Rise and Fall and Rise of Submerged Aquatic Vegetation in the Tidal Freshwater Hudson River

Rise and Fall and Rise of Submerged Aquatic Vegetation in the Tidal Freshwater Hudson River Stuart Findlay – Cary Institute of Ecosystem Studies Maile Neel – University of Maryland

NEXT IRENE KNOWLEDGE The Science Underpinnings – Why SAV?

Submerged Vegetation 1. Inventory 2. Function 3. Change It hits the fan

I t

h i t s

t h e

f a n

From TechEBlog.com Mohawk River at the Cohoes Falls Quick History

• SAV (Vallisneria) ~ 6% River bottom area in 2007 • Almost completely extirpated by TS Irene Aug 2011 • Loss documented by volunteers and others • Loss due to burial of overwintering “buds” [Hamberg et al 2017, Restoration Ecology] Polgar Fellowship Status 2017 • Volunteer observations – good for year-to-year, small-scale change – Two examples to follow

• Other reports [targeted observations]

• Aerial inventories 2014, 2016 (Sarah Fernald, HRNERR)

SAV Present (%) 50SAV Volunteer Data

10 Irene 5

2010 2011 2012 2013 2014 2015 2016 Near Stockport Flats 2016 and 2017

Mostly “No”

1 km

Mostly “Yes”

7 Catskill 2017

YES NO

50 m Irene did not “impair” SAV Habitat Good News – Don’t have to repair the physical template • Light is key – Water column turbidity • HRECOS (Schodack, Norrie) • Point measures at SAV sites • Main channel clarity (LTREB stations)

– Depth of prior SAV documented habitat • Site bathymetry Turbidity 2011 HRECOS Schodack Island

Norrie Point Norrie Point 2012 Summer turbidity back to “normal” 2011 • “Keystone” sites monitored since 2001 • Plant presence, water quality

Peekskill

Iona

Esopus Meadows

Cruger Island

GreenFlats

9.5 Post - Irene Dissolved Oxygen Iona 9 Green Flats 8.5

8 “Recovered”

7.5

6.5 Still Absent 6

5.5

5 2000 2005 2010 2015 2020

SAV Site Attributes Point measures of water quality in SAV sites 35 before/after Irene DO2000 vs DO2015 NTU2000 vs NTU2015 30 Z dm 2000 vs z dm 2015 1:1 25

2015 15

10 DO, turbidity and depth at 5 bed sites similar for 2000 vs 2015 0 0 5 10 15 20 25 30 35 2000 Bathymetry South of Iona Island (no SAV in 2017, planting was successful, A. Carew Polgar) Trial Plantings Largely Successful • Locations (positive results) – Behind Cruger I. [Hamberg – 2014] – Norrie (R. Houser/C. Bowser) – Iona, Esopus Meadows

• Locations (unsure/negative) – Mill Creek (?) – Kingston-Rhi Bridge (?) – N Germantown (?) – Stuyvesant - negative SAV Can Come Back! (slowly) Maile Neel – University of Maryland

NEXT IRENE KNOWLEDGE Rise and Fall and Rise of Submerged Aquatic Vegetation in the Tidal Freshwater Hudson River

Maile Neel Katia Engelhardt The potential for resiliency

Acclimation Phenotypic plasticity

Adaptation Phenotypic variation Genetic variation Natural dispersal Relocation Local Restoration Assisted migration Extirpation Maile Neel, UMCP Katia Engelhardt, UMCES, Appalachian Lab

Michelle Gray PLSC

David Thulman GWU Annie Carew M.S. Student MEES Polgar Fellow ‘17 Natalia Noyes Carrie Perkins BISI/Math Ph.D. Student PLSC Calm Before the Storms - 2011 Brittany Marsden

Collected V. americana samples at 5 sites for genetic analysis and propagation for experiments.

The storms provide a natural experiment in which we can quantify the effects of extreme population reductions on genetic diversity

Revisit 2011 sites to see if individuals persisted and to quantify overall levels of diversity. Quantify genetic diversity in areas not sampled in 2011. Determine the sources of recovering sites. NB Bridge Assess the role of environmental characteristics in Bannerman Island structuring genetic variation. Garrison Propagate additional samples for use in future Iona experiments, restoration, and outreach. Croton Calm Before the Storms - 2011 Brittany Marsden

Collected V. americana samples at 5 sites for genetic analysis and propagation for experiments.

The storms provide a natural experiment in which we can quantify the effects of extreme population reductions on genetic diversity

Revisit 2011 sites to see if individuals persisted and to quantify overall levels of diversity. Quantify genetic diversity in areas not sampled in 2011. Determine the sources of recovering sites. NB Bridge Assess the role of environmental characteristics in Bannerman Island structuring genetic variation. Garrison Propagate additional samples for use in future Iona experiments, restoration, and outreach. Only one site from 2011 supported Croton Vallisneria in 2015 Stillwater Mechanicville Sampling 2015 Sampled from 24 sites

Albany

Stockport Brandow Pt

Cheviot Tivoli

Rondout Creek

NB Bridge Bannerman Island Garrison Iona Annsville Creek

Croton Nyack Stillwater Mechanicville Sampling 2015 Sampled from 24 sites

Albany

Stockport Brandow Pt In Type of Sample Genotyped Propagation Cheviot 2015 Plants 328 328 Tivoli 2015 Leaves 620 0 Rondout Creek 2015 Floating plants 7 0

NB Bridge Bannerman Island Garrison Iona Annsville Creek

Croton Nyack Stillwater Mechanicville Sampling 2015 Sampled from 24 sites

Albany

Stockport Brandow Pt In Type of Sample Genotyped Propagation Cheviot 2015 Plants 328 328 Tivoli 2015 Leaves 620 0 Rondout Creek 2015 Floating plants 7 0 NY DEC 2014 Samples 23 0

NB Bridge Bannerman Island Garrison Iona Annsville Creek

Croton Nyack Stillwater Mechanicville Sampling 2015 Sampled from 24 sites

Albany

Stockport Brandow Pt In Type of Sample Genotyped Propagation Cheviot 2015 Plants 328 328 Tivoli “Natural” 2015 Leaves 620 0 Rondout Creek 2015 Floating plants 7 0 NY DEC 2014 Samples 23 0 Nursery plants 2015 32 32 Nursery experiment 2014 44 0 NB Bridge 2015 Total 1,054 360 Bannerman Island 2011 samples 140 93 “Natural” Garrison Grand Total 1,194 453 Iona Annsville Creek

Croton Nyack Collect georeferenced leaves and shoots Turning Point

740 “base” samples collected at the same distance interval as the 2011 samples:

328 plants from 9 sites Plants Leaves 412 leaves from 15 sites

Antiona 208 leaf samples collected within 1-2 m of other plants to better quantify extent of genotypes within sites

7 floating plants Collect georeferenced Propagate collected plants and leaves plants

Process samples

Genotype each sample at 10 microsatellite loci

Extract DNA Microsatellite Alleles

IDNum Site aagx051 aagx030 m49 aag004 aagx012 aag002 atg002 aagx071 m13 m16 NYA24 Iona 178:181 159:159 376:391 168:168 214:214 322:325 160:160 233:233 271:271 187:187 Ho = 0.3 Heterozygous Homozygous

University of Nebraska Heterozygosity is important because of the well known effects of inbreeding Microsatellite Alleles

IDNum Site aagx051 aagx030 m49 aag004 aagx012 aag002 atg002 aagx071 m13 m16 16 NYA24 Iona 178:181 159:159 376:391 168:168 214:214 322:325 160:160 233:233 271:271 187:187 Ho = 0.3

Alleles at all loci combined the Multi-Locus Genotype Microsatellite Alleles

IDNum Site aagx051 aagx030 m49 aag004 aagx012 aag002 atg002 aagx071 m13 m16 16 NYA24 Iona 178:181 159:159 376:391 168:168 214:214 322:325 160:160 233:233 271:271 187:187 Ho = 0.3 16 NYA25 Iona 178:181 159:159 376:391 168:168 214:214 322:325 160:160 233:233 271:271 187:187 Ho = 0.3

Alleles at all loci combined the Multi-Locus Genotype Microsatellite Alleles

Alleles at all loci combined the Multi-Locus Genotype

How many alleles? How are alleles organized into individuals (i.e. heterozygosity and genotypes)? How many genotypes? How are alleles and genotypes distributed in the River? How different are populations from one another? How are those differences distributed in the River? Allelic Diversity Total number of alleles 65 (n=1,119)

Group N Alleles 2015 Base 740 60

Compare to 2011 Group N Alleles Hudson 140 44 Potomac 150 58 Kennebec River 136 62 Stillwater Mechanicville Allelic Diversity Total number of alleles 65 (n=1,119)

Upper Group N Alleles Rarefied to 140 2015 Base 740 60 54.7 Albany

Compare to 2011 Stockport Brandow Pt Group N Alleles Mid Hudson 140 44 Cheviot Potomac 150 58 Tivoli Kennebec River 136 62 Rondout Creek

Lower NB Bridge Bannerman Island Garrison Iona Annsville Creek

Croton Nyack Stillwater Allelic Diversity Total number of alleles 65 (n=1,119)

Group N Alleles Rarefied to 140 2015 Base 740 60 54.7

Compare to 2011 Stockport Greenport Group N Alleles Hudson 140 44 Potomac 150 58 Kingston Kennebec River 136 62 Esopus Meadows

NB Bridge

West Point Garrison

DEC Nyack Stillwater Mechanicville Observed Heterozygosity

Albany Observed heterozygosity is high!

 No indication of inbreeding. Stockport Brandow Pt Heterozygosity is lowest in the non-tidal sites, but Cheviot even those sites are not extremely low. Tivoli Rondout Creek Lack of inbreeding confirmed when observed heterozygosity is compared with expected.

Bannerman Island Garrison Iona

Croton Nyack Genotypic Diversity 1,111 natural samples with no missing data yielded 470 distinct genotypes

#퐺푒푛표푡푦푝푒푠−1 GD = = 0.42 #푆푎푚푝푙푒푠−1

903 total base + DEC +2011 samples yielded 422 genotypes, GD=0.47

32 Nursery samples yielded 1 genotype! Genotypic Diversity 2011 vs. 2015 Group N # MLGs GD 2015 Base 740 348 0.47

Group N # MLGs GD 2011 Hudson 140 74 0.52 2011 Potomac 150 129 0.86 2011 Kennebec 136 109 0.80 Genotypic Diversity 2011 vs. 2015 Group N # MLGs GD 2015 Base 740 348 0.47 2015 Rarified 140 95 0.68

Group N # MLGs GD 2011 Hudson 140 74 0.52 2011 Potomac 150 129 0.86 2011 Kennebec 136 109 0.80

Different parts of the River have different clonal diversity Stillwater Genotypic Diversity by Population 2015 2011

Median GD 2015 = 0.48 Stockport 2011 = 0.53 Greenport

Cheviot

Median # Genotypes Bannerman Island 2015 = 13 Garrison 2011 = 12 George’s Island

Nyack How Big are the Genotypes?

472 total genotypes

129 were detected more than once

 Indicates 27% of the genotypes contribute to vegetative reproduction at the scale we sampled.

No difference between base samples 2015 (25.5%) and 2011 (28%) How Big are the Genotypes?

129 genotypes detected more than once

For each genotype calculate Maximum distances among samples Minimum convex hull encompassing samples Genotypic Diversity The vast majority of genotypes are small

median maximum distance =22 m

But some genotypes are of considerable size

median area =649 m2

22 genotypes were found at multiple sites Stillwater Mechanicville Genotypic Diversity

Genotypes were shared between sites in three zones

 Indicates vegetative dispersal or prior continuous connected distributions Nutten Hook  Distance and water depth make continuous Hudson distributions unlikely

Tivoli

7 clonal genotypes from 2011 were found again in 2014 or 2015 NB Bridge 4 were 2011 clones from Croton that were detected at Croton again in 2015

George’s Island 3 were found at different sites! Croton Is this a huge population with lots of adaptive potential or one individual?

Low numbers of genotypes yield lower - growth and survival (Williams 2001, Hughes et al. 2009) - resistance to disturbance (Hughes and Stachowiz 2004) - growth (Engelhardt et al. 2014) Diversity effects Effects of particular individuals Reproduction may be reduced or precluded when large clones dominate sites

Engelhardt et al. 2014 Female plant Male plant Stillwater Mechanicville Differentiation Among Sites

Albany Overall Fst: 0.13

Compare to 2011 Brandow Pt Roger’s Island Hudson = 0.12 Cheviot Potomac River = 0.03 Tivoli Kennebec River = 0.16 Rondout Creek

Indicates limited gene flow among at least some sites. … But which ones? West Point

Iona

Croton Nyack Stillwater Mechanicville Spatial Structure

Used the Bayesian approach in the program STRUCTURE to Albany 1) Identify the most likely number of genetic populations based on minimizing deviations from Hardy-Weinberg Equilibrium using the 373 unique genotypes Brandow Pt Roger’s Island 2) Assign individuals to those populations Cheviot Tivoli

Rondout Creek

West Point

Iona

Croton Nyack Stillwater Mechanicville Spatial Structure

Used the Bayesian approach in the program STRUCTURE to Albany 1) Identify the most likely number of genetic populations based on minimizing deviations from Hardy-Weinberg Equilibrium using one sample of each unique genotype in each Brandow Pt Roger’s Island population. Cheviot 2) Assign individuals to those populations Tivoli

Rondout Creek

West Point

Iona

Croton Nyack Stillwater Mechanicville Spatial Structure

Rondout Creek

West Point

Iona

Croton 2011 samples not yet analyzed Nyack Preliminary Hudson River Vallisneria Findings

• Relatively low genotypic and allelic diversity – but already low in 2011. • Nursery plants are all one genotype! • Vegetative reproduction common – little observed fruit production • No significant inbreeding, in fact observed heterozygosity is high. • Restricted gene flow is indicated • Differentiation among sites yields 3 “genetic populations” • Detected 3-4 “zones” of vegetative dispersal • Role of adaptation versus dispersal limitation remains to be determined. Acknowledgments Funding: Hudson River Foundation

• David Fischer • Gavin Lemley (NY-DEC) • Kara Hawkins (Appalachian Lab) • Hudson River National Estuarine Research Reserve • Hudson River Estuary Program • NY Department of Environmental Conservation

Recommendations to Date

• Write Management Plan based on what we learned

• We need to be better poised to react quickly storm events (similar to Irene/Lee/Sandy or otherwise)

• Maintain genetic repository for rapid restoration response with local genotypes Open Questions

• Is partial recovery over 6 years “acceptable”?

• Functional equivalence of various genotypes?

• Merits of “locals only” versus “mix ‘em up and let nature sort it out” approaches

• Consequences for fishes? Next steps

Under Existing Project Better quantify relationships between genetic diversity and salinity  Statistical analysis with current data to disentangle spatial proximity and environment  Analysis of single nucleotide polymorphisms to detect selection. Future Work  More reciprocal transplant experiments to quantify fitness differences in different environments.  Continue to monitor recolonization and sources using aerial survey data to target genetic samples.  Sample immediately after catastrophic events to better quantify survival and timing of regrowth.