1. THE BENEFITS AND CONSEQUENCES OF RESTORATION USING
SELECTIVELY-BRED, DISEASE-TOLERANT OYSTERS
Matthew P. Hare and Colin Rose
Department of Biology
University of Maryland
Abstract: (250 words or less)
Restoration efforts in Chesapeake Bay are now focused on targeted population supplementation using artificially selected, disease tolerant C. virginica to combat high mortalities from parasitic diseases. This tactic has potential benefits and considerable risks. I will argue that genetic testing of recruitment is necessary to evaluate and manage the genetic risks of supplementation with inbred oysters, and under some circumstances it also can provide a much needed direct measure of overall restoration efficacy. Results will be presented from a highly collaborative effort to genetically monitor DEBY-strain restoration plantings in two Chesapeake subestuaries. The results indicate that the DEBY broodstock currently used to produce restoration oysters is genetically bottlenecked and contributing to severe inbreeding in those tributaries where restoration has had the greatest numerical impact. The consequences of this inbreeding will be discussed and recommendations will be offered for the precautionary use of artificially selected strains in restoration.

2. Talk Outline
Rationale behind supportive breeding using artificially-selected oyster strains
pros and cons
Risks are now predictable
Modeling impacts of supportive breeding
Inbred broodstock – a critical factor
Recommendations
Gone from ignorance to one of the best examples of genetic analysis of supportive breeding efficacy and risks

3. The Experts Weighed In, 2000

4. Supportive Breeding Goals and Impacts
Increase population size
limit early mortality in hatchery, then release juvenile “seed” oysters
Hatchery broodstock and restoration seed represent a genetically bottlenecked subset of population
Genetic diversity summarized by inbreeding effective population size, Ne
Planting density important for sessile/sedentary spp with external fertilization

5. Consensus Points from Allen & Hilbish 2000
“stocking programs [supportive breeding] will be important…” for restoration success
“Diseases…are a major limitation” for restoration
“Continued use of selected disease resistant stocks is warrented…”
Benefits
Genetic measurement of restoration efficacy
Disease tolerance  greater longevity of seed oysters
“genetic rehabilitation”, infusing desirable alleles
Risks
Increase overall inbreeding
Decrease genetic variation
Lower mean fitness of population
“Effective population size of wild populations is an essential parameter to predict genetic effects, but is unknown”

6. Broodstock Sources and Numbers of Seed Planted in Virginia Supportive Breeding
3.7x107
Oysters Planted
1999 workshop where it was decided that there’s not much to lose taking this risky approach
1. Live longer on reef, 2. inject alleles for disease-tolerance into wild pop, 3. be able to genetically identify progeny of restoration oysters
T. Leggett, Chesapeake Bay Foundation

7. Risks of Supportive Breeding
Can severely reduce total Ne
total Ne with % hatchery contribution x:
1 x (1-x)2
Ne(tot) Ne(hatchery) Ne(wild)
No estimates for these parameters in 1999 when ‘genetic rehabilitation’ recommended
― = ―
Ryman & Laikre 1991
Single generation Ne, assuming equal fitness of hatchery seed and wild oysters

8. Great Wicomico River, 2002
Tidal physics suggest ‘trap-like’ for retention of larvae

9. Predicted Consequences
Wang & Ryman 2001
Closed hatchery line
Ne(wild) = 1500
95% CI = 
Rose et al. 2006
5-10% contribution from supportive breeding in 2002, Great Wicomico, VA
Hare et al. 2006
5% contribution graphed
N(census)  annually
If overall population size does not increase then Ne will not recover – it will asymptote at a low level
Census N = 20 million
This applies Wang & Rymans MODEL 1, random selection of broodstock from Bay, so doesn’t incorporate added inbreeding in closed hatchery line.
Hare & Rose in prep.

10. Effective Population Size – Hatchery
Hatchery amplification
spawn broodstock from primary line
create lots of DEBY juveniles
plant these “seed” oysters
WILD oysters collected in DElaware BaY
and exposed to disease
= DEBY strain
annual restoration broodstock
hatchery amplification represents bottleneck that would occur if wild broodstock were used. Additional reduction in Ne expected from selection and drift in selection line.
Disease tolerant oysters selected, = “PRIMARY” DEBY line
6 generations of selection

11. Effective Population Size – Hatchery
Hatchery amplification
spawn broodstock from primary line
create lots of DEBY juveniles
plant these “seed” oysters
WILD oysters collected in DElaware BaY
and exposed to disease
= DEBY strain
annual restoration broodstock
hatchery amplification represents bottleneck that would occur if wild broodstock were used. Additional reduction in Ne expected from selection and drift in selection line.
Disease tolerant oysters selected, = “PRIMARY” DEBY line
6 generations of selection

12. hatchery amplification of DEBY strain
Loss of Alleles, locus 2j24
BEFORE
Wild population
n = 300
Allele
AFTER
hatchery amplification of DEBY strain
n = 96 seed oysters
Cannot distinguish between affects of selection versus hatchery amplification here

13. Effective Population Size - Hatchery
Population bottlenecks generate ephemeral correlations among alleles at unlinked genes
gametic phase disequilibrium
Magnitude of correlations depends on bottleneck Ne
Waples 1991 method; Nb(LD) =
8 microsatellite loci, binned into biallelic data n
Wild Primary LCR02 DEBY GWR02 DEBY LCR04 DEBY
LD(28)
Nb(LD)
95% CI   1
3 x (r2 – 1/S)
R2 is the arithmetic mean of squared correlations between alleles at all pairwise combinations of loci, and S is the harmonic mean of sample sizes
Nb is effective number per year, Ne is effective number per generation (accounting for repro output in >1 year). No simple relationship between the two, but if multiple estimates of Nb are consistent, it suggests that harmonic mean Nb = Ne would be similar to Nb

14. Effective Population Size - Hatchery
Waples 1991 method
8 microsatellite loci, binned into biallelic data n
Wild Primary LCR02 DEBY GWR02 DEBY LCR04 DEBY
LD(28)
Nb(LD)
95% CI  

15. Effective Population Size - Hatchery
Waples 1991 method
8 microsatellite loci, binned into biallelic data n
Wild Primary LCR02 DEBY GWR02 DEBY LCR04 DEBY
LD(28)
Nb(LD)
95% CI  

16. Consequences of Supportive Breeding
Closed hatchery line
Ne(hatchery) = 5
Initial reduction in Ne of 75 – 95%
Recovery is slower than implied here, assumes constantly growing census N
If overall population size does not increase then Ne will not recover – it will asymptote at a low level
Census N = 20 million
This applies Wang & Rymans MODEL 1, random selection of broodstock from Bay, so doesn’t incorporate added inbreeding in closed hatchery line.
Hare & Rose in prep.

17. What’s wrong with small Ne ??
No problem with longstanding small Ne
Problems caused by Ne
Severity of problems depends on genetic load
8 to 14 highly deleterious recessives per genome in Pacific oyster, C. gigas Ne
Reduced survivorship, fecundity

18. Inbreeding Depression in Oysters
Common view: high fecundity buffers oysters from inbreeding depression
High early mortality allows ‘purging’ of load
Correct view: even slight inbreeding causes measurable reduction in fitness-related traits
Unrelated parents, F = 0 by definition, First cousin mating is F = , sibling cross F = 0.203
Evans et al. 2004
Aquaculture 230:89-98
40 families, n = 402

19. Recommendations I
Two of three motivations for using selected strains were speculative, still no data on DEBY fitness in restoration setting
Spawn wild oysters and use procedures that  Ne(hatchery)
Continue use of DEBY oysters only in rivers where genetic monitoring can be used to test efficacy of changing planting strategies
Except for genetic tracking, the benefits are speculative possibilities – much more speculative than the risk of reduced Ne in the event that restoration is highly successful

20. Recommendations II
There is no empirical justification for using DEBY x wild broodstock crossings
Speculative benefits
Disables genetic monitoring of restoration
Known inbreeding risks still apply

21. Collaborators C. Rose Department of Biology & K. Paynter
D. Meritt
S.K. Allen, Jr.
M.D. Camara
R. Carnegie
M. Luckenbach
K.S. Reece
Department of Biology &
Horn Point Laboratory,
University of Maryland
Virginia Institute of Marine Science
College of William and Mary, Virginia

22. Oyster Disease Research Program, NOAA, Sea Grant
with help from…
Maryland
Oyster Recovery Partnership
Charlie Frentz
Virginia
Chesapeake Bay Foundation
Rob Brumbaugh
Tommy Leggett
VA Marine Resources Division
Jim Wesson
VIMS:
Cheryl Morrison
Wendy Ribeiro
Missy Southworth
Hare Lab, UMD:
Jenna Murfree
Paulette Bloomer
Natasha Sherman
Gang Chen
Maria Murray
Peter Thompson
Safra Altman
Kristina Cammen
Andrew Ascione
Ninh Vu
Katie Shulzitski
Oyster Disease Research Program, NOAA, Sea Grant