COMPARING COMMUNITY COMPOSITION AND DYNAMICS BETWEEN IN-PLACE AND SURFACE-COLLECTED SAMPLES

There are two methods of collecting fossil communities: surface collection of individual fossils and in-place, bulk collection of sediment or rock. The use of in-place samples tends to be more intensive, but is believed to better represent the fossil community. We compared paleocommunity results of surface-collected to in-place, bulk fossil samples.

We collected 14 4-L in-place samples and 7 surface samples from three outcrops of the Pennsylvanian Finis Shale in Jacksboro, Texas. At the East Spillway locality, samples were divided into three 1 m stratigraphic portions. At the two others, in-place and surface samples were lumped. Two to three people collected each surface sample to reduce bias.

We conducted analyses using abundance counts of brachiopods and mollusks and presence/absence (P/A) of all taxa. In both datasets, the surface samples have a significantly greater richness (22 to 43 genera) than the in-place samples (12 to 21) (p < 0.001). Surface samples are also significantly more even (p = 0.002) (surface mean 0.75; in-place 0.56).

Using NMDS ordination, there is strong separation between surface and in-place samples for both abundance and P/A data; permanova shows a significant difference (p = 0.002 abundance; p < 0.001 P/A). For the Spillway East locality, the differences between in-place and surface samples decrease up section. Evenness increases up section for the in-place samples (0.33, 0.50, 0.67) but decreases up section for the surface samples (0.83, 0.78, 0.55). Decreasing relative abundance of the brachiopod Crurithyris drives the in-place evenness.

Potential explanations for these differences include: 1) Preservation: large or fragile specimens may be damaged in disaggregation of bulk samples, but survive natural weathering. 2) Volume: surface specimens weathered from a greater volume of original rock, so more taxa may be present. 3) Collector bias: human collectors may pick up larger, better preserved, or rare specimens. Efforts to avoid bias toward abundant taxa could inflate evenness. 4) Down-slope weathering: at the East Spillway locality, true diversity patterns may have been “smeared” by fossils from higher beds.

In-place samples likely better represent the once-living community. When surface collecting is preferred, it may be wise to collect an in-place sample to assess biases.