Quantifying Landscape Pattern in

1
Quantifying Landscape Pattern in 7
Metrics Malik Marjan, Magan Mazzarino, Bill
DeLuca
2
Framing Question/Objective

• Can we establish gradients in forest extent and
  fragmentation using a suite of metrics for the
  purpose of exploring the relationship b/t
  breeding birds and forest loss fragmentation?

3
Metric Selection
?
4
Metric Selection
5
Gradients of Habitat Loss Fragmentation
6
Habitat loss Gradient
7
Fragmentation Gradient
0.8839
0.5938
0.8829
0.5999
0.8316
0.6043
0.6753
0.7052
0.7326
0.7416
0.7435
0.8219
8
Core Area Gradient
9
Gradients of Habitat Loss Fragmentation
Pland vs Core Area (awm)
Core Area (awm)
Pland
10
Gradients of Habitat Loss Fragmentation
Core Area AWM
High Habitat Loss Low Habitat Loss
11
Gradients of Habitat Loss Fragmentation
Pland vs Clumpiness
Clumpiness
Pland
12
Gradients of Habitat Loss Fragmentation
Clumpiness
P Land
High Habitat Loss Low Habitat Loss
13
Gradients of Habitat Loss Fragmentation
Pland vs Total Edge Contrast Index
TECI ()
Pland
14
Gradients of Habitat Loss Fragmentation
Total Edge Contrast Index
High Habitat Loss Low Habitat Loss
15
Gradients of Habitat Loss Fragmentation
Pland vs Similarity Index (awm)
Similarity Index (awm)
Pland
16
Gradients of Habitat Loss Fragmentation
Pland vs Number of Patches
of Patches
Pland
17
Gradients of Habitat Loss Fragmentation
Pland vs Number of Disjunct Core Areas
of Disjunct Core Areas
Pland
18
(No Transcript)
19
Grid 2108 PLAND 21.5
Grid 1943 PLAND 22.5
20
Grid 574 PLAND 32.5
Grid 1655 PLAND 30.75
21
Grid 1892 PLAND 38.5
Grid 1727 PLAND 40
22
Grid 2155 PLAND 47.25
Grid 1927 PLAND 51.5
23
Grid 1957 PLAND 69.5
Grid 1522 PLAND 71.5
24
Connectivity Index
25
Connectivity Index
c
3
1) Total Core Area X
2
b
a
2) Euclidean Distance NN a 150m (150m
through Res) b 600m (150m For, 150m Res,
300m Urb) c 100m (100m Urb)
1
3) Weigh distances based on class Patch 1 Patch
2 Patch 3 Forest 0 a 225 a
225 b 825 Res 1.5 b
825 c 200 c 200 Dev 2
Patch Index 525 212.5 512.5
4) Class index 416.67
5) Landscape Index Mean for each class
26
Connectivity Index
Patch 1 Patch 2 Patch 3
a 225 225 b 300 300 c
225 225 Mean 262.5 262.5 225
250
Patch 1 Patch 2 Patch 3
b
2
a 1225 1225 b 300 300 c
1225 1225 Mean 762.5 762.5 1225
1
a
c
3
758.33
27
Connectivity Index
c
3
2

• Functional
• Landscape mosaic
• Patch, Class,or Landscape

b
a
1

• Strengths Limitations
• - Fairly interpretable - Mean of means
• Incorporates corridors - Patch 0 1 are
• - Barriers - ???????
• - Comparable across
• similar landscapes
• - Computationally simple

28
Framing Question/Objective

• Can we establish gradients in forest extent and
  fragmentation using a suite of metrics for the
  purpose of exploring the relationship b/t
  breeding birds and forest loss fragmentation?

29
Sampling Analysis Method

• Grain 50 m cell size (0.25 ha)
• Extent 14 1-km square landscapes
• Thematic Resolution 6 Broad Classes
• From the perspective of the organism

Hermit Thrush
Winter Wren
Black-Throated Green Warbler
30
Perspective

• Landscape mosaic model
• Organism centered- breeding birds
• Process oriented Over what scales are the
  processes causing fragmentation occurring?
• What scale best describes the system we are
  interested in studying?

31
Scaling Limitations

• Grain (Thompson McGarigal 2002)
• a. Start small (lt0.25 ha)
• b. Increase in small increments
• c. at each unit, perform chi-square test of
  independence

50m grid size may be appropriate for nest site
selection
32
Sampling Analysis Method

• 2. Extent (Thompson McGarigal 2002)
• a. Start at a reasonable extent
• b. increase extent in increments
• c. Index of habitat quality for each extent
  (composition and the species preference)

33
Scaling Limitations - Conclusions

• Processes causing fragmentation (human
  development) are acting at scales greater than 1
  km squared
• Organism perspective (extent) is greater than
  what we are considering
• Grain size may be appropriate
• Thematic resolution should be finer.

34
(No Transcript)
35
Connectivity Index
b
2
1) Total Core Area 50
1
a
c
3
2) Euclidean Distance NN a 150m (150m
through Res) b 200m (100m Res, 100m Dev)
3) Weigh distances based on class Patch 1 Patch
2 Patch 3 Forest 1.0 a 0
(barrier) b 80 a 0 Res .5
b 80 c 0 c 0 Dev .3
Patch Index 40 40 0 Barrier 0 or 1
4) Class index 26.67
5) Landscape Index Mean for each class
36
Connectivity Index
c
3
2
b
a
1
Patch 1 Patch 2 Patch 3
a 75 75 b 405 405 c
30 30 Total 480 105 435 Mean 240 52.5 217
.5
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)
40
(No Transcript)
41
1522
880
574
1655
1454
1957
2108
1993
2155
1943
1927
1727
1892
1831