DISCUSSION PAPER
ALTERNATIVE STRATEGIES FOR MANAGING LONG ISLAND'S HARD CLAM RESOURCE
Copyright August 30, 1981
Marian N. Steinberg
Department of Public Administration
Graduate School of Public Affairs
State University of New York
Albany, New York
Abstract
Dramatic declines in harvests strengthen the assumption that Long Island's
hard clam fishery may be heading for collapse. A family of prey-predator models
has been developed to test and evaluate alternative strategies to reverse the
decline in hard clam harvests and/or stabilize the clam population. Harvesting
is simulated as a fixed percent of standing stock and the behavior of baymen in
response to price and supply of clams is not included in the models.
Five types of policies are evaluated: closed season, maximum size limit,
hatchery seeding, bounty on predators, and nursery sanctuaries (closed areas).
Effectiveness is judged for both the short term (ten years) and the long term
(eleven to twenty years after a policy was instituted). While seeding options
produce modest short term improvement in annual value (8.0 to 10.8 percent),
only the two bounty policies produce significant improvement in both the short
term (17.0 and 72.6 percent) and the long term (20.4 and 66.4 percent).
The results of this model reflect the influence of specific management
policies on the biological system alone. A later version, incorporating the
behavior of the baymen, will introduce key social and economic factors.
The Problem
Concern about Long Island's hard clam fishery, for several years the
province of fisheries experts and baymen, has gone public. A recent article in
NEWSDAY, one of Long Island's leading newspapers, sums up the problem
succinctly:
In the past five years, the clam catch has plummeted 46 per
cent, and, although experts say they do not know precisely why,
there are plenty of theories--including predators, poaching,
overfishing and a changed environment. [1]
The sharp decline in clam harvest recalls the collapse of populations of
one species of fish after another in New York's marine waters over the past
hundred years {2}, and has implications at both the local and national levels.
Long island's hard clam fishery is of both local and national importance. It
employs over 7UUU licensed commercial fishermen, who add more than $100 million
annually to Long Island's economy [3], as well as providing food for many
non-commercial (recreational) harvesters. Until recently, half the littlenecks,
anereyéeswes aya otiowders sold across the county originated in Long Island's
Great South Bay.
A variety of solutions for reversing the decline of Long Island's hard
clams has been suggested, including size limits, sanctuaries, hatchery seeding,
and a bounty on predators. Neither scientists nor policy-makers agree on what
action, if any, to take. Theories abound, but the effectiveness of specific
strategies for managing the hard clam resource is unknown.
Purpose ,
One group which is particularly interested in this problem is the New York
Sea Grant Institute, which has been supporting a multi-disciplinary research
program on the hard clam fishery of Great South Bay. In response to the needs
of that organization, a series of models has been developed to test and evaluate
alternative strategies to reverse the decline in hard clam harvests and/or
stabilize the clam population. Five types of policies were evaluated: closed
season, maximum size limit, hatchery seeding, bounty on predators, and nursery
sanctuaries (closed areas).
Model Structure
The model looks at a smail (1000 m*) area, assumed to be in the midst of
Great South Bay. It consists of three sectors--a clam sector, a predator
sector, and a harvesting sector. ‘The clam sector is represented by a five-level
aging chain (larvae, juvenile clams, littlenecks, cherrystones and chowders),
while the predators are patterned after one specific species, the whelk, which
is representative of low-metabolism, slow-growing, long-lived clam predators.
The predator population is divided into juvenile and adult stages. Clams and
predators are assumed to be homogeneously dispersed within the model area.
Harvesting is simulated as a fixed percentage of the. standing crop of hard
clams. The behavior of baymen in response to price and supply of clams is not
included in this model.
The Clam Sector. The clam population is divided into five age
groups--larval clams, juveniles, littlenecks, cherrystones and chowders. Larvae
begin life as fertilized eggs. They float in the water column until they are
about two weeks old, when their shells become large enough to pull them to the
floor of the bay. sawentte clams are sexually immature, but otherwise resemble
adult clams. At about three years they reach sexual maturity, and,
coincidentally, minimum legal harvestable size. The adult stages are divided
into littlenecks (3 year olds), cherrystones (4 to 7 years), and chowders (8 to
25 years). Figure 1 shows juveniles, littlenecks, cherrystones and chowders
represented as levels. Larval clams (LC1) were modelled as an auxiliary in
order to reduce computation costs.
An earlier version of this model contains a full description of the
structure and behavior of the clam sector [4]. The constant clam fertility has
Jc 43 LN 45 cs 47 CHDR 49
juvenile littlenecks cherrystones chowders
RLCG clams .
larval ? J 7
growth I I
rate I i] I
42 | | I
A l
: I |
“7 : MTJ reid \ MTLN ‘od | mTCs" | MTD cee
e juvenile \ littleneck I cherrystone / average-time
maturation rate “ —\ maturation rate | maturation rate / to death
. 7
MR
maturation
rate of
larvae
114
BRI
birth rate
ofclams
112
effect of
ae density on
reproduction survival
rate of
clams 53
Figure l. CLAM SECTOR
been replaced with seasonal fertility, based on what is known about the hard
clam in Long Island's Great South Bay. The relationship between fertility and
time of year is shown in Figure 2. Data are based on discussions with Robert
Malouf and Monica Bricelj, Marine Sciences Research Center, SUNY/Stony Brook
[5]-
The Predator Sector. The main difference between this model and the work
reported earlier (6] is the division of the predator population into two age
groups, or levels--juvenile clam predators (JCP), which range in age from birth
to twelve months and eat only juvenile clams, and large clam predators (LGCP),
which are sexually mature, twelve to 36 months of age, and eat both littlenecks
and cherrystones (but not chowders, which are considered to have outgrown
natural predation).
The growth of the predator population is governed, in large part, by the
size of the clam population. Four factors affect the size of the predator
population (see Figure 3): the juvenile predator's death rate (JPDR), the
juvenile predator's maturation rate (RPM), fecundity (FECUND), and the adult
predator's lifetime (PLTM). The influence of clams on the predator population
is shown in Figure 4. The larger the population of juvenile clams (JC), the
greater the density of juvenile clams (DJC). This greater density reduces the
predator's “search time", and results in the location of more food per day
(JCEIN), and thus a higher nutritional level for the predator. The higher the
nutritional level has two consequences. First, it reduces the death rate of the
juvenile predator (STaRV), and second, the better-fed predators mature more
quickly (JCPGRF). The converse is also true. A sparse juvenile clam population
will result in a higher juvenile predator death rate and a slower maturation
time.
The abundance of large clams affects the large predators in a similar
CFERT
Clam
Fertility
0.5
iz i j _L
a]
1
| 2
Figure 2.
1
3 4 5 67 8 9 10 1
MONTH (month of year)
Seasonal Clam Fertility Table
12
7 T _ juvenile predator
+, | ive predator s 14. juvenile predator's} | death rate
density |] Stella >) ‘nutritional level
juvenile | +, | juv. clam
clams .
‘+ 4 juvenile predator
maturation rate
; ago + fecundity
+ + | orge procators |+ | large predator's}
large | *__,| large clam |" lease of locating |p!
clams density food nutritional level
+ lifespan
Figure 3: Factors Affecting
the Size of the Predator
Population
8
Figure 4. PREDATOR SECTOR
-_-—_——
- ~
7 ~~
D.
JCP 10 ., | LGCP 12
juvenile SZ large
predators predators
'RPD
14 » deaths
NPRR JPDR 27 13
deaths CO
normal predator un
reproduction /\ JoruT') \ PTDN \
rate» \ . 15 | normal lifespan \
/ \ maturation 16 \
’ \
f \
/ \
/ \
/ \
\
/ \
/
/ \
\
a
26 No
| JCPGRF FECUND Pau
| Juv. pred. fecundity litem
\ growth 33,34 ae
\ /
PFERT p
seasonal Yn.
fertility
SCETN LGCETN
juvenile large clams
clams eaten eaten
NS 7
DJC
density of
juvenile
clams
20
DLGC
density of
large clams
28
Jc 43 LN 45 cs 47 :
——— > vv. clams |» littleneck [——— cherrystone |
fashion. The more large clams (LN and CS), the greater their density (DLGC),
and the greater the large clam predator's ease of locating food (LGCEIN). The
easier it is to locate food, the higher the large predator's nutritional level.
The higher the nutritional level of the adult predators, the higher their
fecundity (FECUND), and the longer they will live (PLTM). Data on the
consumption of clams by predators were derived from research currently underway
by Mary Gibbons at the Marine Sciences Research Center, State University of New
York at Stony Brook {7}. Other relationships between clams and predators are
based on consultations with Dr. Robert Malouf, Shellfish Biologist, also at
Stony Brook [8j.
The Harvesting Sector. Figure 5 shows that only littlenecks, cherrystones
and chowders are subject to harvesting, since younger clams are generally below
minimum legal size. Harvesting is simulated as 6 percent of the standing crop
per month, or about 80 percent per year, a high figure supported by biological
surveys of the clam population of Great South Bay [9]. This simplified
representation of harvesting permits a clearer understanding of the impact of
alternative management policies on the basic prey~predator system. It ignores,
however, the potentially powerful influence of the baymen. A new set of models
is therefore being developed which incorporate the behavior of the baymen.
Effect of Predators and Fishing on Clams. The effect of both predation and
fishing on the clam population is displayed in Figure 6. Juvenile predators
reduce the juvenile clam population through the rate of predation on juvenile
clams (KPJC), which is a function of the effect of clam density (DJC) on the
number of juvenile clams eaten (JCEIN), and the number of juvenile predators.
Littlenecks and cherrystones are similarly reduced by the large clam predator
through the rate of predation on littlenecks (RPLN) and the rate of predation on
cherrystones (RPCS).
LN 45
littleneck
7
7 HPRED \
RLNH
harvest
rate 91
95
—_— ~—
LNHVST 101
annual LN
harvest
b-—~
YLNS 102
sold
LBNU
littleneck
bushels/yr.
* ture 5.
83
LNVLU
littleneck
value/yr.
87
HARVEST SECTOR:
84
CSVLU
cherrystone
value/yr.
88
TTLVLU
total
value/yr.
90
cs 47 S CHDR 49
cherrystone tt chowders
~~
N
RCSH iRcHDAR /
harvest py harvest
rate 92 HPRED2 oan rate 03 /|N
4 Y
CSHVST 103 GHDAH. 105
annualCS PSL CHDR
harvest harvest
CSBU
CSS 104 cherrystone CHDRS
sold bushels/yr. sold 116
CHDRBU
chowder
bushels/yr.
85
CHDVLU
chowder
value/yr.
89
OT
Figure 6. EFFECT OF PREDATORS. OFISHING ON CLAMS
JCP LGcP
juvenile bP aauit :
predator 10] | predator 12
juvenile
DLGC
large clam
density
28
juvenile
clams
20
RPLN 18
predation
Jc
juv. clam
43 LN 45 | cs 47
littleneck
RDC
RCSH 92
harvest
harvest
TT
b TZ cHDR 43| ~7
cherrystone chowders
12
Policy Alternatives
In all cases the policies are inserted into a prey-predator system in
equilibrium. . While such an equilibrium is clearly unnatural, it permits the
understanding of the impact of various policies under controlled conditions.
This approach is particularly useful at this stage of the model-building
process, allowing the researcher to gain better insight into the underlying
behavior of predator and prey to external manipulation [10].
Closed Season. One approach used to manage fisheries is the closed season.
This concept is simulated by cutting off all harvesting during the clam's
\ spawning season. The confluence of biology and regulation suggests that this
“might be a useful, strategy. Biologically, clams reach sexual maturity about
their third year. However, regulations written many years ago set a 1l~inch
minimum.legal size for hard clams, which most reach at about three years of age.
Thus -many LTittlenecks, = three year olds, may never have an opportunity to
epaua beford being caught. A closed season would permit littlenecks at least
one spawning season.
Maximum Size Limit. ; Jon Conrad, a resource economist at Cornell
University, has suggested that imposing a maximum size limit might increase net
economic return from the clam fishery: “These cohorts are more valuable in
Great South Bay as spawning stock than in the market." [11]. His reasoning is
based on the chowders's Tow market value ($11/bu vs. $22-23/bu for cherrystones
and $63-65/bu for littlénecks) combined with their high fecundity. This policy
tests the impact of curtailing all harvesting of chowders.
Seeding. Stocking lakes and rivers with hatchery raised fish is another
common management practice. Baymen strongly favor expanding existing programs
which add small clams, called "seed", from hatcheries to augment: the wild clam
stock, a practice known as “seeding.” Seeding is simulated as the annual
13
addition of an amount of hatchery~-raised seed equal to a normal year's natural
spawning from the bay--a level far beyond the hatcheries's realistic output.
Since an earlier version of this model [12] indicated that seeding alone
has little or no effect on increasing stocks of hard clams, several modified
seeding strategies are examined. (1) Time of seeding. There is some biological
evidence that the time that seed is added to wild stocks affects their survival.
Two approaches were therefore tested for comparison, seeding in the fall (after
the predators have spawned), and seeding in the spring (permitting a full
season's growth). (2) Growing seed on racks. Michael Castagna, Virginia
Institute of Marine Sciences, has found that the survival of seed introduced
from hatcheries can be raised from near zero to over 90 percent by protecting
the young seed from predators [13]. One way of protecting the seed is to grow
them on racks above the bay bottom, out of reach of most predators.
Bounty on Predators. The idea of instituting a bounty comes from
management policies applied to other species. A classic System Dynamics
exercise, for example, looks at the impact of a bounty on mountain lions on the
deer population of the Kaibab Plateau. The high bounty assumes that baymen will
harvest predators with the same intensity as clams. The low bounty assumes that
baymen will merely treat predators as an incidental catch, no longer throwing
them back into the bay.
Sanctuary. Conventional wisdom and sampling studies [14] support the idea
that the density of clams is far greater in portions of the bay closed to
harvesting for many years because of pollution than in “open”, or generally
harvested, areas. The closed areas also contain a much larger proportion of
older, more fertile, clams. This policy tests the effectiveness of setting
aside portions of the bay as natural breeding sanctuaries, an idea analogous to
the medieval “three field system" of agriculture. Figure 7 shows how the
Sanctuary
(—
PREDATOR SECTOR IL
CLAM SECTOR IL
NE HARVEST! RWS
Figure 7.
Open Area
| (— PREDATOR SECTOR I >
SPOWn CLAM SECTORI
larvae >
\ HARVESTING SECTOR I
Modelling the Sanctuary Policy
vT
15
sanctuary policy was modelled. The original model (open area) was duplicated to
represent a second 1000 m* area adjacent to the original open area. The
expanded sanctuary model assumes that clams spawned from the two. sectors will
mix and the resulting larvae will be evenly divided among the open and closed
areas.
Results Policy effectiveness was judged in terms of both short term
(arbitrarily set as the first ten years after policy implementation) and long
term (ten to twenty years following the sustained use of a given policy) effects
on the annual value of clam harvests. Results are given in Table 1. In the
short term exceptional improvement was shown only with a high bounty on
predators (72.6 percent increase in total value). Two of the policies (maximum
size limit and the sanctuary), were ineffective in increasing market value.
Five alternatives produced modest increases (8.0 to 17.U percent). Instituting
a closed season led to a loss of over one-fourth in total value in both the
short term and the long term. Only the two bounty policies produced significant
increases in the value of clam harvests in the long term (2U.4 percent for the
low bounty and 66.4 percent for the higher bounty.)
16
Table 1. CHANGES IN TOTAL ANNUAL VALUE OF CLAM HARVESTS
FOR ALTERNATIVE MANAGEMENT STRATEGIES
Years 1-10 Years 11-20
Percent Change Percent Change
Management Option Over Base Year* Over Base Year
1. Closed Season -26.0 -28.3
2. Maximum Legal Size - 1.7 - 0.3
3. Seeding
-spring seeding 10.4 1.6
-fall seeding 8.0 - 0.1
-spring seeding on racks 10.6 1.9
-fall seeding on racks 10.8 2.6
4. Bounty
-low bounty 17.0 20.4
-high bounty 72.6 66.4
5. Sanctuary 2.4 1.8
*Base Year = Year U
Figure 8 iaddeates a the behavior of several of the management policies
over the 20-year period. While data in Table 1 indicate that the high bounty
lead to the highest average increase in total value, the graph shows that the
high bounty sie introduces considerable instability into the system. On the
other hand, the sanctuary option, which has but a modest effect on total value,
causes little disturbance to the system.
Future Directions
Work is currently underway to introduce the social and economic behavior of
the baymen into the model {15]. The new version replaces the current harvesting
method-~80 percent of the standing crop per year--with a set of complex
relationships between clam abundance, price, fishing effort and number of
baymen. In addition, since both biologists and baymen recognize that
high-metabolism, short-lived predators such as green crabs or mud crabs are
responsible for much of the mortality of juvenile clams, they, too, will be
added in the next version.
17
Figure 8. Annual Value of Harvest for Selected Policies
700 4
wo high pounty
xO
600 4 ‘
_ 880 4 ‘
3 ‘ a
3 ; ‘
2 ; ‘
. , x
3 : ‘
Py 500 - ‘
§ ; ‘
3 : 4
3 40 / 4
z ‘ :
s :
z i ‘
8 ! ‘
° 400. 7] ¥ ‘
4 J spring seeding low bounty \
350 spring seeding ,@ On racks of. ‘ y
¢ git —— . *
300
250
maximum size
e timtt
200 . closed season
°
3 oe 2 © © © © © © © ew © wo © ew ew ew
r T + “+r T T “T- r v
2 6 8 10 12 14 16 18 20
TIME (Years)
18
Finally, the difference between the information in Table 1, which indicates
that a high bounty will clearly increase annual revenues over the first and
second ten yéars following its ‘introduction, and Figure 8, which shows that the
same policy creates significant instability in annual income, suggests that
evaluating the effectiveness of alternative policies is, in itself, a complex
problem. Clearly there is a need to continue the analysis of system dynamics
output beyond the interpretation of model behavior. Therefore, major attention
will be given to the question of evaluating the “effectiveness” of alternative
management policies for hard clams, a problem neglected in System Dynamics, pat
of crucial importance to shellfish managers.
References
1. NEWSDAY, June 26, 1981, p. 7. Garden City, New York.
2. McHugh, J. L. 1972. “Marine Fisheries of New York State,” in
Fishery bulletin, V. 70, No. 3, 585-610.
3. Nassau~Suffolk Regional Planning Board. 1974. Guidelines for
the Management of Long Island's Hard Clam Resources.
4. Steinberg, Marian N. 1980. “A Preliminary System Dynamics Model of the
Effectiveness of Shellfish Hatcheries on Increasing Harvestable Yields."
Proceedings of the 1980 IEEE Conference.
5. Malouf, Robert (Shellfish Biologist) and Monica Bricelj (Doctoral
Candidate) , Marine Sciences Research Center, SUNY/Stony Brook.
Personal communication.
6. Steinberg, op. cit.
7. Gibbons, Mary. Doctoral Candidate, Marine Sciences Research Center,
SUNY/Stony Brook. Personal communication.
8. Malouf, op. cit.
9. WAPORA, Inc. Estuarine Impact Assessment (Shellfish Resources) for the
Nassau-Suffolk Streamflow Augmentation Alternatives.
lu. A technical report describing the effect of these policies on the basic
prey-predator system is currently being drafted.
ll. Conrad, John M. 1981. “Management of a Multiple Cohort Fishery:
The Hard Clam in Great South Bay." Unpublished manuscript.
Department of Agricultural Economics, Cornell University.
12. Steinberg, op» cit.
13..Castagna, Michaél, Virginia Institute of Marine Sciences. Personal
communication.
14. WAPORA, op. cit.
15. Steinberg, Marian N. Memorandum, ."Expanded Harvesting Sector.”
May 16, 1961.
- 19-
i AP PEND ER s RQuation.List,Bage Model . soc sowie we
SSRIT FOS TAT EG VN AHOLD YAY OS
PT RUNS “R VESSi91 Deut .
OAST RUN -=" CONTINUOUS HARVESTING 7 ~ 7
& CLAM*MOLELS.HVST
~S COPYRIGAT APRIL 6, 1957 ~ oo °
* MAQLGS He STEINSE-Gy GSP, SUNY/ILSANY
ROM ONTHOY CATCUCATICN, STASCYAL FERTILITY — .
R PER.KLILGCP KENPRRAFECUND MPFERT OK
“A PFERTVKETASLE (PFT MONTHS sa SaT e465) ee ~
T PFTZS/V/O/O/ DI Io 2840297029165 025/028
KIT USS a2 FT 7 — _ woe ee ~
Lo SCP AKIUCP JHU STII ESR oe JK RPMs UK-UPO Re IK) .
“RB REMGKLIC UCP OKFICPRTISICRGRF OO
L LGCP.K= SCP oe JH(UT) (RPh eo JK -RPD eK)
R RPO. KL= LOCPeK/(PTON*PL TMG KD
JCETN«KXJCP OK
SO Sa Rayieee) oh ern
TABHL ASCE TNT UC on 392594)
T 1251 S31 L526 SS eh S 84/305 OOO
A STARVITABRLIDAMT,SCETW Kyo g4y 05)
TORNTEIS Sif Bel OS elle Se 2/6 2 e.
A JCPGAF Sn =TAEHLOUCPGRT, USE 2
T
RK
A
5) ° ° =e
JCPGR
JPOT AKL ASCP eK USTERV AK ee ~
DLGC.K=ICS eK FLUNK /ARZA
TRTUTC The w= TAOALILE CATT, CLGT eK, 75275 8)
LUCATI=Z ss ot/ 1 ef 1302/3065
PLTM KI TAGHLOPLTI'T LGCITN Ky Ty by 0 SD
PLIMTID/ I eo eTos ee FI ele 93 699/1
KITAELECPES TL OCE TK yy 4y oS)
Libel e2al Salat Fd eFI/)
3 LSoS- 2 GA
NOTE CLAN SECTCR MONTHLY, SCASCHAL SPAWNING
KR RLCGeKL >
L Ne dA-RICU SK AAP IHD
Lie KEL FEST) 2 UCA RPL Hane tL Gh AMEN edd
& IO VALILN KANT
L a
Sogel? GLPO De gyorg Sy eee jyrg 2a ae
COMKETON KR ETT
EuLS KE TACHL SURV COeK ody 2 Uy) : . a
SUQVIIS OF bei ase ST eee
TOT BASE RU == CONTINUCUS KAOVESTING eee ee
“TTUCTALE4 IG wes mee
NI
32 = aS = np Gr Baten eee
Sf
oC <7 wee le _ .
is sack
c 1s
é T PHONTH OF yo al
~T / Deda 2ES o1R/ 63 - . —
x /
i d
oa ' eKSINTVL MITIAL)
OU NITIALE oe “ wee
C InTVt=12
NOT wee Taka alee POLICY Beak He —_
Selick Silane att
FISHING AER He BOR Seat a
A
A NEU SKACSEUSKACHO NEU CH
B EU en st
3 CSEUn*1I
A CHOVLUSKICHUSE Ute
BOTTLE URED VLU eh ee K ~
RORLN eK
" a
OF CHAO %,> CUTSNTT
RAC aN
-21-
SREENAVST SOM (OT CRO CH SUK -CTS TUK)
SHVSTeJ#(OT)(2C cHedSK-CSSeuK) —
STabwsle
AORH UFC TI CR CHS AS UK =CHE RS UK) a oe
CUCnDH
A BRIZILNRAC2SCS eK IFO eCHDOR KID
SRLIRE aban Geviett AS noes ola i a “ wriciuns
rs
ENT PT TSS CONTINUOUS HARVESTING
=CHOPHeh/ 12 ~ oo _ oe _
UNRVST,CSSU,CSHVET,CH by CHOL Hy AX¥CELyLNVLESCSVLU,CHL VLU,
Frist UCPyLOCR ys LNylLClyJC yCS CHER
FLOT LNTUDE, CSP UL2, CHET SUS TITV LIPS XTEULTOC, 20) .
PLOT SUPZSyLGOCPH4 (7 RUGI/LN=ELOT SL TMUIACSE 20 SOOT SIC EIEIO Ty
x DIV7CHOR= 312,250) . ~~
FLOn LNVLUZLO 3,252) /CSVLUEZ (5, 25 IPS CHOVLEZS O25) TIL VLUS EET 4
SPECTUT=IVES/PRTPOS=LI/9L TRPESZO/CENS THD ER ° —
SUN BAS! Us
COMPILED “= or uh MaRS MEE s eet oe
CC BY-NC-SA 4.0