Abstract
AbstractSite-selectivity analysis of drilling predation traces may provide useful behavioral information concerning a predator interacting with its prey. However, traditional approaches exclude some spatial information (i.e., oversimplified trace position) and are dependent on the scale of analysis (e.g., arbitrary grid system used to divide the prey skeleton into sectors). Here we introduce the spatial point pattern analysis of traces (SPPAT), an approach for visualizing and quantifying the distribution of traces on shelled invertebrate prey, which includes improved collection of spatial information inherent to drillhole location (morphometric-based estimation), improved visualization of spatial trends (kernel density and hotspot mapping), and distance-based statistics for hypothesis testing (K-, L-, and pair correlation functions). We illustrate the SPPAT approach through case studies of fossil samples, modern beach-collected samples, and laboratory feeding trials of naticid gastropod predation on bivalve prey. Overall results show that kernel density and hotspot maps enable visualization of subtle variations in regions of the shell with higher density of predation traces, which can be combined with the maximum clustering distance metric to generate hypotheses on predatory behavior and anti-predatory responses of prey across time and geographic space. Distance-based statistics also capture the major features in the distribution of traces across the prey skeleton, including aggregated and segregated clusters, likely associated with different combinations of two modes of drilling predation, edge and wall drilling. The SPPAT approach is transferable to other paleoecologic and taphonomic data such as encrustation and bioerosion, allowing for standardized investigation of a wide range of biotic interactions.
Highlights
Drilling predation on shelled marine invertebrates represents a unique opportunity to quantify biological interactions in ancient and modern marine environments
When point patterns for fossil L. latilirata were grouped by geologic age and examined through time, areas of higher concentration of predation traces derived via kernel density were irregularly distributed on the prey skeleton (Fig. 2A–C)
In the point pattern of predation traces derived from feeding trials of N. unifasciata preying on I. subrugosa, we found a single area of higher concentration of drillholes as well as a single hotspot placed at the umbo (Fig. 3B) but slightly closer to the shell edge than in the beach-collected samples
Summary
Drilling predation on shelled marine invertebrates represents a unique opportunity to quantify biological interactions in ancient and modern marine environments. Most of the methods for analyzing site selectivity in drilling predation were primarily developed to test the null hypothesis of a random distribution of traces across arbitrarily defined sectors of the prey skeleton using goodness-of-fit, chi-square or KolmogorovSmirnov statistics (Kelley 1988; Kowalewski 1990; Anderson et al 1991), and diversity metrics such as the Shannon-Weaver index (Dietl et al 2001) These approaches fail to exploit the high-resolution information locked up in the spatial relationship between drillhole locations and depend critically on an arbitrary grid system used to divide the prey skeleton into sectors (Johnson 1984; Kowalewski 2004). This approach was not designed to identify patterns that may occur at different spatial scales or distances measured on the prey skeleton, for example, aggregation at short distances and segregation at large distances
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
