Palynological Study of Lake Sediment Profiles from Southwestern New Brunswick: Discussion

Abstract

Mott (1975) is to be congratulated on his paper for two reasons: for adding materially to the sparse literature on Quaternary palynology in the Maritimes, and for documenting so clearly some of the problems of the radiocarbon dating of calcareous materials. It is the second of these points on which this discussion centers. The old carbonate effect, whereby old carbon is incorporated into the organic and inorganic carbon cycle on limestone terrain and gives rise to spurious radiocarbon ages, has been described by Deevey et al. (1954), and by Broecker and Walton (1959), as well as in several more recent papers by various authors, (e.g. Ollson 1974). Although this hazard has been known since the early years of radiocarbon dating, its effects have seldom been documented from stratigraphic sequences, and have commonly been disregarded by those using radiocarbon dates. A stratigraphic succession of dates showing a roughly constant hard-water error of 1700 years has been documented by Shotton (1972). Mott clearly shows the fallacy of using a sediment surface sample age (or the Ambrosia pollen horizon in order to avoid recent bomb fallout effects) as a constant correction factor for old carbonate error in a lake bottom sediment core. Ogden (1965a,b) and Davis (1969) questioned the application of such constant corrections to the sediment column, but few workers seem to have been troubled by the assumptions involved. One such assumption which troubles us is that the chemical environment has remained constant throughout the history of small lakes. It would seem most likely that major chemical changes have taken place during the 12 000 or more years since deglaciation, and some of these have been documented (Kemp 1969; Mackereth 1966). Newly-deglaciated terrain would be unweathered; in areas of carbonate bedrock, the exposed surface would tend to yield carbonate-rich sediments. As soil formation progressed contemporaneously with vegetational succession, the supply of carbonate-rich sediments would tend to lessen. It follows that small lakes would tend to have carbonate carbon-rich sediments in their early stages, followed by organic carbon accumulation in later stages. Such a sequence is common in bogs and small lakes in southwestern Ontario, where peat and/or gyttja often overlie a layer of marl or marl-like sediment formed during an early lake stage. Carbon analyses carried out on Louise Lake sediments (Fig. I), for example, show highest carbonate carbon and lowest organic carbon percentages throughout the period preceding the rise in spruce pollen. With the invasion of spruce-dominated vegetation, there is a noticeable decline in carbonate carbon and a corresponding sharp rise in organic carbon. Subsequently, carbonate carbon decreases slightly upward to a minimum while organic carbon increases to maximum percentages in mid-postglacial time. The resultant effect of such trends should be greater incorporation of old carbon in the early stages of lake development, and lesser in later stages. The effect on radiocarbon dates would be to yield apparent ages more in error at the base of the sequence than at the top. Replotting of Mott's dates to show the amount of error versus depth shows just such a relationship (Fig. 2, 3a).