Doctoral Dissertations

Date of Award

12-2003

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Geology

Major Professor

Michael L. McKinney

Abstract

The fossil record remains the primary tool to understand macro-evolutionary processes over time. Unfortunately, much debate has centered upon sampling biases that pervade the rock and fossil record. A relative metric is presented that reduces the effect of uneven distribution of fossils across time and space. The Diversity Index of Growth (DIG) metric measures relative change of biodiversity because it examines the four key parameters of biodiversity: New, Extinct, Stable and Total, as obtained from biostratigraphic data. In this study, DIG was utilized to examine biodiversity trends of a Cenozoic Caribbean database consisting of genus-level biostratigraphic occurrences of corals, echinoids, bivalves and gastropods. The results of this study indicate that the individual class-level trends observed from the database are consistent with previously interpreted trends of the four clades. Key features of the observed biodiversity trends of the Caribbean include three peaks of diversification, including increased biodiversity near the Paleocene-Eocene Thermal Maximum (PETM). Marine invertebrate biodiversity appears to be linked to a variety of environmental processes, including volcanism and climate change. The timing of biodiversity increases corresponds to increases in sea-level and three major volcanic episodes in the Caribbean. The trends of the total biodiversity follow sea-level change, especially when sea-level data are reduced to relative change and compiled into time increments equal to the fossil data. The positive relationship between sea-level increases and biodiversification is observed at several spatial-temporal scales, and indicates that relative biotic diversity change may be useful for following highstand facies or for sequence stratigraphy applications. Climate effects, as interpreted from isotopic data and terrestrial data, also suggest a potential agreement between climate, sea-level and biodiversity. Also investigated in this study are the effects of scale and hierarchy from temporal, spatial and taxonomic perspectives. Longer time intervals result in vastly different biodiversity trends than shorter time intervals over the same data set. Taxonomic hierarchy was investigated by comparing trends of species-level and genus-level biodiversity, and it was observed that similar biodiversity trends did not consistently cascade across taxonomic hierarchies. Also demonstrated in this study is the problem of comparing data of differing levels of spatial-temporal resolution. Not only do such comparisons stymie statistical correlations, but they can also yield conflicting interpretations of the causes of biodiversity. A new metric is introduced to understand and quantify the relative degree of resolution achievable by a database. Sensitivity of Data (SOD) gives a measurement of the spatial, temporal, and observational extent over which the interpretations gleaned from a database are valid. In many instances, the potentially differing interpretations made from the same or similar data may be attributed to variation in the spatio-temporal sensitivity of the data, or the environmental processes to which they are compared. Thus SOD allows an evaluation of the temporal, spatial and therefore, hierarchical issues that affect interpretations of geologic data. Higher SOD values indicate finer-resolution data while lower SOD values indicate lower-resolution data. High SOD values are shown here in close agreement to isotopic or sea-level data that are sub-regional to basinal in scale. The Caribbean database used here has a low SOD value and conforms better to regional to global scale data. I hypothesize that the sensitivity of the data contributes to the assignment of various environmental processes as higher or lower hierarchies that impart differential control to the biosphere. These higher- and lower-orders of processes likely may originate only because of variation in the quality and quantity of observations used to interpret past events. Thus, the relativity of time, data, and trends of the incomplete rock record suggest the need to investigate applications of relative metrics to not just the biostratigraphic data but to geologic data, as a whole.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS