1985, collected sediment at a rate of 1.2 kg/m2 per day; if we assume this rate applies for a 90-day period, then the ''annual" accumulation in the inner basin is 110 kg/m2.
These estimates (Figure 7.11) reflect current conditions and a fiord dominated by fluvial discharge. However, at 8 to 10 kaBP the fiords of eastern Baffin Island contained large tidewater glaciers and the transition from tidewater to terrestrial glaciers probably occurred ca. 7 kaBP (Andrews, 1990)(see also Figure 7.3A).
In order to determine sediment flux we need information on (1) sediment bulk density, and (2) numeric ages at different levels in cores. The first requirement has been met by on-board sampling of sediment in 10 cm Lehigh cores (1 to 3 m length) (Reasoner and Hein, 1984), or by determinations on the piston cores. In the latter we have worked on material specifically sampled for mass physical properties and, in addition, we used sediment collected in 3.15 cc paleomagnetic boxes (cf., Andrews et al., 1986; Andrews and Jennings, 1987) to derive wet volume density (Figure 7.12), which was then converted to dry volume density estimates (Figure 7.13). The second requirement is not trivial because the majority of cores from the fiords lacked calcareous foraminifera, so we are unable to date the cores by the accelerator mass spectrometry (AMS) radiocarbon method on 2 to 4 mg samples of hand-picked foraminifera (cf., Andrews et al., 1989a). We had to rely on AMS dates on the acid-insoluble organic matter (AIOM) fraction. We knew from experiments (Andrews et al., 1985b) that these dates were too old; however, a statistical relationship did exist between paired dates on shells and the AIOM fraction. We have used this correction and investigated its applicability by analysis of the geomagnetic secular variations (Andrews et al., 1986) of piston cores. The results indicated that the corrected 14C dates provided a "reasonable" chronology from 7 to 12 ka. This conclusion was also supported by the pollen and dinoflagellate biostratigraphy (Short et al., 1989). The raw dates and further discussions are included in Jennings (1986), Andrews (1987b; 1990), and Andrews et al. (1989a).
Most cores have three 14C dates that cover the past 10 ka, hence we feel it is most appropriate to consider the sediment flux in appropriate 2 to 5 ka time-slices. Analysis of the contributions of the various measurement errors to errors in estimation indicates that the 95 percent confidence limit in estimating the sediment accumulation (age ± 20 percent and dry volume density ± 10 percent) is ± 14 percent, and no worse than ± 20 percent (Taylor, 1982).
On Figure 7.14 we show a generalized profile for McBeth and Itirbilung fiords (Figure 7.1) and the Fn↓ of sediment in the fiord to shelf transects. The data have been divided into four intervals, i.e., the past 10 ka, and then the periods between 8 to 10 kaBP (the glacial/deglacial transition), 5 to 8 kaBP (paraglaciation, i.e., massive meltwater delivery), and 0 to 5 kaBP (neoglaciation). These broad intervals are appropriate given the number of dates per core (about 3) and the problems in radiocarbon dating.
We are now in a position to discuss changes in categories of Fn↓; that is, we can convert percentage data on grain size—clay- and silt-size mineralogy (Andrews et al., 1989b)—into fluxes and examine variations during the last deglacial cycle. A critical question is: Can we differentiate the sediment sources by considering changes in specific grain-size, mineral suite, or sediment type?
There are few data on the grain-size characteristics of suspended sediment in glacier meltwater (e.g., Church and Gilbert, 1975; Pfirman, 1985; Gurnell and Clark, 1987; Syvitski et al., 1987b) but an indicator of glacial meltwater activity may be the percentage of coarse silt (Figure 7.15), although this can be influenced by a number of other factors. Coarse silt percentages vary between ca. 2 and 25 percent; there is a consistent difference in the coarse silt seafloor flux between McBeth and Itirbilung fiords (Figure 7.15), although the amount of coarse silt in IT3.1 is markedly different from all other cores. The sample interval employed for the grainsize data on Figure 7.15 is too coarse to be a measure of glacial meltwater variations on a 100-year interval.
Several studies have shown that detrital carbonate is significantly higher on the shelf than in the fiords (Jennings, 1986; Andrews et al., 1989b) and that this reflects the transport of ice eroded sediment from Paleozoic basins in northwest Greenland and the high Canadian Arctic. Carbonates are absent within the fiords and detrital carbonate may reflect a shelf → fiord sediment transfer (e.g., Jennings, 1986; Andrews et al., 1989).
The 100-km transect from McBeth Fiord (Figure 7.14) includes the outer basin and adjoining shelf. Four cores are available (Table 7.1); the shelf core (HU78-36) has a basal date on shell of 11,770 ± 550 yr (GX-6280) and appears to