Fish Stock Assessment Manual

FAO
FISHERIES
TECHNICAL
PAPER

393

by
Emygdio L. Cadima
Consultant
FAO Fisheries Department

DANIDA

Rome, 2003

PREPARATION OF THIS DOCUMENT

The author, Emygdio Cadima, now retired, was an FAO scientist in the Fisheries Department
until 1974, when he returned to the Instituto de Investigao das Pescas e do Mar (IPIMAR) in
Portugal, having also been a Professor at the University of Algarve until 1997. At the end of
1997 he was the lecturer of a course in Fish Stock Assessment in IPIMAR, which became the
basis for the preparation of this manual, requested and supported by the Project FAO/DANIDA
GCP/INT/575/DEN. This manual also incorporates notes from courses in Fish Stock Assessment
held at several different venues in the world, mainly in Europe, Latin America and Africa. These
courses had an active collaboration of fisheries scientists from all over the world, especially
Portugal. These scientists are also co-responsible for the orientation, for the matters treated and
particularly for the elaboration of the exercises.
This manual aims to present the basic knowledge on the problems and methods of fish stock
assessment to young scientists, post-graduate students, and PhD students. This is a scientific area
in permanent development, where the knowledge of fisheries biology is applied in order to make
a rational and sustained exploitation of the fishing resources.
The Manual of Fish Stock Assessment is mainly concerned with the theoretical aspects of the
most used models for fish stock assessment. The practical application (i.e. the exercises solved in
a spreadsheet), is considered as a complementary part to help the understanding of the theoretical
matters.
The editing of the manuscript was made by Siebren Venema, manager of
Project GCP/INT/575/DEN and Ana Maria Caramelo, Fishery Resources Officer in the FAO
Fisheries Department.

Distribution:
DANIDA
Fisheries Education Institutes
Marine Research Institutes
National and International Organizations
Universities
FAO Fisheries Department

iii

Cadima, E.L.
Fish stock assessment manual.
FAO Fisheries Technical Paper. No. 393. Rome, FAO. 2003. 161p.
ABSTRACT
The manual follows the same order of the lectures in the last course held in IPIMAR
(November/December 1997). It starts with an introduction to the mathematical models applied in
Fish Stock Assessment and some considerations on the importance of fisheries. The need for a
rational management of the fishing resources is then stressed, this being indispensable for an
adequate exploitation, aiming at conservation, to occur. The basic assumptions about a model
and the concepts of different variation rates of a characteristic in relation to time (or to other
characteristics) are presented, highlighting the most important aspects of the simple and
exponential linear models which are used in the chapters that follow. After some considerations
on the concept of cohort, models for the evolution in time of the number and weight of the
individuals that constitute the cohort are developed, including models for the individual growth
of the cohort. In the chapter concerning the study of the stock, the fishing pattern and its
components are defined, the most used models for the stockrecruitment relation are presented,
as well as the short and long term projections of a stock. With regard to fishing resources
management, the discussion is focused on the biological reference points (target points, limit
points and precautionary points) and fisheries regulation measures. The last chapter, which
presents and discusses theoretical models of fish stock assessment, deals with production models
(also designated as general production models) and with the long and shortterm projections of
the catches and biomasses. Finally, the general methods of estimating parameters are described
and some of the most important methods are presented, with special relevance to the cohort
analysis by age and length. Then a solution of the exercises from the last course held in IPIMAR,
is presented by the author and the scientist Manuela Azevedo.

iv

TO MY FIRST MASTERS AND OLD-TIME FRIENDS

Ray Beverton
John Gulland
Gunnar Stersdal

PREFACE
This work is essentially orientated to present an introduction to the mathematical models applied
to fisheries stock assessment.
There are several types of courses about the methods used in fish stock assessment.
One type considers practical application as the main aspect of the course, including the use of
computer programs. The theoretical aspects are referred to and treated as complementary
aspects.
A second type is mainly concerned with the theoretical aspects of the most used models. The
practical application, considered as the complementary part, facilitates the understanding of the
theoretical subjects.
In this work, the second type was adopted and exercises were prepared to be solved in a
worksheet (Microsoft Excel). The table of contents indicates the exercises corresponding to each
subject.
This manual is the result of a series of courses on Fish Stock Assessment held in the following
places. Portugal : Instituto de Investigao das Pescas e do Mar IPIMAR (ex-INIP) in Lisbon,
Faculdade de Cincias de Lisboa, University of Algarve and Instituto de Cincias Biomdicas de
Abel Salazar in Oporto. Other courses were held at Instituto de Investigao das Pescas in Cape
Verde, at the Centro de Investigao Pesqueira in Angola, at the Instituto de Investigao das
Pescas in Mozambique, at the Centro de Investigacion Pesquera CIP in Cuba, at the Instituto
del Mar del Per IMARPE in Peru, at the Instituto Espaol de Oceanografa IEO (Vigo and
Mlaga Spain). It is also a result of some lectures integrated into cooperation courses held in
several countries and organized by FAO, by SIDA (Sweden), by NORAD (Norway) and by
ICCAT.
Other fisheries scientists cooperated in these courses and they are also co-responsible for the
orientation of the subjects studied and very particularly for the elaboration of the exercises and
the editorial work. With no particular criterium, these are some of the collaborators to whom I
express my appreciation: Ana Maria Caramelo, Manuel Afonso Dias, Pedro Conte de Barros,
Manuela Azevedo Lebre, Ral Coyula, Renato Guevara.
Lisbon, December 1997
E. Cadima

CONTENTS
Glossary of technical terms used in the manual
Bibliography

Page
xi
xvii

1.
1.1
1.2
1.3
1.4

INTRODUCTION
The Importance of Fisheries
Fisheries Resources Management
Fisheries Resources Research
Fish Stock Assessment

1
1
1
2
3

2.
2.1
2.2
2.3
2.4

MODELS AND RATES

Models
Rates
Simple Linear Model
Exponential Model

6
6
8
12
14

3.
3.1
3.2
3.3
3.4
3.5
3.6
3.7

COHORT
Cohort Introduction
Evolution of the number of a cohort, in an interval of time
Catch, in number, over an interval of time
Individual Growth
Biomass and Yield, during the interval Ti
Cohort during the exploitable life
Simplification of Beverton and Holt

18
18
19
23
24
31
32
35

4.
4.1
4.2
4.3
4.4
4.5

STOCK
Stock over a one year period
Fishing pattern over a one year period
Shortterm projections of the stock
Longterm projections of the stock
Stockrecruitment (SR) relation

4.6

Relation between R and B (RS relation)

38
38
41
42
42
46
50

5.

BIOLOGICAL REFERENCE POINTS AND REGULATION

MEASURES
Biological reference points for the management and conservation of
fisheries resources
Biological target reference points (Fmax, F0.1 , Fmed and FMSY)
Biological limit reference points (Bloss, MBAL, Fcrash and Floss)
Precautionary reference points Fpa, Bpa
Fisheries regulation measures

5.1
5.2
5.3
5.4
5.5

vii

53
53
55
62
63
64

CONTENTS
6.
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8

PRODUCTION MODELS
Basic assumption about the evolution of the biomass of a non exploited
stock
Exploited stock
Variation of the biomass in the interval Ti

Page
67
67
67
68

Long term projections (LT) (Equilibrium conditions)

Biomass and fishing level indices
Biological target reference points (TRP)
Types of production models
Short term projections

69
69
70
70
77

7.
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8

ESTIMATION OF PARAMETERS
Simple Linear Regression Least squares method
Multiple Linear Regression Least squares method
Non-linear model Method of Gauss-Newton - Least squares method
Estimation of growth parameters
Estimation of M Natural mortality coefficient
Estimation of Z Total mortality coefficient
Estimation of the parameters of the stock-recruitment (S-R) relation
Estimation of the matrix [F] and of the matrix [N] Cohort analyses
AC and LCA

83
83
86
89
92
94
97
103
104

8.

EXERCISES

113

8.1

Mathematical Revision

113

8.2

Rates

115

8.3

Simple linear model

116

8.4

Exponential model

117

8.5

Cohort Evolution in number

118

8.6

Cohort Catch in number

119

8.7

Individual growth in length and weight

121

8.8

Cohort during all life Biomass and catch in weight

124

8.9

Cohort during its life Simplification of Beverton and Holt model

126

8.10

Stock Short term projection

127

viii

CONTENTS
Page
8.11

Stock Long term projection

129

8.12

Stock Recruitment relation

131

8.13

Fmax

132

8.14

F0.1

133

8.15

Fmed and FMSY

134

8.16

MBAL and Bloss

136

8.17

Floss and Fcrash

137

8.18

139

8.19

Production models (equilibrium) Schaefer

Production models (equilibrium) Abundance and fishing level
indices

8.20

Production models Short term projection

141

8.21

Simple linear regression Estimation of the parameters of the

WL relation and growth parameters (FordWalford, Gulland and Holt
and Stamatopoulos and Caddy)

142

8.22

Multiple linear model Revision of matrices Estimation of the

parameters of Fox integrated model (IFOX)

144

8.23

Non linear regression Estimation of the growth parameters and of

the SR relation (GaussNewton method)

147

8.24

Estimation of M

148

8.25

Estimation of Z

150

8.26

Age cohort analysis (CA)

152

8.27

Length cohort analysis (LCA)

155

8.28

Examination Written test (Lisbon, Dec. 1997)

158

ix

140

GLOSSARY OF TECHNICAL TERMS USED IN THE MANUAL

Abundance index (U) A characteristic preferably proportional to the available biomass of the
resource. The catch per unit effort, cpue (especially when the effort is expressed in
appropriate units) is an important index..
Biological Limit Reference Point (LRP) Biological reference point indicating limits of the
fishery exploitation with regard to stock self-reproduction, aiming at conservation of the
resource.
Biological Precautionary Reference Point (PaRP) biomass levels (Bpa) and fishing levels
(Fpa), established under the precautionary principle, concerning the reproduction of the
stock, aiming at conservation of the resources. The assumptions and methods used to
determine the PaRPs should be mentioned.
Biological Reference Point (BRP) Values of F and B, taking into consideration the best
possible catch and/or ensuring the conservation of the fishery resource. There are BRPs
based on long term projections (LP), BRPs based on values observed during a certain
period of years and BRPs based on the two previous criteria. The BRPs can be TargetPoints (TRP), Limit-Points (LRP), and Precautionary Points (PaRP). In this manual the
following biological reference points are referred to: Fmax, F0.1, Fhigh, Fmed, FMSY, Floss,
Fcrash, Bmax, B0.1, Bmed, BMSY, Bloss, MBAL. Other biological reference points, used in
management, like F30%SPR, are not mentioned in this manual.
Biological Target Reference Point (TRP) Biological reference point indicating long term
objectives (or targets), for the management of a fishery, taking into consideration the best
possible catch and ensuring the conservation of the stock.
Biomass (B) Weight of an individual or a group of individuals contemporaneous of a stock.
Capturability Coefficient (q) Fraction of the biomass that is caught by unit of fishing effort.
Carrying capacity (k) Capacity of the environment to maintain the stock living in it. It is,
theoretically, the limit of the non exploited biomass (see intrinsic gross rate of the
biomass, r).
Catch in number (C) Number of individuals caught.
Catch in weight or Yield (Y) Biomass of the stock taken by fishing. Yield does not
necessarily correspond to landed weight. The difference between the two values, yield
and landings, is mainly due to rejections to the sea of part of the catch which, for some
reason (price, quality, space problems or even legal reasons), is not landed.
Cohort Set of individuals of a fishery resource born from the same spawning.
Exploitation pattern of a gear (s) Fraction of the individuals of a given size, available to the
gear, which is caught. Also designated by Selectivity or partial recruitment.
Individual growth coefficient (K) Instantaneous relative rate of change of a function of the
individual weight, w, that is, H(w)-H(w), where w is the asymptotic individual weight
and H(w) is a function of w (frequently a power function, including the logarithmic
function). The adopted models for the function H(w) have two constants, w and K.
Some models introduce one more parameter, b, which is used to obtain a general relation
xi

to include the most common individual growth relations. The constant K has the physical
dimension of time 1.
Individual Quota (IQ) Quota attributed to a vessel.
Individual Transferable Quotas (ITQ) System of fisheries management characterized by the
sale, at auctions, of the fishing annual vessel quotas.
Minimum Biomass Acceptable Level (MBAL) Biological reference limit point that indicates
a spawning biomass level under which the observed biomasses during a period of years,
are small and the associated recruitments are smaller than the mean or median
recruitment.
Number of individuals of a cohort or of a stock (N) Number of survivals of a cohort (or a
stock) at a certain instant or over an interval of time.
Partial recruitment (see exploitation pattern)
Precautionary principle This principle establishes that a lack of information does not justify
the absence of management measures. On the contrary, management measures should be
established in order to maintain the conservation of the resources. The assumptions and
methods used for the determination of the scientific basis of the management should be
presented.
Production models Models that consider the biomass of the stock as a whole, that is, they do
not take into consideration the age or size structure of the stock. These models are only
applied in analyses that consider fishing level changes, as they do not allow the analysis
of the effects of changes in the exploitation pattern, on catches and biomasses.
Quota (Q) Each of the fractions in which the TAC was divided.

RATES
Absolute Instantaneous Rate of y, air(y) Velocity of the variation of the function y(x), at the
instant x.
Absolute Mean Rate of y, amr(y) Mean velocity of the variation of the function y(x), during
a certain interval of x.
Annual Survival Rate (S) Mean rate of survivals of a cohort during one year, relative to the
initial number.
Exploitation Rate (E) Ratio between the number of individuals caught and the total number of
individuals dead, over a certain period of time, that is, E = C/D.
Fishing mortality instantaneous rate (F) (Fishing mortality coefficient) Relative
instantaneous rate of the mortality of the number of individuals that die due to fishing .
Intrinsic rate of the biomass growth (r) Constant of the Production models that represents
the instantaneous rate of the decreasing of the function H(K)-H(B), where B is the
biomass, H(B) is a function of the total biomass, usually a power-function, (including the
logarithmic function that can be considered a limit power function) and k is the carrying
capacity of the environment. Some models introduce one more parameter, p, which is
used to obtain a more general relation.

xii

Natural mortality instantaneous rate (M) (Natural mortality Coefficient) Instantaneous

relative rate of the mortality of the number of individuals that die due to all causes other
than fishing.
Relative instantaneous rate of y, rir(y) Velocity of the variation of the function y(x), relative
to the value of y, at the instant x.
Relative mean rate of y, rmr(y) Mean velocity of the variation of the function y(x) relative to
a value of y, during a certain interval of x.
Total mortality instantaneous rate (Z) (Total mortality coefficient) Relative instantaneous
rate of the mortality of the number of individuals that die due to all causes. Z, F and M
are related by the following expression : Z=F+M .
Recruitment to the exploitable phase (R) Number of individuals of a stock that enter the
fishery area for the first time each year.
Selectivity (see exploitation pattern)
Spawning or adult biomassa (SP) Biomasss of the stock (or of a cohort) which has already
spawned at least once.
Stock Set of survivals of the cohorts of a fishery resource, at a certain instant or period of time.
It may concern the biomass or the number of individuals.
Stock-Recruitment (S-R) relation Relation between the parental stock ( spawning biomass)
and the resulting recruitment (usually the number of recruits to the exploitable phase).
The models have two constants, and k. The constant k has the physical dimension of
weight and has the dimension of weight-1. Some models introduce one more
parameter, c, which is used to obtain a general relation that includes the most common
relations.
Structural models Models that consider the structure of the stock by ages or sizes. These
models allow one to analyse the effects on catches and biomasses, due to changes in the
fishing level and exploitation pattern.
Total Allowable Catch (TAC) Management measure that limits the total annual catch of a
fishery resource, aiming to indirectly limit the fishing mortality. The TAC can be divided
into Quotas (Q) using different criteria, like countries, regions, fleets or vessels.
Total number of deaths (D) Total number of individuals that die during a certain period of
time..
Virgin biomass (VB) Biomass of the stock not yet exploited.

xiii

SYMBOLS
Symbols

Indicating :

Constant of the simple linear model (intercept of the straight line )

Constant of the Stock-Recruitment relations (limit value of R/S when S0)

amr(y)

Absolute mean rate of variation of y

air(y)

Absolute instantaneous rate of variation of y

Constant of the simple linear model (slope of the straight line)

Biomass

SP, SB

Spawning Biomass

Catch, in number

Constant of the Stock-Recruitment relations (generalizes the models)

Total number of deaths

Exploitation rate

Fishing mortality coefficient

Cconst

Non defined constant

Cte

Non defined constant

General power function

ITQ

Individual Transferable Quotas

Constant of the individual growth models (associated to growth rate)

Constant of the Stock-Recruitment relations

Constant of the production models (Carrying capacity)

L,L

Total length of an individual

MBAL

Minimum Biomass Acceptable Level (biological reference limit point)

Natural mortality coefficient

Number of individuals of a cohort

Constant of the Production models (generalizes the models)

Capturability coefficient

Constant of the Production models (intrinsic rate associated with the biomass
growth)

R2

Determination coefficient

Recruitment to the exploitable phase

rmr(y)

Relative mean rate of variation of y

rir(y)

Relative instantaneous rate of variation of y

xiv

Symbols

Indicating :

Annual Survival rate

Adult or total biomass (in the relations S-R)

Exploitation pattern (selectivity)

SQ

Sum of the squares of the deviations

S-R

Stock-Recruitment relation

Instant of time

Interval of time between 2 instants

TAC

Total Allowed Catch

TRP

Biological Target Reference Point

Stock abundance index

Function to be maximized for the determination of F0.1

Individual weight

Catch in weight

Total mortality coefficient (total mortality instantaneous rate)

xv

SUBSCRIPTS
The characteristics of this glossary are usually shown with indices; that is why it was considered
necessary to present the meaning of those subscripts.
Subscripts

Indicating :

Economical value of the respective characteristic of the cohort

Maximum age

0.1

Value of F (and of other characteristics of the cohort) corresponding to the

air of the biomass equal to 10 percent of the virgin biomass

Recruitment to exploitable phase

crash

Value of F which, at long term, corresponds to the collapse value of the

spawning biomass

Value of the characteristics of the cohort corresponding to an equilibrium

point

Age

infl

Value of the characteristic corresponding to an inflection point of any

relation between that characteristic and other variable.

Length

lim

Value of B or of F corresponding to a biological reference limit point

loss

Value of B or of F corresponding to the minimum spawning biomass

observed

Max

Value of F (and of other characteristics of the cohort) where the yield per
recruit is maximum

Med

Value of F (and of other characteristics of the cohort) which, at long term,

will produce a spawning biomass per recruit equal to the median value of the
spawning biomasses per recruit observed during a certain period of years

MSY

Value of F (and of other characteristics of the stock) where the long term
total yield is maximum

Recruitment to the exploitable phase

xvi

BIBLIOGRAPHY
Bertalanffy, L. Von 1938. A quantitative theory of organic growth. Hum. Biol., 10 (2): 181-213.
Beverton, R.J.H. & Holt, S.J. 1956. On the dynamics of exploited fish populations. U.K. Min.
Agric. Fish., Fish. Invest (Ser. 2) 19: 533p.
Caddy, J.F.& Mahon, R. 1995. Reference points for fishery management. FAO Fish. Tech. Pap.
349: 80p.
Cadima, E. 1991. Some relationships among biological reference points in general production
models. ICCAT, Coll. Vol. Sc. Papers, (39):27-30.
Cadima, E. & Pinho, M.R. 1995. Some theoretical considerations on non equilibrium production
models. ICCAT, Coll. Vol. Sc. Papers, (45):377-384.
Cadima, E. & Palma, C. 1997. Cohort analysis from annual length catch compositions. WD
presented to the Working Group on the assessment of the Southern Shelf Demersal
Stocks. Copenhagen, 1-10 September, 1997.
Cadima, E. & Azevedo, M. 1998. A proposal to select reference points for long term fishery
management objectives. ICES,C.M. 1998/T:9, 18p.
Cardador, F. 1988. Estratgias de explorao do stock de pescada, Merluccius merluccius L., das
guas Ibero-Atlnticas. Efeitos em stocks associados. Dissertao apresentada para
provas de acesso categoria de Investigador Auxiliar. IPIMAR (mimeo)
CE 1994. Report of the southern hake task force. Lisbon, 10-14 October, 1994.
Clark, W.G. 1991. Groundfish exploitation rates based on life history parameters. Can. J. Fish.
Aquat. Sci., 48: 734-750.
Clark, W.G. 1993. The effect of recruitment variability on the choice of target level of spawning
biomass per recruit. pp 233-246 In: Kruse, G.; Eggers, D.M.; Marasco, R.J.; Pautzke, C.
& Quinn, T.J. (eds.). Proceedings of the International Symposium on Management
Stategies for Exploited Fish Populations. Alaska Sea Grant College Program Report
N 93-02, Fairbanks, University of Alaska.
Cushing, D.H. 1996. Towards a science of recruitment in fish populations. In: Excelence in
Ecology, Book 7, Ecology Institut, Oldendorf/Scuhe, Germany.
Deriso, R.B. 1980. Harvesting strategies and parameter estimation for an age-structured model.
Can. J. Fish. Aquat. Sci., 37: 268-282.
Duarte, R; Azevedo, M. & Pereda, P. 1997. Study of the growth of southern black and white
monkfish stocks. ICES J. mar. Sci., 54: 866-874.
FAO 1995. Code of Conduct for Responsible Fisheries, Rome, FAO, 41p.
FAO 1996. Precautionary approach to fisheries. FAO Fish. Tech. Pap. 350 (2): 210p.
Ford, E. 1933. An account of the herring investigations conducted at Plymouth during the years
from 1924-1933. J. Mar. Biol. Assoc., N.S., 19: 305-384.
Fox, W.W. Jr 1970. An exponential surplus-yield model for optimizing exploited fish
populations. Trans. Am. Fish Soc., 99: 80-88.
Garrod, G.J. 1969. Empirical assessment of catch/effort relationships in the North Atlantic cod
stocks. Res. Bull ICNAF, 6: 26-34.

xvii

Gompertz, B. 1825. On the nature of the function expressive of the law of human mortality, and
on a new mode of determining the value of life contingencies. Phil. Trans. Royal Society,
115 (1): 513-585.
Gulland, J.A.1959. Fish Stock Assessment: A manual of basic methods. FAO/Wiley series,
223p.
Gulland, J.A. 1969. Manual of Methods for Fish Stock Assessment - Part 1. Fish Population
Analysis. FAO Manuals in Fisheries Science No. 4.
Gulland, J.A. 1983. Fish stock assessment A manual of basic methods. FAO/Wiley Ser. on Food
and Agriculture, Vol 1: 233 pp.
Gulland, J.A. & Boerema, L.K 1973. Scientific advice on catch levels. Fish.Bull. 71 (2):
325-335
Gulland, J.A. & Holt, S.J. 1959. Estimation of growth parameters for data at unequal time
intervals. J. Cons. ICES, 25 (1): 47-49.
Gunderson, D.R. 1980. Using r-K selection theory to predict natural mortality. Can. J. Fish
Aquat. Sci. 37: 2266-2271.
Hilborn, R. & Walters, C.J. 1992. Quantitative Fisheries Stock Assessment: Choice, Dynamics
and Uncertainty. New York, Chapman and Hall, 570p.
ICES 1996. Report of the Northern Pelagic and Blue Whiting Fisheries Working Group. Bergen,
23-29 April 1996. ICES CM 1996/Assess:14
ICES 1997a. Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks.
Copenhagen, 3-12 September 1996. ICES CM 1997/ASSESS:5
ICES 1997b. Report of the Working Group on the Assessment of Mackerel, Horse Mackerel,
Sardine and Anchovy. Copenhagen, 13-22 August 1996. ICES CM 1997/ASSESS:3
ICES 1997c. Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks.
Copenhagen, 3-12 September 1996. ICES CM 1997/ASSESS:5
ICES 1997d. Report of the Comprehensive Fishery Evaluation Working Group. Copenhagen,
25 June - 04 July 1997. ICES CM 1997/ASSESS:15.
ICES 1998a. Report of the Study Group on the Precautionary Approach tro Fisheries Management.
Copenhagen, 3-6 February 1998. ICES CM 1998/ACFM: 10
ICES 1998b. Report of the Working Group on the Assessment of Mackerel, Horse Mackerel,
Sardine and Anchovy. Copenhagen, 9-18 September 1997. ICES CM 1998/ASSESS:6
ICES 1998c. Report of the Working Group on the Assessment of Southern Shelf Demersal Stocks.
Copenhagen, 1-10 September 1997. ICES CM 1998/ASSESS:4
ICES 1998d. Report of the Working Group on the Assessment of Northern Shelf Demersal Stocks.
Copenhagen, 16-25 June 1997. ICES CM 1998/ASSESS:1
Jones, R. 1961. The assessment of the long term effects of changes in gear selectivity and fishing
effort. Mar. Res. Scot., 2, 19p.
Jones, R. & van Zalinge, N.P. 1981. Estimates of mortality rate and population size for shrimp in
Kuwait waters. Kuwait Bull. Mar. Sci., 2: 273-288.
Lotka, A.J. 1925. The Elements of Physical Biology. Baltimore, Williams and Wilkins.

xviii

Marquardt, D.W. 1963. An algorithm for least squares estimation of non-linear parameters.
J. Soc. Ind. Appl. Math., 2: 431-441.
Mattos e Silva, G.O. 1995. Aplicao de modelos de produo geral em condies de noequilbrio para a avaliao do manancial de gamba Parapenaeus longirostris (Lucas,
1846) da costa sul portuguesa. Dissertao apresentada para obteno do grau de Mestre
em Estudos Marinhos e Costeiros. UAL, Unidade de Cincias e Tecnologias dos
Recursos Aquticos, Faro, 96p.
Megrey, B. 1989. Review and comparison of age-structured stock assessment models from
theoretical and applied points of view. Am. Fish. Soc. Symp., 6: 8-48.
Paloheimo, J.E. 1961. Studies on estimation of mortalities. Comparison of a method described
by Beverton and Holt and a new linear formula. J. Fish. Res. Bd. Can., 18 (5): 645-662.
Pauly, D. 1980. On the interrelationships between natural mortality, growth parameters and
mean environmental temperature in 175 fish stocks. J. Cons. Int. Explor. Mer, 39: 175192.
Pella, J.J. & Tomlinson, P.K. 1969. A generalized stock production model. Bull. Inter. Am. Trop.
Tuna Comm., 13: 419-496.
Pestana, G. 1989. Manancial Ibero-Atlntico de sardinha, Sardina pilchardus, Walb., sua
avaliao e medidas de gesto. Dissertao original para provas de acesso categoria de
Investigador Auxiliar. IPIMAR, 192p. (mimeo).
Pope, J.G. 1972. An investigation of the accuracy of virtual population analysis using cohort
analysis. Res. Bull. ICNAF, 9: 65-74.
Prager, M.H. 1994. A nonequilibrium surplus-production model. Fish. Bull. 92 (2): 372-389
Prager, M.H. 1995. Users Manual for ASPIC: a stock-production model incorporating
covariates, program version 3.6x. Miami Lab. Doc. MIA-92/93-55
Richards, F.J. 1959. A flexible growth function for empirical use. J. Exp. Bot., 10: 290-300.
Ricker, W.E. 1954. Stock and recruitment. J. Fish. Res. Bd. Can., 11: 559-623.
Ricker, W.E. 1958. Handbook of computation for biological statistical of fish population. Bull.
Fish. Res. Bd. Can., 119: 300p.
Ricker, W.E. 1969. Effects of size-selective mortality and sampling bias on estimates of growth,
mortality, production and yield. J. Fish. Res. Bd. Can., 26: 479-541.
Ricker, W.E. 1975. Computation and interpretation of biological statistics of fish population.
Bull. Fish. Res. Bd. Can., 191: 382p
Rikhter, J.A. & Efanov, V.N. 1976. On one of the approaches to estimation of natural mortality
of fish population. ICNAF 76/VI/8, 12p.
Rosenberg, A.A.; Kirkwood, G.P.; Crombie, J.A. & Beddington, J.P. 1990. The assessment of
stocks of annual squid species. Fish. Res. 18:335-350.
Stersdal, G. 1984. Investigao, gesto e planificao pesqueiras. Revista de Investigao
Pesqueira, 9. Instituto de Investigao Pesqueira. Maputo. R.P.M.: 167-186.
Schaefer, M. 1954. Some aspects of the dynamics of populations important to the management
of the commercial marine fisheries. Bull. Inter. Am. Trop. Tuna Comm., 1 (2): 27-56.

xix

Shepherd, J.G. 1982. A versatile new stock-recruitment relationship for fisheries, and the
construction of sustainable yield curves. J. Cons. Int. Explor. Mer, 40 (1): 67-75.
Sparre, P. & Venema, S.C. 1997. Introduo avaliao de mananciais de peixes tropicais. FAO
Doc. Tc.Pescas, 306/1 Rev 2. (Parte 1 & 2): 404 & 94 pp.
Stamatopoulos, C. & Caddy, J.F. 1989. Estimation of Von Bertalanffy growth parameters: a
versatile linear regression approach. J. Cons. Int. Explor. Mer, 45: 200-208.
Tanaka, S. 1960. Studies on the dynamics and the management of fish populations. Bull. Tokai.
Reg. Fish. Res. Lab., 28: 1-200.
Volterra, V. 1928. Variations and fluctuations of the number of individuals in animal species
living together. J. Cons. Int. Expl. Mer, 3 (1): 3-51.
Yoshimoto, S.S. & Clarke, R.P. 1993. Comparing dynamic versions of the Schaefer and Fox
production models and their application to lobster fisheries. Can. J. Fish. Aquat. Sci. 50:
181-189.
Walford, L.A. 1946. A new graphic method of describing the growth of animals. Biol. Bull. Mar.
Biol. Lab. Woods Hole, 90: 141-147.

xx

CHAPTER 1 INTRODUCTION

1.1

THE IMPORTANCE OF FISHERIES

The importance of fisheries in a country cannot only be measured by the contribution to

the GDP, but one must also take into consideration that fisheries resources and products
are fundamental components of human feeding and employment.
Another aspect that makes fisheries resources important is the self renewable character.
Unlike mineral resources, if the fishery resources or any other biological resources are
well managed, their duration is pratically unlimited.
An important conclusion is that the fundamental basis for the conservation and
management of fisheries resources stems from the biological characteristics. (This does
not mean that social, economic or any other effects are not important for management).
In Portugal, the fisheries contribution to the GDP is less than 1.5 percent. However, with
regard to food, the annual consumption value of 60 kg of fish per person, is very high.
Only countries like Iceland, Japan and some small insular nations reach a higher value.
We still have to consider that of the total amount of protein necessary in our food
consumption, 40 percent comes from fisheries. This corresponds to 15 percent of the total
amount spent on food by the Portuguese population.
From a social point of view, we estimate that there are, at present, 34 000 fishermen in
Portugal. Assuming that each job at sea generates 4 or 5 jobs on land (canning, freezing
and fish meal industry, commercialization, administration, research and training, etc.) one
can estimate that about 150 000 Portuguese work in the several sectors of fisheries.
Consequently, taking a minimum of 3 people per family, it is not unreasonable to say that
about half a million Portuguese people depend on fisheries activities for their livelihoods.

1.2

FISHERIES RESOURCES MANAGEMENT

Stersdal (1984) defined a general principle of fisheries management as :

to obtain the BEST POSSIBLE utilization of the resource

for the benefit of the COMMUNITY
It will be necessary to define, in each particular case, what best, possible and community
mean.

In fact, best can be taken as :

Bigger yield

Bigger value of the catch

Bigger profit (difference between the value of the landing and the costs of
exploitation)

More foreign currency

More jobs, etc.

Community may also be taken as :

The population of the world

The European Community

A country

A region

Groups of interests (fishermen, shipowners, consumers, )

Possible
reminds us that we cannot forget the self renewable character of fisheries resources and
consequently, that the conservation of the fishery resource must be guaranteed in order to
allow the application of the general principle for a long period of time. This statement
means that conservation of an ecosystem does not imply that one should attribute the
same importance to all its components.

1.3

FISHERIES RESOURCES RESEARCH

Figure 1.1 shows that the research on fisheries resources covers several sectors of the
fishing activity.The assessment models are the main concern of this manual. Among the
several works and books on fish stock assessment, the books and/or manuals by Beverton
& Holt (1956), Ricker (1958, 1975) and Gulland (1969, 1983) are historical standing
references.

Gears
Transformation
Legislation
Economical aspects
Etc

FISHERIES RESEARCH

FISHERIES RESOURCES
RESEARCH

For management

Data sources :
For exploitation
strategies

Data Bases

Commercial fisheries
(commercial statistics of catches,
catch per effort and fishing effort)

Biological sampling on landings

(and/or catches)

Data Analyses and

Parameters Estimation

Biological sampling of the stocks

Application of
Assessment Models
Environmental data
Others

Conclusions

Figure 1.1 The several sectors of fishing activity

1.4

FISH STOCK ASSESSMENT

The following are necessary to assess a fish stock:

The appropriate data bases

Analyses of the available data

Short and long-term projections of the yield and biomass

To determine long-term biological reference points

To estimate the short and long-term effects on yield and biomass of different
strategies of the fishery exploitation

The different steps to assess a stock can be summarized as follows :

a) To define the objectives of the assessment according to the development phase of
the fisheries and the available information.

b) To promote the collection of information :

Fisheries commercial statistics : total and by resource landings, catch per effort,
fishing effort (number of trips, days, tows, time spent fishing, etc.), and
characteristics of the gears used.
Types of operation of the fleets and of its fishing gears, etc.
Biological sampling in the landing ports.
Biological sampling (and information about the fishing operation) on board
commercial vessels.
Biological sampling on board research vessels.
c) To analyse the stocks
The knowledge gained about the resource and the available basic data, determine the type
of models that should be used and consequently the type of analyses that can be done. As
an illustration, let us look at some general situations :

Fishery resource with little information

Analyses using particular methods to estimate biomasses and potential yields.

Fishery resource with data on catches and catch per effort (CPUE) or stock abundance
indices during several years

Analyses using production models in order to make projections of yield and catch per
effort.

Fishery resource with information collected over several years on :

 Biological distribution of the catches by species, by length, by ages, etc.
 Commercial catches
 Fishing effort or CPUE
 Research cruises (distribution of the stock by areas, by length, by ages, etc.)

Historical analysis of the stock (VPA)

Long and short-term projections with different conditions (scenarios)

Comments
1. The lack of information may prevent certain projections, but allows other types of
analyses.
2. The Precautionary Principle forces one to estimate and to project catches and
biomasses, even if they are not very precise. This will be discussed later.

CHAPTER 2 MODELS AND RATES

2.1

MODELS

Science builds models or theories to explain phenomena. One observes phenomena and then
looks for relations, causes and effects. Observations are made about the evolution of a magnitude
(characteristics) with time (or with other characteristics) and possible causes (factors) are
explored. Examples:
Physics phenomenum of the movement of the bodies (characteristics distance related to
time spent)
Biology phenomenum of growth (characteristic length or weight, related to time).

2.1.1 STRUCTURE OF A MODEL

Basic assumptions
The assumptions to serve as a basis for a model should :
simplify reality
be simple and mathematically treatable (manageable)
not be contradictory
not be demonstrated
be established with the characteristics
Usually basic assumptions are related to the evolution of the characteristics. So, they are
established on the variation rate of those characteristics and they do not need to be proved.

Relations (properties)

they are deduced from the basic assumptions by the laws of logic (mathematics).
The properties are also designated by
results or conclusions of the model.

Verification

the results of the model must be coherent (to agree) with reality.
This implies the application of statistical methods and sampling techniques
to check the agreement of the results with the observations.

Improvement
if agreement is approximate, it is necessary to see if the approximation is enough or not.
if the results do not agree with reality, then the basic assumptions have to be changed
the changes can aim to the application of the model to other cases.

Advantages
it is easier to analyse the properties of the model than the reality.
the models produce useful results.
they allow analysis of different situations or scenarios by changing values of the factors.
to point out the essential parts of the phenomenon and its causes.
they can be improved in order to adjust better to the reality.

2.1.2 SOME TYPES OF MODELS USED IN STOCK ASSESSMENT

Production Models
The production models are also designated as General Production models, Global models,
Synthetical models or Lotka-Volterra type models. These models consider the stock globally, in
particular the total abundance (in weight or in number) and study its evolution, the relation with
the fishing effort, etc.. They do not consider the structure of the stock by age or by size.

Structural Models
These models consider the structure of the stock by age and the evolution of the structure with
time. They mainly recognize that the stock is composed of individuals of different cohorts, and,
consequently, of different ages and sizes. So, they analyse and they project the stock and the
catches for the coming years, by following the evolution of its different cohorts.
This manual will not follow the chronological construction of the models. It was thought to be
more convenient to deal firstly with the structural models and afterwards with the production
models.

2.2

RATES

The basic assumptions of a model, for the evolution of a characteristic, require the concept of
variation rate of the characteristic related to time (or to other characteristics).

Figure 2.1

Evolution of the length (L) of an individual with time (or age) (t)

In order to generalize the study of the rates, the characteristic L in the example above will be
substituted by y, and the associated variable will not be time, t, but the variable x. To study the
stock assessment models and to make this study easier, it will be considered that the function y
will only assume real and positive values.

2.2.1 ABSOLUTE MEAN RATE amr (y)

Consider y a function of x and the interval i with the limits (xi, xi+1)

Figure 2.2

Function y= f(x) with variation in the interval i

Let :
xi = x i+1 xi be the size of the interval
yi = the value of y when x = xI
yi+1 = the value of y when x = xi+1
The variation of y in the interval xi will be yi = y i+1 yi
The absolute mean rate, amr (y), of the variation of y within the interval xi, will be :
amr (y ) =

y i
x i

Graphically :

Figure 2.3

Absolute mean rate of the variation of y within the interval xi

y i
= amr (y) during xi
x i
Note: amr (y) is known in physics as the mean velocity of the variation of y with x, in the interval
xi.

Slope of the secant =

2.2.2 ABSOLUTE INSTANTANEOUS RATE air (y)

Let y be a function of x
The absolute instantaneous rate of y at the point x = xi is the derivative of y in order to x at that
point.
dy
air (y )x = x i =
dx x = x i

Graphically :

air ( y) x = x i

equal to the slope of the

tangent to the curve at the point x=xi

Figure 2.4

Absolute instantaneous rate of y at point xi

Note: air(y) is known as the instantaneous velocity of variation of y with x at the point x.

Properties
1. Given the value of air(y) the calculation of the function y is obtained, by integration, being
y = f(x) + Constant, where f(x) = Primitive of air (y) and Constant is the constant of
integration.
If the initial condition x* , y* is adopted, where y* is the value of y corresponding to x = x*,
eliminating the Constant, then one can write y = y* + f(x)-f(x*)
2. The angle made by the tangent to the curve y with the xxs axis is designated by inclination.
The trigonometric tangent of the inclination is the slope of the geometrical tangent.
air (y) = derivative of y = slope = tg (inclination)
3. If, at point x :
air (y) > 0 then y is increasing at that point
air (y) < 0 then y is decreasing at that point
air (y) = 0 then y is stationary at that point (maximum or minimum)
4. If air (y) is constant (= const) then y is a linear function. From property 1, it will be :
y = Constant + const. x
y = y* + const.(x-x*)
5.

or
and vice versa

If y(x) = u(x) + v(x) then air (y) = air(u) + air(v)

6. If factors A and B cause variations in y, then factors A and B considered simultaneously

cause a variation of y with:
air (y) total = air (y) causeA + air (y) causeB

air (air ( y )) =

d2y
= acceleration of y at the point x
d 2x

7. If the acceleration at the point x is increasing, then air (y) is positive and if that acceleration
is decreasing, then air (y) will be negative.

2.2.3 RELATIVE MEAN RATE - rmr (y)

Consider y a function of x and the interval (xi, x i+1)
Let :
xi = x i+1 - xi = the size of the interval
yi = value of y when x = xI

10

yi+1 = value of y when x = x i+1

xi* = a certain point in the interval (xi, x i+1)
yi* = value of y when x = xi*
xi* can be xi , x i+1 , x centrali , etc.

The mean rate of y relative to yi* will be :

1 y i
.
y *i x i
1
rmr ( y) = * .amr ( y)
yi

rmr( y) =

or

Comments
1. rmr (y) is associated with the mean rate of the variation of the percentage of y in relation to
y*, that is:
yi
)
y *i
x i

2. Let

x centrali be x centrali = x i +

x i 1
= .( x i + x i +1 ) = x i
2
2

3. It is convenient to designate by y centrali the value of y in the interval (xi , xi+1) when

x = x centrali .
Notice that y centrali can be different from the mean,

( y i + y i+1 )
2

4. It is frequent to calculate rmr(y) in relation to y centrali of the interval.

2.2.4 RELATIVE INSTANTANEOUS RATE - rir(y)

Let y be a function of x.
The relative instantaneous rate of y at the point x = xi is
rir ( y) =

1 dy

y i dx x = x i

or

11

rir ( y) =

air ( y) x =x i
yi

Properties
1. Given rir(y), the calculation of the function y is obtained by integration, being
y = f(x) + Constant, where f(x) = Primitive of rir(y) and C is the constant of integration.
If one accepts the initial condition x* , y* , where y* is the value of y corresponding to
x = x*, one will get, eliminating the Constant, y = y* + f(x) f(x*)
2. If, at a point x :
rir (y) > 0
rir (y) < 0
rir (y) = 0

then y is increasing at that point

then y is decreasing at that point
then y is stationary at that point (maximum or minimum)

3. rir(y) = air (lny) as can be deduced from the derivation rules.

4. If rir (y) = constant = (const) then y is an exponential function of x, that is,
y = Constant . e const.x

or

y = y* . e const.(x-x*)

and vice-versa

5. If y(x) = u(x).v(x) then rir(y) = rir(u) + rir(v)

6. If the factors A and B cause variations in y, then simultaneously,
factors A and B cause a variation in y, with:
rir (y)total = rir (y)cause A + rir (y)cause B

2.3

SIMPLE LINEAR MODEL

Let y = f(x)

Basic assumption of the model

air(y) = Constant = b

in the interval (xi, xi+1) with

xi = xi+1 xi

Initial Condition
x* = xi y* = yi

Figure 2.5

Graphical representation of a simple linear model

12

Properties
y = y i + b (x x i ) ; y = a + bx

1.

General expression

2.

Value yI+1 at the end of the

interval, xi

y i +1 = y i + b x i

3.

Variation, yi , in the interval,

xi

y i = y i +1 y i = b.x i

4.

Central value, y centrali

of the interval, xi

5.

Cumulative value, y cumi

during the interval, xi

= y i + b.( x centrali x i ) = y i + b

ycentrali

y cumi =

x i
2

x i +1

y.dx =x (a + b x )
i

xi

or from the Property 1

y cum i = x i [y i + b (x i x i )]
6.

Mean value, y i , in the

interval, xi

yi =

y cum i

yi =

y cum i

x i

x i

= a + b xi

where

= y i + b (x i x i )

Other useful expressions

7.

Cumulative value, y cumi

y cumi = x i y i

during the interval, xi

8.

Mean value, y i , during the

interval, xi

y i = y i + b.( x i x i ) where y i = a + bx i

9.

Mean value, y i , in the

interval, xi

yi = yi + b

10.

Mean value, y i , during the

interval, xi

y i = y central i

11.

Relation between amr(y) et

air(y)

amr (y i ) =

13

x i
2

y i
= b = air (y )
x i

12.

If y i < 0 then b < 0 et vice-versa

13.

In the linear model, the arithmetic mean of y i and y i+1 is equal to the mean value, y i ,
and equal to the central value y centrali

Important demonstrations
General expression
Property 1

If tia(y) = b in the interval xi then y is linear with x and considering

the initial condition it will be: y = yi+ b.(x-xi)

Central value
Property 4

x
x

y centrali = y i + b.(x centrali x centrali ) = y i + b ( x i + i ) x i = y i + b. i

2
2

Cumulative value
Property 5

from the definition of the cumulative value :

y cum i =

x i +1

xi

(a + bx ) dx

x i2+1 x i2
= a (x i+1 x i ) + b

2
2

it will be necessary to use the factorization of the difference of two

squares, that is :
x2i+1 - x2 = (xi+1 - xi).(xi+1 + xi) = xi . (xi+1 + xi)
and then:

y cum i = axi + bxi .xi = xi (a + b.xi)

Property 10

2.4

x
x
x
y i = y i + b. x i - x i = y i + b i +1 - i = y i + b. i = y centrali
2
2
2

y i et y centrali

EXPONENTIAL MODEL

Let y = f(x)

Basic assumption of the model

rir(y)=Constant=c in the interval (xi, xi+1), with xi = xi+1 xi

Initial condition
x* = xi

y* = yi
14

Properties
rir(y) = air(lny) means that the exponential model of y against x is equivalent to the linear model
of lny against x. So being, its properties can be deduced by backwards application of logarithm
rules to the properties of the linear model of lny against x.

Figure 2.6

Figure 2.7

Graphical representation of the exponential model

Graphical representation of the linear model of lny against x

Exponential model of y
(y against x)

Linear model lny

(lny against x)

1.

General expression

y = y i e c( x x i )

lny = lnyi+c(x-xi)

2.

Value of yi+1 at the end of the

interval, xi

y i+1 = y i e cx i

lnyi+1= lnyi+cxI

3.

Variation, yi , during the interval,

xI

4.

Central value, y centrali , in the interval

xi

y i = y i +1 y i = y i e cx i 1 calculated from 1

y centrali = y i .e

c. x i
2

y centrali = (yi . yI+1)1/2

( y centrali = geometric
mean of the extremes yi
and yi+1)

15

ln y centrali = ln y i +

cx i
2

ln y centrali = (lnyi+lnyi+1)/2

5.

6.

Cumulative value, y cumi , during the

interval, xi

y cumi =

Mean value, y i , during the interval,

xI

yi =

x i +1
y
y.dx = c i
xi

1 y
= . i
x i
c x i

y cumi

yi = yi .
yi =

e cx i 1
c x i

(replacing yi using
Propriety 3)

y i+1 y i
ln y i+1 ln y i

y i y centrali

Other useful expressions

7.

Expressions of variation, yi

yi = c. y cumi
y i = c.y i x i
y i
= c.y i
x i

8.

Expression of amr (y)

amr ( y) =

9.

Expression of rmr (y)

in relation toyi

rmr ( y) in relation

10.

Expression of rmr (y)

to y i

= c = tir(y)

rmr(y) = amr (lny) =

ln y i
=c
x i

y decreases

11.

c<0
amr (y ) < 0
If y i < 0 alors

rmr (y ) < 0

y cumi > 0

y > 0
i

16

and vice-versa

12.

In the exponential model, the geometric mean of y i and y i+1 is equal to the central value,
y centrali (Prop. 4) and approximately equal to the mean value, y i (Prop. 6), been the
approximation better when x i is smaller.

Demonstrations
Cumulative value
Property 5

y cumi =

x i +1

xi

y dx =

x i +1

xi

yi e

c( x x i )

Relation between y i
and y centrali
Property 6
4th expression

h
From the approximation e 1 eh / 2
h

with h = c.xi

and from property 6-2nd expression, will be:

yi yi . ecx i / 2

Finally, by property 4-1st expression, one can conclude that:

y i y centrali

17

i +1
1
y
dx = = y i
c
c xi

CHAPTER 3 COHORT
3.1

COHORT INTRODUCTION

A cohort or annual class or a generation, is a group of individuals born in the same spawning
season. The following scheme illustrates the different phases of the life cycle of a cohort:

.
PRE-EXPLOITABLE PHASE
egg

larvae

EXPLOITABLE PHASE
Adults

Recruitment

1st spawning

juvenile

2nd spawning
time

new cohorts

Figure 3.1

Cycle of life of a cohort

Let us start, for example, with the egg phase. The phases that follow will be larvae, juvenile
and adult.
The number of individuals that arrive in the fishing area for the first time is called recruitment
to the exploitable phase. These individuals grow, spawn (once or several times) and die.
After the first spawning the individuals of the cohort are called adults and in general, they
will spawn again every year, generating new cohorts.
The phases of life of each cohort which precede the recruitment to the fishing area (egg,
larvae, pre-recruits), are important phases of its life cycle but, during this time they are not
usually subjected to exploitation. The variations in their abundances are mainly due to
predation and environmental factors (winds, currents, temperature, salinity,). In these non
exploitable phases mortality is usually very high, particularly at the end of the larvae phase
(Cushing, 1996). This results in a small percentage of survivors until the recruitment. Notice
that this mortality is not directly caused by fishing.
The recruitment of a cohort during the exploitable phase, may occur during several months in
the following schematic ways :

number of recruits

1 year

Figure 3.2

1 year

1 year

Types of annual recruitment to the exploitable phase

18

With some exceptions, the forms of recruitment can be simplified by considering that all the
individuals are recruited at a certain instant, tr called age of recruitment to the exploitable
phase. It was established that recruitments will occur on 1 January (beginning of the year in
many countries). These two considerations do not usually change the results of the analyses,
but simplify them and agree with the periods of time to which commercial statistics are
referred.
It should be mentioned that not all the individuals of the cohort spawn for the first time at the
same age. The proportion of individuals which spawn increases with age, from 0 to
100 percent. After the age at which 100 percent of the individuals spawned for the 1st time, all
the individuals will be adult. The histogram or curve that represents these proportions is
called maturity ogive.
In certain cases, the maturity ogive can also be simplified supposing that the 1st spawning
occurs at the age tmat designated as age of 1st maturity. This simplification means that the
individuals with an age inferior to tmat are considered juveniles and those with the same age or
older, are considered adults.
Figure 3.3 represents a maturity ogive with the shape of a histogram or curve :

Proportion

100
80
60
40
20
0
0

3.2

age

t mat

Figure 3.3

Maturity ogive

EVOLUTION OF THE NUMBER OF A COHORT,

IN AN INTERVAL OF TIME

Consider the interval (ti, ti+1) with the size Ti = ti+1 - ti of the evolution of a cohort with time
and Nt the number of survivors of the cohort at the instant t in the interval Ti (see
Figure 3.4).
The available information suggests that the mean rates of percentual variation of Nt can be
considered approximately constant, that is, rmr (Nt) constant.

19

Basic assumption

Numbers

The relative instantaneous rate of variation of Nt , in the interval Ti is :

rir (Nt) = constant negative = - Zi

Ni = number of survivors
at the beginning of the
interval (ti, ti +1)
Ni+1 = number of survivors
at the end of the interval
(ti, ti +1)

Ni
N
Ni+1

ti

t t i+1
Ti

Figure 3.4

time

Evolution of N in the interval Ti

The model of the evolution of Nt, in the interval Ti, is an exponential model (because rir(Nt is
constant). This model has the following properties:

Properties
1.

General expression. From the basic assumption

rir(Nt) = -Zi
with the initial condition that, for t = ti it will be Nt = Ni then:

N t = N i e Z( t t i )
2.

Number of survivors, Ni+1 , at the end of interval Ti

N i+1 = N i e

3.

ZT i

Number of deaths, Di , during the interval Ti

D i = N i N i +1
D i = N i (1 e Zi Ti )
(notice that Di is positive but the variation Ni = Ni+1- Ni is negative)

20

4.

Cumulative number of survivors, Ncumi , during the interval Ti

N cumi =

Di
Zi

N cumi = N i

5.

1 e Zi .Ti
Zi

Approximate central value , Ncentrali , in the interval Ti

N centrali N i e

6.

Zi Ti

Mean number, N i, of survivors during the interval Ti

Ni =

N cumi
Ti

1 e Zi Ti
Ni = Ni
Zi Ti
Ni =

Di
Zi Ti

Ni =

N i N i+1
lnN i lnN i +1

Ni Ni e

Zi Ti

(Ricker)

N i N centrali when Zi Ti is small (Zi.Ti < 1)

2

Comments
1. The basic assumption is sometimes presented in terms of absolute instantaneous rates, that
is:
air (Nt) = - Zi .Nt

{ air (Nt) proportional to Nt} or

air (lnNt) = -Zi

Zi = mortality total coefficient, assumed constant at the interval Ti
Notice that:
+Zi = rir of total mortality of Nt
Zi = rir of variation of Nt

2. Unit of Zi

21

From the definition, it can be deduced that Zi is expressed in units of [time]-1. By

agreement, the unit year-1 has been adopted, even when the interval of time is smaller or
bigger than a year.
The following expressions show, in a simplified way, the calculation of the unit of Zi,
with the rules and usual symbols [..] of dimension in the determination of physical units .

[1] [dN t ] = [ Z ]
i
[ N t ] [dt ]
1
number

= +[Z i ]
number time
3.

then

[Zi ] = time1

Annual survival rate, Si

When Ti =1 year, it will be :
N cumi = N i = Di
Zi

and also

N i+1 = N i e Zi
Si = Annual survival rate in the year i
(or percentage of the initial number of individuals that survived at the end of the year).
Si =

N i +1
Ni

Si = eZi

1 - Si = Annual mortality rate in the year i

The percentage of the initial number of individuals that die during the year is, by
definition, the relative mean rate rmr (Nt) of mortality of Nt , over one year, in relation to
the initial number, NI

1 Si =

N
N N i+1
Di
= 1 i +1 = i
Ni
Ni
Ni

1 S i = 1 e Zi
4. Absolute mean rate

amr ( N t ) = Z i .N i

5. Relative mean rate

rmr( N t ) = Z i in relation to N i

22

6. Notice that Si takes values between 0 and 1 , that is:

0 Si 1

but

Zi can be > 1

7. If the limits of the interval Ti were (ti, ) then it would be:

Ti =
Ni+1 = 0
Di = Ni
N cumi =

Ni
and
Zi

Ni = 0

3.3

CATCH, IN NUMBER, OVER AN INTERVAL OF TIME

The causes of death of the individuals of the cohort due to fishing will be separated from all
other causes of death. These other causes are grouped together as one cause designated as
natural mortality. So, from the properties of the exponential model, the result will be
rir(Nt) total = rir(Nt) natural + rir(Nt) fishing
Supposing that, in the interval Ti, the instantaneous rates of mortality due to natural causes
and due to fishing are constant and equal to Mi and to Fi, respectively, then
Zi = Fi + Mi
Multiplying both factors of the previous equation (equality) by .N cumi , then:
Z i N cum i = Fi N cum i + M i N cumi

Z i .N cum i

is the number, Di , of deaths due to total mortality,

D i = Z i .N cumi

In the same way Fi .N cumi will be the number of dead individuals due to fishing, that is, the
Catch, Ci, in number, and then:
C i = Fi .N cumi
Notice too that, M i .N cumi will be the number of dead individuals due to natural causes.
The exploitation rate, Ei , during the interval Ti was defined by Beverton and Holt (1956) as:
23

and then,

Ei =

number capture C i
=
number dead
Di

Ei =

Fi .N cumi

Ei =

Z i .N cumi

Fi
Zi

or

Fi
Fi + M i

The capture in number, Ci, in the interval Ti, can be expressed in the following different ways:
C i = Fi .N cumi
Ci = Fi .Ni . Ti
C i = E i .D i

Ci =

Fi
.D i
Zi

Ci =

Fi
N i 1 e ( Fi + Mi )Ti
Fi + M i

Comments
1. Ricker (1975) defines the exploitation rate, Ei*, as the percentage of the initial number
that is captured in the interval Ti, that is: Ei* = Ci / Ni.
a) Rickers definition may be more natural, but mathematically Beverton & Holts
definitions are more useful.
b) It is easy to verify that E *i = E i .(1 e Zi .Ti )
2. The exploitation rate, Ei , does not have any unit, it is an abstract number.
3. The possible values of Ei are between 0 e 1, being 0 when there is no exploitation and 1
when the capture Ci is equal to the number of total deaths Di, that is, when Mi = 0.

3.4

INDIVIDUAL GROWTH

In order to study the evolution of the biomass of a cohort, one can use the model of the
evolution of a cohort, in number, and combine it with a model of the evolution of the mean
weight of an individual of the cohort. In effect, the biomass Bt is equal to Nt.Wt where Wt is
the individual mean weight at the instant t.

24

To define a model for the individual growth weight Wt, there are then two possibilities :

ALTERNATIVE 1:
A) To define a model for the mean individual growth in length, Lt
B) To define the relation Weight-Length.
C) To combine A) with B) and obtain a mode for the mean individual growth in
weight, Wt

ALTERNATIVE 2:
D) To define directly growth models for Wt and Lt.

ALTERNATIVE 1
A) Model for the individual growth by length
The models that are used in fisheries biology are valid for the exploitable phase of the
resource. The most well known is the von Bertalanffy Model (1938) adapted by Beverton
and Holt (1957). The existing observations suggest that there is an asymptotical length, that
is, there is a limit to which the individual length tends.

Figure 3.5

von Bertalanffy Model

t - age
Lt - individual mean length at the age t
L - asymptotical length
tr beginning of the exploitable phase
So, Lt presents an evolution where :
air(Lt) is not constant (because growth is not linear)
rir(Lt) is not constant (because growth is not exponential)
However, it can be observed that the variation of the quantity (L - Lt) (which we could call
what is left to grow), presents a constant relative rate and can be described by an
exponential model. So, we can adopt the:

25

Basic assumption
rir (L - Lt) = - K = negative constant during all the exploited life
where K is the growth coefficient (attention: the growth coefficient K is not the velocity of
growth but the relative velocity of what is left to grow !!).
The properties of this model can be obtained directly from the general properties of the
exponential model. The initial condition :
t = t a L t = La
where ta (and La), that corresponds to an instant within the exploitable phase, will be adopted.
The properties of the model of individual growth by length by Beverton & Holt (1957)
deduced from the exponential model of ( L L t ) are summarized as :
Properties of the exponential model
for (L L t )

1.

von Bertalanffy Model

for L t

General expression

(L L t ) = (L La ) e K(tt )

L t = L (L La ) e K( t ta )

Parameters
L = asymptotic length
K = growth coefficient

Parameters
L = asymptotic length
K = growth coefficient

t a
=
L a

t a
=
L a

initial condition

initial condition

For ta = t0 and La = 0 :

For ta = t0 and La = 0 :

(L L t ) = L e K (t t

L t = L L e K (t t 0 )

L t = L . 1 e K .( t t 0 )
2.

Value at the end of the interval Ti

(L L i+1 ) = (L L i ) e

KTi

L i +1 = L (1 e KTi ) + L i e KTi
Ford-Walford expression (1933-1946)

26

3.

Variation during the interval Ti

) (
) (
)
= (L L i ). e k .Ti (L L )

As (L L i ) = L i

L L i = L L i +1 L L i

It will be::

) [ k.Ti 1]

4.

L i = (L L i ) 1 e KTi

= L L e
i

Cumulative value during the interval Ti

(L Li )cum

(L Li ) Li
=
K
K

Li
Lcumi = L . T i K
from:
(L Li ) cum = L . T i Lcum =
i
i

Li

5.

Mean value during the interval Ti

(L L i ) =

(L L i )cum

Ti

L i
K Ti

Li = L

L i
K Ti

Gulland e Holt expression (1959)

from: (L L i ) = L L i
6.

Central value of the interval Ti

(L L )central

= (L L a ) e

K t central t a
i

L central = L (L L a ) e
i

(L L )central

L Li

and

and

K ( t central t a )

L central L i

B) Relation Weight-Length
It is common to use the power function to relate the individual weight to the total (or any
other) length. Then :
Wt = a L t

where constant a is designated as condition factor and constant b as allometric constant. This
relation can be justified accepting that the percentage of growth in weight is proportional to
the percentage of growth in length, otherwise, the individuals would become disproportionate.
Thus, the basic assumption is :
rir (W ) = b rir (L )

where b is the constant of proportionality.

27

C) Combination of A) and B) and comments:

1.

From the combination of W = a Lb with L t = L (1 e K ( t t 0 ) ) we have

Wt = W (1 e K ( t t 0 ) ) b with W = a Lb
This relation of growth in weight is designated as the Richards equation (1959).
When b=3 the equation is the von Bertalanffy growth equation (1938).

2.

From, W = a Lb we have, by definition, Wcentrali = a Lbcentrali , where Wcentrali is the value

corresponding to L centrali

3.

Let Wi* be the weight corresponding to Li , that is,

As, Li L centrali then, Wi* a Lbcentrali = Wcentrali

Wi

= a (Li )

In practice, L centrali and Wcentrali are preferred to the mean points

4.

The Richards and von Bertalanffy models are not the only models used for the evolution
of Wt. Other models which have also been used in stock assessment are :
Gompertz model (1825) and Ricker model (1969). (see Alternative 2 property 3)

5.

Historically, the von Bertalanffy model was developed from the basic assumption
2
3

tia (Wt ) = Cte1 .W Cte 2 .W

Where Cte1 and Cte2 were designated by von Bertalanffy as the anabolism and the
catabolism constants, respectively.
Adopting the relation W = a L3 the basic assumption will become

air(Lt) = Cte1 - Cte2 . L

(where Cte1 e Cte2 are other constants).
The solution of this diferential equation gives the von Bertalanffy equation.

ALTERNATIVE 2

D) Model directly for Wt and Lt

In alternative 1, a model was developed for growth in length, and then a model was built for
growth in weight, using the relation between weight and length.
The basic assumption adopted by alternative 1 used the characteristic (L L t ) instead of
weight in order to be able to have a characteristic with a rate rir constant, that is, an
28

exponential model. The relation W=a.Lb was adopted to obtain the model of growth in
weight. Notice that it can be said that L was considered as a function of W, that is, L=
(W/a)1/b. It will then be possible to adopt, instead of that function of W, another function of
the weight H (Wt ) , in order to be able to formulate directly the basic assumption:
air[H(W ) H(Wt )] = K = constant
with the initial condition
t = t a Wt = Wa

Properties
The properties of this model (once it is an exponential model) can be obtained directly from
the general properties of the exponential model. It is particularly interesting to derive the
general expression Wt resulting from different choices of function H.
1.

2a.

General expression

[H ( W ) H ( Wt )] = [H ( W ) H ( Wa )].e K ( t ta )

or

H (Wt ) = H (W ) [H (W ) H (Wa )] e K( t t a )

Richards equation in weight

The result will be the general
expression:

Adopting the following function H (Wt) = Wt1/b

1

Wtb = Wb ( Wb Wab ) e K ( t ta )

that is, the Richards equation; and when b=3, this is the equation of von Bertalanffy, so:
2b.

von Bertalanffy equation in

weight, will be :

3.

Gompertz equation in weight

1
3
t

1
3

W =W

1
3

1
3
a

( W W ) e K ( t ta )

Adopting the function H(Wt) = ln Wt

The result will be the general expression:
ln Wt = ln W (ln W ln Wa ) e K ( t ta ) )

4.

The respective equations in length can be obtained by adopting other functions of

H(Wt) :

4a.

von Bertalanffy equation in

length

4b.

Gompertz equation de
in length

Adopting

H(Wt) = Lt it will be:

Lt = L - (L-La).e-K.(t-ta)
Adopting H(Wt) = ln Lt it will be:
lnL t = lnL (lnL lnL a ) e

29

K .( t ta )

5.

5a.

Simplified equations
The individual growth equations, both in length and in weight, are simplified when
one selects H(Wa) = 0 for ta = t*

So, the simplified general

expression will be reduced to :

H(Wt ) = H(W ).(1 e K( t t*) )

Simplified Richards equation,

in weight, will be :

Wt = W 1 e K( t t 0 )

where t* was represented by t0 because H(Wa) = 0 in Richards model means that Wa

will also be zero.
5b.

5c.

5d.

Simplified Gompertz equation,

in weight, will be:
Simplified Richards equation
in length, will be:
Simplified Gompertz equation,
in length, will be:

ln Wt = ln W 1 e K (t t

In this case H(Wa) = 0 corresponds to Wa = 1

Lt

= L 1 e K(t t 0 )

(with La = 0 for ta = t0)

lnL t = lnL 1 e K(t t*)

(with La = 1 for ta = t*)

Comments
1. Gompertz equation , in weight, is similar to Gompertz equation, in length, but, in their
simplified forms, t* represents different ages, because they will correspond, respectively,
to Wa = 1 and to La = 1.
2. Gompertz equation, in length, is similar to von Bertalanffy if Lt is substituted by ln Lt. In
practice, this fact allows the utilization of the same particular methods to estimate the
parameters in both equations, using L in the von Bertalanffy expression and lnL in the
Gompertz expression. (See Section 7.4)
3. It is important to notice, once again that, in practice, Lcentrali and Wcentrali are used instead
of the mean values,Li andWi.
4. Gompertz growth curve in length, has an inflection point, (tinfl, Linfl), with:
tinfl = ta+(1/K).ln(ln(L/La)) Linfl = L / e
5. Gompertz growth curve in weight has an inflection point, (tinfl, Winfl) with:
tinfl = ta+(1/K).ln(ln(W/Wa)) Winfl = W / e
6. Richards growth curve in length does not have an inflection point but the growth curve in
weight has an inflection point, (tinfl, Winfl).

30

1
b

t infl = t a K .ln
1
W b
1 a
W

1 b
Winf l = (1 ) . W
b

In the particular case of the von Bertalanffy equation it will be :

Winfl = (8/27).W

and

tinfl = t0+(1/k).ln3

7. Some authors refer Gompertz equation in other ways, for example, using the inflection
point tinfl and the asymptotic weight W, instead of the parameters ta and Wa.
It will then be

w t = w . exp e k ( t tinf ) or

L t = L . exp e k ( t t inf )

Sometimes the length expression is presented in its general form: L t = a. exp(b.e c. t )

The parameters of the length model will then be: L = a; k = -c et tinfl = (1/c).ln(-b)

8. The growth in length presents an inflection in fish farming, where the study of growth
covers very young ages and it is common to use the Gompertz equation. In fisheries, the
tradition is to use the von Bertalanffy equation.
9. A model that can sometimes be useful, is the Ricker model (1975). This model is valid
for a certain interval of time Ti and not necessarily for all the exploitable life of the fishery
resource. In fact, the model is based on the basic assumption that the individual growth is
exponential in the interval Ti.
It will be, for example, L t = L i .e Ki .( t t i ) where Ki can be different from one interval to the
next.

3.5

BIOMASS AND YIELD, DURING THE INTERVAL Ti

1. Biomass

Theoretically, it could be said that the biomass at the instant t of the interval Ti is given by:
Bt = Nt . Wt
Thus, the cumulative biomass during the interval Ti would be:
t i +1
B cumi = B t .dt

ti
and the mean biomass in the interval Ti would be:
Bi =

B cumi
Ti

31

In the same way, the mean weight of the cohort, Wi , in the interval Ti would be:
Wi =

B cum i
N cum i

The biomass can be obtained by dividing both terms of the fraction by Ti, as
Bi = N i .Wi

2. Yield

The yield, Yi , during the interval Ti will be expressed as the product of the catch in numbers,
times the individual mean weight :

Yi = C i W i

Comments
In practice W i is considered approximately equal to Wcentrali

at the interval Ti .

Other expressions of Yi will also be :

Yi = Fi .N cum i .Wi
= Fi .B cumi

= Fi .N i .Wi .Ti
= Fi . Bi .Ti

3.6

COHORT DURING THE EXPLOITABLE LIFE

Consider the evolution of a cohort during the exploitable life, beginning at age tr, and intervals
of time, Ti, covering all the exploitable phase (frequently the intervals are of 1 year...).
Figure 3.6 illustrates the evolution of the number of survivors of the cohort, N, and the
evolution of the catches in number, C, which are obtained during the successive intervals of
time Ti.

32

Number

N
C

tr

Figure 3.6

ti

Ti

t i+1

tim e

tx

Evolution of the number of survivors of the cohort, N, and the catches in

number, C

weight

Figure 3.7 illustrates the evolution of the biomass of the cohort, B, and the evolution of the
catches in weight, Y, which can be obtained during the successive intervals of time Ti.

tr

ti

Ti

t i+1

time

tx

Figure 3.7 Evolution of the biomass of the cohort, B, and the catches in weight, Y
Values of the most important characteristics of the cohort,
during all the exploitable phase

Duration of the life of the cohort

= Ti

Total number of deaths

D = Di

Cumulative number of survivors

Ncum = cum i

Mean number of survivors

N = N cum /

33

(= R (recruitment) when all the

individuals die)

Cumulative biomass of survivors

Bcum = Bcum i

Mean biomass of survivors

B = Bcum /

Catch in number

C = CI

Catch in weight

Y = YI

Mean weight of the cohort

Wcoorte = Bcum / N cum = B / N

Mean weight of the catch

Wcapt = Y / C

Critical age

tcrtical = age where B is maximum, when

the cohort is not exploited.

Comments
1. At first sight it may seem that the values of the characteristics of a cohort during all the
exploitable phase are of little interest, because, very rarely is fishing applied to an
isolated cohort. At each moment, the survivors of several cohorts are simultaneously
present and available for fishing.
2. Despite this fact and for reasons which will be mentioned later, it is important to analyse
the characteristics of a cohort during all its exploitable life. Knowledge of the evolution of
a cohort, in number and in biomass, and particularly the critical age, is important for the
success of the activities in fish farming. As Bt = Nt . Wt , the critical age, tcrtical , will be
the age t in the interval Ti where
rir (Wt) = - rir(Nt)=M because, the critical age
derivative of B will be equal to zero.

being the maximum biomass, the

3. Notice that Ncum can be expressed in function of the recruitment as

Ncum = R.{...}
where {...} represents a function of biological parameters and annual fishing mortality
coefficients Fi during the life of the cohort. Bcum can also be expressed as:
Bcum = R .{...}
where the function{...} also includes growth parameters.

34

3.7

SIMPLIFICATION OF BEVERTON AND HOLT

Beverton and Holt (1957) deduced algebraic expressions for the characteristics of a cohort
during the exploited life, adopting the simple assumptions:
1. The exploited phase of the cohort is initiated at age tc and is extended to the infinite.
2. The natural mortality coefficient, M, is constant during all the exploitable phase.

Figure 3.8

Life of a cohort during the exploitable phase

3. The fishing mortality coefficient, F, is constant during all the exploited phase.
4. Growth follows the von Bertalanffy equation with for L a = 0 for t a = t 0

Basic equations referring to the exploited phase

1.

c (ratio between length at age t c and

the asymptotical length)

c = Lc = 1 e k( t c t o )
L

2.

Recruitment

R c = R e M (tc t r )

3.

Cumulative number

Rc
Ncum = Z

4.

Catch in number

C = E Rc

5.

Cumulative biomass

1
(1 c ) + 3 (1 c )2 (1 c )3
=

3
R
W

c
Bcum
M+F+K
M + F + 2K M + F + 3K
M + F

35

Bcum = R c .W .[...]

It can just be written

6.

Mean weight in the catch

W = Bcum = Z . W .[...]
Ncum

7.

Catch in weight

Y = C W = F . Bcum = F . Rc . W .[...]

8.

Mean age in the catch

1
t = tc +
Z

9.

Mean length in the catch

L = L (L Lc).

10.

Critical age

Y
tcritique = t o

Z
Z+K

1 M
.ln

K M + 3K

Comments
1. The simplification of Beverton and Holt allows the calculation of any characteristic of the
cohort during its life with algebric expressions, avoiding the addition of the values of the
characteristic in the successive intervals Ti. This was useful for calculations in the 60-70s
when computers were not available. It is also useful when the only available data is
natural mortality, M, and growth parameters.
2. At present, the simplified expressions are also useful to study the effects on the biomass,
on yield and on the mean weight of the catch due to changes in the fishing mortality
coefficient, F, and/or on the age of first capture, tc. These analyses are usually illustrated
with figures. For example, Figure 3.9 exemplifies the analyses of the biomasses and of
the catches in weight, obtained from a cohort during the exploited life, subjected to
different fishing mortality coefficients, assuming a fixed age tc .
Age of recruitment at the exploited phase, tc, fixed
tc=const

Figure 3.9 Evolution of the biomass and of the catch in weight from a cohort subject
to different fishing mortality coefficients and fixed tc (notice that the Figure
illustrates only the analysis and does not take into consideration the scales of
the axis).

36

3. Notice that the forms of the previous curves, Y and Bcum against F, do not depend on the
value of the recruitment and so, they are usually designated as curves of biomass and
yield per recruit, B/R and Y/R, respectively. The calculations are usually made with
R=1000.
4. The mean weight, the mean age and the mean length of the catch do not depend on the
value of the recruitment. The curves of the characteristics of a cohort during its life
against the fishing level, F, or against the age of first catch, tc, deserve a careful study,
for reasons which will be stressed in the chapter concerning the long-term projections of
the stock .
5. Bcum was calculated as:

B cum = N t . W t .dt
tc

where
Nt = R c .e

( M + F ).( t tc )

and

W t = W . 1 e

K .( t to ) 3

The calculations can also be made using other values of the constant b, from the relation
W-L, different from the isomorphic coefficient, b=3, using the incomplete mathematical
function Beta (Jones,1957).
6. The meansL andt can be calculated from the cumulative expression

Lcum = N t . Lt .dt

and

L = Lcum / Ncum

and

tcum = N t .t.dt as

tc

tc

t = t cum / N cum

These means are designated as weighted means, being the weighting factors, the number of
survivals, Nt , at each age t.

37

CHAPTER 4 STOCK
4.1

STOCK OVER A ONE YEAR PERIOD

4.1.1 EVOLUTION OF THE AGE STRUCTURE OF THE STOCK

Let us consider the evolution of a stock, with several cohorts, over the period of one year.
Figure 4.1 shows the structure of the stock at the beginning of the year, the mean
characteristics during the year and the age structure surviving at the end of the year.

Begin

During

End
N0

ages

N0

M0 F0 W0

N1

Recruits of year a+l

N1

cohort year a

i
N2

cohort year a-l

Ni

Mi Fi Wi
N i+1
cohort year a-i

Structure of the stock

at the beginning of the year a

Figure 4.1

Survivals at the end

of year a

Structure of the stock at the beginning and end of the year

Let i = 0,1,2,3, ... be the ages

4.1.2 CHARACTERISTICS OF THE STOCK AT THE BEGINNING OF

THE YEAR
Let :
Ni number of individuals at age i, at the beginning of the year.
wi - individual weight at age i, at the beginning of the year.
Bi - total biomass of the individuals at age i, at the beginning of the year.
N - total number of survivors of the stock at the beginning of the year.
B - total biomass of the stock at the beginning of the same year.

38

Then :
N = Ni
i

B = B i = (N i .Wi )
i

4.1.3 CHARACTERISTICS OF THE STOCK DURING THE YEAR

Let :
Mi, Fi e Zi

Then:

total mortality (natural and by fishing) coefficients,

at age i during the year

Wi

individual mean weight at age i during the year

mean number of individuals during the year

mean biomass during the year

catch, in number, during the year

catch, in weight, during the year

Wcatch

mean weight of the individuals caught during the year

Wstock

mean weight of the individuals of the stock during the

year

Ncum = Ncum i
i

Ncum
T
C = Ci
N=

(where T=1 year)

Bcum = Bcum i
i

B = Bcum
T
Y = Yi

(where T=1 year)

Y
C
B
=
N

Wcatch =
Wstock

with Ncumi, Ci, Bcumi and Yi calculated for all the ages i according to the expressions given
previously in Chapter 3 - Cohort.

39

4.1.4 CHARACTERISTICS OF THE STOCK AT THE END OF THE

YEAR
The number of individuals at beginning of age i+1 will be:

Ni +1 = Ni . e Zi

Let :
R - recruitment of the cohort in the following year
Then the number and the biomass of the stock at the beginning of the following year will be :
N = R + Ni +1

and

B = R. W1 + Bi +1
i

where the product R.W1 is the biomass of the recruitment of the following year.

Comments
1.

The end of the year coincides with the beginning of the following year. So, the number of
survivors of age i at the end of the year will be Ni+1, with age i+1.

2.

The sum of the total number of survivors of the stock at the end of the year is not equal
to the number of individuals of the stock at the beginning of the following year, because
one has to count the recruits entering that year.

3.

The total number of deaths, D, during the year, would be D = D i

i

4.

As the interval of time is 1 year, the cumulative values will be equal to the mean values,
that is :
N cumi = N i
Bcumi = Bi

N = Ni
i

B = Bi
i

5.

The utilization of the same symbols N, B, D, etc., for the stock and for the cohort should
not create any confusions.

40

4.2

FISHING PATTERN OVER A ONE YEAR PERIOD

4.2.1 FISHING LEVEL AND EXPLOITATION PATTERN

The direct action of fishing a stock can be represented by the coefficients of mortality by
fishing Fi. These coefficients are associated with the quantity of effort, with the disponibility
of the individuals of different sizes or ages, i, and with the fishing gears used by the vessels
during the year.
It is usual to separate the coefficients of mortality by fishing into two components:
One is designated as the level of intensity of mortality by fishing, F , during the year, called
fishing level, F. The level is associated with the quantity of fishing effort, (the number of
vessels fishing), the number of days, hauls, fishing hours during the year, and with the
efficiency or the fishing power of the vessels or gears. Another component, designated as
exploitation pattern, si, is associated with the selective properties of the fishing gears relative
to the sizes or ages of the individuals available to be captured, during that year.
The combined set of the fishing level (a unique value for all ages) and the exploitation pattern
(different values according to the size or the age), is designated as fishing pattern or fishing
regime. The designation of fishing pattern may cause some confusion with the exploitation
pattern, and the designation of fishing regime may be confused with what the economists and
managers call fishing regime.
The fishing pattern, Fi, of one age i during one year, is equal to the product of the fishing
level of that year, F , times the exploitation pattern of each age, si. That is:
Fi = F.s i
To analyse the effects of the coefficient of mortality by fishing, F , on the characteristics of
the resource and on the catches, the exploitation pattern is generally taken as constant from
one year to the next. Sometimes one analyses the effects of the changes of the exploitation
pattern maintaining a fixed fishing level, but one can also analyse the combined effects of the
two components.
The fishing level, F , is often represented by F.

41

4.3

SHORTTERM PROJECTIONS OF THE STOCK

Knowing the structure of the stock at the beginning of one year, it is possible to estimate the
characteristics of the stock during that year and project the structure of the stock for the
beginning of the next year (with an exception to the recruitment of that year), for different
values of the fishing level, F , (and for the exploitation pattern, si).
It is, of course, necessary to know the biological parameters of growth, maturity and the
natural mortality coefficients, in each age during the year.
Adopting one value for the recruitment of the following year, the projection could be repeated
for one more year and so on. The inconvenience in projecting the stock for several years will
be that those projections will depend more and more on the adopted annual recruitment
values. That is why, in practice, the stock and the catches are projected for one, or at the most,
two years. In Section 4.5, the estimation of recruitments will be discussed.
Notice that to project a stock for the following year, it is necessary to have data from the
previous year. So, the stock is firstly projected for the current year and then the catch and the
biomass are projected for the following year.
Let us suppose, for example, that in 1997 one wants to project the characteristics of the stock
for 1998. As the available data will be, in the best hypothesis, referring to 1996, one will have
to project the stock of 1996 for the beginning of 1997 and together with the recruitment of
1997, project the stock until the end of 1997 and only after that, can one project to 1998,
estimating previously the recruitment of 1998.

4.4

LONGTERM PROJECTIONS OF THE STOCK

Let us consider the vector N with the components N1, N2, ..., Ni, ... representing the structure
of the stock at the beginning of a year. Notice that N1, N2, ..., Ni, ... are the survivors of the
different cohorts of the stock, at the beginning of the year.
Let Mi and Fi be the natural and fishing mortality coefficients of age i, assumed to be constant
in the future years.
Figure 4.2 illustrates, with a theoretical stock, the projections of the numbers of survivors of
the different cohorts at the beginning of 1980, for the years to come, from 1981-1986 (the
values are expressed in million of individuals).
Notice that the recruitments are missing during the years 1981 to 1986, as they have not yet
occurred. So, it is clear that the respective survivors are also missing during those years.
Let us suppose now that the recruitments of the same period of years were equal to those of
1980, that is, 440 million of individuals. Figure 4.3 shows the projections in future years. It
can be seen that the values of the age structure of the stock in 1986 are equal to the annual
evolution of the cohort in 1980.

42

Year
1980 1981 1982 1983 1984 1985 1986
Age

0
1
2
3
4
5
6
Figure 4.2

440
995
367
68
245
345
76

326
680
229
41
149
209

223
425
139
25
90

139
258
84
15

85
156
51

51
95

31

Structure of the stock at the beginning of 1980 and projections of the

number of survivors (million of individuals) of the different cohorts
at the beginning of the years 1981-1986

Year
1980 1981 1982 1983 1984 1985 1986 1987

Age
0
1
2
3
4
5
6
Figure 4.3

440
995
367
68
245
345
76

440
326
680
229
41
149
209

440
326
223
425
139
25
90

440
326
223
139
258
84
15

440
326
223
139
85
156
51

440
326
223
139
85
51
95

440
326
223
139
85
51
31

440
326
223
139
85
51
31

Longterm projections of the cohort of 1980 and structure of the stock

assuming a constant annual recruitment of 440 million of individuals

One practical conclusion seems to be that to obtain the structure of the stock in 1986 it would
be enough to follow the evolution of the cohort of 1980 and then, it would not be necessary to
have the complete structure of the stock at the beginning of the year. It would be enough to
adopt one value for the recruitment ( R ) of a cohort (and, of course, assume that the
biological parameters would be stable in the following years). An advice would be to make
the calculations adopting a Recruitment of 1000 individuals (with computer software a
recruitment equal to 1 is adopted).
Except for the mean values age, length, weight in the catch, etc, which are independent of
the adopted value of R, the characteristics of the stock in the longterm, under the previous
conditions, are proportional to the recruitment. So, these projections are also designated as
projections by recruit , by LT projections or even as equilibrium projections.
The longterm projections do not only concern the survivors at the beginning of the year, but
also the values of the mean abundances, N i during the year, and the catches in number, Ci .

43

One can also project the total biomasses, B or B, the spawning biomasses, SB or SB , as
well as the catches in weight, Y, knowing the initial individual weights, Wi, the mean
weights, W , and other parameters like those of maturity.
It is important to verify that the cumulative values, Ncumi e Bcumi, of a cohort during 1 year
are equal to the mean values,Ni e Bi , during that year. The longterm projections can be
calculated as:
Nstock = Ncumi = Ni
Bstock = Bcumi = Bi

( SB stock = SB i)

Wstock = Bstock /Nstock

C = Ci
Y = Yi
Wcatch = Y / C

Several longterm projections can be made with different values of Fi, that is, with several
fishing levels, F , and/or with several exploitation patterns, si.
As mentioned before, the analyses of the effects of the fishing pattern on the catches and
stocks can be done with the two components (fishing level and exploitable pattern), separated
or combined.
Figure 4.4 (A-C) illustrates several types of yield per recruit curves, maintaining a fixed
exploitation pattern. The curves are different for other exploitation patterns.

(A)

Figure 4.4

(B)

(C)

Examples of types of curves of the Yield per Recruit (Y/R) against F,

given the exploitation pattern: (A with a maximum, B - flat-top,
C asymptotic)

44

Figure 4.5 (A-E) illustrates the relations between the most important characteristics of the
stock and fishing level, maintaining a stable exploitation pattern:
(A)

(B)

(C)

(D)

Figure 4.5

(E)

Relations between the longterm characteristics of the stock against the

fishing level F, (A mean number/R, B catch in number/R, C - mean
weight in the catch, D mean biomass /R, E yield/R)

A more detailed analyses of these curves will be presented later on, in Chapter 5.

Conclusion
The longterm structure of the
stock by ages during 1 year

Evolution of the cohort

during its life

Comments
1.

One projects a cohort during its life in order to obtain the longterm projection of the
stock for one year, assuming the annual recruitments to be constant.

2.

It is necessary to know Mi for all ages of the cohort, as well as Wi , W , si

and the fishing level, F, that is assumed to be constant in the years that follow.

3.

Any recruitment size can be used. Adopt R = 1000 (or R = 1) with worksheets on
your computer.

4.

The five most important characteristics of the stock are Nstock, Bstock, C, Y and
W catch (see previous pages in this chapter for the respective expressions of
calculation)

5.

A characteristic of the stock that is also important is the spawning biomass, SB. To
calculate SB, it is necessary to know the maturity ogive (or histogram).

6.

Longterm projections are also designated as equilibrium situations.

7.

Longterm projections are useful to define the longterm management objectives.

45

8.

Annual Wcatch , are independent of the recruitment size (such as L catch and t catch ).

9.

Economists transform the total yield, Y. into value Y$, the mean weight of the catch,
Wcatch into price of the catches, W$ catch , the catch per vessel (or the cpue) into value
of the production by vessel, U$, and the fishing level, F into costs of exploitation,
F$. The difference between the value of the catch and the cost of exploitation, Y$-F$,
is the profit, L$. Figure 4.6 illustrates an example of the LT relations between those
characteristics used by the economists against the fishing level, F.

Y$=value
F$=cost

Figure 4.6

4.5

Longterm relations between the value of the catch, the cost of exploitation and the profit against the fishing level

STOCKRECRUITMENT (SR) RELATION

The stockrecruitment relation, known by SR relation, associates the size of the stock, for
one year, with the recruitment which results from the stock spawning normally during that
year. The recruitment can be the recruitment at the exploitable phase and the stock can be the
spawning stock. This recruitment may occur one or more years after the spawning.
The problem with the relation between the parental population and the new generation is not
a special case of the fisheries resources, it is common to all the self renewable populations.
It is important to determine, in each case, the stock and the recruitment to be used. In fact,
that stock can be the total number of individuals (at the beginning of the year or the mean
value during the year), the number of adult individuals of the stock, the number of adult
females, etc. One can also adopt, not the abundances in number, but the corresponding
biomasses. The decision will depend on the type of resource and on the available data. It is
necessary to define if the recruitment is in weight or in number and if the recruitment is to the
fishing area or to the exploited phase.
In this manual, stock (S) will be considered as the spawning biomass and the recruitment (R)
will be expressed in number.
After the spawning, the individuals of the new generation will have to go through different
phases of the biological cycle: eggs, larvae and juveniles, before they become recruits. These
phases, which, in some species can take years, are not directly submitted to the fisheries
exploitation. That is why fishing in those years does not directly affect the size of the new
46

recruitment. It is true that fishing acts on the size of the parental biomass, but that does not
happen in the pre-recruit phases, and it is precisely for that interval of time that the relation
SR is applicable.
There is, in these pre-recruit phases, a great mortality due to climate and environmental
factors (winds, currents, temperatures, etc.) as well as due to biological factors (available
food, predation and others).
A great variety of factors (besides fishing) cause great fluctuations to the recruitment sizes,
therefore, the relation SR is a complex one. In conclusion, the relation SR, between the
stock and the resulting recruitment, can be considered independent of fishing.

.
.

. .
.
.
.

.
.

.
.

.
.
.

0
0
S

Figure 4.7

Example of the dispersion of the points in the relation SR

Figure 4.7 shows the type of dispersion of values in the plot of recruitment (R) against
parent stock (S).
Despite the difficulties, some models have been proposed for the relation SR. One of the
reasons for this is that there must be a relation. One point of the relation (S = 0, R = 0) is
even known already.

4.5.1 BEVERTON AND HOLT MODEL (1957)

Beverton and Holt, when analysing plaice fishery in British waters proposed one model,
which can be re-written as:
R=

.S
S
1+
k

where and k are constants.

47

Figure 4.8 Beverton and Holt Relation SR

This model shows a curve where R is asymptotic when S .

4.5.2 RICKER MODEL (1954)

Ricker, studying cod fishery in Canadian waters proposed another model that can be written
as:
R = .S.e-S/k
where and k are constants.
R

Figure 4.9 Ricker Relation SR

This model presents a maximum resulting recruitment at an intermediate value of the parental
stock, S.

4.5.3 OTHER MODELS

Other models of SR have been proposed, like the Deriso generalized model (1980), which
can be re-written (Hilborn & Walters, 1992) as :

S
R = .S.(1 c. )1 / c
k
(the model is the Beverton and Holt for c = -1 and when c 0 it is the Ricker model)
or Shepherds generalized model (1982), as:

48

R=

.S
S
1+
k

(the model is the Beverton and Holt when c = 1 and when c = 2 the curve is an pproximation
of the Ricker model)
Notice that in every model presented, S=0 implies R=0, as expected and the slope of the
tangent to the curve at that point, (0,0), is equal to the parameter . Sometimes, the relations
SR are presented as R/S in function of S, so the Beverton and Holt equation would be :

=
showing that the inverse of R/S is linear with S.
S 1+ S
k
With the Ricker model the relation between R/S against S would be:
S
R
= .e k
S

that is, R/S is exponential negative with S (or ln(R/S) is linear with S).
With the Deriso model it would be (R/S)c linear with S.
Mathematically, these linear relations can be useful to estimate the parameters and k, but
statistically they cause some problems, because the variable S appears in the response
variable and in the auxiliary variable.

Comments
1.

Remember that relations SR are independent of the fishing level.

2.

Relations SR may be introduced in the calculations of the stock projections. In that

case, the projections will have to be made every year and they will require the structure
of the stock at the beginning of the initial year.

3.

It must not be forgotten that there is a great dispersion of the observed points (S, R)
around the model.

4.

To determine the limit of the Deriso equation when c 0 it is enough to remember that
limit (1+A/n)n = eA when n and A is a constant.

5.

The Beverton and Holt model presents an asymptote, R .k when S . When S=k it
will be R = k/2.

The Deriso model presents the maximum:

Smax = k / (1+c), Rmax = .k / (1+c)1+(1/c)

49

The Shepherd model presents the maximum:

Smax = k.(c-1)-1/c, Rmax = ( /c).k.(c-1)1-(1/c)

4.6

RELATION BETWEEN R ANDB (RS RELATION)

Up to now the discussion have been centred on the relation SR, that is the relation between
the biomass S and the resulting recruitment R. There is another relation (which has already
been referred to in Section 4.4 Longterm projections of the stock, particularly in the
conclusion about the structure of the stock and the evolution of a cohort during all its life) that
could be called the relation RS, that is the relation between the recruitment R and the
resulting cumulative biomass, Bcum of a cohort during all its life for a given fishing pattern.
The cumulative biomass (or spawning biomass) of a cohort during its life, Bcum , is, as it has
already been seen, equal to:
Bcum= R . {function of biological parameters and of the fishing pattern, F}
As mentioned before, the cumulative biomass, Bcum , of a cohort during its life is equal to the
longterm mean biomass,B , of the stock during one year, so, one can refer to the Bcum asB.
It can then be said that in the longterm, if the biological parameters are considered constant,
then for a certain fishing pattern, the mean biomass of a stock during one year is proportional
to the recruitment. Figure 4.10 illustrates the proportionality.
LP

Longterm

The slope of the line, B/R

is a function of the selected F
R

Figure 4.10

Illustration of the proportionality in the longterm between the mean

Biomass B of a stock and the Recruitment (R), for a certain fishing level,
F, assuming that the exploitation pattern and the biological parameters
are constant

Notice that in Figure 4.5-D, the relation B/R against F was shown. The two curves 4.5-D and
that of Figure 4.10 are different representations of the same situation. While in Figure 4.5-D,
a value of F corresponded to a value of B/R given by the curve, in Figure 4.10 the value of F
will correspond to a slope B/R of a given straight line.
For other values of F, the straight line of Figure 4.10 will have different inclinations.
Figure 4.11 shows several lines according to the different values of F.

50

F=0

LP

LT

B
F smaller than F1

F1
F bigger than F1
0
0

Figure 4.11

Illustration of four lines corresponding to different fishing levels, F

Figure 4.12 shows the overlapping of the Ricker curve SR (Figure 4.9) and the line in
Figure 4.10, the axes of this last figure having been switched. To avoid confusions, it is
convenient to refer the slope of the line corresponding to a value F in relation to the axis R.
Longterm
S-R

Fgiven

R-S

0
0

Figure 4.12

Overlapping of the curve SR with the line R-S for a given F

One can start from a value of biomass, and, through the relation SR, determine the future
recruitment, R. That R will give a resulting biomass, B , through the straight line. This process
can be repeated until a situation of equilibrium is found.
It will be possible to illustrate the combined analysis through the two relations for a certain
fishing level.
For example, let us select one valueB1. The curve SR allows the calculation of the resulting
value R2. Through the line (of a given F), the resulting value ofB2 will correspond to that
value of R2, and so on. Figure 4.13 shows that the process reaches the equilibrium point
(RE,BE). This point would correspond to a value of the fishing level F, determined by the
slope of the line in relation to the R axis.

51

LT
S-R
R
LTLLLLLLLLLLLLLLLLLLL
RE
R2

Fgiven

R-S

0
0

Figure 4.13

B1

B2 BE

Illustration of the process that, starting from a biomassB1, theoretically

leads to the equilibrium point (RE, BE)

Notice that the intersection of the two relations does not always lead to an equilibrium point.
The existence of an equilibrium point depends on the angle between the straight line and the
curve at the intersection point.

52

CHAPTER 5 BIOLOGICAL REFERENCE POINTS

AND REGULATION MEASURES
5.1

BIOLOGICAL REFERENCE POINTS FOR THE MANAGEMENT

AND CONSERVATION OF FISHERIES RESOURCES

The longterm objectives for fisheries management should take into consideration scientific
fishing research and population dynamics, as well as the climatic changes that may affect the
stocks.
In order to define these longterm objectives we have to consider the values of the fishing
level, which allow bigger catches in weight, whilst also ensuring the conservation of the
stocks. The extreme values of the biomass or the fishing level, which might seriously affect
the self renovation of the stocks, also have to be considered. These fishing level values, of
catch and biomass are designated as biological reference points (BRP). In this manual some
of the different types of BRP will be considered (Caddy, & Mahon, 1995; FAO, 1996 and
ICES, 1998).
The Target Reference Points, TRP are BRP defined as the level of fishing mortality or of the
biomass, which permit a longterm sustainable exploitation of the stocks, with the best
possible catch. For this reason, these points are also designated as Reference Points for
Management. They can be characterized as the fishing level Ftarget (or by the Biomass, Btarget).
The most well known Ftarget is F0.1 but other values, like Fmax, Fmed, and FMSY will also be
studied.
For practical purposes of management, the TRP will be converted, directly or indirectly, into
values of fishing effort, given as percentages of those verified in recent years.
The Limit Reference Points, LRP are maximum values of fishing mortality or minimum
values of the biomass, which must not be exceeded. Otherwise, it is considered that it might
endanger the capacity of self-renewal of the stock.
In the cases where fishing is already too intensive, the LRP can be important to correct the
situation or to prevent it from getting worse.
The LRP are limit values, mainly concerned with the conservation of marine stocks and they
are therefore also referred to as reference points for conservation (this designation does not
imply that the Ftarget are not concerned with conservation).
Several LRP have been suggested, which will generally be referred to as Flim or Blim. In this
manual the levels of biomass Bloss and MBAL will be referred to as Blim and the fishing levels
Floss and Fcrash as Flim .
The Precautionary Principle, proposed by FAO in the Conduct Code for Responsible
Fisheries (FAO, 1995), declares that the limitations, uncertainties or lack of data for the
assessment or for the estimation of parameters, cannot be justification for not applying
regulation measures, especially when there is information that the stocks are over-exploited.

53

From this point of view, it is important to make clear which basic assumptions are necessary
in order to estimate the consequences on the catches and on the abundance of the stocks.
The uncertainties associated with the estimation of Flim , and Blim , therefore lead us to
determine new reference points, called Precautionary Reference Points, Fpa or Bpa.
The assumptions and the consequences of adopting alternative hypotheses about the stock and
fishing characteristics should always be presented to justify the estimated values of Fpa (or
Bpa).
The new limits (Fpa or Bpa) due to the application of the Precautionary Principle, will be
more restrictive than the LRPs. The practical consequences of these new limits are the
regulation measures designed to control the fishing effort which are more severe than in those
cases where there is appropriate data.
It can be said that this is the price to pay for not having the appropriate conditions to make
available reliable data and information.
The Precautionary Approach, suggests that the results of fisheries research should be adopted
by management with regard to the formulation of the regulation measures and that these
measures should also take into consideration the socio-economic and technical conditions of
fishing (FAO, 1996).
A final remark about all the Biological Reference Points mentioned above :
The evaluation of the biological reference points has to be updated, taking into consideration
the possible changes in the biological parameters or any other necessary correction of the
exploitation pattern. This fact is important because the new biological reference points will be
different from the previous ones.

54

5.2

BIOLOGICAL TARGET REFERENCE POINTS

(Fmax , F0.1 Fmed and FMSY)

5.2.1 Fmax
Definition
1.

Consider the longterm yield per recruit, Y/R, as a function of F, for a certain
exploitation pattern.
Fmax is the point of the curve, Y/R against F, where Y/R is maximum.
Figure 5.1 shows a curve of Y/R against F.

Figure 5.1

Y/R as function of F for a certain tc constant,

showing Fmax and Ymax

In Chapter 4 it was mentioned that to estimate the longterm projections one

could assume that the recruitment is constant and equal to 1 (R=1). In this way,
the mathematical expressions are sometimes written with Y instead of Y/R.
2.

Mathematically, at point Fmax, the derivative of Y/R against F is equal to zero, that
is,
For

F = Fmax will be

dY
=0
dF

air (Y ) = 0 (Value of Y is maximum)

For

F < Fmax will be

dY
>0
dF

air (Y ) > 0 (Y is increasing with F )

For

F > Fmax will be

dY
<0
dF

air (Y ) < 0 (Y is decreasing with F)

Geometrically, the slope of the tangent to the curve is equal to zero for F = Fmax ,
positive for F < Fmax and negative for F > Fmax .

55

Comments
1.

Bmax /R and Ymax/R are the values at Fmax.

It is also convenient to analyse the situation of B /R at the points F Fmax.

F < Fmax corresponds to B > Bmax
F > Fmax corresponds to B < Bmax
Point Fmax does not depend on the value of the recruitment.
2.

For another relative pattern of exploitation there will be another Fmax .

3.

All the points of the curve Y/R against F, are longterm points or equilibrium
points.

4.

When the level, F, is bigger than Fmax it is said that there is growth overfishing.
It is convenient to present the two curves Y/R and B /R against F, in the same
graph (usually with different scales).

Figure 5.2

Longterm curves of Y/R andB/R against F, given an

exploitation pattern

Fmax was adopted by the majority of the International Fisheries Commissions as a

longterm objective of management (1950-1970).
Even today Fmax is used as a target-point having been proposed as a Limit
Reference Point (LRP) in some cases.
The flat-top and asymptotical curves do not allow the determination of an Fmax .
The definition of Fmax does not consider the appropriate level of spawning
biomass.
only indicates the value of F which gives the maximum possible yield per
recruit from a cohort during its life, for a given exploitation pattern.
Fmax

The analyses of these longterm curves, mainly of B , Y and W catch against the
fishing level, give information about the abundance of the resource (or catch per
vessel), total yield of all the fleet and mean catch weight for different fishing
levels.

56

5.2.2 F0.1
1.

Definition
Consider the longterm yield per recruit, Y/R, as a function of the coefficient of fishing
mortality, F. One value of air(Y/R), corresponds to each fishing level, F. The air(Y/R) is
maximum when F = 0 and decreases, being zero when F = Fmax.
The point F0.1 is the value of F where air(Y/R) is equal to 10 percent of air(Y/R)
maximum.
The figure 5.3 illustrates this situation.

Figure 5.3

2.

Y/R showing the reference target point F0.1

For F = 0, the biomass per recruit, B/R will be B0 /R, also designated as Virgin
Biomass or Non-exploited Biomass. The air(y) at F=0 is also equal to B0
In fact,

Y = F B implies

Then, for

F = 0,

dY
dB
= B+ F
dF
dF
dY
air(Y) =
= B0
dF F=0

So, from the definition given in point 1 one can also say that F0.1 is the value of F where
air(Y) = 10 percent of the virgin biomass.
3.

Calculation of F0.1

Let the function V = Y 0.1. B0 .F

It can be proved that the function V is maximum when F = F0.1
dV
In fact, V is maximum when
= 0 , then:
dF

dV dY
dY
=
0.1 B0 = 0
or
= 0.1 B0
dF dF
dF
Therefore, the value of F corresponding to the previous dY/dF is the value of F0.1.
F0.1 can then be calculated by maximizing the function V = Y 0.1.B0 .F
57

B0 can be calculated, for example, from the longterm relation ofB against F,
when F = 0.

Graphically it will be:

longterm

Figure 5.4

4.

Curve Y/R showing the maximum of the function V

Why adopt air(Y/R) equal to 10 percent and not any other percentual value, for example,
20 percent?
Gulland and Boerema (1969) presented some arguments, including financial arguments.
Some countries (like South Africa) adopt the value of 20 percent with a resulting
reference point F0.2 that is more restrictive than F0.1.

5.

Figure 5.5 illustrates the two biological reference points Fmax and F0.1.

Figure 5.5

Longterm variation of Y/R and B/R against F

and points corresponding to Fmax and F0.1

B0.1 and Y0.1 are the values of B and Y corresponding to F0.1

is always smaller than Fmax
B0.1 is always larger than B
Y0.1 is always smaller than Ymax, although, in practice, the difference is not large.
F0.1

The second sentence above indicates the advantages of B0.1 over Bmax. The last sentence
shows that Y0.1 is not the largest possible catch, but is acceptable as a target point of
management. The fact that B0.1 is larger than B max suggests that the fishing level F0.1 is
preferable to Fmax as TRP.

58

Notice that F0.1 can be calculated even when the curve is asymptotical or flat-top.
Another value of F0.1 will be obtained if the exploitation pattern changes.
In the years 1960-70 F0.1 started to be preferred to Fmax as a target point by resource
managers, and it was adopted in the 80s, as a longterm objective by many International
Fisheries Commissions and by the EEC.

5.2.3 Fmed
1.

Definition
This target point considers the relation S-R between the stock and the resulting
recruitment, in order to avoid the assumption of a constant recruitment.
Let (the spawning or total) biomasses and the resulting recruitments for each year of a
certain period of time be known. In this case, one can calculate the median value of the
ratios between the annual spawning biomasses and the corresponding recruitments.
Fmed is defined as the F value corresponding to the median B/R ratio in the longterm
B/R relation against F.
Usually, Fmed is illustrated by considering the graph of the points corresponding to pairs
of values of parental biomass (total or spawning),B, during that year and the resulting
recruitment , R. Figure 5.6 shows this situation.
R
resultant

m edian

.
.
.
.

. line
.

.
.

0
0

B
parental

Figure 5.6 Illustration of a median line

The marked line is a line passing through the origin, which separates the total number of
points in equal parts, that is, 50 percent of the points are in the upper part and 50 percent
are in the lower part of the line. This line is designated as the median line, or 50 percent
line, which can be explained as follows: in 50 percent of the years of the considered
period the values of R were smaller than the values of R which were estimated by the
median line (or, in 50 percent of the years of the referred period the values of R were
bigger than the values of R estimated by the median line).
As seen in Section 4.5 the slope (B/R) of each line marked in the graph, is associated
with a certain value of the fishing level, F. The value of F associated with the median line
is then, the median target point, Fmed.

59

It can be said that, given a certain level of parental biomass and knowing the B / R
corresponding to Fmed, then there is a 50 percent probability that the resulting recruitment
will be less than (or greater than) the value indicated by the median line.

2.

Calculation of Fmed
In order to determine the value of Fmed it is necessary to consider the longterm relation
between the resulting biomass per recruit and the fishing level, F, (Section 4.4, Figure 4D).
The determination of Fmed can be done mathematically or graphically.
To make the mathematical calculation of Fmed the ratioB/R has to be determined for each
pair of values (B, R). Those values have to be ordered and then the median
value,B/Rmedian can be calculated.
In the longterm relation of the Biomass per Recruit against F, the value of Fmed is the
value of F that corresponds to the median value previously obtained.
To make the graphical calculation, notice that in Figure 5.6 the slope of the median line
against axis R is equal toB/R. This value will be the basis for the calculation of value
Fmed in the graph ofB/R against F, in the longterm projection. Figure 5.7(A-B)
illustrates the calculation.
A)
B)
B/R

median line

.
.
slope

.
.

.
.

slope

Figure 5.7

Fmed

Illustration of the graphical calculation of Fmed

Other straight lines, corresponding to probabilities other than 50 percent, can be marked.
Figure 5.8 illustrates the graphical calculation of F10%. The line marked on Figure 5.8A,
separates the points, in such a way that 10 percent of them remain below the line (or
90 percent of the points stay above the line). So, this line has been designated as the
10 percent line.
Notice that the slope of the 10 percent line with axis R is larger than the slope of the
median line, as shown in Figure 5.9B. In this way, F10%, or Flow, is smaller than Fmed.

60

A)

B)
B/R

.
.
.

.
.
.

slope

slope

10% line

Figure 5.8

Illustration of the graphical calculation of F10%

A)

B)
B/R

median lines

F10%

.
.

slopes .

.
.

slope

10% line

.
.

10%

slope
50%

Figure 5.9

F10% Fmed

Graphical illustration of the slopes and corresponding Fs

for the median and 10 percent lines

Comments
1.

The target point Fmed intends to ensure an acceptable level of biomass based on the
empirical relation S-R.

2.

Other percentages can also be adopted, corresponding to straight lines which estimate
different probabilities of recruitments which are less than those indicated by the median
line. So, Fhigh would be the fishing level corresponding to the 90 percent line, for which
the recruitments of 90 percent of the observed years would be less than those estimated
by the line.

3.

F90%, as will be seen in the following Section, is also considered as a Limit Point (LRP).

4.

Notice that the slope of the 10 percent line with axis R is larger than the slope of the
median line and therefore, F10% (or Flow) is smaller than Fmed (Figure 5.9B).

5.

Fmed was used in management, in recent years, particularly with Iberic sardines.

6.

The biomass used can be the total biomass,B, but is frequently the spawning biomass,
SP.

7.

If the median line does not pass through a marked point in the scatter plot (which
happens when there is an even number of points) then one can use any straight line
passing between the two central points, for instance, the mid-line. In any case, Fmed is
always an approximate value.

61

5.2.4 FMSY
Definition
FMSY is defined as being the value of F which produces the maximum yield in the longterm. It
is necessary to select an S-R relation to estimate FMSY. This point is different from Fmax.

5.3 BIOLOGICAL LIMIT REFERENCE POINTS

(Bloss, MBAL, Fcrash and Floss)
There are several proposals for Flim and Blim. For each stock, the adopted values of Flim and
Blim depend on the characteristics of the stock and on its exploitation. What is important is
that the adopted LRP be a value that allows an exploitation level which avoids dangerous
situations of stock renewal.
Some of these points are derived from the observed values of Biomass and of Resulting
Recruitment. Some examples of this type are Bloss and MBAL. These LRP are also usually
classified by some authors as non-parametric, because their determination does not depend on
any particular model of the S-R relation.
Another category of LRP points, classified as parametric, is derived from S-R models.
Fcrash will be mentioned.
Let us also mention the category of LRP points involving observed values and values
obtained by the application of S-R models, like, for example, Floss.

5.3.1 Bloss
Bloss is the smallest spawning biomass observed in the series of annual values of the spawning
biomass (Lowest Observed Spawning Stock).

5.3.2 MBAL
More satisfactory is the LRP designated as Minimum Biological Acceptable Level, MBAL.
In fact, this LRP is a spawning biomass level below which, observed spawning biomasses
over a period of years, are considered unsatisfactory and the associated recruitments are
smaller than the mean or median recruitment.

5.3.3 Fcrash
The name itself shows that it is a limit that corresponds to a very high value of F, showing a
great probability of collapse of the fishery.
Fcrash is the fishing level F which will produce a longterm spawning biomass per recruit
(S/R) equal to the inverse of the instantaneous rate of variation of R with the biomass, at the
initial point (S = 0, R = 0). With the S-R models of Section 4.5 that value is the parameter 1/
of the models.
In order to make the graphical determination of this LRP one can start by obtaining the slope
of the angle that the tangent to the S-R curve makes with the R axis at the origin. Afterwards,
and starting from the relationB/R against F, the value of F that corresponds to the value
62

B R indicated by that slope is obtained. Fcrash will then be the value of F corresponding to
B R equal to that slope, in the long-term relation B / R against F.

5.3.4 FLOSS
Floss is usually defined as the fishing level F which will produce a longterm spawning
biomass per recruit (S/R) associated to Bloss .
To determine this limit point, first obtain the value of R corresponding to Bloss on the adjusted
curve S-R. Then, calculate Bloss/R and find the value of F, in the longterm relation B/R
against F.
Most of the Limit Points shown have been criticised for depending on the observed values or
on the adjustment of the S-R relation.

5.4

PRECAUTIONARY REFERENCE POINTS - Fpa, Bpa

As previously mentioned, the Precautionary Principle recommends that the assessments

should be done even when the basic data presents some gaps. This recommendation implies
that, in this case, the determination of the Biological Reference Points will not be very
precise. The uncertainities of the estimates should be calculated, and it is necessary to
mention the assumptions and models which have been used.
One suggestion to determine Fpa and Bpa might be to estimate Flim or Blim and from these
values, to apply the following empirical rules :
Fpa = Flim.e-1.645.

and

Bpa = Blim.e+1.645.

The constant is one measure associated with the uncertainity in the estimation of the fishing
mortality level, F. The values obtained in several fisheries indicate that values of are within
the interval (0.2, 0.3) (ICES, 1997). In practice, it can be said that Fpa is between 0.47Flim and
0.61Flim ,and Bpa is between 1.39Blim and 1.64Blim.
It is important to make clear that the target points may also, in certain cases, be considered as
limit or precautionary points depending on the combined analyses of the exploitation of the
stock and of the biological reference points obtained.

63

5.5

FISHERIES REGULATION MEASURES

The regulation measures aim to control the fishing level and the exploitation pattern applied
to the stock for an adequate exploitation.
The most common regulation measures to control fishing levels are:

Limitation of the number of fishing licences.

Limitation of the total fishing effort each year (limiting fishing days, number of trips,
number of days at sea, etc.).
Limitation of Total Allowable Catch (TAC)

TAC is a measure that directly controls the catch and, indirectly, the fishing level. It is
convenient to combine the TAC with the allocation of quotas of this total TAC for each
component of the fleet. In this way, the competition between vessels to fish the maximum
possible catch, as quickly as possible, before the TAC is reached, can be avoided.
The system of quotas allocated to each vessel is called Individual Quotas (IQ).
The regulation measures to correct the exploitation pattern are usually called technical
measures. Some of these measures are :

The minimum size (or weight) of the landed individuals.

The minimum mesh size of the fishing nets.

The prohibition of fishing in spawning.

The closed areas and periods for the protection of juveniles.

The fishing management have the duty to promote legislation and the application of the
regulation measures. (In the particular case of the EU, and for the stocks of the Economic
Exclusive Zones (EEZs) of the member states, the Commission decides on the measures to
be taken). In any case, management needs the analyses on the state of the stock and its
exploitation and on the effects of the recommended measures. That study must be done by the
fisheries scientists of each country or region and their Fishing Research Institutes, who will
have to calculate the projections of the stocks. The International Council for the Exploitation
of the Sea (ICES) analyses the assessments and recommends regulation measures and the
expected effects of the application of those measures to the Commission, as well as the
consequences of their non-application.
The shortterm projections, as well as the regulation measures, only make sense if the
longterm objectives of fisheries management are previously analysed and defined.
Shortterm projections of the stock and of its fishing must also be made by the scientists.

Comments
1.

Management needs to define the fishing objectives, based on the longterm projections.
Those objectives are valid for a period of years, even if they can be adjusted every year.
64

2.

The regulation measures, on the contrary, have to be established every year, although
some of them may be valid for more than one year. Some technical measures, like the
minimum mesh sizes of the fishing nets or the minimum size of the landed individuals,
are valid for several years.

3.

All the measures have advantages, difficulties and disadvantages regarding the purposes
they intend to reach.
The concession of fishing licences is a common practice almost everywhere, with a
limited total number.
TACs and quotas, because they control the catches, have caused misleading declarations
about catches.
The direct limitation of the total fishing effort (f), is based on the assumption that the
measure causes a similar limitation on the fishing mortality coefficient (F). However, this
relation may not be proportional. In the first place, it is difficult to measure the fishing
effort of the different fishing gears and of all the involved fleets and it is also difficult to
express it in units that respect the proportionality between F and f. Secondly, the
capturability of several gears may increase (and consequently increase F) without
increasing the fishing effort. Finally, the expected proportionality between F and f may
not be true. In any case, what matters, is not to forget that there is a relation between F
and f.

4.

The protection of the juveniles should be carried out during the whole year and will
preferably control the fishing mortality throughout the year. The occasional measures,
like areas and periods when fishing is forbidden for the protection of juveniles, require
annual investigations in order to discover whether there are exclusive concentrations of
juveniles, to assess the effects of that occasional interdiction, and to find out the
consequences of the interdiction on other species, etc. The minimum size of the landed
individuals, does not mean that smaller individuals are not caught, but only that they are
not landed. The difference between the catch and the landing is the so-called rejections to
the sea. It is clear that if the individuals are caught and rejected to the sea, the fishing
mortality is larger than the one suggested by the landings. The minimum landing size of
the individuals may have the effect of dissuading fishermen from catching small
individuals. Currently, some countries are forcing the landing of all fish caught.

5.

The closed spawning areas and seasons, to save the spawning biomass and indirectly
protect recruitment, is far from effective in the latter objective. In fact, large spawning
biomasses correspond to a large number of eggs, but that does not necessarily imply
bigger recruitments, as seen in Section 4.5. It is also not always true that forbidding
fishing during the spawning, and not forbidding before (or after) the spawning, protects
the spawning biomass. The only way to protect the spawning biomass will be to control
fishing level during the whole year. Finally, it has to be said that, in any case, the
interdiction of fishing in the spawning area and period, or on any other occasion, always
represents a reduction of the fishing effort. This is not a major inconvenience and in
some cases, may even be beneficial.

6.

It has to be stressed that no regulation measure will accomplish its objectives without
observing two conditions :
The understanding of the fishermen (broadly speaking) that the measure is good for the
fishery. Hence, it is important to discuss the scientists' conclusions, their objectives, their
reasons and the expected effects.
65

An efficient fiscalization in the ports and at sea! The 200-mile exclusive zone may be
very vast and the fiscalization very expensive, but it is not necessary to fiscalize the
whole area intensively. It is enough to control the areas of larger catches more intensively
and the remaining areas less so.
During the last few years, new ways of controlling access to fisheries resources and
exploitation levels are being implemented. Some examples are the establishment of
Individual Transferable Quotas systems (ITQ), co-management systems or even the
system of regional or municipal management, where some management responsibilities
are attributed to the resource users themselves.
7.

The ITQ management system is based on the abusive assumption that only the
economically efficient and profitable vessels deserve to be active in fishing. So, TAC's
are divided into individual quotas, to be auctioned for the best offer.
The co-management system delegates a great part of the responsibility of management to
those who directly exploit the fishing resources - managers, fishermen and their
professional organizations or unions. With this system neither are quotas sold in auctions,
nor are the fishing licences lost.
These are the most well known systems.
The ITQ system presents the following inconveniences : permanent loss of the titles of
quotas and of fishing licences; concentration of quotas in the hands of a small group of
people (who may not even belong to the fishing sector or are even foreigners); and
underestimation of the social, human and cultural aspects, in favour of economic
efficiency criteria.
On the contrary, the co-management system, is concerned with the social aspects of the
people involved; it seeks their direct and conscientious co-participation with government
authorities in the management responsibilities, including fiscalization.

66