Pergamon Quaternary Science Reviews, Vol. 15, pp. 913-937, 1996. Copyright 0 1996 Published by Elsevier Science Ltd. Printed in Great Britain. All rights reserved. PII: SO277-3791(96)00058-3 0277-3791196 $32.00 FORAMINIFERAL EVIDENCE FOR THE AMOUNT OF COSEISMIC SUBSIDENCE DURING A LATE HOLOCENE EARTHQUAKE ON VANCOUVER ISLAND, WEST COAST OF CANADA JEAN-PIERRE GUILBAULT,* JOHN J. CLAGUEf and MARTINE LAPOINTES *BRAQ-Stratigraphie, 10.545 Meilleur, Montkal, Q&bee H3L 3K4, Canada lGeologica1 Survey of Canada, Suite 101605 Robson Street, Vancouver, British Columbia V6B 5J3, Canada SGEOTOP, Universite’ du Qukbec ci Montrkal, C.P. 8888, Succursale Centre-Ville, Montrkal, Q&bee H3C 3P8, Canada Abstract - Foraminiferal data from two sites, 6 km apart, on the shores of an inlet near Tofino on the west coast of Vancouver Island, British Columbia, allow estimates to be made of the amount of coseismic subsidence during a large earthquake 100400 years ago. The sampled sediment succession at the two sites is similar; peat representing a former marsh surface is abruptly overlain by intertidal mud grading upward into peat of the present marsh. At one of the sites, a layer of sand, interpreted to be a tsunami deposit, locally separates the buried peat from the overlying intertidal mud. The abrupt peat-mud contact records sudden crustal subsidence during the earthquake. The paleoelevation of each fossil sample was estimated by comparing its foraminiferal assemblage with modem assemblages of known elevation. The modem assemblages were obtained from surface samples collected along transects across the marsh near the fossil sample sites. Comparisons were made statistically using transfer functions. Estimates of coseismic subsidence, based on differences in paleoelevations just above and below the top of the buried peat, range from 20 cm to 1 m, with the most likely value in the 55-70 cm range. Post-seismic crustal rebound began soon after the earthquake and may have been largely complete a few decades later. Copyright 0 1996 Published by Elsevier Science Ltd QSR INTRODUCTION Foraminiferal data presented in this paper are used to estimate the amount of subsidence that accompanied a large earthquake on western Vancouver Island, British Columbia, less than 400 years ago. This study builds on, and refines, the previous study of Guilbault et al. (1995) in the same area. The work is founded on the principle that foraminiferal assemblages in tidal marshes vary with elevation and that a change in elevation of as little as 5- 10 cm may produce a recognizable change in the foraminiferal assemblage at a given site. This is the essence of the conclusions of Scott and Medioli (1980), based on their investigation of Nova Scotia tidal marshes. This paper and others concerned more specifically with marsh foraminifera from the northwest coast of North America (Phleger, 1967; Williams, 1989; Patterson, 1990a; Jennings and Nelson, 1992; Jonasson and Patterson, 1992) and from Chile (Jennings et al., 1995) provide a basis for discriminating upper marsh from lower marsh environments on the British Columbia coast. Broad limits can then be placed on paleoelevations of samples from intertidal sedimentary sequences using foraminiferal data (Jennings and Nelson, 1992), provided the vertical limits of the marsh zones are known and have not changed with time. The originality of our method lies in the use of transfer functions to estimate paleoelevation (Imbrie and Kipp, 1971). This approach allows a more objective treatment of foraminiferal assemblages than is possible through visual inspection and can be applied to large, multivariate data sets. It also provides a paleoelevation for every sample in a sequence, thus improving the resolution. In this paper, we compare fossil foraminiferal assem- blages at two sites, 6 km apart, near Tofino, British Columbia, with modern assemblages from the adjacent marsh. At both sites, a peat layer representing a former marsh is abruptly overlain by mud that grades upward into peat of the present marsh. The abrupt change from peat to mud has been attributed to sudden subsidence during a large earthquake 100-400 years ago (Clague and Bobrowsky, 1994a). The earthquake may have resulted from the rupture of the boundary between the North America and Juan de Fuca plates along the Cascadia subduction zone (Fig. 1). Samples collected through the sediment sequence at each site were compared with samples taken from a nearby modem transect no more than 150 m away. This was done to minimize the 913