Ships that pass in the night: Competition in the fossil record
Summary
A lab activity using the Paleobiology Database to produce and interpret diversity curves for brachiopods and bivalves that tests the hypothesis of competitive displacement, as first demonstrated in Gould and Calloway's (1980) classic study "Clams and brachiopods: ships that pass in the night." A "hidden" motive is for students to recognize that the fossil record can yield incorrect conclusions if not interpreted properly. The activity is written as two complementary activities: Activity A is intended for introductory courses and Activity B is intended for advanced courses.
Collapse
Context
Audience
Activity A: Introductory historical geology or paleontology course for majors/non-majors; Activity B: Advanced undergraduate majors course in paleontology or paleobiology.
Skills and concepts that students must have mastered
For Activity A, students should be familiar with the geologic time scale, the nature of the fossil record and its biases, major mass extinctions, and how to read bivariate graphs.
For Activity B, students should be comfortable downloading diversity data from the Paleobiology Database, plotting graphs of that data in Excel, manipulating data in Excel, calculating the correlation coefficient in Excel, and understanding what this statistic means.
For Activity B, students should be comfortable downloading diversity data from the Paleobiology Database, plotting graphs of that data in Excel, manipulating data in Excel, calculating the correlation coefficient in Excel, and understanding what this statistic means.
How the activity is situated in the course
Activity A is well suited to be used midway through a semester. Activity B is well suited to be used toward the end of the semester. Each is a stand-alone exercise that can be completed during a lab session.
Goals
Content/concepts goals for this activity
- Interpret how diversity changes through time. (The roles of mass extinctions, ecological competition, among others.)
- Evaluate how sampling biases and other artifacts affect diversity curves.
- Understand the significance of large fossil datasets such as the Paleobiology Database for testing paleontological/earth historical hypotheses.
Higher order thinking skills goals for this activity
- Generate predictions for how diversity would change when competition occurs between two taxa.
- Using actual data to test paleontological hypotheses.
Other skills goals for this activity
- Search the Paleobiology Database to compile appropriate data.
- Download diversity data from the Paleobiology Database.
- Read and interpret diversity curves.
- For Activity B: Use Excel to graph diversity curves and bivariate plots, calculate first differences and the correlation coefficient.
- Orally communicate interpretations with other students.
- Communicate their final interpretations in writing.
Description and Teaching Materials
This exercise is conducted on computers with Internet access using the Paleobiology Database (https://www.paleobiodb.org/). Students can complete the exercise without a partner, though they can also be partnered at the instructor's discretion. To complete the assignment, they will need to navigate the website, and for Activity B they will need to download data that they will graph and analyze in Microsoft Excel.
Activity A is designed for introductory-level students and can be used for non-majors; it requires less software experience and statistical proficiency. Activity B is intended for majors in a paleontology, paleobiology, or quantitative methods course. Each activity can stand alone, or elements of each may be merged at an instructor's discretion.
Activity A begins with an introduction to the Paleobiology Database (PBDB) and the concept of "Big Data" in paleontology. Students begin with a guided exploration using the database's "Navigator" interactive map to discover the functions of the PBDB and the types of data available. This can be customized for the local geographic context by tasking students to discover the age range of local rocks, most common local fossils/state fossil, or other relevant observations. Next, students locate and compare the faunas of two sites, one Paleozoic in age, the other Cenozoic. In doing so, their observations demonstrate the distinction between two of Jack Sepkoski's "evolutionary faunas," the characteristic marine faunas of different phases of life's history. (The sites examined can also be customized for other locations.) This sets up a motivating research question. Why did the animals of the "Modern evolutionary fauna" replace those of the "Paleozoic evolutionary fauna"?
Students investigate this question by focusing on the hypothesis of competitive replacement in the context of brachiopods and bivalves. [This brachiopod-bivalve example was famously examined by Stephen Jay Gould and C. Bradford Calloway in the 1980 paper, "Clams and brachiopods—ships that pass in the night", hence the title of the exercise.] In order to test this hypothesis, they first develop predictions for fossil diversity patterns if bivalves outcompeted brachiopods. (This can be expanded on if students have been introduced to the ecology of the two groups.) They should arrive at something like a "double-wedge" model in which brachiopod diversity decreases as bivalve diversity increases. Next, they examine real-world data from the PBDB to see whether the diversity of each group over the Phanerozoic matches their predictions. For introductory students, this plot can be provided (see uploads). Otherwise, students can download the data and create a graph in Excel, as described for Activity B.
The diversity curves for brachiopods and bivalves present a superficial appearance of a double-wedge pattern. Students are encouraged to look closer at the data to see if the pattern is genuine. When bivalve diversity increases from one data point to the next, does brachiopod diversity decrease, and vice versa? (This is a simple, qualitative approach to the more sophisticated first-differences method described below in Activity B for advanced majors.) Finding that brachiopod and bivalve diversity are positively correlated, opposite the prediction of the competitive replacement hypothesis, students reject that hypothesis and seek a new one. Having noted that the changeover from high brachiopod diversity to high bivalve diversity occurs at the Permian – Triassic boundary and with prior knowledge of the "Big 5" mass extinctions, they consider the role of mass extinctions in major faunal change. In doing this, students gain ownership of the pattern (two evolutionary faunas), the causative process (mass extinction as opposed to competitive replacement), and the hypothesis-testing approach to link the pattern to the process. At the end, students are encouraged to brainstorm other questions about the fossil record and how to test them using the PBDB.
Activity B is designed for students with greater statistical proficiency and ability in Microsoft Excel, including knowledge of the correlation coefficient and its calculation in Excel. It can be done instead of Activity A, or it can include the introductory and exploratory components but replace the interpretation of the diversity plot. As before, students are introduced to the competitive exclusion hypothesis for the replacement of brachiopods by bivalves. The double-wedge model is framed explicitly as a statistical negative correlation between the groups. Students download and plot the diversity data from Fossilworks, a companion site to the PBDB, to create a diversity time series. They next create a plot of brachiopod diversity against bivalve diversity, using Excel to calculate the trendline, correlation coefficient, and r2 value. This reveals a negative correlation between the two that, at first glance, appears to support the hypothesis.
Students are then introduced to the problem of autocorrelation in time-series data; each time interval is not statistically independent of the intervals before and after, which is a violation of standard statistical techniques. After learning about this phenomenon, students learn to address the problem using first differences/detrending. In this case, the data analyzed is the change in diversity from one time interval to the next (or the slope), rather than the raw number of genera present. (This is the quantitative equivalent of the direction of change observed in Activity A.) Students use Excel to calculate first differences for brachiopods and bivalves, then plot them against one another, finding the trendline, correlation coefficient, and r2 as before. In this case, they discover a positive correlation, opposite the prediction of the competition hypothesis. They revisit the hypothesis and can now reject it. As in Activity A, they have applied hypothesis testing to a proposed pattern-process relationship while also discovering how an understanding of statistical independence and methods of statistical analysis are essential to arriving at the correct interpretation of the data. At the end, they are encouraged to brainstorm other tests they could apply to this research question.
Uploaded resources:
1. Activity A: Paleobiology Database fossil diversity/competition exercise designed for introductory students and non-majors.
2. Activity B: Paleobiology Database fossil diversity/competition exercise designed for advanced paleontology students with greater software and statistical experience. Can be done instead of or merged with elements of Activity A.
3. Graph: Previously prepared generic diversity curve for brachiopods and bivalves in the Phanerozoic. Intended for use with Activity A; in Activity B, students create the diversity curves for themselves.
4. Key: Excel spreadsheet ("TimeSeries_Key.xlsx") giving diversity data for bivalves and brachiopods (downloaded from Paleobiology Database via FossilWorks.org on April 27, 2018) and key for how to detrend the time series using first-differences. The key includes graphs of bivalve and brachiopod diversity curves through time, raw and detrended brachiopod vs. bivalve bivariate graphs, and calculation of correlation coefficients for both raw and detrended data.
5. Appendix: "AppendixI_DiversityCurves.pdf" provides instructions how to use FossilWorks.org (companion site to the Paleobiology Database) to create diversity curves.
Activity A for introductory students/non-majors (Microsoft Word 2007 (.docx) 16kB Apr29 18)
Activity B for advanced students (Microsoft Word 2007 (.docx) 40kB Apr29 18)
Graph handout: diversity of brachiopods and bivalves through time (Acrobat (PDF) 48kB Apr29 18)
Key for Excel spreadsheet in Activity B (Excel 2007 (.xlsx) 27kB Apr29 18)
Appendix, How to download data from Fossilworks.org (Acrobat (PDF) 218kB Apr29 18)
Activity A is designed for introductory-level students and can be used for non-majors; it requires less software experience and statistical proficiency. Activity B is intended for majors in a paleontology, paleobiology, or quantitative methods course. Each activity can stand alone, or elements of each may be merged at an instructor's discretion.
Activity A begins with an introduction to the Paleobiology Database (PBDB) and the concept of "Big Data" in paleontology. Students begin with a guided exploration using the database's "Navigator" interactive map to discover the functions of the PBDB and the types of data available. This can be customized for the local geographic context by tasking students to discover the age range of local rocks, most common local fossils/state fossil, or other relevant observations. Next, students locate and compare the faunas of two sites, one Paleozoic in age, the other Cenozoic. In doing so, their observations demonstrate the distinction between two of Jack Sepkoski's "evolutionary faunas," the characteristic marine faunas of different phases of life's history. (The sites examined can also be customized for other locations.) This sets up a motivating research question. Why did the animals of the "Modern evolutionary fauna" replace those of the "Paleozoic evolutionary fauna"?
Students investigate this question by focusing on the hypothesis of competitive replacement in the context of brachiopods and bivalves. [This brachiopod-bivalve example was famously examined by Stephen Jay Gould and C. Bradford Calloway in the 1980 paper, "Clams and brachiopods—ships that pass in the night", hence the title of the exercise.] In order to test this hypothesis, they first develop predictions for fossil diversity patterns if bivalves outcompeted brachiopods. (This can be expanded on if students have been introduced to the ecology of the two groups.) They should arrive at something like a "double-wedge" model in which brachiopod diversity decreases as bivalve diversity increases. Next, they examine real-world data from the PBDB to see whether the diversity of each group over the Phanerozoic matches their predictions. For introductory students, this plot can be provided (see uploads). Otherwise, students can download the data and create a graph in Excel, as described for Activity B.
The diversity curves for brachiopods and bivalves present a superficial appearance of a double-wedge pattern. Students are encouraged to look closer at the data to see if the pattern is genuine. When bivalve diversity increases from one data point to the next, does brachiopod diversity decrease, and vice versa? (This is a simple, qualitative approach to the more sophisticated first-differences method described below in Activity B for advanced majors.) Finding that brachiopod and bivalve diversity are positively correlated, opposite the prediction of the competitive replacement hypothesis, students reject that hypothesis and seek a new one. Having noted that the changeover from high brachiopod diversity to high bivalve diversity occurs at the Permian – Triassic boundary and with prior knowledge of the "Big 5" mass extinctions, they consider the role of mass extinctions in major faunal change. In doing this, students gain ownership of the pattern (two evolutionary faunas), the causative process (mass extinction as opposed to competitive replacement), and the hypothesis-testing approach to link the pattern to the process. At the end, students are encouraged to brainstorm other questions about the fossil record and how to test them using the PBDB.
Activity B is designed for students with greater statistical proficiency and ability in Microsoft Excel, including knowledge of the correlation coefficient and its calculation in Excel. It can be done instead of Activity A, or it can include the introductory and exploratory components but replace the interpretation of the diversity plot. As before, students are introduced to the competitive exclusion hypothesis for the replacement of brachiopods by bivalves. The double-wedge model is framed explicitly as a statistical negative correlation between the groups. Students download and plot the diversity data from Fossilworks, a companion site to the PBDB, to create a diversity time series. They next create a plot of brachiopod diversity against bivalve diversity, using Excel to calculate the trendline, correlation coefficient, and r2 value. This reveals a negative correlation between the two that, at first glance, appears to support the hypothesis.
Students are then introduced to the problem of autocorrelation in time-series data; each time interval is not statistically independent of the intervals before and after, which is a violation of standard statistical techniques. After learning about this phenomenon, students learn to address the problem using first differences/detrending. In this case, the data analyzed is the change in diversity from one time interval to the next (or the slope), rather than the raw number of genera present. (This is the quantitative equivalent of the direction of change observed in Activity A.) Students use Excel to calculate first differences for brachiopods and bivalves, then plot them against one another, finding the trendline, correlation coefficient, and r2 as before. In this case, they discover a positive correlation, opposite the prediction of the competition hypothesis. They revisit the hypothesis and can now reject it. As in Activity A, they have applied hypothesis testing to a proposed pattern-process relationship while also discovering how an understanding of statistical independence and methods of statistical analysis are essential to arriving at the correct interpretation of the data. At the end, they are encouraged to brainstorm other tests they could apply to this research question.
Uploaded resources:
1. Activity A: Paleobiology Database fossil diversity/competition exercise designed for introductory students and non-majors.
2. Activity B: Paleobiology Database fossil diversity/competition exercise designed for advanced paleontology students with greater software and statistical experience. Can be done instead of or merged with elements of Activity A.
3. Graph: Previously prepared generic diversity curve for brachiopods and bivalves in the Phanerozoic. Intended for use with Activity A; in Activity B, students create the diversity curves for themselves.
4. Key: Excel spreadsheet ("TimeSeries_Key.xlsx") giving diversity data for bivalves and brachiopods (downloaded from Paleobiology Database via FossilWorks.org on April 27, 2018) and key for how to detrend the time series using first-differences. The key includes graphs of bivalve and brachiopod diversity curves through time, raw and detrended brachiopod vs. bivalve bivariate graphs, and calculation of correlation coefficients for both raw and detrended data.
5. Appendix: "AppendixI_DiversityCurves.pdf" provides instructions how to use FossilWorks.org (companion site to the Paleobiology Database) to create diversity curves.
Activity A for introductory students/non-majors (Microsoft Word 2007 (.docx) 16kB Apr29 18)
Activity B for advanced students (Microsoft Word 2007 (.docx) 40kB Apr29 18)
Graph handout: diversity of brachiopods and bivalves through time (Acrobat (PDF) 48kB Apr29 18)
Key for Excel spreadsheet in Activity B (Excel 2007 (.xlsx) 27kB Apr29 18)
Appendix, How to download data from Fossilworks.org (Acrobat (PDF) 218kB Apr29 18)
Teaching Notes and Tips
The activities are written to allow flexibility in analyses and to foster discussion of broader conceptual issues relevant to scientific analyses.
Common hurdles:
For Activity B, the most common hurdles include downloading the data from the Paleobiology Database (or companion website Fossilworks.org), importing the downloaded data into Microsoft Excel, combining the bivalve and brachiopod diversity time series into a single spreadsheet, calculating first differences, and graphing the diversity data. Depending on time and student's prior experience with these skills, a pre-downloaded diversity spreadsheet (such as found on left side of online key) can be provided by the instructor. Instructions to download data and generate diversity curves from FossilWorks can be found as online Appendix I.
Students struggle to match up the time series (i.e., so that the Cenozoic 1 bivalve diversity value is in the same row as the Cenozoic 1 brachiopod diversity value and so that Cambrian 2 matches Cambrian 2.) This is most troublesome for the Cambrian intervals, where some Cambrian intervals have brachiopods but lack bivalves (such that these bivalve intervals are missing). It is critical that the intervals lacking bivalves be manually added as blank rows (with values of "0" genera). The key provided illustrates a correct alignment.
Students with less experience with Excel will benefit from additional instructions and one-on-one troubleshooting when calculating first-differences (change in diversity / million years) between adjacent time intervals. (The solution is mathematically equivalent to the slope between adjacent data points.) The provided key gives the correct solution. When students "click and drag" to copy the equation for other time intervals, they often copy for all time intervals, forgetting that "differences" (i.e., slopes between adjacent points) are an "n-1" problem. In other words, if there are 49 time intervals, there can only be (49 – 1 =) 48 first differences to calculate. The last row of the calculated first differences must be uncalculated (or missing, because otherwise the cells are calculating the change between diversity in Cambrian 1 interval and an unspecified "missing" time interval with age of 0).
Extensions:
This exercise can be modified in several ways if the instructor chooses:
(a) Geographic context: For students in different regions, different locations can be selected in the exploratory portion of Activity A. For example, the Silica Formation, Martin-Marietta Quarry of southern Michigan and the Onondaga Formation, Kingston, New York, both Devonian-aged, are good representatives of typical Paleozoic faunas, as are any Ordovician samples in the Cincinnati, OH area. Other sites will serve as well provided that they contain predominately representatives of typical "Paleozoic fauna" or "Modern fauna" animals.
(b) Alternative scenarios of competitive replacement: Although brachiopods vs. bivalves is the textbook example of perceived competitive replacement, other possibilities could also be examined. Ideas include fish vs. cephalopods, dinosaurs vs. mammals, pterosaurs vs. birds, trilobites vs. fish, or rodents vs. plesiadapiforms. However, be warned that the authors of this exercise have not scrutinized the data on those groups to confirm their suitability for this exercise. Proceed with appropriate caution.
Broader concepts:
Causation vs. correlation: Both activities involve the concept of correlation (conceptually in Activity A and statistically in Activity B). It is always worth reiterating to students that a strong correlation between two variables does not necessarily imply a causative relationship between them. Although the hypothesis of competitive displacement assumes a causal relationship between the variables (in this case, diversification of bivalves drives the extinction of brachiopods), the correct analysis of the data (using either graphical analysis in Activity A or statistical analysis in Activity B) shows a weak positive correlation. The presence of this positive correlation should not be interpreted as a causal one, in which the diversification of bivalves positively caused diversification of brachiopods. The more likely interpretation is that both groups of animals were similarly responding to changes in habitable area and sea level change, climate and primary productivity conditions, and other geological changes. Without formal confirmatory analyses, these "correlational" activities cannot formally explain what these causes actually were.
Observing competition and clade-replacement in the fossil record: These activities are useful to help students better conceptualize how to think about competition and clade replacement in the fossil record, advancing from simple (and possibly misleading) interpretations of diversity curves to more intellectually sophisticated interpretations made by paleontologists. Once students appreciate how initial interpretations can be misleading, it may be worth guiding class discussions regarding better ways to test these hypotheses. For example, paleontologists rarely actually use diversity curves to formally test claims of ecological competition because generic diversity is not the appropriate scale at which ecological competition actually occurs (which occurs among individuals within local communities). Better ways to test competitive interactions should include analyses of well preserved (i.e., taphonomically autochthonous) fossil assemblages where evidence of direct competitive interactions (such as overgrowth of adjacent bryozoan colonies or preservation of predatory trace fossils) occurs.
These activities are also helpful in spurring broader discussions of the role of mass extinctions (and other disturbances) in leading to the replacement of certain clades by others, a concept well discussed in the class Gould and Calloway (1980) study these activities replicate. The concept of "incumbent replacement" (sensu Rosenzweig and McCord 1991) is also relevant to these discussions.
Common hurdles:
For Activity B, the most common hurdles include downloading the data from the Paleobiology Database (or companion website Fossilworks.org), importing the downloaded data into Microsoft Excel, combining the bivalve and brachiopod diversity time series into a single spreadsheet, calculating first differences, and graphing the diversity data. Depending on time and student's prior experience with these skills, a pre-downloaded diversity spreadsheet (such as found on left side of online key) can be provided by the instructor. Instructions to download data and generate diversity curves from FossilWorks can be found as online Appendix I.
Students struggle to match up the time series (i.e., so that the Cenozoic 1 bivalve diversity value is in the same row as the Cenozoic 1 brachiopod diversity value and so that Cambrian 2 matches Cambrian 2.) This is most troublesome for the Cambrian intervals, where some Cambrian intervals have brachiopods but lack bivalves (such that these bivalve intervals are missing). It is critical that the intervals lacking bivalves be manually added as blank rows (with values of "0" genera). The key provided illustrates a correct alignment.
Students with less experience with Excel will benefit from additional instructions and one-on-one troubleshooting when calculating first-differences (change in diversity / million years) between adjacent time intervals. (The solution is mathematically equivalent to the slope between adjacent data points.) The provided key gives the correct solution. When students "click and drag" to copy the equation for other time intervals, they often copy for all time intervals, forgetting that "differences" (i.e., slopes between adjacent points) are an "n-1" problem. In other words, if there are 49 time intervals, there can only be (49 – 1 =) 48 first differences to calculate. The last row of the calculated first differences must be uncalculated (or missing, because otherwise the cells are calculating the change between diversity in Cambrian 1 interval and an unspecified "missing" time interval with age of 0).
Extensions:
This exercise can be modified in several ways if the instructor chooses:
(a) Geographic context: For students in different regions, different locations can be selected in the exploratory portion of Activity A. For example, the Silica Formation, Martin-Marietta Quarry of southern Michigan and the Onondaga Formation, Kingston, New York, both Devonian-aged, are good representatives of typical Paleozoic faunas, as are any Ordovician samples in the Cincinnati, OH area. Other sites will serve as well provided that they contain predominately representatives of typical "Paleozoic fauna" or "Modern fauna" animals.
(b) Alternative scenarios of competitive replacement: Although brachiopods vs. bivalves is the textbook example of perceived competitive replacement, other possibilities could also be examined. Ideas include fish vs. cephalopods, dinosaurs vs. mammals, pterosaurs vs. birds, trilobites vs. fish, or rodents vs. plesiadapiforms. However, be warned that the authors of this exercise have not scrutinized the data on those groups to confirm their suitability for this exercise. Proceed with appropriate caution.
Broader concepts:
Causation vs. correlation: Both activities involve the concept of correlation (conceptually in Activity A and statistically in Activity B). It is always worth reiterating to students that a strong correlation between two variables does not necessarily imply a causative relationship between them. Although the hypothesis of competitive displacement assumes a causal relationship between the variables (in this case, diversification of bivalves drives the extinction of brachiopods), the correct analysis of the data (using either graphical analysis in Activity A or statistical analysis in Activity B) shows a weak positive correlation. The presence of this positive correlation should not be interpreted as a causal one, in which the diversification of bivalves positively caused diversification of brachiopods. The more likely interpretation is that both groups of animals were similarly responding to changes in habitable area and sea level change, climate and primary productivity conditions, and other geological changes. Without formal confirmatory analyses, these "correlational" activities cannot formally explain what these causes actually were.
Observing competition and clade-replacement in the fossil record: These activities are useful to help students better conceptualize how to think about competition and clade replacement in the fossil record, advancing from simple (and possibly misleading) interpretations of diversity curves to more intellectually sophisticated interpretations made by paleontologists. Once students appreciate how initial interpretations can be misleading, it may be worth guiding class discussions regarding better ways to test these hypotheses. For example, paleontologists rarely actually use diversity curves to formally test claims of ecological competition because generic diversity is not the appropriate scale at which ecological competition actually occurs (which occurs among individuals within local communities). Better ways to test competitive interactions should include analyses of well preserved (i.e., taphonomically autochthonous) fossil assemblages where evidence of direct competitive interactions (such as overgrowth of adjacent bryozoan colonies or preservation of predatory trace fossils) occurs.
These activities are also helpful in spurring broader discussions of the role of mass extinctions (and other disturbances) in leading to the replacement of certain clades by others, a concept well discussed in the class Gould and Calloway (1980) study these activities replicate. The concept of "incumbent replacement" (sensu Rosenzweig and McCord 1991) is also relevant to these discussions.
Share your modifications and improvements to this activity through the Community Contribution Tool »
Assessment
Formative: In-class discussions, faculty/TA check-ins during problem solving.
Summative: In Activity A, students are evaluated based on their written responses to prompts. They explain the reasoning behind their hypothesis and their ultimate support or rejection of it. They are evaluated on their successful completion of the assigned tasks, demonstration of critical thinking in the process of hypothesis testing, and their original creative thinking on other applications of the methods. In Activity B, students deliver a correctly formatted Excel document with their completed exercise that shows that the data were downloaded correctly, along with the requested plots showing trendlines, correlation coefficients, and r2 values. They then interpret these results using open-ended questions.
Summative: In Activity A, students are evaluated based on their written responses to prompts. They explain the reasoning behind their hypothesis and their ultimate support or rejection of it. They are evaluated on their successful completion of the assigned tasks, demonstration of critical thinking in the process of hypothesis testing, and their original creative thinking on other applications of the methods. In Activity B, students deliver a correctly formatted Excel document with their completed exercise that shows that the data were downloaded correctly, along with the requested plots showing trendlines, correlation coefficients, and r2 values. They then interpret these results using open-ended questions.
References and Resources
Websites used:
Paleobiology Database (https://paleobiodb.org/#/): Public database of fossil occurrences referenced by organism, location, and age. Students use this site to explore fossil occurrences in space and time.
Fossilworks.org: Companion site to the Paleobiology Database from which students in Activity B download raw diversity data for brachiopods and bivalves.
References:
Gould, S. J., and C. B. Calloway. 1980. Clams and brachiopods: ships that pass in the night. Paleobiology 6(4): 383-396. https://doi.org/10.1017/S0094837300003572
Rosenzweig, M. L., and R. D. McCord. 1991. Incumbent replacement: evidence for long-term evolutionary progress. Paleobiology 17(3):202-213. https://doi.org/10.1017/S0094837300010563
Paleobiology Database (https://paleobiodb.org/#/): Public database of fossil occurrences referenced by organism, location, and age. Students use this site to explore fossil occurrences in space and time.
Fossilworks.org: Companion site to the Paleobiology Database from which students in Activity B download raw diversity data for brachiopods and bivalves.
References:
Gould, S. J., and C. B. Calloway. 1980. Clams and brachiopods: ships that pass in the night. Paleobiology 6(4): 383-396. https://doi.org/10.1017/S0094837300003572
Rosenzweig, M. L., and R. D. McCord. 1991. Incumbent replacement: evidence for long-term evolutionary progress. Paleobiology 17(3):202-213. https://doi.org/10.1017/S0094837300010563