Hawaiian Quaternary Paleoenvironments: a Review of Geological, Pedological, and Botanical Evidence1

Hawaiian Quaternary Paleoenvironments: A Review of Geological, Pedological, and Botanical Evidence1

ROBERT T. GAVENDA 2

ABSTRACT: Climates in Hawaii during glacial periods were relatively wetter and cooler than interglacial climates . Eolian deposits indicate that northeasterly trade winds predominated during glacial periods. Orographic rainfall patterns were probably similar to those oftoday except that they were shifted downward in response to lowered sea levels and a depressed inversion level. Botanical evidence indicates that some areas probably received more than double their current annual rainfall. Greater rainfall during glacial periods was probably responsible for the formation ofhighly weathered soils that are now in semiarid climates. More intense periglacial processes may have operated during glacial periods. Snowline on Mauna Kea was depressed about 900 m and glaciation may have occurred because of lower air temperature and greater cloudiness. Ocean temperature was probably also slightly cooler. At low elevations, interglacial climates were drier than glacial climates because of the influence higher sea levels had on orographic rainfall distribution. Trade winds still predominated but the inversion level was higher, which may have caused greater rainfall at high elevations. Pedological evidence indicates a highly erosive environment before the formation of the Kaena shoreline at about 650,000 yr ago. Climatic conditions at that time are not known. Subsequent environmental conditions have not been as conducive to erosion, and the past several hundred thousand years have witnessed relative landscape stability.

THEEVIDENCEFOR former climatic regimes on or can be explained by conditions other than a worldwide scale is so compelling that Wright former climatic regimes. and Frey (1965) have considered climatic Very little research has been conducted with change to be the "dominating circumstance" the intent of elucidating former climatic re ofthe Quaternary. Although paleoclimates of gimes in Hawaii. The intent ofthis paper is to the middle and high latitudes and continental summarize the geological, pedological, and glaciation have been studied extensively, only botanical research conducted in Hawaii that recently have the various impacts of paleo sheds light on former climatic regimes. This is climates in the tropics received much atten not an in-depth review of each discipline but tion (Street 1981). an attempt to pull together widely scattered The evidence of paleoenvironments in Ha information and to integrate the information waii includes geologic deposits and landforms, on paleoclimatic regimes to the extent that relict soils, and fossil plant remains. Some the evidence permits. evidence indicates the direction and/or magni tude of climate change. Other geologic and Present Hawaiian Climate pedologic features have poor time constraints Price (1983) characterized Hawaii's climate as the interaction of latitude, the surrounding I Manu script accepted 27 August 1991. ocean , the islands' terrain, and Hawaii's loca 2 Department of Agron omy and Soil Science, Univer sity of Hawai i at Manoa, 1910 East-West Road, Hono tion relative to the storm tracks and the Pacific lulu, Hawaii 96822. anticyclone. These factors produce a mild 295 296 PACIFIC SCIENCE, Volume 46, July 1992 climate with moderate humidity at low eleva tion patterns, winter precipitation is due as tions, persistent northeasterly trade winds, much to the increase and intensity of trade infrequent severe storms, and considerable wind rain as it is to the increased frequency of variation in rainfall and temperature as in cyclonic rain (Yeh et al. 195Ib). fluenced by topography. Using a modified In summary, the present Hawaii climate Thornthwaite system, Jones and Bellaire is characterized by an orographic-convective (1937) identified a total of 15 climatic regimes cell that produces localized, intensive rainfall on Kauai, Oahu, Maui, and Hawaii. All centers from trade winds near mountain sum rainfa ll types except E (arid) and all tempera mits that serve as focal points for rainfall ture types except F' (perpetual frost) are activity regardless of wind direction (Ruhe found on the island of Hawaii. 1964). A nearly uniform daylength throughout the year is one factor that causes low seasonal Geological Evidence ofPaleoclimates variation in air temperature (Price 1983). In addition to supplying moisture to the air, the SEA LEVEL FLU CTUATIONS. Former climatic ocean moderates air temperatures in Hawaii. regimes have had the greatest impact on The seasonal extremes in sea surface tempera Hawaiian geology through glacio-eustatic ture range from 22°C in March to 26°C in fluctuations of sea level. These fluctuations October. The extremes in average monthly were the result ofglobal climatic changes that means for air temperatures coincide with the caused water either to be removed from the extremes in sea surface temperature, indicat oceans and tied up in advancing glaciers or to ing the overriding importance of the ocean in be released into the oceans as retreating gla controlling air temperatures. ciers melted . Sea level fluctuations resulting During the summer months the Pacific from tectonic activity are without climatic anticyclone dominates and produces the per significance other than how raising or lower sistent northeasterly trade winds (Price 1983). ing a landmass may influence weather pat As these winds rise in response to the topogra terns. phy, they cool and the moisture they carry Numerous papers have documented emerg precipitates, producing orographic rainfall ed and submerged reef deposits in the Hawai patterns where rainfall increases rapidly from ian Islands (Stearns and Vaksvik 1935, Ste the coastlines to the mountaintops. On the arns 1935, 1961 , 1974, 1978, Ruhe et al. 1965a, low islands (below 1500 m elevation), the Coulbourn et al. 1974). Stearns (1978) sum maximum rainfall occurs immediately to the marized the various relict shorelines in the lee of the mountain crests. The temperature Hawaiian Islands (Figure I). The shorelines inversion, which characterizes the trade wind range in elevation from about - 1100 to + 365 periods, limits the highest convective clouds m. Information on the principal shorelines is to about 2100 m. Mountains between 1500 presented in Table I. and 2100 m elevation receive maximum rain fall near their summits. The mountains that rise above the inversion receive their maxi TABLE I mum precipitation between 900 and 1300 m, whereas the mountaintops are semiarid. S UMMARY OF ELEVATIONS AND AGES OF TH E PRINCIPAL Isohyetal rainfall maps of Hawaii have been SHORELINES IN THE HAWAIIAN ISLA NDS compiled by Giambelluca et al. (1986). SHORELINE ELEVATION (m) AGE* Similar rainfall patterns occur in the winter , but cyclonic sto rm rainfall occurs when the Mamala -106 12 southerly portions of frontal systems pass Waimanalo + 7.5 125 over Hawaii (Leopold 1951).There is a mark Waipio -106 Illinoian ed correlation between winter precipitation Kaena + 30 650 and the latitude and speed of the jet stream Source: Stearns 1978. (Yeh et al. 195Ia) . Under the current circula- • Thousands oryears ago. Hawaiian Quaternary Paleoenvironments- GAVENDA 297

INTERGL ACIALSTAGES:_ AFTONIAN YARMOUTH SANG AM ON

Fe e" '00' OSC/L iAnNa ..,) 0 - Anc ient Snoretin« in Howaii SEA LEvE"( \ H - H Olocene 20 0 ' ,, 10 0' ,

0 ' "'\ wIDO· ;.,

-200'

-300'

-400' \\, I '7/

GLACI AL STA GES: NEBRASKAN KAN SAN IL LI NO IAN WIS CON SI N A N

r-- -?------PLE/STOCENE------*

FIGURE I. Graphic representation of glacio-eustatic sea level Iluctuations and shorelines as proposed by Stearns (from Stearns 1978, fig. I, reprinted by permission of Bishop Museum Press).

Glacio-eustatic level could be lowered by ofalluvial deposits that were graded to a high about 130 m during a glacial maximum and sea level (probably the Kaena + 30 m) during could be raised about 43 m if all existing an interglacial period and were subsequently glacial ice melted (Donn et al. 1962, Flint eroded when sea level lowered . Waimea Val 1971). The alleged shorelines above 43 mare ley on Oahu is an example of a stream that problematical because they lie above the theo was incised during a low stand of the sea retical sea level during an intense interglacial (probably - 100 m) and was then filled with period. Because subsidence is the dominant sediment as the sea rose to its present level. tectonic activity in Hawaii (Moore 1987), it The lochs of Pearl Harbor are stream valleys seems unlikely that tectonic uplift is responsi that were cut when sea level was lower, then ble for shorelines above 43 m. Very large drowned as sea level rose (Stearns and Vak tsunamis due to debris avalanches could leave svik 1935). Subsidence of the Pearl Harbor deposits that could be mistaken for shorelines area was responsible for some changes in base (Moore and Moore 1984, Lipman et al. 1988, level (Macdonald et al. 1983), but at least the Moore et al. 1989). Ku et al. (1974) thought last major cycle of alluviation and erosio n was that shorelines between 7.6 m and current sea probably caused by glacio-eustatic sea level level could have been formed by storm waves changes. or weathering unrelated to glacio-eustatic sea Regardless of the cause, sea level fluctua levels. The sequence ofsubmarine terraces off tions around Hawaii have resulted in the Oahu is so complicated that Coulbourn et al. relative raising or lowering of the existing (1974) expressed doubts that the history of topography relative to the orographic-con Hawaiian Quaternary sea levels can be un vection cells produced by the trade winds raveled. The picture is complicated by tec (Ruhe 1964, 1965, 1975). Equations that quan tonic activity , which can alter the position of titatively describe precipitation relative to ele reefs (Moore and Fornari 1984). vation and distance from the summit on the Stream terraces along Kalalau Stream, leeward Koolau Range have been derived Kauai, and lao Strea m, Maui, are examples from meteorological statistics that character- 298 PACIFIC SCIENCE, Volume 46, July 1992

MAMALA SEA STAND OAHU. HAWAII 25 - 50% > present rainfall

50 - 100% > present rainfall

> 100% > present rainfall 10 kilometers I

. N 20' 40'

FIGURE 2. Estimated rainfall relative to present rainfall during the Mamala, - 106 rn, sea stand , Oahu, Hawaii (modified from Ruhe 1964, fig. 3. Reprinted by permission of American Journ al of Science).

ize the present climate on Oahu (Ruhe 1964). cate that they formed during a glacial low These equations predict that a lowering ofsea stand of the sea and that northeasterly trade level by 100m could have produced more than winds predominated at that time (Stearns a 100% increase in rainfall (relative to present 1940, 1947, Stearns and Macdonald 1942, rainfall) in parts of the Wahiawa Basin (cen 1947, Macdonald et al. 1960). tral Oahu) (Figure 2). A 30-m rise in sea level The persistence ofnortheasterly trade winds may have caused a 50% decrease in rainfall in during glacial periods is also shown by the some areas (Figure 3). asymmetrical morphology of pyroclastic cones on Oahu. The southwestern rim of most WIND DIRECTION DURING GLACIAL PERIODS . of these cones is higher than the northeastern The main assumption underlying Ruhe's rim because trade winds carried the pyro (1964) paleoclimate estimates is that the gen clastics toward the southwest (Wentworth eral circulation pattern during both glacial 1926). Ulupau Head (Kahipa stand) and and interglacial periods basically resembled Diamond Head (Waipio stand) (Winchell that oftoday. That is, precipitation related to 1947) are two prominent examples of asym trade winds is responsible for most rainfall. metrical cones deposited during low sea levels. The orientation of lithified sand dunes and Orientation of sand dunes at the Mauna their occurrence below current sea level indi- Kea-Mauna Loa saddle, and tephra plumes Hawaiian Quaternary Paleoenvironments-i-GxVENDA 299

KAENA SEA STAND OAHU , HAWAII IZI 50% of present rainfall

~ 50-75% of present rainfall

EJ > 75% of present rainfall 10 kilometers I I

o N

o N 20' 40'

FI GURE 3. Estimated rainfall relative to present rainfall during the Kaena , +30 m, sea stand, Oahu , Hawaii (modified from Ruhe 1964, fig. 4. Reprinted by permission of American Journal of Science). on Mauna Kea also indicate a northeasterly The CLIMAP Project members (1976) con trade wind pattern during the middle of the cluded that during the last glacial maximum last glacial period (Porter 1979). polar frontal systems were displaced toward Molina-Cruz (1977) interpreted abun the equator, and the central gyre of the sub dances of pelagic quartz, opal, and radiolarian tropical Pacific maintained nearly stable posi assemblages as indicating increased trade tion and temperature. A computer simulation wind velocity in the equatorial Pacific during model of a tropical climate during a glacial glacial periods. The grain size ofpelagic eolian period (Manabe and Hahn 1977) predicted sediments indicates greater variability in wind that the zonal mean rate ofprecipitation over velocity in the central equatorial Pacific prior oceans at 5°_30° N latitude would be consid to 300,000 years ago (Chuey et al. 1987). A erably greater as compared with the present general correlation between glacial maxima climate. The model also predicted glacial and increased wind velocity is also indicated period aridity in continental areas and this is by these sediments. supported by geological evidence (Damuth and Fairbridge 1970, Williams 1975, Street GLACIAL CLIMATIC PATTERNS. Climate-sen and Grove 1976). sitive marine sedimentary parameters indicate climatic fluctuations across the equatorial OCEAN TEMPERATURE. CLIMAP Project Pacificduring the Quaternary (Valencia 1977). members (1984) reported that the last inter- 300 PACIFIC SCIENCE, Volume 46, July 1992 glacial sea surface temperature of about least at low elevations), but the inversion 125,000 yr ago was probably very similar to level may have been higher during interglacial present sea surface temperature. Research on climates (cf. Selling 1948), which may have marine fauna by the CLIMAP Project mem allowed more precipitation to reach the bers (1976) indicated that August sea surface Haleakala summit. temperatures near Hawaii during the last glacial maximum were about 1-2°C cooler MAUNA KEA GLACIATION. On Mauna Kea , than today. Fossiliferous strata of the Ewa the only summit in the tropical mid-Pacific coastal plain do not indicate that Pleistocene Ocean Basin to show evidence of glaciation, climatic fluctuations had any specific effect on there are four glacial drift sheets (Wentworth reef and lagoonal microfauna (Resig 1969). and Powers 1941, Porter 1979). The oldest Gregory and Wentworth (1937) claimed and largest of these glaciations began about that alternating cool and warm climates have 280,000 yr ago and covered an area of nearly left their mark on the floral and faunal com 150 km 2 (Porter 1979). Glaciation on Mauna position of coral reefs, and they stated that Kea may have been due to an increase of various lines of evidence show that glacial winter snowfall and a decrease in ablation ocean temperature around Hawaii was be rates resulting from lower air temperatures tween 13°Cand 20°e. They did not present and increased cloudiness. supporting evidence for either assertion. Based on the extent and thickness of the most recent ice cap on Mauna Kea, Porter TOPOGRAPHY-INFLUENCED RAINFALL. East (1979) estimated the equilibrium line altitude Molokai, which now receives about 3000 mm of rain annually, may have received over (ELA) of the glacier under steady state con ditions. The ELA marks the point on a glacier 12,000 mm of rainfall annually before sub where the net mass balance is zero. This sidence of the island (Stearns and Macdonald climate-sensitive parameter is dependent on 1947). Before the summit of the Koolau Range, Oahu, was lowered by erosion, the various accumulative and ablation-related processes. When the ELAs for the four glacia northwestern part of the range was more tions are corrected for the subsidence of the sheltered from trade wind rains than now island , a picture ofthe relative drop in temper (Stearns and Vaksvik 1935). The relatively ature for the four glacial periods emerges. A youthful stream development stage in that comparison of these ELAs with oxygen iso part of the range , furthermore, may be relict tope ratios and eustatic sea level fluctuations from a time when the trade winds were more shows a close correlation. easterly than at present. Before its lowering by From the ELAs, Porter (1979) estimated erosion and subsidence, the summit of the that the snowline during the last glacial maxi Koolau Range may have been about 1500 m mum was depressed about 935 m. If snowline high and it may have received double its depression was only due to a reduction of air present rainfall (Stearns and Vaksvik 1935). The large valleys of the leeward Koolau temperature, and lapse rates were similar to those oftoday, then July temperatures near Range were incised during that period of higher rainfall. the summit of Mauna Kea would have been about 5°C lower during a glacial maximum. INVERSION LEVEL. Stearns (in Stearns and If mean air temperatures were only reduced Macdonald 1942), writing about the forma 1-2°C (such as the change in ocean tem tion ofHaleakala Crater, Maui, thought that perature) then an increase in precipitation the "erosion of the loosely knit cinder cones would be required to account for glaciation and lavas ofthe summit was probably acceler on Mauna Kea . Other researchers (e.g., ated by more precipitation there during the Krauss 1973) believe that a slight cooling of warm interglacial epochs than at present." tropical surface waters amplifies the cooling The sum ofthe evidence reviewed in this paper of air at higher elevations. In a computer indicates that glacial periods in Hawaii were simulation of the effects of tropical deforesta probably wetter than interglacial periods (at tion (which might be analogous to the reduc- Hawaiian Quaternary Paleoenvironments-e-GxVENDA 301 tion of rainforest area during glacial periods), affected the amount of rainfall received and Potter et al. (1975) found that their model the distribution of vegetation (Ruhe 1964). predicted a steepening of lower troposphere The pattern of soils in the Pearl Harbor area lapse rates. If this were the case, then there reflects the complex interaction of these fac would be no need to invoke greater amounts tors of soil formation. of precipitation to explain the glaciations on The lower end of the Wahiawa Basin above Mauna Kea . Pearl Harbor is marked by a sea cliff formed during the Kaena + 30 m stand of the sea of PERIGLACIAL FEATUR ES. Porter (1979) re about 650,000 yr ago. The soils above this cliff ported solifluction lobes covering a postgla are highly weathered and are classified as cial soil above 3300 m on Mauna Kea , which Oxisols (Foote et al. 1972). Below the cliff the indicates a change to unstable slope condi soils are developed in alluvium and marine tions. Solifluction activity in association with clay (Ruhe et al. 1965b) and are classified as an archeological site at 3790 m on Mauna Kea Inceptisols, Mollisols, and Vertisols, indicat indicates that solifluction has been an active ing a less-weathered mineralogy. All of these process for alI or part of the past 800 yr. soils now occur in a semiarid climate. Sher Gregory and Wentworth (1937) documented man and Alexander (1959) postulated that other periglacial features on Mauna Kea . the highly porous nature of the basalt and /or Noguchi et al. (1987) observed small stone volcanic ash parent material of the Oxisols stripes near the summit of Haleakala and allowed the formation ofthese soils under this attributed them to active periglacial proces s semiarid climatic regime. Uehara and Sher es. A former, more severe periglacial environ man (1956) considered the possibility of an ment is suggested by even larger stone stripes ash parent material, but they also postulated and stone-banked terrace s near the Haleakala that the Oxisols may have formed under a summit (Noguchi et al. 1987). wetter climate . The apparently incongruous occurrence of Pedological Evidence ofPaleoclimates Inceptisols, Mollisols, and Vertisols adjacent to Oxisols can be explained by the geologic BURI ED SOILS. The occurrence of buried and climatic history of the Pearl Harbor area. soils, especialIy between basalt flows, has Whereas the Inceptisols and Mollisols de been reported in the geologic literature of veloped in basaltic alluvium and the Vertisols Hawaii. For example, Stearns (1947)described developed in marine clay (Ruhe et al. 1965b), a highly weathered soil buried by lithified recent research (Ga venda 1989) has shown dune s of pre-Waimanalo age. In general , in that soil derived from volcanic ash forms the suffi cient information is given in these geo upper portion ofthe highly weathered soils of logic references to make paleoclimatic infer the Wahiawa Basin. According to Ruhe et al. ences. Beckmann (1963) attempted to deduce (1965b), the Oxisols are developed on pre paleoclimatic conditions from buried soils in middle Pleistocene surfaces. The Vertisols are central Oahu but conceded that the main developed in marine clay deposited during obstacle in his reconstructions was the prob Kaena time, and the Inceptisols and Mollisols lem of separating the effects of climate and are developed in alluvial deposits that are time. post-Kaena in age. The Oxisols were exposed SOILS OFTHE PEARL HARBOR AREA AND THE for a longer duration and to more climates WAHIAWA BASIN. In the Pearl Harbor area, that were wetter than those that the other soils Oahu, formation of parent materials, changes experienced . in topography and vegetation, and duration The montmorillonitic mineralogy in the of exposure to subaerial weathering were all Vertisols of the Ewa coastal plain is inherited affected by glacio-eustatic sea level fluctua from basaltic detritus that underwent little tions during the Quaternary (Ruhe 1965).The diagenic alteration while in the marine en relative height of the Koolau Range, as in vironment (Ruhe et al. 1965b). Hus sain and fluenced by sea level fluctuations, probably Swindale (1974) postulated that the high 302 PACIFIC SCIENCE, Volume 46, July 1992

water table in these soils is promoting the been observed in soils on Molokai and Lanai formation of montmorillonite. The persis (Gavenda, unpublished data) and has also tence and/or formation of montmorillonite been reported on Kauai, Maui, and Hawaii during one glacial and two interglacial cli (R. V. Ruhe, unpublished data). Some of mates indicate that conditions conducive to these surfaces may be contemporaneous, the the alteration ofmontmorillonite to kaolinite result of a regional climatic regime conducive have not existed for any great period of time to erosion. on the Ewa coastal plain since the transgres The botanical evidence (Lyon 1930) and sion of the Kaena sea. Impeded drainage, paleoclimate models (Ruhe 1964, Manabe however, rather than dry climates, may be and Hahn 1977) indicate that glacial periods responsible for the persistence of montmoril were wetter than interglacial period in the lonitic mineralogy. Salt Lake area, Oahu. Vegetative cover on the Wahiawa Basin was therefore probably more PALEOSOLS AND EROSION SURFACF-S. The soils dense during glacial periods, which suggests of the Wahiawa Basin are composed of three that the erosion surface was created during a distinct layers (Gavenda 1989). Below a layer semiarid interglacial period when there was of topsoil is a layer of volcanic ash-derived insufficient vegetative cover to reduce run-off. soil. The lowest layer is a truncated paleosol The soils of the Wahiawa Basin record that contains an argillic horizon (i.e., a subsoil other erosional events that took place after the characterized by an accumulation of clay formation of the pre-Kaena erosion surface. translocated from higher in the soil pro In the soil exposure along Kunia Road north file). Argillic horizon formation is promoted of Huliwai Gulch there are two ash deposits by alternating wet and dry soil conditions that rest on the pre-Kaena erosion surface rather than continuously wet or dry condi (Figure 4). These deposits are separated by an tion s (Eswaran and Sys 1979). A seasonally erosional unconformity, and both deposits wet and dry soil moisture regime occurs over are truncated below the present topsoil, which much of central Oahu today. The ash-derived formed in tropospheric dust (Gavenda 1989). soil overlying the paleosol, however, has mini The topsoils in the Wahiawa Basin contain mal expression ofclay translocation, whereas a considerable amount of tropospheric dust clay translocation in the truncated paleo that originates from continental Asia (Rex et argillic horizon is strongly expressed, espe- al. 1969, Jackson et al. 1971, Dymond et al. cially in the old Waianae alluvium. . 1974). Atan accumulation rate of about I mm The erosion surface (truncated paleosol) per 1000 yr, 200 mm of topsoil consisting of and the ash-derived soil predate the Kaena relatively unmixed dust indicates that the shoreline (650,000 yr ago) because neither the geomorphic surfaces of the Wahiawa Basin erosion surface nor the ash mantle is known have been accumulating dust, and have there to occur below it. There should have been fore been stable, for at least several hundred sufficient time for an argillic horizon to form thousand years. There have been one or more if conditions were favorable. Perhaps the glacial/interglacial cycles since the start of climate since Kaena time has been too dry to dust accumulation, but this series of climates produce much clay translocation. On the did not produce the massive erosion that other hand, the minimal argillic expression occurred in pre-Kaena time. in the ash-derived soil could be due to poor dispersion ofthe soil, which inhibits the trans SOIL STRIPPING. Stearns attributes "soil location of the clay minerals. stripping" at relatively high (> 75 m) eleva The erosion surface that truncates the paleo tions on Molokai, Lanai, Kahoolawe, and argillic horizon in central Oahu appears to Maui to marine erosion during high stands of have been caused by a basin-wide event when the sea (Stearns 1940,1947, Stearns and Mac stream channels on the leeward Koolau slopes donald 1942, 1947). As discussed earlier, it is were in their early stages of entrenchment highly unlikely that high interglacial seas (Gavenda 1989). An erosion surface has also exceeded about 40 m elevation. The erosion Haw aiian Quaternary Paleoenvironments-e-Gxveret», 303

FI GURE 4. Exposure of soil along Kunia Road north of Huliwai Gulch , Oahu . The truncated paleosol developed in old Waianae alluvium is at a depth ofabout I m. The paleosol is overlain by two ash deposits, which are separated by an erosional unconformity. Both of these ash deposits were truncated and are mantled by aerosolic dust from cont inental Asia.

that Stearns described can be accounted for Zone I pollen accumulated during the last by giant tsunamis, exceptional subaerial ero glacial period and is characterized by a domi sion under former climatic regimes, or oro nance of xerophytic subalpine species, with graphic climate/vegetation patterns similar rainforest vegetation being more restricted to those under persistent northeasterly trade than at present. Selling interpreted this distri winds. bution of plants as indicating the depression of the inversion layer. This interpretation basically agrees with Porter's (1979) estimate Botanical Evidence ofPaleoclimates of treeline depression on Mauna Kea to an FOSSIL POLLEN. Palynological evidence from altitude of 2000 m during that time . Africa, South America, and New Guinea Selling's Zone II pollen reflects the domi demonstrates that equatorial vegetation has nance ofrainforest vegetation. Zone II pollen not been immune to the climatic fluctuations was interpreted as representing the period of of the Quaternary (Flenley 1979). In Hawaii, maximum postglacial warmth when rainfor Selling (1948) examined the pollen found in est vegetation and the inversion level probably the mountain mires of Kauai, Molokai, and extended several hundred meters higher than Maui. The pollen represents only mountain today. Zone II was also interpreted to be a vegetation because pollen of lowland species time of pronounced trade wind rains and does not appear in the montane peat. The possibly high rainfall. bogs and swamps that contain the pollen Zone III represents the change in vegetation range in elevation from 1200 to 1765 m and to its present composition. Rainforest vegeta currently receive high precipitation (ca. 5000 tion is not well represented, but conditions mm /yr) . Selling recognized three major pollen are not as dryas conditions represented by zones (Figure 5). Zone I pollen. Selling's interpretation is that 304 PACIFIC SCIENCE, Volume 46, July 1992

SUMMIT BOG SWAMP BELOW SUMMIT

Xerophytic Xerophyt ic (and ± mesaphytic) Hydrophytic Pollen (and ± mesophytic) Hydrophyt ic pollen (%) pollen (%) zone pollen (%) pollen (%) 50 o 50 100 00 50 o 50

III , _ JQ 20

30 s: II - 40 _1Q 0. / C1l / / o 5 / 50 , / / 60 / / I / /

F IGURE 5 . Pollen diagrams and zones from bog sediments at about 1700 m on West Maui volcano (from Porter 1979, fig. 15; data from Selling 1948, pI. 24 and 25. Reprinted by permission of the author, Bishop Museum Press, and Academic Press). the inversion layer is lower than it was dur Easton (1987) reported that lava interca ing Zon e II, and the vegetation displacement lated with and immediately overlying Pahala may be due to increased trade wind rains. ash and Pohakaa ash on Kilauea Volcano The more xerophytic vegetation representing contains molds oftree ferns. He stated that an Zone s I and III may indicate the decreased annual rainfall of 2000 mm is needed for fern influence of the North Pacific high-pressure forests , but this area now receives only 500 area . mm per year . This indicates that the climate Bennett (1985) examined the fossil pollen was wetter at the time of ash deposition. preserved in nearshore marine sediments on Because the bulk of Pahala ash was deposited the Ewa coastal plain but did not attempt between 24,000 and 10,000 yr ago, this may a paleoecological reconstruction because of also indicate that glacial climates in Hawaii uncertainty regarding the source area of the were wetter than interglacial climates . pollen. LITERATURE CITED BURIED PLANTREMAINS. The Salt Lake Tuff was deposited during the Waipio (350,000 yr BECKMANN, G. G. 1963. The genesis ofcertain ago) low stand of the sea (Hay and Iijima Hawaiian palaeosols and their alteration 1968). The tuff contains many fossil plants following burial. Ph.D. diss. University of found in an upright position, which indicates Hawaii at Manoa, Honolulu. that they were not transported there by stream BENNETT, T. M. 1985. Palynology of selected action. A total of fifteen species including horizons from the Ewa coastal plain, Oahu, koa, ohia, and loulu palm has been identified Hawaii. M.S. thesis. University of Hawaii (Lyon 1930). These trees, which are found in at Manoa, Honolulu. high rainfall areas today, were collected from CHUEY, J., D. REA, and N . PISIAS. 1987. Late an area that now receives 500 mm of rain Pleistocene paleoclimatology of the central annually. Based on paleoclimatic estimates of equatorial Pacific: A quantitative record Ruhe (1964) , rainfall in the Salt Lake area at of eolian and carbonate deposition. Quat. the time of tuff deposition could have been Res. 28: 323-339. 1000 to 1750 mm annually. CLIMAP PROJECT MEMBERS. 1976. The sur- Hawaiian Quaternary Paleoenvironments-s-GxVENDA 305

face of the ice-age Earth. Science (Wash., HAY, R . L., and A. IIJIMA . 1968. Nature and D .C.) 191: 1131-1144. origin of palagonite tuffs of the Honolulu ---. 1984. The last interglacial ocean. Group on Oahu, Hawaii. Geol. Soc . Am. Quat. Res . 21 : 123-224. Mem.116:331 -376. COULBOURN, W. T ., J. F . CAMPBELL, and R . HUSSAIN, M. , and L. SWINDALE. 1974. The MOBERLY. 1974. Hawaiian submarine ter physical and chemical properties of the races, canyons, and Quaternary history Gray Hydromorphic soils of the Hawaiian evaluated by seismic-reflection profiling. Islands. Soil Sci. Soc. Am. Proc. 38: 935 Mar. Geol. 17: 215-234. 941. DAMUTH, J. F., and R . W . FAIRBRIDGE. 1970. JACKSON, M . L., T . W. M . LEVELT, J. K . Equatorial Atlantic deep-sea arkosic sands SYERS, R . W. R EX, R. N . CLAYTON, G . D . and ice-age aridity in tropical South Amer SHERMAN, and G . UEHARA. 1971. Geomor ica . Geol. Soc . Am. Bull. 81 : 189-206. phological relationships of tropospheric DONN, W. L., W . R. FARRAND, and M . ally derived quartz in the soils ofthe Hawai EWING. 1962. Pleistocene ice volumes and ian Islands. Soil Sci. Soc . Am . Proc. 35: sea level lowering. J. Geol. 70: 206-214. 515-525. DYMOND, J., P. E. BISCAYE, and R. W. REX. JONES, S. B., and R . BELLAIRE. 1937. The 1974. Eolian origin of mica in Hawaiian classification ofHawaiian climates. A com soils . Geol. Soc. Am . Bull. 85 : 37-40. parison of the Koppen and Thornthwaite EASTON, R . M . 1987. Volcanism in Hawaii. systems. Geogr. Rev. 27 : 112-119. Stratigraphy of Kilauea Volcano. U .S. KRAUSS, E. 1973. Comparison between ice age Geol. Surv. Prof. Pap. 1350 :243-260. and present general circulations. Nature ESWARAN, H. , and C. Svs, 1979. Argillic (Lond.) 245 : 129-133. horizon in LAC (low activity clay) soils. Ku, T. L., M . A. KIMMEL, W. H . EASTON, Formation and significance to classifica and T . J . O'NEIL. 1974. Eustatic sea level tion. Pedologie 29: 175-190. 120,000 years ago on Oahu, Hawaii. FLENLEY, J. R. 1979. The late Quaternary Science (Wash., D .C.) 183 :959-962. vegetational history of the equatorial LEOPOLD, L. B. 1951. Hawaiian climate-its mountains. Prog. Phys. Geogr. 3 :488-509. relation to human and plant geography. FLINT, R . F . 1971. Glacial and Quaternary Meteorol. Monogr. 1(3): 1-6. geology. J. Wiley and Sons, New York. LIPMAN, P. W ., W . R. NORMARK, J. G. MOORE, FOOTE, D., E. HILL, S. NAKAMURA, and F . J. B. WILSON, and C. E. GUTMACHER . 1988. STEPHENS. 1972. Soil survey of the islands The giant Alika debris slide, Mauna Loa, of Kauai, Oahu, Maui, Molokai, and La Hawaii. J. Geophys. Res . 93: 4279 -4299. nai . State of Hawaii. U .S. Department LYON, H. L. 1930. The flora of Moanalua of Agriculture. Soil Conservation Service. 100,000 years ago (abstract). Pages 6-7 U.S. Government Printing Office, Wash in Proc. Hawaii. Acad. Sci. Fifth Annual ington, D .C. Meeting. Bernice P. Bishop Mus. Spec. GAVENDA, R . T. 1989. Soil genesis and land Publ. 16. scape evolution in central Oahu, Hawaii. MACDONALD, G . A ., D . A. DAVIS, and D. C . Ph .D . diss. University ofHawaii at Manoa, Cox. 1960. Geology and ground-water re Honolulu. sources of the island of Kauai, Hawaii. GIAMBELLUCA , T. W ., M. A. NULLET, and Hawaii Div. Hydrogr. Bull. 13. T . A . SCHROEDER. 1986. Rainfall atlas of MACDONALD, G . A., A. T. ABBOTT, and F. L. Hawai'i, Report R76 . State of Hawaii. PETERSON. 1983. Volcanoes in the sea, 2d Department of Land and Natural Re ed. University of Hawaii Press, Honolulu. sources. Honolulu, Hawaii. MANABE, S., and D . G . HAHN. 1977. Simula GREGORY, H. E., and C. K . W ENTWORTH. tion of the tropical climate of an ice age. J. 1937. General features and glacial geology Geophys. Res . 82: 3889- 3911. of Mauna Kea, Hawaii. Geol. Soc. Am. MOLINA-CRUZ, A . 1977. The relation of the Bull. 48 : 1719-1742. southern trade winds to upwelling pro- 306 PACIFIC SCIENCE, Volume 46, July 1992

cesses during the last 75,000 years. Quat. RUHE, R. V., J. M . WILLIAMS, R. C. SHUMAN, Res. 8: 324- 338. and E. L. HILL. 1965b. Nature of soil MOORE, J. G. 1987. Volcanism in Hawaii. parent material in Ewa-Waipahu area, Subsidence of the Hawaiian Ridge. U.S. Oahu, Hawaii. Soil Sci. Soc. Am. Proc. Geol. Surv. Prof. Pap. 1350: 85-100. 29 :282-287. MOORE, J. G ., and D . J. FORNARI. 1984. SELLING, O. H . 1948. Studies in Hawaiian Drowned reefs as indicators of the rate of pollen statistics. Part III. On the late Qua subsidence ofthe island ofHawaii. J. Geol. ternary history ofthe Hawaiian vegetation. 92 : 752-759. Berni ce P. Bishop Mus. Spec. Publ. 39. MOORE, J. G ., and G . W. MOORE. 1984. SHERMAN, G . D. , and L. T. ALEXANDER. 1959. Deposit from a giant wave on the island of Characteristics and genesis of Low Humic Lanai, Hawaii. Science (Wash., D .C.) 226: La to sols. Soil Sci. Soc. Am. Proc. 23 : 168 1312-1315. 170. MOORE, J. G ., D . A. CLAGUE, R. T . HOLCOMB, STEARNS, H . T . 1935. Pleistocene shorelines P. W. LIPMAN, W. R. NORMARK, and M. on the islands of Oahu and Maui, Hawaii. E. TORRESAN. 1989. Prodigious submarine Geol. Soc. Am. Bull. 46: 1927-1956. landslides on the Hawaiian Ridge. J. --, - . 1940. Geology and ground-water re G eophys. Re s. 94 : 17, 465-17, 484. sources of the islands of Lanai and Ka NOGUCHI , Y., H. TABUCHI, and H. HASEGAWA. hoolawe, Hawaii. Hawaii Div. Hydrogr. 1987. Physical factors controlling the for Bull. 6. mation of patterned ground on Haleakala, ---. 1947. Geology and ground-water re Maui. Geogr. Ann. 69A: 329-342. sources of the island of Niihau, Hawaii. PORTER, S. C. 1979. Hawaiian glacial ages. Hawaii Div. Hydrogr. Bull. 12. Qu at. Res. 12: 161-187. ---. 1961. Eustatic shorelines on Pacific POTTER , G. L., H . W. ELLSAESSER , M. C. islands . Z. Geomorphol. Suppl. 3: 3-16. MACCRACKEN, and F . M . LUTHER. 1975. ---. 1974. Correlation of Pleistocene Possible climatic impact oftropical defores shorelines in Gippsland, Australia, and tation . Nature (Lond.) 258 : 697-698. Oahu, Haw aii: Discussion. Geol. Soc. Am. PRICE, S. 1983. Climate. Pages 59-66 in R. W. Bull. 86 : 1189. Arm strong, ed. Atlas of Hawaii, 2d ed. ---. 1978. Quaternary shorelines in the U niversity of Hawaii Press, Honolulu. Hawaiian Islands. Bernice P. Bishop Mus. RESIG, J. M. 1969. Paleontological investiga Bull. 237. tion s of deep borings on the Ewa Plain, STEARNS, H. T. , and G. A. MACDONALD. 1942. Oahu, Hawaii. Hawaii Inst. Geophys. Rep. Geology and ground-water resources of HIG-69-2. the island of Maui, Hawaii. Hawaii Div. REX, R . W., J. K. SYERS, M . L. JACKSON, and Hydrogr. Bull. 7. R. N. CLAYTON. 1969. Eolian origin of --- . 1947. Geology and ground-water re qu artz in soils of the Hawaiian Islands sources of the island of Molokai, Hawaii. and in Pacific pelagic sediments. Science Hawaii Div. Hydrogr. Bull. II. (Wash., D.C.) 163: 277-279. STEARNS, H. T. , and K . N. VAKSVIK . 1935. RUHE, R. V. 1964. An estimate of paleo Geology and groundwater resources of the clim ate in Oahu, Hawaii. Am. J. Sci. 262 : island of Oahu, Hawaii. Hawaii Div. Hy 1098-1115. drogr. Bull. I. - - - . 1965. Relation of fluctuations of sea STREET, F . A. 1981. Tropical paleoenviron level to soil genesi s in the Quaternary. Soil ments. Prog. Phys. Geogr. 5: 157-185. Sci. 99 : 23-29. STREET, F. A., and A. T. Grove. 1976. Envi ---. 1975. Geomorphology. Houghton ronmental and climatic implications oflate Mifflin Co., Boston. Quaternary lake-level fluctuations in Af RUHE, R. V., J. M . WILLIAMS, and E. L. HILL. rica. N ature (Lond.) 261: 385-390. 1965a. Shorelines and submarine shelves, UEHARA, G. , and G. D . SHERMAN. 1956. The Oahu, Hawaii. J. Geol. 73:485-497. nature and properties ofthe soils of the red Hawaiian Quaternary Paleoenvironments-i-GxVENDA 307

and black complex ofthe Hawaiian Islands. WINCHELL, H . 1947. Honolulu Series, Oahu, Hawaii Agric. Exp. Stn. Tech. Bull. 32. Hawaii. Geol. Soc. Am . Bull. 58 : 1-48. VALENCIA, M . J. 1977. Pacific Pleistocene WRIGHT, H . E., and D. G . FREY. 1965. Pref paleoclimatic stratigraphies: A compara ace. In H. E. Wright and D. G . Frey, tive analysis of results. Quat. Res . 8 : 339 eds . The Quaternary of the United States. 354. Princeton University Press, Princeton. WENTWORTH, C. K. 1926. Pyroclastic geology YEH, T. C., J. E. CARSON, and J. J. MARCIANO. of Oahu. Bernice P. Bishop Mus. Bull. 30. 1951a. On the relation between the circum WENTWORTH, C. K. , and W. E. POWERS. 1941. polar westerly current and rainfall over the Multiple glaciation ofMauna Kea, Hawaii. Hawaiian Islands. Meteorol. Monogr. 1(3): Geol. Soc. Am. Bull. 52 : 1193-1218. 47-55. WILLIAMS, M. A. J. 1975. Late Pleistocene YEH, T. c., C. C. WALLEN, and J. E. CARSON. tropical aridity synchronous in both hemi 1951b. A study of rainfall over Oahu. spheres? Nature (Lond.) 253: 617-618. Meteorol. Monogr. 1(3): 34-46.