Censusing the Sea in the 21 st Century

Transcription

1 Censusing the Sea in the 21 st Century Nancy Knowlton & Matthieu Leray Photo: Ove Hoegh-Guldberg Smithsonian s National Museum of Natural History

2 Estimates of Marine/Reef Species Numbers (Millions) Marine Habitat Global Ocean Data Type Author, Date 10 deep sea coral reefs coral reefs * Coral reefs 0.7* coral reefs * coral reefs coral reefs box cores, 21 m 2 Grassle & Maciolek, * tropical forests, species-area * 5 m 2 reef samples, species-area European brachyuran crabs Reaka-Kudla, 1997 Small et al., 1998 Bouchet, ratios of taxa Mora et al., expert opinion Appeltans et al., * expert opinion Fisher et al *Assumes 31% of marine species live on reefs

3 Why so hard? Most marine species are small and rare One third < 4.2 mm Example: Mollusks (Philippe Bouchet) Rank Abundance doubletons singletons 33-91% of all marine species still to be named

4 Global analyses ignore almost all biodiversity Some fish 1700 spp Corals 804 spp Some snails 662 spp Lobsters 69 spp Coral Triangle Roberts et al All??? 3235 spp (~250,000 descr.)

5 Census of Marine Life - CReefs Artificial Surfaces (ARMS) DNA Natural Surfaces (dead heads) Standardization - Allows results of different studies to be compared Automation - Identify species using genetics rather than names Scalability - World-wide sampling strategy is feasible

6 Number of species First Studies: Barcoding Crustaceans The Coral Triangle Bali Lizard Island Heron Poc Heron ARMS Ningaloo Moorea Line Islands FFS- Hawaii Number of Samples Plaisance et al. PLoS One 2011 Unpublished data

7 Put Another Way. Just 6.3 m 2 contained the equivalent of 80% of the described crab diversity of European seas!

8 Barcoding works well for the charismatic microfauna Photos: Matthieu Leray

9 What about the sessile stuff? Indonesia (Photos: David Littschwager)

10 Or the fairly small stuff? (<2mm) Photos: Matthieu Leray

11 From Barcoding to Metabarcoding Bulk Sample DNA Amplified from Entire Community Next-Gen Metabarcodes (easier said than done)

12 ARMS 1) sessile 4 Fractions (+ vouchers) COI Metabarcode 2) > 2mm 4) μm 3) 2mm 500 μm COI barcode

13 Oyster reefs (6-month deployment) Leray & Knowlton, PNAS ARMS (3 sets of 3) at each location Virginia(VA) ~100 m Florida (FL) ~100 m S ~100 m Sample Size Total 7.8 m m 3 ~2 m At what geographic scale are communities different?

14 Chesapeake Bay, Virginia (9 ARMS) Ft. Pierce, Florida (9 ARMS) SHARED 4% barcode 21% meta -barcode #sequences: 572, ,613 % matched: % unknown: % singletons: Leray & Knowlton, PNAS 2015

15 No. of OTUs Virginia Diversity Total = Total = No. of OTUs Florida >2mm 2mm 500um Fraction um sessile >2mm 2mm 500um Fraction um sessile Sub-tropical Florida is more diverse 2/3 of diversity is < 500μm

16 Community similarity Presence-absence Relative abundance VA FL FL VA Different locations and metabarcoding communities are distinct

17 Proportion of OTUs >2mm Taxonomic composition 2mm 500um um Fraction 100 Virginia Echinodermata Mollusca Bryozoa Chordata Cnidaria Porifera Other animals Plantae Ph Amoebozoa Chromista Fungi 100?? Phylum_ReOrderedVA 75???? Annelida Arthropoda Proportion of OTUs Proportion of OTUs Unidentified 0 sessile >2mm >2mm 500um Florida 2mm 500um 106um um Fraction sessile sessile Major Groups Phylum_ReOrderedFL Phylum_ReOrderedFL Annelida Arthropoda Echinodermata Mollusca Bryozoa Chordata Cnidaria Porifera Other animals Plantae Amoebozoa Chromista Fungi Annelida Arthropoda Echinodermata Mollusca Bryozoa Chordata Cnidaria Porifera Other animals Plantae Amoebozoa Unidentified Chromista Fungi Unidentified Leray & Knowlton, PNAS animal phyla - Arthropods most diverse

18 More Sampling Needed (similar pattern for sampling effort measured by # sequences)

19 Leray & Knowlton, PNAS 2015 Jackknifed UPGMA tree based on relative abundance Fine-scale geographic structuring VA FL Adjacent samples cluster together at the meter scale

20 Number of sequences per OTU Abundance information as well as presence-absence Virginia Florida Leray & Knowlton, PNAS 2015 Undetected Amount of DNA per OTU Amount of DNA per OTU 65.3% to 91.7% of OTUs successfully recovered per sample Linear relationship between number of reads and amount of DNA

21 Number of sequences per phylum Even stronger relationship for functional groups Virginia Florida Leray & Knowlton, PNAS 2015 Amount of DNA per phylum Amount of DNA per phylum Similar results for sessile community calibrated by % cover

22 Not Just an Abstract Exercise Marine Biodiversity in the Anthropocene

23 Example: Effects of Ocean Acidification Biodiversity sampling from PNG CO 2 seeps Photos: Laetitia Plaisance

24 Example: Effects of Ocean Acidification Normal (ph 8.0) Low (ph 7.7) 284 individuals >2mm 82 OTUs Phoos: Laetitia Plaisance 114 individuals >2mm 43 OTUs

25 Smallest Fraction Less Sensitive to Acidification? All metabarcode μm 500μm- 2mm Sessile Plaisance, unpubl.

26 Looking Forward Multi-gene PCR-based to increase taxonomic coverage Shotgun metagenomics approach to - avoid PCR bias - capture metabolic genes and look at local adaptation Global deployments and analyses: MarineGEO/PBM?

27 What are some big questions/challenges? What are the really dark taxa? (need branches, not twigs) How does biodiversity scale geographically by taxonomic group? by body size? as a function of location? What is the nature of rarity? Are there biodiversity collapse thresholds? How is biodiversity changing? (need community vouchers) What lives in the sea (or even a bay)?

28 Photo: Barry Brown Thanks Questions?

29 Bioinformatics pipeline WORKFLOW Initial quality filtering PROGRAMS Mothur Alignment to ref. barcode & Removal of non functional sequences based on amino acid translations MACSE Chimera removal UCHIME OTU clustering Taxonomic assignments CROP Blast & SAP Individual- & sample-based rarefactions OTU table PCoA and jacknifed clustering analysis EstimateS QIIME R (Vegan)