WKURXJK PLUH RQ /ODQTXLKXHPRUDLQH$ 5 5HGDWLQJ RI VDPSOH $$$ ‘ %DVDO J\WWMD IURP P FRUH WDNHQ EHVLGH FRUH RI $$$ ‘ %DVDO J\WWMD IURP P FRUH WDNHQ EHVLGH FRUH RI $$$$ ‘ 6PDOO ZRRG SLHFH IURP ZLWKLQ J\WWMD DW EDVH RI P FRUH RI $$ 7KLV ZRRG LV UHWUDQVSRUWHG IURP QXPHURXV IRVVLO ZRRG SLHFHV LQ6HQR 5HORQFDYt EDVLQ EHFDXVH LW LV HQFORVHG ZLWKLQ J\WWMD GDWHG WR & \U %3 $ 2UJDQLF VDQG IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ/ODQTXLKXH PRUDLQH EHOW$ 5 2UJDQLF VDQG IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ/ODQTXLKXH PRUDLQH EHOW $$ ‘ %DVDO J\WWMD IURP FRUH WR EDVH RI PLUH RQ /ODQTXLKXH PRUDLQH $ 5 2UJDQLF IUDJPHQWV LQ VDQG PDWUL[ DW P GHSWK LQ P FRUH WREDVH RI PLUH RQ /ODQTXLKXH PRUDLQH $ 5 2UJDQLFULFK VDQG IURP P GHSWK LQ P FRUH WR EDVH RI PLUHRQ /ODQTXLKXH PRUDLQH $ 5 ‘ 3HDW IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ/ODQTXLKXH PRUDLQH226GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTSiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description $ 5 ‘ 2UJDQLF VLOW IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ/ODQTXLKXH PRUDLQH$ ‘ 2UJDQLF VLOW IURP P GHSWK LQ P FRUH WKURXJK PLUH RQ/ODQTXLKXH PRUDLQH$ 2UJDQLF VLOW IURP P GHSWK LQ P FRUH WKURXJK PLUH RQ/ODQTXLKXH PRUDLQH $ 1HDU &DOEXFR 2UJDQLF FODVW UHZRUNHG LQWR FRDUVH IRUHVHW EHGV RI DSURJODFLDO ODFXVWULQH GHOWD WKDW LV FDSSHG E\ ORGJHPHQW WLOO 0D[LPXPDJH IRU JODFLDO DGYDQFH DFURVV VLWH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $ ‘ 6KHOOV LQ PDULQH VLOW DQG FOD\ EHG PHFKDQLFDOO\ GLVORFDWHG LQWR WKH FRUHRI D PRUDLQH DQG QRZ P DERYH VHD OHYHO $$ ! 7DLTXHPy :RRG\ SHDW IURP P GHSWK LQ P FRUH WR EDVH RI7DLTXHPy PLUH RQ RXWHUPRVW /ODQTXLKXH PRUDLQH 7KLV GDWH LV WKH ROGHVWRI D VHULHV RI FRQVLVWHQW GDWHV LQ WKLV FRUH +HXVVHU HW DO $$ ‘ 7HJXDFR 0DFURIRVVLOV IURP RUJDQLF WUDVK OD\HU RQ VXUIDFH RI FPWKLFN J\WWMDDQGRUJDQLF VLOW EHG VHDOHG ZLWK JODFLRODFXVWULQH XQLW 'DWHVIORRGLQJ RUJDQLF EHG E\ SURJODFLDO ODNH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ :RRG SLHFH HQFDVHG LQ ODPLQDWHG JODFLRODFXVWULQH VLOW WKDW UHVWV RQRUJDQLF EHG DQG VHDOV WKH VXUIDFH RUJDQLF WUDVK OD\HU 7KH ZRRGFROOHFWHG ZLWKLQ FP RI WKH RUJDQLF WUDVK OD\HU 7KH ZRRG ZDVSUREDEO\ ZDVKHG LQWR WKH SURJODFLDO ODNH DOWKRXJK LW PD\ KDYH EHHQIORDWHG IURP WKH ODQG VXUIDFH E\ ULVLQJ ODNH ZDWHUV 'DWHV IORRGLQJ E\SURJODFLDO ODNH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ :RRG SLHFH HQFDVHG LQ PDVVLYH ODFXVWULQH VLOW DQG ILQH VDQG WKDW RYHUOLHVWKH ODPLQDWHG JODFLRODFXVWULQH VLOW 7KH ZRRG ZDV SUREDEO\ ZDVKHG LQWRWKH ODNH4/ ‘ *\WWMD IURP XSSHU FP RI FPWKLFN VLOW\ J\WWMD EHG VHDOHG E\JODFLRODFXVWULQH VHGLPHQWV$ 2UJDQLF VLOW IURP ORZHUPRVW FP RI FPWKLFN RUJDQLF VLOW DQG J\WWMDEHG$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH4/ ‘ 2UJDQLF VDQG IURP WKH ORZHUPRVW FP RI D FPWKLFN RUJDQLF VDQGXQLW WKDW XQGHUOLHV WKH SURPLQHQW FPWKLFN VLOW\J\WWMD RUJDQLF EHG $$ ‘ 'DOFDKXH :RRG SLHFHV XS WR FP ORQJ O\LQJ RQ XSSHU VXUIDFH RIRUJDQLF OD\HU KHUH FP WKLFN WKDW ILOOV D VPDOO EDVLQ RQ D WLOO VXUIDFH7KHVH ZRRG SLHFHV DQG DVVRFLDWHG SODQW ILEHUV UHSUHVHQW DQ LQWDFW ROGODQG VXUIDFH SUHVHUYHG EHQHDWK JODFLDO GHSRVLWV WKDW LQILOO WKH EDVLQ 7KHFROOHFWLRQ VLWHV DUH DOO ZLWKLQ PRIORFDWLRQ $ LQ )LJ GeograÞska Annaler á 81 A (1999) á 2227


G.H. DENTON ET AL.SiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ /DUJH ZRRG SLHFH FP RQ ROG ODQG VXUIDFH RQ VXUIDFH RI 'DOFDKXHRUJDQLF EHG DW ORFDWLRQ $8*$ ‘ 6DPH ZRRG SLHFH DV 8*$8*$ ‘ 6DPH ZRRG SLHFH DV 8*$8*$ ‘ 2UJDQLF VLOW DQG ILQH VDQG HQFDVLQJ ZRRG RI 8*$8*$ ‘ 8SSHUPRVW FP RI RUJDQLF VLOW IURP 'DOFDKXH RUJDQLF EHG P VRXWK RIORFDWLRQ $ )LJ 8*$ ‘ 2UJDQLF VLOW IURP FP GHSWK LQ 'DOFDKXH RUJDQLF EHG P VRXWK RIORFDWLRQ $ )LJ 8*$ ‘ 2UJDQLF VLOW IURP FP GHSWK LQ 'DOFDKXH RUJDQLF EHG P VRXWK RIORFDWLRQ $ )LJ 8*$ ‘‘2UJDQLF VLOW IURP FP GHSWK LQ 'DOFDKXH RUJDQLF EHG P VRXWK RIORFDWLRQ $ )LJ &RXQWHG WZLFH8*$ ‘ 2UJDQLF VLOW IURP FP DW EDVH RI 'DOFDKXH RUJDQLF EHG P VRXWK RIORFDWLRQ $ )LJ 8*$ ‘ 2UJDQLF VLOW IURP FP DW EDVH RI 'DOFDKXH RUJDQLF EHG P VRXWK RIORFDWLRQ $ )LJ 8*$ ‘ 2UJDQLF VLOW IURP FP DW EDVH RI 'DOFDKXH RUJDQLF EHG P VRXWKORFDWLRQ $ )LJ 8*$ ‘ 2UJDQLF VLOW IURP FP GHSWK DW ORFDWLRQ $ LQ 'DOFDKXH RUJDQLF EHG )LJ 6HH +HXVVHU HW DO 8*$ ‘ 2UJDQLF VLOW IURP FP GHSWK DW ORFDWLRQ $ LQ 'DOFDKXH RUJDQLF EHG)LJ 6HH +HXVVHU HW DO 8*$ ‘ 2UJDQLF VLOW IURP FP GHSWK DW ORFDWLRQ $ LQ 'DOFDKXH RUJDQLF EHG)LJ 6HH +HXVVHU HW DO 8*$ ‘ 2UJDQLF VLOW IURP FP GHSWK DW ORFDWLRQ $ LQ 'DOFDKXH RUJDQLF EHG)LJ 6HH +HXVVHU HW DO 8*$ ‘ 2UJDQLF VLOW IURP FP GHSWK DW ORFDWLRQ $ LQ 'DOFDKXH RUJDQLF EHG)LJ 6HH +HXVVHU HW DO $ ‘ 2UJDQLF VLOW IURP XSSHUPRVW FP RI 'DOFDKXH RUJDQLF EHG DW ORFDWLRQ %LQ )LJXUH 6HH +HXVVHU HW DO $ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 2UJDQLF VLOW IURP FP GHSWK LQ WKH 'DOFDKXH RUJDQLF EHG DW ORFDWLRQ% )LJ 228GeograÞska Annaler á 81 A (1999) á 2


LLANQUIHUE DRIFTSiteNo.Lab. No.Age14 C yr BPd 13 C(ä)Description8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH &ROOHFWHG WR FKHFN FRQVLVWHQF\ RI GDWHVRI RUJDQLF VLOW VDPSOHV FROOHFWHG DW VDPH GHSWK8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$ ‘ 2UJDQLF VLOW IURP FP GHSWK DW WKH EDVH RI WKH 'DOFDKXH RUJDQLF EHGDW ORFDWLRQ % )LJ $ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH$$ ‘ 2UJDQLF VLOW IURP FP GHSWK LQ WKH 'DOFDKXH RUJDQLF EHG DW ORFDWLRQ % )LJ 7KLV VDPSOH DQG DOO WKH RWKHUV WDNHQ EHORZ DUH IURP D VHFWLRQ FXW EDFN P IURP WKH RULJLQDO IDFH IURP ZKLFK DOO WKH VDPSOHV OLVWHG DERYH IURPORFDWLRQ % ZHUH FROOHFWHG +HUH WKH RUJDQLF EHG LV FP WKLFN$$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ 5 ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ 5 ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ 5 ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$ 'LWWR H[FHSW VDPSOH IURP FP GHSWK$$ ‘ 'LWWR H[FHSW VDPSOH IURP FP GHSWK $5 D 2UJDQLF VLOW ZLWK PDFURIRVVLOV IURP P GHSWK LQ P FRUHWKURXJK PLUH LQ NHWWOH RQ XSSHU OLS RI LFHFRQWDFW VORSH 0LQLPXP DJHIRU LFH UHFHVVLRQ IURP PRUDLQH RQ WRS RI LFHFRQWDFW VORSH4/ E 2UJDQLF FODVW UHZRUNHG LQWR LFHFRQWDFW VWUDWLILHG GULIW RQ XSSHU OLS RILFHFRQWDFW VORSH 0D[LPXP DJH IRU LFH DGYDQFH WR WRS RI LFHFRQWDFWVORSH4/ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH8*$ ‘ 'LWWR H[FHSW VDPSOH VSOLW DQG GDWHG WZLFH‘ 8*$ ‘ 'LWWR H[FHSW GLIIHUHQW VDPSOH $$ ‘ 0D\RO &OD\ J\WWMD IURP P GHSWK LQ P FRUH WR EDVH RI PLUH RQ/ODQTXLKXH PRUDLQH +HXVVHU HW DO 78D$ ‘ (VWHUR +XLWDQTXH 2UJDQLF PDWWHU IURP QHDU EDVH RI FRUH WKURXJK PLUHRQ /ODQTXLKXH PRUDLQH +HXVVHU HW DO %HWD ‘ ¥ 'LWWR H[FHSW GLIIHUHQW VDPSOH 7 $ ‘ 3HDW IURP EDVH RI P SHDW GHSRVLW LQ WURXJK EHWZHHQ WZR /ODQTXLKXHDJH WLOO IOXWHV /RZHOO HW DO %HWD ‘ ¥ %DVDO GDWH RI P FRUH DW 3XHUWR &iUPHQ 9LOODJUiQ %HWD ‘ ¥ 'DWH RI SHDW IURP DERXW P GHSWK LQ D P FRUH DW /DJXQD &KDLJXDWD9LOODJUiQ 5/ ‘ ¥ &XHVWR 0RUHQJD 3HDW RYHU GULIW 0LQLPXP GDWH IRU GHJODFLDWLRQ RI VLWH+HXVVHU GeograÞska Annaler á 81 A (1999) á 2229

More magazines by this user
Similar magazines