Ancient DNA Megafauna Collapse Timeline Project Ideas

ISEF Category: Animal Sciences

Subcategory: Systematics and Evolution  ·  Difficulty: Advanced  ·  Setup: Home Setup  ·  Time: Full Year

The Hook

A single fossil can hide thousands of years of change. Ancient DNA lets you see when a megafauna population started to fade, not just when the bones stopped showing up. That turns extinction from a guess into a timeline you can test. Climate shifts leave a trail too, so you can compare the two stories side by side.

What Is It?

Ancient DNA is DNA recovered from old bones, teeth, or other preserved remains. For Pleistocene megafauna, that means animals like mammoths, mastodons, and giant ground sloths. You can think of each sample like one page from a very long book. One page does not tell the whole story, but enough pages can show when the population starts to thin out.

Re-analyzing published datasets means you work with sequence data, sample dates, and site information that other researchers already made public. You then ask a new question, such as whether population size dropped before, during, or after major climate changes. In genetics, effective population size means the number of individuals that were really passing genes to the next generation. It is not the same as the headcount you would see in the field, but it gives a useful picture of long-term population health.

Why This Is a Good Topic

This is a strong science fair topic because the data already exist, the question is testable, and the analysis forces you to make real choices about dating, controls, and statistics. You can connect a biological pattern to a real-world problem, climate stress and extinction risk. You will learn how to clean metadata, compare timelines, and tell the difference between a real decline and a sampling artifact.

Research Questions

• How does the inferred collapse date change when you use only well-dated ancient-DNA samples?
• What is the effect of sample coverage on the shape of the population-size curve?
• Does the collapse timeline shift when you compare mitochondrial data with nuclear data?
• To what extent do different climate proxy records line up with the start of the decline?
• Which megafauna species show the earliest drop in effective population size?
• What is the effect of excluding sites with many repeated samples on the inferred trend?

Basic Materials

• Laptop with at least 16 GB RAM.
• Stable internet connection for downloading public datasets.
• Spreadsheet software for cleaning sample metadata.
• R or Python installed for analysis and plotting.
• Text editor or notebook app for notes and code.
• External drive or cloud storage for dataset backups.
• Access to public sequence databases such as NCBI or ENA.

Advanced Materials

• Workstation with at least 32 GB RAM.
• Access to a university cluster or cloud compute for heavier models.
• Curated ancient-DNA sample tables with radiocarbon dates and metadata.
• Public sequence alignments or FASTA files from NCBI, ENA, or published supplements.
• Climate proxy datasets from NOAA Paleoclimatology or related archives.
• Version-controlled project folder for scripts, inputs, and outputs.

Software & Tools

• R: Fits trend tests, change-point models, and confidence intervals for the timeline.
• Python: Cleans metadata, joins tables, and automates repeatable analysis steps.
• Google Colab: Runs notebooks in the browser when local compute is limited.
• BEAST: Estimates time-scaled demographic patterns from dated sequence data.
• FigTree: Reads tree output and helps you inspect the result visually.

Experiment Steps

1. Choose one species or a tight species set, and define what counts as a collapse in your analysis.
2. Build a clean metadata table with sample age, location, coverage, and source fields.
3. Match the DNA timeline to a climate proxy or dated environmental record that covers the same window.
4. Pick one main model for population change, then decide how you will compare it with a simpler baseline.
5. Set up sensitivity checks for uneven sampling, site clustering, and low-quality sequences.
6. Plan one comparison that tests whether the pattern repeats across regions, species, or time slices.

Common Pitfalls

• Mixing calibrated and uncalibrated dates, which shifts the collapse window by centuries.
• Combining samples with different coverage thresholds, which makes one species look more stable than another.
• Treating every published sample as independent, which can overcount the same site or excavation.
• Using a climate proxy that does not overlap the DNA dates, which breaks the timeline comparison.
• Reading a noisy drop in sample count as a real population crash, which confuses sampling bias with biology.

What Makes This Competitive

A strong entry will not just plot sample counts. It will separate real demographic change from sampling bias, then test whether the collapse stays put when you change dating filters, geographic bins, or model choice. The best projects compare more than one megafauna species and line the DNA trend up with an external climate proxy, so the story has both depth and a check against false patterns.

Project Variations

• Focus on one region, such as North America or Eurasia, and compare collapse timing across local megafauna.
• Swap in a different proxy, such as temperature, ice volume, or vegetation change, and see whether the DNA trend still matches.
• Compare mitochondrial and nuclear datasets for the same species to test whether the inferred decline depends on marker choice.

Learn More

• NCBI PubMed: Search review articles on ancient DNA, megafauna extinction, and demographic inference.
• NCBI GenBank and SRA: Find sequence records and linked study data for public ancient-DNA projects.
• European Nucleotide Archive: Browse public sequencing datasets and associated metadata.
• NOAA Paleoclimatology: Download climate proxy records that overlap Pleistocene timelines.
• Paleobiology Database: Check fossil occurrence timing and geographic context for megafauna.
• R Project: Read the free manuals and package documentation for statistical analysis.