Community Disassembly by an Invasive Species

1
Community Disassembly by an Invasive Species

• Sanders et. al (2003)
• PNAS of the United States

Jonathan L. Bowers
2
Effect of Invasion on Community Structure

• Not well understood
• Many studies of invasive processes and their
  effects are done on the fly (reactionary)
• No pre-invasion community data
• Only qualitative estimates of effects of
  invasion need for long-term study of community
  structure pre-, during, and post-invasion

3
Arguments against Community Disassembly Studies

• Previous studies attempting to explain overall
  effects of invasive species on community
  structure and function are retroactive analyses
  with no pre-invasion data to comparatively
  analyze current network assembly and function
• Fukami and Morin 2003 suggest that low
  productivity and low biodiversity in systems may
  occur as a result of the history of community
  assembly in a specific system
• This is one of a large number of potential
  factors
• Disturbance occurrence and frequency
• Spatial scale
• Niche specialization
• Some studies have not comparatively analyzed
  species composition and community assembly in
  intact regions where invasion has not been as
  successful

4
Sanders et. al (2003)

• Test subject
• Linepitheda humile (Argentine ant)
• Native to Argentina, Uruguay, Paraguay, and parts
  of Brazil
• Usually displaces many or all native ant species
  during invasion
• 7-year study conducted on the effect of Argentine
  ant invasion on native ant community composition
  and structure

5
Study Sites

• Research conducted at Jasper Ridge Biological
  Preserve (northern CA)
• Sample plots of 20 meter radii were monitored
  twice a year (7 years total)
• Located in the center of over 100 1-hectare
  quadrats
• Observation either included a 5 minute search or
  stimulus to draw ants (honey) in the absence of
  ants during an active search

6
Analysis

• Simulated communities were analyzed using a
  Standardized Effect Size (SES)
• SES is a measure of the effects observed across
  simulated control and treatments groups
• Comparatively analyze control vs. treatment on
  the same scale
• Species Co-Occurrence
• To quantify this effect, species
  segregations/aggregations were observed in intact
  and invaded plots
• L. humile not incorporated into co-occurrence
  model, but was instead just the indicator of
  invasion

7
Results

• Analysis of pre- and post-invasion data showed
  trends from segregation in pre-invasion plots to
  a community structure of random assignment
  (aggregated sub-communities) in post-invasion
  plots
• Disassembly of intact community (network)
  structure
• Series of indirect cascades likely to lead to
  lowered species abundance and a decrease in
  biodiversity as indicated in previous
  observations (Japan, New Zealand, Easter Island)
• Concluded that this disassembly is rapid in
  nature, usually within 1 year of invasion
• Rapid effect on community structure through
  interspecific resource competition

8
Results (Cond)
Table 1. Intact (non-invaded) communities
(hollow circles) vs. Invaded (L. humile detected)
(filled circles)
Table 2. Preinvasion (hollow circles) vs.
Post-invasion (filled circles) data for each plot
9
Discussion

• By using pre- and post-invasion data, results
  indicate community disassembly induced by
  competition involving the invasive ant (L.
  humile)
• Diamonds assembly rule
• Some pairs of species never coexist, either by
  themselves or as part of a larger combination
• Communities post-invasion tended to be aggregated
  assemblages of a subset of species relative to
  intact (non-invaded) plots
• Different effects yield different assemblages
  based on interspecific competition or other
  species interactions

10
What About the Predator????

• Invasions have been shown to alter specialist
  predator/prey interactions
• In turn can alter interspecific competition
• Bottom-up and Top-down trophic cascade can result
• Insects as middle rungs on the trophic ladder

11
Further Discussion

• Argentine ant is not normally distributed
• Since its presence and abundance varies across
  spatial scales, you would also expect their
  direct and indirect influence to vary as well
• Spread of L. humile varies across ranges for
  other reasons
• Abiotic factors
• Climatic variation

12
Conclusions

• Invasive species affect levels of biodiversity
• In addition to this, they also alter established
  community assemblages
• Prior to invasion, communities exhibited
  segregated structure typical of dynamic
  competition
• Post-invasion, aggregation and original community
  disassembly was observed.

13
Arguments to Research

• Each matrix shows the presence or absence of ant
  species within each plot as binary input (0 for
  no presence, 1 for presence)
• As in other network and community models, binary
  matrices indicate observed links but do not
  necessarily equate to observed indirect effects
• Quantification of these relies on weighted direct
  effects
• Very difficult to obtain
• Very little information in the literature shows
  community assemblages as weighted networks of flow