15
2006 SAV Forecast Michael Williams (UMCES/CBPO) TMAW, 12/7/06

2006 SAV Forecast Michael Williams (UMCES/CBPO) TMAW, 12/7/06

  • Upload
    lucas

  • View
    26

  • Download
    0

Embed Size (px)

DESCRIPTION

2006 SAV Forecast Michael Williams (UMCES/CBPO) TMAW, 12/7/06. 2006 Forecast. Case studies for 2006 forecast : Susquehanna Flats, Tangier Sound, Choptank River, Potomac River, York/Mobjack River, Patuxent River March - May 2006 field trips to assess the spring abundance of SAV - PowerPoint PPT Presentation

Citation preview

Page 1: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

2006 SAV ForecastMichael Williams (UMCES/CBPO)

TMAW, 12/7/06

Page 2: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

2006 Forecast

• Case studies for 2006 forecast:– Susquehanna Flats, Tangier Sound, Choptank River,

Potomac River, York/Mobjack River, Patuxent River

• March - May 2006 field trips to assess the spring abundance of SAV

• Development of site-specific relationships April-May

• 2006 SAV forecast ready by late May

Page 3: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

Dependent variables

• SAV by area coverage, % change from previous year, and density

Independent variables

• 1) Previous year’s growing season and 2) Spring:

-flow, N and P loads (RIMP and BFL point sources), flow-adjusted loads, TSS, Chl a, Secchi, salinity, surface water temp, rate of change in spring surf water temp, PAR, and wind speed and direction

-min, max, median, and variance were also analyzed for some parameters

Page 4: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

• Results (TANMH)– Correlations

• PY Secchi (r=0.66, p=0.002)

• PY wind speed (r=-0.60, p=0.018)

– Univariate (area coverage)• PY Secchi

– Stats: r2 = 0.55, p = 0.0016

– Stepwise MLR (at p<0.05)• Same as univariate model

Page 5: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

• Results (Potomac)– Correlations

• Flow-adj. TN (r=-0.59, p=0.011)

• BFL TN loads (r=-0.47, p=0.047)

• Surf. Temp (r=-0.41, p=0.083)

– Univariate (area coverage)• Flow-adj. TN load

– Stats: r2 = 0.34, p = 0.01

– Stepwise MLR (at p<0.05)• Same as univariate

– Class 3 density• Secchi

– Stats: r2 = 0.32, p = 0.03

Page 6: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06
Page 7: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06
Page 8: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

• Results (CB1TF)– Correlations

• Flow-adj. TN (r=-0.66, p=0.002)

• PY wind speed (r=-0.68, p=0.001)

– Univariate (area coverage)• Flow-adj. TN load

– Stats: r2 = 0.46, p = 0.002

– Stepwise MLR (at p<0.05)• Same as univariate

– Class 2 density• Secchi

– Stats: r2 = 0.73, p < 0.0001

Page 9: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

r2=0.46, p=0.002

Page 10: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

Relationships of area coverage (AC) and various independent

parameters for the six SAV communities used in this analysis.

Site

Univariate model

Coef. of determination

& p Multiple linear

regression

Coef. of determination

& p

Susquehanna Flats (CB1TF)

PY Flow-adj. TN load

PY wind speed at BWI

Class 3 density vs PY Secchi

0.46 (0.002)

0.52 (0.0004)

0.73 (0.0001)

PY Flow-adj. TN load 0.46 (0.002)

Patuxent (PAXTF, OH, and MH)

PY BFL PS TN loads

0.67 (0.0004) Same as univariate Same as univariate

Potomac (POTTF, OH, and MH)

Flow-adj. TN loadClass 3 density vs

PY Secchi

0.34 (0.01)0.32 (0.03)

Same as univariate Same as univariate

Potomac MH (POTMH)

PY BFL PS TN loads

0.85 (0.001) Same as univariate Same as univariate

Choptank(CHOMH1)

PY Max PAR 0.35 (0.007) PY Max PAR + PY BFL TN load

0.65 (0.01)

Tangier Sound (TANMH)

PY Secchi 0.55 (0.0016) Same as univariate Same as univariate

Mobjack Bay(MOBPH)

PY TP BFL 0.37 (0.013) PY BFL TP + PY BFL TN + Apr-Oct PAR

0.79 (0.026)

Page 11: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

2006 Forecast• Univariate and/or multiple linear regressions explain 32% to

85% of SAV variance• Models indicate that most case study areas will be similar or

decrease in SAV area; Pax & MH Potomac will increase• These are somewhat in contrast to predictions from field

observations• Effect of reduced N loads needs to be viewed with caution

(i.e., is N a causative factor?). However, it is interesting that areas less directly affected by N loading (i.e., Tangier and Mobjack) do not show the increases in SAV area observed in several other case study areas

Page 12: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06
Page 13: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06
Page 14: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

December Update• High salinity areas (e.g., Tangier): Decrease due to 2005

dieback• Mid-salinity areas: Mixed bag. Even within a single system

SAV is both up and down. Decrease in Honga and increase off Barren Island. Increase at Bloodsworth and South Marsh islands (widgeongrass)

• Lower salinity areas: Abundant around Flats, Elk and Bohemia. Big decrease in the Chester and around Eastern Neck Islands

• TF Potomac: Great but still little regrowth from the WW bridge to DC. Good in St Mary's , Breton and St Clements bays

• Choptank: Gone in north parts during flights (Harris and Broad creeks) but abundant in the south side around Chapel Creek – (JJ’s comment: “go figure - of course it’s widgeongrass again!”)

• Severn and Magothy rivers: abundant SAV, but milfoil increase (2005 Patapsco) gone in 2006 (higher spring salinities?)

Page 15: 2006 SAV Forecast Michael Williams (UMCES/CBPO)  TMAW, 12/7/06

Potential Improvements

• New case study areas

• Other parameters (independent & dependent)– SAV cover = Density_class_1_area * 0.189 +

Density_class_2_area * 0.369 + Density_class_3_area * 0.538 + Density_class_4_area * 0.755

• Others?