Title: Dynamics of Giant Kelp Forests: The Engineer of California
1Dynamics of Giant Kelp Forests The Engineer of
Californias Nearshore Ecosystems Dave Siegel,
Kyle Cavanaugh, Brian Kinlan, Dan Reed, Phaedon
Kyriakidis, Stephane Maritorena, Steve Gaines,
Kristin Landgren UC Santa BarbaraDick
Zimmerman Victoria HillOld Dominion University
2Macrocystis pyrifera Giant Kelp
- High economic ecological importance
- Ecosystem engineer of the nearshore ecosystems
- Source of natural products
- Dominant canopy forming macroalga in So Cal
- Highly dynamic
- Plant lifespans 2.5 years
- Frond life spans 4 months
- Fronds growth can be 0.5 m/day
3Macrocystis Fish Stocks
- Growth and mortality regulated by water temp,
nutrients, light, depth, bottom type, predation,
wave action
PDO Shift
El Nino
El Nino
Kelp biomass data from Kelco visual estimates
Fish observations from Brooks et al 2002
Reed et al. 2006
4Macrocystis Dynamics
- Growth
- Nutrients seawater temperature
- Mortality / Disturbance
- Wave action (esp. storms), senescence, predation,
DOC release, etc. - Colonization
- Spore dispersal, benthic light levels, depth,
substrate type, etc.
5(No Transcript)
6Research Goals
- Understand variability of giant kelp canopy cover
carbon biomass - High resolution satellite imagery (SPOT, AVIRIS,
etc.) informed by SBC-LTER observations - Develop models of kelp forest dynamics
- Light utilization gross / net primary
production - Patch dynamics models of canopy cover
7Research Area
8Remote Sensing of Macrocystis with Multispectral
Imagery
- Surface canopy of giant kelp exhibits high near
infrared (NIR) reflectance - SPOT imagery well suited to differentiate kelp
9Methods Canopy Cover
- Perform dark pixel atmospheric correction
- Principal components analysis to separate
residual surface signal (PC1) from kelp (PC2)
PC band 1
- Positive contribution from all 3 bands
- Glint, sediment loads, atmosphere variations,
etc.
False color SPOT image (8/15/2006)
PC band 2
- High NIR, low green and red reflectance
- Kelp
10Methods Canopy Cover Classification
- Minimum kelp threshold value selected from
99.9th-tile value of offshore (35-60 m) pixels
11Validation Canopy Cover
- Cover measurements compared with high resolution
2004 CDFG aerial kelp survey
SPOT Oct 29, 2004
CDFG Sept-Nov 2004
r2 0.98 p 0
12Kelp Occupation Frequency Jan 2006- May 2007
- 8 image dates
- 39 of occupied pixels were present in at least
half the scenes - 4 of pixels were present across all dates
13Kelp Forest Biomass
- Useful for understanding modeling ecosystem
interactions - NPP, turnover, export, etc.
- Difficult to measure directly
- Time and effort intensive
- BUT SBC-LTER does monthly surveys
14Research Area
15SBC-LTER Diver Surveys
- Monthly measurements of kelp forest attributes at
Arroyo Quemado, Arroyo Burro Mohawk Area - Assessment of areal kelp biomass, frond/blade
density, net primary production, etc. - Sampling for 160 m2 transect
- About 16 SPOT 5 pixels
16Seasonal kelp biomass changes along 3 LTER
transects
- Maximums in late 2002
- Wave driven seasonality apparent
17Methods Biomass
- Normalized Difference Vegetation Index (NDVI)
- (NIR-RED)
- (NIRRED)
- Calculated for areas of kelp cover
NDVI Transform
18Empirical Estimation of Kelp Biomass from SPOT
- Provides path to the remote estimation of kelp
biomass (kg/m2) - Enables
- regional assessment
- high temporal resolution views with multiple
scenes
r2 0.71
r2 0.71 n 37
19Seasonal Kelp Forest Changes
20Regional Kelp Biomass
- Created from biomass-NDVI relationship for areas
of kelp cover
21Validation using Visual Biomass Observations
r2 0.73 p lt 110-7
22- Spectra obtained from airborne inaging
spectrometers are similar to lab measures of
individual kelp blade reflectances
Optical estimates of kelp physiological state
23Area and Productivity Estimates Depend on Spatial
Resolution
- Bias increases as spatial resolution decreases
- Not a linear function of spatial resolution
- Resolution classes result from
- inherent scale of kelp patches
- spectral averaging as pixel size increases
24Metapopulation Modeling
Patch 16
Patch 17
gt500 m
Patch 18
Patch 19
Bed 28
Bed 27
25Next Steps
- Acquire as much imagery as possible
- Characterize kelp forest variability
- Patch-level description of occupancy, etc.
- Estimate regional scale kelp forest NPP
- Assess disturbance factors (waves, etc.)
- Space/time modeling of kelp cover biomass
- Predict probability of where / when kelp changes
- Driven by substrate / disturbance / etc.
- Compare kelp gross photosynthesis to NPP
26Thank You!!
27ISP Alginates Visual Kelp Biomass
Kelp canopy biomass, 34-year monthly time series
1000
100
10
0
Canopy Biomass (tons/km coast)
Raw data provided by D. Glantz, ISP Alginates,
Inc. Santa Barbara Coastal LTER
28Regional Kelp Biomass
11/2004 15000 metric tons
UCSB
11/2006 7800 metric tons
04/2007 22358 metric tons