Guild Structure

1
(No Transcript)
2
Guild Structure

• Cody (1975) demonstrated that the division of
  available resources amongst sets of insectivorous
  birds in shrub grasslands results in the
  projection upon each resource dimension of almost
  identical niches within Californian chaparral,
  Chilean matorall and South African machia

3
Guild Structure

• While a set of resources may be divided by
  consumers in an infinite number of ways, in
  reality the realized niches of each species
  showed a remarkable similarity in these three
  shrubland systems despite the fact each is
  colonized by a taxonomically distinct set of
  species

4
Guild Structure

• The match is so close that the different species
  of the analogous communities even display
  morphological convergence in their adaptations to
  apparently parallel niches

5
Guild Structure

• Similar constancy of niche structure in parallel
  communities is recorded among assemblages of
  montane lizards in Chile and California (Fuentes
  1976), coral reef fish assemblages of the
  Atlantic and Pacific (Gladfelter et al. 1980) and
  finch communities (Schluter 1986)

6
(No Transcript)
7
Guild Structure

• What is a guild?
• Any group of species that exploit the same
  class of environmental resources in a similar
  way

8
Guild Structure

• Many guilds have been documented
• e.g. nectar feeders, desert lizards, terrestrial
  salamanders, insectivorous birds
• However, one can subdivide groups
• E.g. Insectivorous birds become
• foliage gleaners, flycatchers, bark gleaners, or
  ground gleaners

9
Guild Structure

• Furthermore, you do not have to group species
  together in terms taxonomic positions, but rather
  on their niche requirements

10
Guild Structure evidence

• The basic idea is that within a community there
  are clusters of species interacting among
  themselves more strongly than with other species
  in the community

11
Guild Structure problems

• There are problems with the guild concept
• There is no objective criteria for assigning
  guild membership
• Limits on membership not clearly defined
• Causes of guild structure unresolved
• Most studies do restrict analysis to a single
  taxonomic assemblage

12
Guild Structure evidence

• Inger and Colwell (1977) made the first attempt
  at objective identification of guild structure
  within community matrices by seeking sharp
  discontinuities in the arrangement of resource
  use curves along resource axes (e.g. sudden
  increases in the variance of mean overlap)

13
Guild Structureobjective grouping

• Joern and Lawlor (1981 Oikos) determined group
  membership of guilds through use of a clustering
  technique, progressively linking together (in
  unidimensional space) pairs and then clusters of
  species with highest overlap

14
Guild Structure

• Guild structure (of grasshoppers) based upon
  resource use
• Guild structure is ascribed from cluster analysis
  of species in relation to expressed overlap in
  resource use

15
Guild Structure

• Joern and Lawlors analysis clusters together
  groups of species whose competitive interactions
  with others are strongest with that same guild
  (greater overlap)
• This approach can be extended to multidimensional
  space using a range of clustering techniques
  (e.g. PCA or FA)

16
Guild Structure

• While this approach effectively defines groups
  and identifies species, it does provide
  statistical significance to the clusters
  identified (but see Jaksic and Medel 1990)

17
Guild Structure null model

• The null model for guild structure is that the
  relative frequency of guilds in the assemblage
  represents a random sample of species from the
  colonizing source pool

18
Guild Structure null model

• Two deviations from the null model are possible
  the difference in guild frequencies between the
  source pool and the assemblage might be unusually
  small or large

19
Guild Structure null model

• When the deviations are large, certain guilds are
  over- or under-represented in local assemblages
• For example, in many land-bridge islands, there
  is a consistent absence of some bird families
  (MacArthur et al. 1972 Ecology)
• Why?

20
Guild Structure null model

• When the variation is unusually small, that would
  indicate the guilds are similar to one another in
  the level of co-occurrence observed

21
Guild Structure null model

• How to achieve a correct null model?
• Solution drawing species randomly from an
  appropriate source pool to evaluating the
  expected amount of variation in such island (or
  small) assemblages

22
Guild Structure null model

• Gotelli and Abele (1982 J of Biogeography) used
  the hypergeometric distribution to test for
  deviations in species richness of West Indian
  landbird families

23
Guild Structure null model

• For each island, the observed number of species
  in each family was compared to the expected
  number if species were drawn equiprobably from
  the archipelago list

24
Guild Structure null model

• Results the number of coexisting species of
  parrots (Psittacidae) was less than expected but
  the number of coexisting pigeons and doves
  (Columbidae) and mockingbirds (Mimidae) was
  greater than expected
• Can you think of what characteristics might
  contribute to over-representation?

25
Guild Structure EcoSim null model

• The guild structure works within the
  co-occurrence analysis option
• One could create a separate presence-absence
  matrix for each guild in the assemblage and
  analyze each matrix separately

26
Guild Structure EcoSim null model

• However, there are also times when one might want
  to test for patterns among the guilds as a group

27
Guild Structure EcoSim null model

• All species are assigned to a single guild
  (user-defined and therefore subject to biologists
  biases)
• This module does not test for classification or
  recognition of guilds, but rather hypotheses
  about them that have been designated a-priori by
  YOU
• Tests for differences among different guilds

28
Guild Structure EcoSim null model

• Species Site1 Site2 Site3 Site4
• SpeciesA 1 1 0 0
• SpeciesB 0 0 1 0
• SpeciesC 0 0 0 1
• SpeciesD 1 1 1 1
• SpeciesE 0 1 1 0
• SpeciesF 0 1 0 0
• SpeciesG 1 1 1 0
• SpeciesH 0 0 0 1
• SpeciesI 1 1 0 1
• SpeciesJ 1 0 1 1
• SpeciesK 1 0 0 1

29
Guild Structure EcoSim null model

• Species Guild Site1 Site2 Site3 Site4
• SpeciesA X 1 1 0 0
• SpeciesB X 0 0 1 0
• SpeciesC Y 0 0 0 1
• SpeciesD X 1 1 1 1
• SpeciesE Y 0 1 1 0
• SpeciesF Y 0 1 0 0
• SpeciesG Z 1 1 1 0
• SpeciesH Z 0 0 0 1
• SpeciesI Y 1 1 0 1
• SpeciesJ X 1 0 1 1
• SpeciesK Y 1 0 0 1

30
Guild Structure EcoSim null model

• Another form of analysis in the co-occurrence
  module is rather than analyze for differences
  among guilds, one can compare among regions (or
  any other type of site grouping)

31
Guild Structure EcoSim null model

• Species Site1 Site2 Site3 Site4 Regions Intact
  Intact Invaded Invaded SpeciesA 1 1 0 0 SpeciesB
  0 0 1 0 SpeciesC 0 0 0 1 SpeciesD 1 1 1 1 Spec
  iesE 0 1 1 0

32
Guild Structure EcoSim null model

• Generates similar indices as the standard
  co-occurrence analysis (e.g. C-score, the number
  of checkerboard species pairs, the number of
  species combinations, and the variance ratio)

33
Guild Structure EcoSim null model

• In addition to grouping by guild or region, there
  is a favored states analysis (based upon the
  hypothesis of Foxs assembly rules)

34
Guild Structure EcoSim null model

• Fox suggested that species are added sequentially
  to a community such that different functional
  groups (or guilds) are represented as evenly as
  possible.
• Communities can then be classified as to whether
  they are in a favored or an unfavored state

35
Guild Structure EcoSim null model

• For example, if a community had 7 species and 4
  guilds, a favored state would have the guilds
  filled with (1,2,2,2) species
• However, an unfavored state would be (1,0,3,3)

36
Guild Structure EcoSim null model

• EcoSim reshuffles the guild labels, then examines
  each column of the matrix and designates it as a
  favored or unfavored state. The number of favored
  states in a matrix is an integer that can range
  from 0 to a maximum of C, the number of columns
  in the data matrix