THE IMPACT OF TAPHONOMY ON RECONSTRUCTIONS OF TERRESTRIAL COMMUNITIES IN TIME AND SPACE

A persistent problem in bridging the gap between modern and ancient ecology is the perception that the fossil record is flawed – subject to missing and biased information that makes it difficult or impossible to reconstruct attributes of past terrestrial ecosystems. Decades of research on taphonomic processes combined with studies of ancient communities using data from the fossil record have demonstrated, to the contrary, that paleontological samples are valuable resources for understanding ecology over a wide range of temporal and spatial scales and are especially suitable for investigating macroecological patterns. It is becoming standard practice to account for taphonomic biases in large species occurrence datasets that are used for paleocommunity analysis. Such studies typically address biases that affect comparability among samples, including whether species richness and co-occurrence patterns in paleocommunities can be validly compared with modern communities. We will discuss approaches to the problem of bias in these variables, including: 1) using subsets that have equivalent taphonomic biases, such as vertebrates of similar body size, 2) taking into account variability in species representation by sampling a limited set of depositional environments and/or taphonomic modes, 3) dealing with uneven taphonomic sampling by adjusting for variable sample sizes (e.g., species under- or over-representation in Pairs Analysis), 4) framing questions that stay within boundaries set by the taphonomy of the fossil samples, e.g., limits due to temporal and spatial averaging, and 5) recognizing that interpretations represent particular circumstances favorable to fossil preservation rather than the paleoecosystem as a whole. The quality of data from the fossil record can exceed what is available from modern communities, not only in terms of temporal range but also species richness and abundance patterns that may be difficult to measure in living ecosystems. With the cumulative knowledge now available about taphonomic processes, modeling the effects of different biases “up” from the past as well as “down” from the present could strengthen understanding of what paleontological data have to offer and help to mitigate misconceptions about the limitations of the fossil record.