Natural History Collections Infrastructure

1
Natural History Collections Infrastructure

• Ricardo Scachetti-Pereira
• The University of Kansas
• Biodiversity Research Center
• Natural History Museum
• KU-BRC/NHM

2
Biological Collections

• Natural History Museums, Herbaria and
  Culture Collections provide fundamental
  resources for biodiversity and ecological
  research
• Development of IT infrastructure to make
  results of those services (specimen data)
  more accessible

3
Collections Data Integration

• Integrate Common Information
• Scientific Name, Taxonomy
• Geography, Locality, GPS coordinates
• Collection Events (Collector) and other
  information
• Geographically Distributed
• Birds of Mexico spread over 43 institutions
  around the world main holder had only
  16 of total specimens
• Heterogeneous Hardware Software
• Database Vendors (Access, Oracle, SQL Server)
• Database Schemas (Table Definitions)
• Software (Specify, Biota, In-house, etc)

4
Collections Data Integration

• Distributed Generic Information Retrieval
  (DiGIR)
• XML based protocol for retrieving structured
  data from multiple, distributed,
  heterogeneous databases over the Internet

5
DiGIR Protocol

• Portal (UI) builds XML query
• Portal broadcasts XML query to providers
• Each provider translates XML query into
  native SQL query (database schema)
• Provider translates results into XML result
  set and send it back to portal
• Portal integrates answers from various
  providers into a single, homogeneous result
  table

6
MaNIS
7
MaNIS
8
MaNIS
9
GBIF
10
GBIF
11
GBIF
12
GBIF
13
Lifemapper
14
Lifemapper
15
Applications of Collections Data Methods

• Prediction of Species Distributions
• Looks for non-random correlations between
  point occurrence and environmental conditions
• Genetic Algorithms (GARP)
• Artificial Neural Networks (ANN)
• Generalized Linear Models (GLM)
• Generalized Additive Models (GAM)
• Many, many others

16
Applications of Collections Data Examples

• Prediction of Species Actual and Potential
  Geographical Distribution
• Invasive Species
• Spread of Diseases
• Evolutionary Biology
• Management
• Monitoring

17
Applications Predicting the distribution of
filovirus disease
Slide by A. Townsend Peterson (KU)
18
Applications Predicting the impact of
species invasions
19
Ecological Data Integration

• Specimen data currently integrated
• Other data required for analysis
• Climate, Relief, Land Cover Use, Remote
  Sense, etc.
• Acquisition, processing and integration is
  still largely manual

20
Ecological Data IntegrationThe SEEK Vision

• Science Environment for Ecological Knowledge
  (SEEK)
• Partners NCEAS, UNM, SDSC, KU

21
SEEK Objectives

• Provide access to biodiversity, ecological
  and environmental data (discovery, sharing
  and reuse)
• Provide scalable and streamlined framework
  for analysis and synthesis
• Use Semantic Mediation to integrate
  heterogeneous data and analytical steps

22
SEEK Overview
Slide by Chad Berkeley (NCEAS)
23
SEEK Ecogrid

• Integrate diverse data networks from
  ecology, biodiversity and environmental
  sciences
• XML based language used for data
  documentation
• Access to computational resources via the Grid

Slide by Matt Jones (NCEAS)
24
SEEK Data Integration
Slide by Matt Jones (NCEAS)
25
SEEK Analysis Modeling
Slide by Matt Jones (NCEAS)
26
SEEK Kepler
27
SEEK Kepler
28
SEEK Taxonomic Object Service
Elliot 1816
R. plumosa

• Taxon concepts change
• over time (and space)
• Multiple competing
• concepts coexist
• Names are re-used for
• multiple concepts

Gray 1834
R. plumosa
Rhynchospora plumosa s.l.
R. Plumosa v. intermedia
R. plumosa v. plumosa
Chapman 1860
R. Plumosa v. interrupta
R. intermedia
Kral 1998
R. plumosa
R. pineticola
Peet 2002?
R. plumosa v. plumosa
R. plumosa v. pinetcola
R. sp. 1
A
B
C
Slide by Bill Michener (UNM)
Information by Robert Peet (UNC)
29
SEEK Road map

• Now into the 2nd year (out of 5)
• Working prototypes for
• Ecogrid Kepler (UI)
• Semantic Mediation System
• Taxonomic Object Service

30
Role of Collections in NEON

• Provide fundamental services for biodiversity
  and ecological research and monitoring
• Collections count on IT infrastructure to
  provide valuable information to NEON
• Will be seamlessly integrated to other
  relevant sources of data

31
Integrating Collections into NEON

• Rate and amount of deposits limited by
• Physical Installations (Storage Facilities)
• Personnel (Allocation and Training)
• Preservation/Storage Processes
• Computerization Process
• Require proper allocation of resources to
  function as part of a monitoring facility