Biodiversity Assessments in Jorka Lakes

1
Biodiversity Assessments in Jorka Lakes
Ashley Steinberg Supervisor Dr. H. MacIsaac
2
Development of Habitat-based Conservation Programs
Protect habitats with most species or most unique
species?
Protect habitats with most species or most unique
species?

• need to determine how many species occur in
  different habitats
• this is usually done through estimations of
  species richness

3
How to Estimate Diversity?

• estimation based upon representative subsamples
• develop saturation curve of species vs.
  samples
• asymptote can be estimated using nonlinear
  regression
• asymptote is the estimated total species
  diversity

different approaches, same saturation
cumulative species
different saturation, same approach
Gotelli Colwell (2001) Ecology Letters
4
Estimators of Species Diversity based upon
Subsamples
Chao2
Jackknife2
where Sobs is observed species, m is total
number of samples, Q1 is frequency of unique
samples and Q2 is frequency of duplicate samples
Smith and van Belle 1984 Colwell and Coddington
1994 Hortal et al. 2006
5
Estimations of Species Diversity

• Used to assess temporal changes in diversity in a
  single system
• Used to assess differences between systems at one
  time

• Estimate considers only level of diversity and
  not changes in community composition (i.e.
  species turnover)
• To consider these changes, must determine
  coefficient of similarity (Sorensens Jaccards
  etc.)
• Evenness of species within community calculated
  using Shannon or Simpson Indices (calculated but
  not shown)

6
Similarity Coefficients

• Coefficient of Sorensen
• Ss 2a
• 2a b c
• where a is the of species common to both
  samples, b is the of species only occurring in
  one sample and c is the species only occurring
  within the other sample
• uses presence and absence data to determine
  similarity of subsamples

• values range from 0 (no similarity) to 1 (total
  species overlap)

7
Phytoplankton of Lakes Biwa (Japan) Baikal
(Russia)
Sorensens Coefficient 0.27
Bondarenko et al. (2006), Hydrobiologia
8
Purpose of Study

• A) Use an extensive data set and Monte-Carlo
  simulations to determine

• theoretical species diversity levels using 2
  different estimators of species diversity
• - Chao2 and Jackknife2
• whether differences are greatest among lakes in
  a given year, or within a lake between years

B) Assess community similarity among lakes and
over time
9
Masurian Lake District, Poland
Majcz Wielki, Glebokie, Inulec, Zelwazek
10
Species Diversity Estimates Across Systems
70
1976
50
Majcz Wielki
30
Zelwazek
Inulec
Cumulative of species
Glebokie
10
70
1997
50
30
10
n
0
5
10
15
20
25
30
35
40
of samples
11
Species Diversity Estimates Across Time
1997
1976
Glebokie
70
50
30
10
Cumulative of species
Zelwazek
Inulec
45
70
35
50
25
30
15
10
5
n
n
25
30
35
40
0
5
10
15
20
25
30
0
5
10
15
20
of samples
12
Statistical Analysis

• Different curves had different sample sizes, thus
  statistical analysis was done on rarefied data
• Rarefaction to lowest common sample size

N14 in my case
Crist Veech (2006) Ecol. Lett.
samples
13
Biodiversity Estimates

• all Pgt0.05
• all lakes were similar with regard to predicted
  diversity, and temporal differences were minor

14
Coefficient of Sorensen
Lake
Majcz Zelwazek Glebokie Inulec
Majcz 0.77 Zelwazek 0.78 Glebokie 0.77
Inulec 0.74
Lake
1976 vs. 1997

15
Coefficient of Sorensen
Lake
Majcz Zelwazek Glebokie Inulec
Majcz 0.77 0.78 0.85 0.80 Zelwazek 0.78 0.80 0.
78 Glebokie 0.77 0.85 Inulec 0.74
Lake
1976 similarities between lakes in 1976
16
Coefficient of Sorensen
Lake
Majcz Zelwazek Glebokie Inulec
Majcz 0.77 0.78 0.85 0.80 Zelwazek 0.73 0.78 0.8
0 0.78 Glebokie 0.75 0.80 0.77 0.85 Inulec 0.72
0.75 0.81 0.74
Lake
1997 similarities between lakes in 1997
17
Results
Overall Diversity Estimation Results

• Similar diversity existed in all four lakes and
  across time
• Shannon Index revealed no clear relation to
  rarefied diversity estimate or evenness across
  years for the 4 lakes
• These patterns are remarkable given the long time
  period that elapsed between sampling (21 years)

Overall Sorensens Result Between years
(0.77 0.02) Among lakes in 1976 (0.81
0.03) Among lakes in 1997 (0.76 0.04)
Communities were as similar in different lakes as
between years in the same lakes
18
Acknowledgements

• Special thanks to
• Dr. Jolanta Ejsmont-Karabin Polish Academy of
  Sciences
• Jim Muirhead, GLIER
• Dr. Hugh MacIsaac
• And extra special thanks to the rest of the lab,
  especially Sandy, Derek, and Meighen