Carbon accumulation in coastal wetlands is normally assessed by extracting a sediment core and estimating its carbon content and bulk density. Because carbon content and bulk density are functionally related, the latter can be estimated gravimetrically from a section of the core or, alternatively, from the carbon content in the sample using the Mixing Model equation from soil science. We analyzed the effect that the choice of corer and the method used to estimate bulk density could have on the final estimates of carbon storage in the sediments.

The choice of corer did not have much influence on the final estimates of carbon density; the main factor in selecting a corer is the operational difficulties that each corer may have in different types of sediments.

Because of the multiplication of errors in a product of two variables subject to random sampling error, when using gravimetric estimates of bulk density, the dispersion of the data points in the estimation of total carbon density rises rapidly as the amount of carbon in the soil increases. For this reason, the estimation of carbon densities in peaty soils with this method can be very imprecise in peaty sediments.

In contrast, the estimation of total carbon density using only the carbon fraction as a predictor is very precise, especially in sediments rich in organic matter. This method, however, depends critically on an accurate estimation of the two parameters of the Mixing Model (the bulk density of pure peat and the bulk density of pure mineral sediment). If these parameters are not estimated accurately, the calculation of total carbon density can be biased.