Evaluation of sampling alternatives to quantify stand structure in riparian areas of Western Oregon forests Theresa Marquardt Oregon State University Paul.

Evaluation of Evaluation of sampling alternatives sampling alternatives

to quantify stand to quantify stand structure in riparian structure in riparian

areas of Western areas of Western Oregon forests Oregon forests

Theresa Marquardt Theresa Marquardt Oregon State UniversityOregon State University

Paul Anderson Paul Anderson USDA Forest Service PNWUSDA Forest Service PNW

June 26, 2007June 26, 2007

OutlineOutline

►IntroductionIntroduction

►MethodsMethods

►Sampling AlternativesSampling Alternatives

►SimulationSimulation

►Preliminary ResultsPreliminary Results

IntroductionIntroduction

IntroductionIntroduction

►What are riparian areas?What are riparian areas?

►What makes them difficult to sample?What makes them difficult to sample?

►How is forest structure defined?How is forest structure defined?

►What is the population of interest?What is the population of interest?

Riparian AreasRiparian Areas

►Three dimensional zones of Three dimensional zones of interaction between terrestrial interaction between terrestrial and aquatic ecosystems extending and aquatic ecosystems extending outward from the channel to the outward from the channel to the limit of flooding and upward into limit of flooding and upward into the canopy of streamside the canopy of streamside vegetation – (Swanson et. al. 1982)vegetation – (Swanson et. al. 1982)

Riparian Areas (Cont’d)Riparian Areas (Cont’d)

►A riparian area is a dynamic ecosystem of vegetation, soils, and living creatures along a river or other water body, where unique ecological conditions exist, mainly due to the interaction and exchanges between the land and water.

(S. Chan, 2004)(S. Chan, 2004)

Riparian areas are Riparian areas are dynamic.dynamic.

Approximate position original bank

Riparian areas are Riparian areas are diverse.diverse.

Reasons for SamplingReasons for Sampling► Discover interactions between aquatic Discover interactions between aquatic

and upland ecosystems.and upland ecosystems.

► Measure important ecological functions:Measure important ecological functions: Wildlife habitatWildlife habitat Stream bank stabilityStream bank stability Nutrient assimilationNutrient assimilation Influence on microclimateInfluence on microclimate Filtration of sediment and debris Filtration of sediment and debris

transported by runofftransported by runoff Large woodLarge wood

► Monitor diversity over timeMonitor diversity over time Complex, dynamic environment serving as Complex, dynamic environment serving as

hotspot of biological diversityhotspot of biological diversity

Stand StructureStand Structure

►Key structural attributes include Key structural attributes include spatial arrangement, canopy spatial arrangement, canopy cover, tree diameter, tree height, cover, tree diameter, tree height, type of foliage, species type of foliage, species composition, deadwood, and composition, deadwood, and understory vegetation.understory vegetation.

(McElhinny et al., 2005)(McElhinny et al., 2005)

Cissel et. al. (2006)

Stream SelectionStream Selection

►Density Management Study (DMS) Density Management Study (DMS) stream reaches in western Oregonstream reaches in western Oregon

►Headwater streams Headwater streams Intermittent, seasonal, perennialIntermittent, seasonal, perennial

Flowing water less than three meters Flowing water less than three meters widewide

Flowing water less than 30 cm in Flowing water less than 30 cm in depthdepth

DMS Site AttributesDMS Site Attributes

►DensityDensity Control: 200-350 TPA. Control: 200-350 TPA. Moderate density: Moderate density: Approximately 80 TPA.Approximately 80 TPA.

►Thinning BuffersThinning Buffers Ranging from 15.24 m to 146 mRanging from 15.24 m to 146 m

ObjectivesObjectives

1.1. Examine the accuracy and suitability Examine the accuracy and suitability of selected sampling methods to of selected sampling methods to quantify forest stand structure and quantify forest stand structure and vegetation of headwater streams.vegetation of headwater streams.

2.2. Examine relationships between Examine relationships between arrangement of forest structure and arrangement of forest structure and microclimate and micro-site microclimate and micro-site attributes.attributes.

3.3. Influence of tree density, slope, and Influence of tree density, slope, and aspect on microclimate in areas of aspect on microclimate in areas of western Oregon.western Oregon.

MethodsMethods

Data CollectionData Collection

►Stream reaches were randomly Stream reaches were randomly selected from a list of headwater selected from a list of headwater streams generated from DMS maps. streams generated from DMS maps.

►Stem MappingStem Mapping Total Station Survey Equipment and SoftwareTotal Station Survey Equipment and Software 72 by 72 m area (0.5184 ha)72 by 72 m area (0.5184 ha) Random start for plot locationRandom start for plot location 9 Stem maps 9 Stem maps

Plot LayoutPlot Layout

Random Starting PointRandom Starting Point

72 m36 m

Stem MapStem Map

Data Collection (Cont’d)Data Collection (Cont’d)

►Attributes Recorded for Each Tree: Attributes Recorded for Each Tree: DBH trees larger than 7.5 cmDBH trees larger than 7.5 cm SpeciesSpecies Canopy Classification (Dominant, Co-Canopy Classification (Dominant, Co-

dominant, Intermediate, Suppressed)dominant, Intermediate, Suppressed) Condition (Dead, Live)Condition (Dead, Live) Decay Class (1, . . ., 5)Decay Class (1, . . ., 5) Crown ClassificationCrown Classification

Sampling Alternatives Sampling Alternatives

Sampling AlternativesSampling Alternatives

► Simple Random SamplingSimple Random Sampling Fixed radius circular plotsFixed radius circular plots

► Systematic Sampling with a random startSystematic Sampling with a random start Fixed radius circular plotsFixed radius circular plots Strip cruiseStrip cruise

► PerpendicularPerpendicular► Perpendicular AlternatePerpendicular Alternate

► StratifiedStratified Strip cruiseStrip cruise

► Parallel Parallel

Sampling Alternatives Sampling Alternatives (Cont’d)(Cont’d)

► Two-Stage SamplingTwo-Stage Sampling Fixed area square plotsFixed area square plots Strip cruiseStrip cruise

► Perpendicular one sidePerpendicular one side

► Horizontal Line SamplingHorizontal Line Sampling

► Variable Width Variable Width (Adapted from Roorbach et al. 2001).(Adapted from Roorbach et al. 2001).

► Each alternative will be sampled at an Each alternative will be sampled at an intensity of 10 and 20 % of the 72 mintensity of 10 and 20 % of the 72 m2 2 area.area.

Simple Random Sampling:Simple Random Sampling:Fixed Radius PlotsFixed Radius Plots

Plots

Stream

36 m

72 m

Intensity Plot Size (m)Number of

Plots10% 5.75 220% 5.75 410% 9 520% 9 10

Systematic Sampling:Systematic Sampling:Fixed Radius PlotsFixed Radius Plots

36 m

72 m

Stream

Intensity Plot Size (m)Number of

Plots10% 5.75 220% 5.75 410% 9 520% 9 10

Systematic Sampling: Systematic Sampling: PerpendicularPerpendicular

36 m

72 m

PlotIntensity Plot Size (m)Number of

Plots10% 3.6 x 72 220% 3.6 x 72 410% 7.2 x 72 120% 7.2 x 72 2

Systematic Random Systematic Random Sampling:Sampling:

Perpendicular Alternate Perpendicular Alternate StripsStrips

36 m

72 m

Intensity Plot Size (m)Number of

Plots10% 3.6 x 36 420% 3.6 x 36 810% 7.2 x 36 220% 7.2 x 36 4

Stratified Sampling:Stratified Sampling: Parallel Strips Parallel Strips

1

1

1

1

2

2

2

2

36 m

36 m

Intensity Plot SizeNumber of

PlotsSamples per Strata

10% 3.6 x 36 4 220% 3.6 x 36 8 410% 9 x 28.8 2 120% 9 x 28.8 4 2

Plot

Two-Stage: Two-Stage: Square PlotsSquare Plots

Plot

36 m

14.4 m

Intensity Plot Size (m)Number of

Plots10% 7.2 x 7.2 1020% 7.2 x 7.2 20

72 m

Two Stage: Two Stage: Perpendicular One SidePerpendicular One Side

Intensity Plot Size (m)Number of

Plots10% 3.6 x 36 420% 3.6 x 36 810% 7.2 x 36 220% 7.2 x 36 4

36 m

72 m

Horizontal Line SamplingHorizontal Line Sampling►Lynch (2006)Lynch (2006)

Use of point sampling along a line Use of point sampling along a line to estimate tree attributes without to estimate tree attributes without land area estimation.land area estimation.

B = Baseline Length

Riparian Area

Sample lines

Horizontal Line Sampling Horizontal Line Sampling (Cont’d)(Cont’d)

► A baseline is used to create a uniform A baseline is used to create a uniform distribution with the following distribution with the following probability density function:probability density function:

► Sampled trees are within a limiting Sampled trees are within a limiting distance of each transectdistance of each transect

otherwise

BxBxf

0

0

/1)(

Horizontal Line SamplingHorizontal Line Sampling

►BAF of 8 and 10 metric BAF of 8 and 10 metric ►Use 1 or 2 transectsUse 1 or 2 transects►Baseline length of 72 mBaseline length of 72 m

B

Variable Width DesignVariable Width Design

Centerline

Random Start

Bankfull Channel Edge

Plot width = core &inner zone width

32.8 ft (10m)

65.6 ft(20m)

98.4 ft(30m)

131.2 ft(40m)

164.1 ft(50m)

0 ft (0m)

►Design adapts to curvature in Design adapts to curvature in the stream and bankfull the stream and bankfull channel edgechannel edge

From Roorbach et. Al. (2001)

Variable Width DesignVariable Width Design► Plot width from stream of 25 mPlot width from stream of 25 m

► Centerline length of 20.8 and 41.6 Centerline length of 20.8 and 41.6 for the 10 and 20% intensity for the 10 and 20% intensity respectivelyrespectively

► Both sides of the stream will be Both sides of the stream will be measuredmeasured

► Use stream center rather than Use stream center rather than bankfull width bankfull width

Analysis MethodsAnalysis Methods

Evaluating Sampling Evaluating Sampling AlternativesAlternatives

► Size ClassesSize Classes Diameter Class Distribution Diameter Class Distribution

► 10 cm classes10 cm classes

► Calculate MSE, RMSE, Bias, Relative Calculate MSE, RMSE, Bias, Relative EfficiencyEfficiency Volume per HectareVolume per Hectare Merchantable Volume per HectareMerchantable Volume per Hectare Basal Area per HectareBasal Area per Hectare Merchantable TPHMerchantable TPH TPHTPH

Analysis MethodsAnalysis Methods

► MSEMSE

valueknowntheisYand

ssimulationofnumbertheis

samplethefromestimatetheisYWhere

YYMSE

k

k

k

30

,

30

^

30

1

2^

Analysis MethodsAnalysis Methods

► BiasBias

valueknowntheisYand

ssimulationofnumbertheis

samplethefromestimatetheisYWhere

YYBIAS

k

k

k

30

,

30

^

30

1

^

Analysis MethodsAnalysis Methods

►Percent ErrorPercent Error

►Nonparametric MethodsNonparametric Methods Kruskal-Wallis TestKruskal-Wallis Test

Preliminary ResultsPreliminary Results

Diameter ClassesDiameter Classes

Diameter Class Range

1 7.5 to 15.5 cm2 15.6 to 25.5 cm3 25.6 to 35.5 cm4 35.6 to 45.5 cm5 45.6 to 55.5 cm6 55.6 to 65.5 cm7 65.6 to 75.5 cm8 75.5 to 100.5 cm

Preliminary ResultsPreliminary Results

Diameter Class RMSE

Total RMSE

Diameter Class RMSE

Total RMSE

1 13.75 1 9.752 13.54 2 11.073 17.87 3 13.414 30.17 4 19.92

5 9.56 5 8.796 4.73 6 2.208 2.52 8 2.83

Parallel 3.6m x 36m 10% Parallel 3.6m x 36m 20%

92.13 67.96

Preliminary Results Preliminary Results (Cont’d)(Cont’d)

Diameter Class RMSE

Total RMSE

Diameter Class RMSE

Total RMSE

1 7.90 1 7.582 25.88 2 13.693 31.13 3 11.844 16.40 4 9.005 14.45 5 9.396 5.26 6 2.488 6.19 8 3.08

Parallel 9m x 28.8m 10% Parallel 9m x 28.8m 20%

107.21 57.07

Preliminary Results Preliminary Results (Cont’d)(Cont’d)

Diameter Class RMSE

Total RMSE

Diameter Class RMSE

Total RMSE

1 12.69 1 7.552 11.69 2 8.903 11.78 3 5.154 18.75 4 11.995 10.77 5 7.266 7.19 6 4.148 3.79 8 1.87

3.6 x 72 Perpendicular 10% 3.6 x 72 Perpendicular 20%

76.65 46.85

Preliminary Results Preliminary Results (Cont’d)(Cont’d)

Diameter Class RMSE

Total RMSE

Diameter Class RMSE

Total RMSE

1 13.35 1 9.812 10.33 2 2.743 14.58 3 11.004 16.48 4 11.075 13.95 5 10.506 4.58 6 4.388 3.42 8 1.58

7.2 x 72 Perpendicular 10% 7.2 x 72 Perpendicular 20%

76.70 51.08

Preliminary Results Preliminary Results (Cont’d)(Cont’d)

Sampling AlternativeTotal

RMSE

Parallel 3.6m x 36m 10% 92.13Parallel 9m x 28.8m 10% 107.21Perpendicular 3.6 x 72 10% 76.65Perpendicular 7.2 x 72 10% 76.7Parallel 3.6m x 36m 20% 67.96Parallel 9m x 28.8m 20% 57.07Perpendicular 3.6 x 72 20% 46.85Perpendicular 7.2 x 72 20% 51.08

Preliminary Results Preliminary Results (Cont’d)(Cont’d)

Diameter Class Distribution 3.6m x 72m Sample v. Expected

0.0

8.0

16.0

24.0

32.0

40.0

48.0

56.0

64.0

72.0

1 2 3 4 5 6 7 8

Diameter Class

Num

ber

of Tre

es .

Sampled Population

Preliminary Results Preliminary Results (Cont’d)(Cont’d)Percent Error Total Number of Trees for Eight Samplling

Alternatives

0

0.05

0.1

0.15

0.2

0.25

3.6x3610%

3.6x3620%

9x28.810%

9x28.820%

3.6x7210%

3.6 x72 20%

7.2x7210%

7.2x7220%

Sampling Alternative

Perc

en

t Err

or

ConclusionConclusion

►In this case, strips parallel to the In this case, strips parallel to the stream had a higher mean square stream had a higher mean square error than those running error than those running perpendicular to the stream.perpendicular to the stream.

►This could account for higher This could account for higher variation from stream to upslope variation from stream to upslope than running parallel to the stream than running parallel to the stream bank.bank.

Questions ?Questions ?

ReferencesReferences► Bruce, D. 1981. Consistent Height-Growth and Growth-Rate Estimates for Bruce, D. 1981. Consistent Height-Growth and Growth-Rate Estimates for

Remeasured Plots. For. Sci. 27(4): 711-725. Remeasured Plots. For. Sci. 27(4): 711-725.

► Cissel, J., Anderson, P., Berryman, S., Chan, S., Puettman, K., and Cissel, J., Anderson, P., Berryman, S., Chan, S., Puettman, K., and Thompson, C. Thompson, C. BLM Density Management and Riparian Buffer BLM Density Management and Riparian Buffer Study: Study: Establishment Establishment Report and Study Plan. U.S. Dept. Interior, Report and Study Plan. U.S. Dept. Interior, U.S. U.S. Geological Survey. 2006. Geological Survey. 2006. 161 p. 161 p.

► Lynch, T. B. 2006. Horizontal line sampling for riparian forests Lynch, T. B. 2006. Horizontal line sampling for riparian forests without land without land area estimation. For. Sci. 52(2): 119-129area estimation. For. Sci. 52(2): 119-129

► McElhinny, C., P. Gibbons, C. Brack, and J. Bauhus. 2005. Forest and McElhinny, C., P. Gibbons, C. Brack, and J. Bauhus. 2005. Forest and woodland woodland stand structural complexity: its definition and stand structural complexity: its definition and measurement. For. Ecol. measurement. For. Ecol. Manage. 218 (1-3): 1-24. Manage. 218 (1-3): 1-24.

► Roorbach, A., Schuett-Hames, D., Haight, R., and McGowan, M.. Roorbach, A., Schuett-Hames, D., Haight, R., and McGowan, M.. Field Field Procedures for the Pilot Study to Validate the DFC Performance Procedures for the Pilot Study to Validate the DFC Performance

Targets for West-Side Riparian Prescriptions in Washington's Targets for West-Side Riparian Prescriptions in Washington's Forest Forest Practices Rules. Northwest Indian Fisheries Commission, 2001.Practices Rules. Northwest Indian Fisheries Commission, 2001.

► Temesgen, H. 2002. Evaluation of sampling alternatives to quantify tree Temesgen, H. 2002. Evaluation of sampling alternatives to quantify tree leaf leaf area. Can. J. For. Res. 33: 82-95.area. Can. J. For. Res. 33: 82-95.