UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Age and growth of the mosshead sculpin Clinocottus globiceps Girard with an assessment of its role in… Mgaya, Yunus Daud 1989

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1989_A6_7 M47.pdf [ 5.87MB ]
JSON: 831-1.0097548.json
JSON-LD: 831-1.0097548-ld.json
RDF/XML (Pretty): 831-1.0097548-rdf.xml
RDF/JSON: 831-1.0097548-rdf.json
Turtle: 831-1.0097548-turtle.txt
N-Triples: 831-1.0097548-rdf-ntriples.txt
Original Record: 831-1.0097548-source.json
Full Text

Full Text

A G E A N D G R O W T H O F T H E M O S S H E A D S C U L P I N CLINOCOTTUS GLOBICEPS G I R A R D W I T H A N A S S E S S M E N T O F ITS R O L E IN P R O D U C T I O N O F T I D E P O O L F I S H E S by YUNUS DAUD MGAYA B.Sc. (Hons.), The University of Dar es Salaam, 1986 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE i n THE FACULTY OF GRADUATE STUDIES Department of Zoology We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September 1989 © YUNUS DAUD MGAYA, 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada Date DE-6 (2/88) A B S T R A C T Age and growth of Clinocottus globiceps G i rard were investigated with the aid of otoliths taken from fish obtained from tidepools at Helby Island, Brit ish Columbia during the period M a y 1988 to June 1989. Length-frequency analysis was used to verify otolithic ageing. Data from fish collected between 1980 and 1987 were also included in this study. Growth rate was estimated using a back-calculation formula and a Gompertz growth curve. Instantaneous growth rates were determined. Results indicated that the C. globiceps population was composed of individuals from less than one year of age to 5 years of age. Growth was faster for younger age groups (about 17 m m per year) and declined at older age groups (about 9 m m per year). Growth was described by a Gompertz model and the following equation was obtained: L t = 26.7mm * exp{1.58(l - exp[-0.30t])}. Instantaneous growth rates were highest for the 0 + age class. The highest levels of instantaneous growth rates occurred during the spring and early summer, the period that water temperatures reach a max imum and food is most abundant. The lowest instantaneous growth rates occurred during the fall and winter months. The age —length relationship for C. globiceps is presented. The length—weight relationship of the species is described. No differences in growth between sexes ii as revealed by length — weight relationship were observed, thus the following expression described the length — weight relationship for C. globiceps population at Helby Island: W = 1.5913 * 1 0 ~ 5 L 3 - 1 5 5 2 Overall tidepool production with regards to C. globiceps was assessed by direct comparison with production of sympatric Oligocottus maculosus Girard, an abundant tidepool cottid. Production was estimated by both the instantaneous growth rate and size-frequency methods. Annual production as computed by the instantaneous growth rate method was 6.9 and 11.0 g/m2/year for C. globiceps and 0. maculosus respectively. Young- age groups (between 1+ and 2 + ) contributed 33 and 65% of production for C. globiceps and O. maculosus respectively. Estimates made by the size-frequency method were higher and it was suggested that these estimates may be more accurate, since the method is not affected by nonsynchronous cohort development. Production was analyzed by zones on the intertidal area and the results reflected the distribution pattern of the two species, i.e., higher production was observed at the upper intertidal pools for 0. maculosus and at the lower pools for C. globiceps. The relationships between the physical characteristics of the tidepools and production of the two species are given. None of the physical variables examined were significant predictors of production. iii T A B L E O F C O N T E N T S A B S T R A C T i i T A B L E O F C O N T E N T S iv L I S T O F T A B L E S v i L I S T O F F I G U R E S v i i i L I S T O F A P P E N D I C E S x A C K N O W L E D G E M E N T S x i I. G E N E R A L I N T R O D U C T I O N 1 1.1 B A C K G R O U N D 1 1.2 D E S C R I P T I O N O F T H E S T U D Y S I T E 5 1.2.1 P h y s i c a l C h a r a c t e r i s t i c s 5 1.2.2 F l o r a a n d F a u n a 7 II. A G E A N D G R O W T H 9 2.1 I N T R O D U C T I O N 9 2.2 M A T E R I A L S A N D M E T H O D S 11 2.2.1 S a m p l i n g 11 2.2.2 L e n g t h M e a s u r e m e n t s 12 2.2.3 W e i g h t M e a s u r e m e n t s 12 2.2.4 S e x D e t e r m i n a t i o n 13 2.2.5 A g e D e t e r m i n a t i o n 13 A g e D e t e r m i n a t i o n u s i n g Oto l i ths 15 A n a l y s i s o f L e n g t h - F r e q u e n c y D a t a 16 A g e - L e n g t h R e l a t i o n s h i p 17 2.2.6 G r o w t h D e t e r m i n a t i o n 17 M e t h o d of B a c k - C a l c u l a t i o n of F i s h L e n g t h 18 Instantaneous G r o w t h Rates 19 T h e o r e t i c a l G r o w t h M o d e l 20 2.3 R E S U L T S 21 2.3.1 S i z e C o m p o s i t i o n 21 2.3.2 Oto l i th D e s c r i p t i o n 21 2.3.3 Oto l i th A n a l y s i s 24 2.3.4 L e n g t h - F r e q u e n c y A n a l y s i s 28 2.3.5 T i m e of A n n u l u s F o r m a t i o n 28 2.3.6 G r o w t h i n L e n g t h 33 B a c k - C a l c u l a t e d L e n g t h s 33 T h e o r e t i c a l G r o w t h C u r v e s 40 L e n g t h - W e i g h t R e l a t i o n s h i p 45 G r o w t h Ra tes 49 2.3.7 A g e - L e n g t h R e l a t i o n s h i p 49 2.3.8 T o t a l L e n g t h - S t a n d a r d L e n g t h R e l a t i o n s h i p 52 2.3.9 S e x Rat io 52 2.4 D I S C U S S I O N 56 iv III. F I S H P R O D U C T I O N 64 3.1 I N T R O D U C T I O N 64 3.2 M A T E R I A L S A N D M E T H O D S 66 3.2.1 C o l l e c t i o n of F i s h e s 66 3.2.2 A b u n d a n c e E s t i m a t e s a n d D e t e r m i n a t i o n o f A g e 66 3.2.3 P r o d u c t i o n 67 Instantaneous G r o w t h Ra te M e t h o d 68 S i z e - F r e q u e n c y M e t h o d 70 3.3 R E S U L T S ,. 73 3.3.1 D i s t r i b u t i o n w i t h i n the S tudy A r e a 73 3.3.2 A g e of Oligocottus maculosus 76 3.3.3 D e n s i t y 76 3.3.4 B i o m a s s 84 3.3.5 G r o w t h 85 3.3.6 P r o d u c t i o n 86 3.3.7 S i z e - F r e q u e n c y E s t i m a t e of P r o d u c t i o n 90 3.3.8 A n n u a l P r o d u c t i o n to B i o m a s s Ra t ios 91 3.3.9 E s t i m a t i o n of P r o d u c t i o n f r o m Ini t ia l B i o m a s s 93 3.4 D I S C U S S I O N 94 IV . G E N E R A L D I S C U S S I O N A N D C O N C L U S I O N S 102 L I T E R A T U R E C I T E D 106 A P P E N D I C E S 120 v LIST OF TABLES Table 1. S u m m a r y statistics for standard lengths and body weight for the Clinocottus globiceps population at Helby Island 23 Table 2. Sample size and mean length (+SE) males, females, and combined sexes of C. globiceps of each age group 27 Table 3. Parameters of the component normal distrbutions identified in the population length distributions by the method of MacDonald & Pitcher (1979) 30 Table 4. S u m m a r y of standard length—otolith radius regression statistics (y = a+bx) and analysis of covariance between male and female C . globiceps 35 Table 5. Back-calculated lengths at each age for combined data of C. globiceps 36 Table 6. Back-calculated lengths at each a'ge for male C. globiceps 37 Table 7. Back-calculated lengths at each age for female C. globiceps 38 Table 8. Parameter estimates, derived from the Gompertz growth model, for C. globiceps 43 Table 9. Analys is of variance statistics for Gompertz model 43 Table 10. Summary of mean observed, back-calculated and theoretical growth in length from the present study for male and female C. globiceps 44 Table 11. S u m m a r y of mean observed, back-calculated and theoretical growth in length for combined sexes 44 Table 12. Length — weight relationship of male, female and combined sample of C. globiceps 47 Table 13. Functional length — weight regression described as L o g W = Log a + (b/r) Log L for C. globiceps 47 Table 14. Analys is of covariance between sexes for length — weight regression and t-test for comparison of calculated regression coefficient with regression coefficient of the cube law 48 Table 15. Instantaneous growth rates of C . globiceps in their age 0 + through age 5 + . Rates were calculated for the intervals between the 1987 and 1988 collections 50 vi Table 16. Summary of age—length regression statistics (y = a+bx) and analysis of covariance between male and female C. globiceps 53 Table 17. Sex ratios of C. globiceps by age groups 53 Table 18. A g e - l e n g t h data from Chadwick's (1976) study and the present study showing the discrepancy in the former study 58 Table 19. Age structure of Oligocottus maculosus from the present study, as determined by age —length regression 77 Table 20. M e a n annual density, initial biomass, mean biomass, annual production and production to biomass ratios for C. globiceps and O. maculosus 79 Table 21. Analysis of variance statistics for growth, density, biomass and production by years and zones for C. globiceps 80 Table 22. Analysis of variance statistics for growth, density, biomass and production by years and zones for O. maculosus 80 Table 23. Summary of regression analysis variables (defined in text). Correlations of each set of variables were tested for significance 83 Table 24. Annua l production, mean biomass and P:B ratios as estimated by instantaneous growth rate method and size-frequency methods 92 vii LIST OF FIGURES Figure 1. M a p showing the general location of study site 6 Figure 2. Photograph of male and female Clinocottus globiceps 14 Figure 3. Length-frequency distribution of C. globiceps 22 Figure 4. (a) Illustration of the labyrinth, showing the position of the otolith chambers, (b) otolith of C. globiceps showing opaque and hyaline zones 25 Figure 5. Length-frequencies of male and female C. globiceps by estimated age groups 26 Figure 6. Length-frequency distribution of C. globiceps sampled between Apr i l 1987 and October 1988 29 Figure 7. Length-frequency distribution of age 0+ through 5+ fish derived from otolithic ageing 29 Figure 8. Length-frequency distributions of C. globiceps showing progression of modes . 31 Figure 9. Otolith marginal increment plotted by month for the three age groups : . . .32 Figure 10. Relationship of saccular otolith radius to fish standard length for C. globiceps 34 Figure 11. Gompertz growth function for C. globiceps 42 Figure 12. Length — weight relationship of C. globiceps 46 Figure 13. Instantaneous growth curves for age 0+ through 3 + C. globiceps .51 Figure 14. Age —length relationship of C. globiceps determined by otoliths 54 Figure 15. Relationship of total length to standard length and vice versa 55 Figure 16. Length-frequency distribution of C. globiceps from tidepools at different levels 74 Figure 17. Length-frequency distribution of O. maculosus from tidepools at different levels 75 Figure 18. Density of C. globiceps and O. maculosus by age group, in each zone for the period 1 9 8 6 - 8 7 to 1 9 8 7 - 8 8 81 viii Figure 19. Age-frequency distribution curves for populations of C. globiceps and O. maculosus from Helby Island 82 Figure 20. Production of C . globiceps and O. maculosus by age group, in each zone for the period 1 9 8 6 - 8 7 to 1 9 8 7 - 8 8 88 Figure 21. Relationship between production and initial biomass of 1987 and 1988 year classes of C. globiceps and O. maculosus 89 Figure 22. Relationship between production and density of 1987 and 1988 year classes of C . globiceps and O. maculosus 97 Figure 23. Relationship between P:B ratio and mean biomass (B) of C. globiceps 100 ix LIST OF APPENDICES Appendix 1. Length-frequency distribution table for combined data of C . globiceps 120 Appendix 2. Length-frequency distribution table for female C. globiceps 121 Appendix 3. Length-frequency distribution table for male sample of C. globiceps 122 Appendix 4. Age—length key for Helby Island C . globiceps, showing the number of fish per age-class by sex 123 Appendix 5. M e a n annual density, initial biomass, mean biomass, mean instantaneous growth rates and annual production of C. globiceps and O. maculosus by age groups as computed by the instantaneous growth rate method 124 x ACKNOWLEDGEMENTS This thesis would not have come about without the help from many people. Special thanks are due to m y supervisor, Dr . Norman J . Wil imovsky, for encouraging me to work on tidepool fishes and for making his data and computer programs available to me. H is support through every phase of this study is highly appreciated. I would like to thank m y supervisory committee, Drs. T imothy R. Parsons, Geoffrey G . E . Scudder and Antony R . E . Sinclair whose guidance has considerably improved this thesis. Thanks are also due to Dr . J a n Heggenes for his assistance in the field. M y sincere thanks are due to Dr . John E . Mclnerney, Director of Bamfield Marine Station of the Western Canadian Universities Biological Society, for extensive logistic support as well as his staff particularly Steve L a C a s s e . I gratefully acknowledge the financial support of a Canadian Commonwealth Scholarship for the entire duration of m y graduate studies in Canada. I am also grateful to Dr . N . J . Wil imovsky and the Bamfield Mar ine Station for their financial support during all field trips. The staff of the University of Bri t ish Columbia Computing Centre and Biosciences Data Centre, especially Jon Nightingale and Alistair Blachford provided essential technical support. Last ly , I would like to thank M s . Matseliso Morapeli for typing the manuscripts of this thesis. xi I. G E N E R A L I N T R O D U C T I O N 1.1 BACKGROUND Cottidae (commonly known as sculpins) is a large diverse family represented by some 50 species in the marine waters of B . C . (Hart 1973). Most members of the Fami ly Cottidae are demersal in habit and occupy diverse habitats ranging from the subtidal to the intertidal zone of the northern hemisphere (Clemens & Wilby 1961). The mosshead sculpin, Clinocottus globiceps G i rard 1857 is an abundant intertidal cottid throughout much of its range (Green 1971b). It is found inhabiting rocky tidepools on the Pacific coast of North Amer ica from the central California coast to the Gulf of A laska (Miller & L e a 1972). Its max imum size is reported to be about 190 m m in total length (Hart 1973). Despite this wide distribution few ecological data have been published outside of manuscript reports with regards to this species. Several investigations have been conducted concerning age, growth, and food habits of a few marine species of Cottidae. Mitchell (1953) analyzed the stomach contents of three species of tidepool sculpins in California. N a k a m u r a (1971) studied food habits of two cohabiting tidepool Cottidae, Oligocottus maculosus Girard and 0. snyderi Greeley, and Yosh iyama (1980) investigated food habits of three intertidal sculpins in central California. 1 2 Mor ing (1979, 1981) analyzed age structure and dynamics of two very similar tidepool cottids, 0. maculosus and O. snyderi. N a k a m u r a (1976a, 1976b) investigated factors that influence vertical distribution of the two species. Other aspects of life history including reproduction and behaviour of tidepool cottids have also been reported. Morr is (1960, 1961) analyzed the distribution of three cottids, with respect to their sensitivity to temperature and salinity gradients. Green (1971a, 1971b, 1971c, 1973) studied movement, distribution, and homing behaviour of tidepool sculpins. Studies on reproductive biology include DeMart in i (1978) on spatial aspects of reproduction in Enophrys bison; DeMart in i & Patten (1979) on reproductive biology of Hemilepidotus hemilepidotus, and Grossman & deVlaming (1984) on female O. synderi. The only publications dealing specifically with Clinocottus globiceps are by Morr is (1960, 1961) who presented data on sensitivity to temperature and salinity gradients in relation to southern limits of this species; Clemens & Wilby (1961) and Miller & L e a (1972) summarized essentially taxonomy and distribution of this species. The homing behaviour, movement, and distribution patterns of the species were established by Green (1971b, 1973). Chadwick (1976a) estimated the age of C . globiceps and 0. maculosus but this has been questioned (Craik 1978; Mor ing 1979; Freeman et al. 1985). Resident tidepool fishes have their home range close to the tidepools (Gibson 1967; Green 1971a, 1971b; Richkus 1978; Grossman 1982) and some, notably O. maculosus and C. globiceps, exhibit strong homing behaviour (Green 1971a, 1973). Based on this information it was considered feasible to launch a study on 3 tidepool fish production dynamics. Production in the ecological sense, is considered a basic parameter in population dynamics and the ecology of ecosystems because it links population density, growth and survival rates (Le Cren 1972). The objectives of this study were two-fold. First , since little was known about the life history of Clinocottus globiceps, investigation on the age and growth of this species was conducted. The second objective was to evaluate production dynamics of C. globiceps in comparison with the widely distributed Oligocottus maculosus. Information on production was combined with data on intertidal distributions of these species to test a hypothesis that rate of fish production would be similar in physically similar tidepools. Two species, C. globiceps and O. maculosus, were chosen for production study. Oligocottus maculosus is a common intertidal fish ranging from northern California to the Bering Sea (Miller & L e a 1972) and is found in great abundance in tidepools at the rocky shore of Helby Island at the west coast of Vancouver Island. Production estimates, by combining age, growth, mortality, density, and biomass statistics, provide the most reasonable means of assessing the performance of a species in its environment (Le Cren 1969). Most fish production studies have been conducted in freshwater ecosystems and the literature in this area is enormous. Although not reviewed extensively, relevant aspects of the methodology used in such studies have been adopted in the present study. Although there is considerable understanding of the structure of marine fish communities, there is very scanty documentation for production estimates 4 completed on individual species or entire fish taxocene (assemblage) in a tidepool environment. For example, Bennett (1984) reported on an energy budget and production of Clinus superciliosus, a tidepool resident of the southwestern coast of South Afr ica . Bennett & Griffiths (1984) gave quantitative estimates (biomass) of an entire fish taxocene in the intertidal rocky pools on the Cape Peninsula, South Afr ica . Apar t from these studies, virtually nothing has been published for the Pacific coast of North Amer ica. In this study, aspects of the biology of C . globiceps including age and growth were examined and formed a basis for the investigation of production dynamics. O. maculosus was included in the production study for comparison. This thesis is composed of two major sections. The first section deals with aspects of age and growth of C. globiceps. In the second section, an attempt is made to estimate annual production of C . globiceps and O. maculosus. Relationship between production and various population attributes (e.g. biomass, density, and growth) and physical parameters (e.g., depth and surface area) of tidepools are analyzed and discussed. F inal ly , a general discussion bringing together the two sections is presented. 5 1.2 DESCRIPTION OF THE STUDY SITE 1.2.1 Physical Characteristics The study site for this investigation was a series of rocky tidepools at Helby Island near the Bamfield Marine Station on the west coast of Vancouver Island. Helby Island is situated in the Deer group of Islands in Barkley Sound, Brit ish Columbia. The site selected for study is a rocky shelf stretching from the northeastern to northwestern beach ( 4 8 ° 5 1 ' N , 1 2 5 ° 10' W ; Figure 1). The area is typical of open to semi-protected rocky intertidal habitats. The substrate consists mainly of boulders and cobblestones although several pools have patches of sandy bottom. A n important advantage of this site is its isolation from human habitation, comparative inaccessibility, and a long series of fish collections. The intertidal zone at the study site was extensively surveyed. A total of 20 bench marks previously established were used to determine the level of tidepools. The height of the bench mark relative to the zero tide level was determined from tide information corrected for barometric pressure. A l l tidepools used in the study were given identifying symbols. The vertical height of each tidepool relative to zero tide level was obtained from levelling data. These data were recorded along with notes describing the general physical features, max imum depth, and surface dimensions (perimeter, width and length) of each tidepool. The study pools fall into four vertical zones along the shore: upper, middle, Figure 1. M a p showing the general location of study site ( a d a p t e d f rom C r a i k 1 9 7 8 ) . 7 lower and base. The terms "base" and "lower" refer to positions from —0.3 to 0.6 m and from 0.6 to 1.9 m respectively, while "middle" and "upper" refer to positions from 1.9 to 2.3 m and from 2.3 to 3.5 m respectively. These tidal elevations are of interest because they represent different degrees of emergence and submergence. The tide is mixed semidiurnal with max imum high tides about 3.96 m and extreme low tides about 0 m. Annua l seawater temperatures range from about 6.7 to 1 2 . 0 ° C . 1.2.2 Flora and Fauna Predominant organisms in the study area were typical rocky shore forms. Moving up the intertidal, Fucus sp. dominates. Amongst the Fucus, Cladophora sp. and Leathesia sp. are present. The predominant algae in the pools is Prionitis spp., corallines, and Fucus spp.; Phyllospadix scouleri also occurs. The lower intertidal is characterized by large beds of the kelp Laminaria sp. and Hedophyllum sp. Invertebrates that occur in the pools include mussels Mytilus californianus and M. edulis; sea anemones Anthopleura xanthogrammica and A. elegantissima; limpets Notoacmaea persona, N. scutum, and Collisella digitalis; gastropod Tegula funebralis; grapsid crab Hemigrapsus nudus; hermit crabs Pagurus spp., and chitons Mopalia spp. In pools located at the upper and middle levels, barnacles Balanus glandula and B. cariosus and littorines are common. Commonly found in tidepools are about 26 species of fish of which about 14 species belong to the family Cottidae. The remaining species are primarily from 8 the families Stichaeidae and Pholidae, with few representatives from the families Liparidae, Gobiesocidae, and Hexagrammidae (Dr. N . J . Wil imovsky, personal communication). The most abundant fish in the upper tidepools is Oligocottus maculosus whereas Clinocottus globiceps was regularly caught and occurred in high numbers in the middle and lower tidepools. II. A G E A N D G R O W T H 2.1 INTRODUCTION Age structure and growth are essential features in the study of fish populations. A s a result, a multitude of literature has been built up on the subject but with a strong bias towards commercial harvestable species. Despite the wide distribution of cottid fishes, information on the age and growth of these fishes on the Pacific coast of North Amer ica is limited. Studies have been conducted on a few species, notably Oligocottus maculosus (Atkinson 1939; Craik 1978; Mor ing 1979), O. synderi (Moring 1981; Freeman et al. 1985), Clinocottus analis (Wells 1986) and C . globiceps Chadwick (1976a). The latter study is questionable due to small sample sizes and lack of confirmation (or validation) of the ageing work. In all investigations of fish populations, a knowledge of age and growth is of great importance. Bagenal & Tesch (1978) stated that "age data, in conjunction with length and weight measurements, can give information on stock composition, age at maturity, lifespan, mortality, growth and production". Age determination in fish has generally been based on the presence of growth rings which appear on scales and other hard structures of the body. Historically, otoliths (calcium concretions deposited in the membraneous labyrinth of the inner ear, Lowenstein 1971) have been utilized to age fish reliably for the following reasons: they are not susceptible to resorption (Mugiya & Watabe 1977); they undergo no chemical alteration once formed (Campana 1983a); and they are available in species which 9 10 lack or have small scales (Six & Horton 1977). O f the 25 different structures tested by Six & Horton (1977), otoliths were the superior structure for age determination. Chadwick (1976a) used vertebrae to assess the age structure of C . globiceps populations at two locations, one on southwestern Vancouver Island and another along northern California. However, Chadwick's study was plagued by relatively limited sample sizes (i.e., 41 at Port Renfrew, B . C . and 10 at Bruels Point, California) that were collected during a one-week period in Ju ly 1973. This study attempts to fill some of the gaps in knowledge concerning the life history of this common intertidal fish, the pr imary objective being to present data on age structure and growth of a C. globiceps population at Helby Island, B . C . 11 2.2 MATERIALS AND METHODS 2.2.1 Sampling Clinocottus globiceps were sampled from a set of 27 tidepools scattered throughout the intertidal zone. F ish specimens were sampled from the tidepools during low tides (usually within an hour of low tide) by using ichthyocides or bailing where possible. Specimens were then collected with dipnets. Sampling was terminated after a complete search of the pool failed to yield additional specimens. In pools with sparse cover or in pools where all rocks could be removed, all of the fish could be collected.' In rocky pools with many molluscs, barnacles and numerous crevices, undoubtedly a few of the fish avoided detection. Collected fish were immediately fixed in 10% buffered formalin and later washed in freshwater and transferred to 37.5% isopropanol for permanent storage in the laboratory. The study pools were sampled between February and October, 1988. Material collected over the past years from the same site by Dr . N . J . Wil imovsky were included in the currrent study. They comprised 1980 (n=85), 1981 (n = 64), 1983 (n = 225), 1986 (n=105) and 1987 (n=164). In addition, 306 specimens were collected in 1988. 12 2.2.2 L e n g t h M e a s u r e m e n t s Length and weight measurements formed the basis for estimating growth and production. Total and standard lengths were measured. Total length was measured as the distance from the anterior most tip to the end of the tail, when pressed to the position of maximum extension; standard length was measured as the distance from the head tip to the posterior end of the hypural plate (Hubbs & Lagler 1958). A l l measurements were made to the nearest 0.1 m m using dial calipers. Shrinkage of fish was considered to be negligible (Parker 1963) since it is less after preservation in alcohol than in formalin (Shetter 1936), and a minor bias when compared to human error in measurement (Balon 1974). Unless specified otherwise, standard length has been used in all analyses throughout this thesis. 2.2.3 W e i g h t M e a s u r e m e n t s Individual fish were weighed using a Sauter balance. Measurements were made to the nearest 0.05 g after blot drying the specimen. A logarithmic regression of weight against length was calculated using regression equation of the form b W = a L , where W is the weight (gm), L is the length (cm), b is the slope (regression coefficient) and a is a constant. The parameters a and b were calculated by using a computer program for nonlinear least squares parameter estimation (Pienaar & Thomson 1969). 13 The length — weight relationship was examined separately for males, females, and for combined sexes. Slopes (b values) of regression equations for males and females were compared by analysis of covariance to test for significant differences between sexes before pooling the data for a common regression equation. A l l statistical inferences were based on a significance level of a=0 .05 . 2.2.4 S e x D e t e r m i n a t i o n Sexes were easily distinguished in this species. Males are characterized by the presence of a genital papilla or penis (Figure 2). However, males smaller than 18.0 m m long (standard length) can sometimes be mistaken for females because of lack of development of the genital papilla. Therefore, sexing was not attempted in smaller specimens. Sex ratios were calculated for all collections of Clinocottus globiceps and tested for significant differences by chi-square analysis. 2.2.5 A g e D e t e r m i n a t i o n Because mosshead sculpins have no scales, their age was determined by a statistical approach based on length-frequency distributions (distribution mixture analysis, MacDonald & Pitcher 1979), and by an otolithic approach based on counting annuli in otoliths (Brothers 1987). The former method was used as a cross-validation technique of the latter method. 14 Figure 2. Photograph of male and female Clinocottus globiceps. 15 Age Determination using Otoliths Although each teleost has six otoliths, three on each side, the one ordinarily used is the sagitta laid down in the sacculus of the inner ear (see Figure 4). A calcareous concretion, the otolith is laid down in concentric layers, a process which is probably continuous. Less known factors, likely dependent on food or seasons, cause slight density variations which produce definite bands (annuli) in most species (Campana 1983b). After removal, otoliths were cleaned of attached tissue before examination. Otoliths from C. globiceps are thin, therefore whole-view examination was preferred to transverse sections. Readings were made with a dissecting. microscope ( X 3 0 ) equipped with an ocular micrometer. The most critical factor was illumination and it was found that transmitted light could not be used as it obliterated the peripheral circuli. In reflected light the periphery exhibited differential opacity, while the central region, though more opaque than when observed with transmitted light, could still be interpreted. Vary ing the illumination frequently improved the readability, although the best resolution was usually achieved by placing the source of light at an angle of 40 or 50 degrees above the otolith against a black background. Both saccular otoliths were examined while immersed in water and the one with greatest definition was used for annuli counting. Opaque bands were counted as annuli. The term "annulus" is used in the remainder of this thesis to refer to rhythmic growth increments or bands on 16 otoliths; it is assumed that the formation of these annuli coincide with annual events. Criteria were established after making a subjective study of 30 otoliths. A l l rings were broad and reasonably distinct on the anterior tip, but narrow and crowded on other parts of the otolith. Wi th the expanded tip as a starting point, it was possible to trace these rings, or at least count the same number at three or four different places on the otolith. Accordingly, all groups were designated as n + in age indicating that some growth beyond the most recent ring had occurred. Ages were determined from three complete series of readings. Disagreements between the readings were resolved by further examinations of both otoliths. To eliminate bias, otoliths were examined without prior reference to length or sex and in random order. The following measurements were taken from each otolith (in micrometer units and converting to millimeters) with an ocular micrometer (X30): 1. Otolith radius — the distance from the nucleus to the outside r im of the rostrum, and 2. Size of annulus — the distance from the nucleus to the outside edge of each growth band. Analysis of Length-Frequency Data A s a further verification of age groups in the population as determined by otoliths, the length-frequency distribution was analysed using the computer 17 program designed to fit normal distributions to polymodal data (MacDonald & Pitcher 1979; MacDonald 1980). The basic assumption of the program is that the length-frequencies are mixtures of normal distributions, for example of different cohorts of individuals. The resulting mixtures were interpreted from age—length data obtained from otoliths collected during the sampling period spanning from Apr i l 1987 to October 1988. Age—Length Relationship 'The age —length relationship allows for age determination of any given length of a species. Ordinary predictive regressions of age on length were calculated separately for sexes from data obtained from otolith readings. Analys is of covariance was performed to determine whether there were significant differences between the sexes, and if not, they were pooled into one age —length regression. 2.2.6 Growth Determination Growth in this study is defined as the change of size of a fish with age. Measurements of growth zones (spacing between annual marks) on otoliths were used in growth determination. Three assumptions inherent in growth studies using skeletal hard parts must be satisfied before a body part can be used for back-calculation of age and growth (Van Oosten 1929): 1. The size of the fish and size of hard part must be closely related throughout its entire lifespan. 18 2. The annuli used for age and growth analysis must be formed once yearly and at approximately the same time. 3. Estimates of length at a given age from different year classes of the population must agree. These assumptions were tested for C. globiceps otoliths in the following way: 1. Standard length was regressed on otolith radius to ascertain proportionality. Regression lines of the form y = a+bx were calculated separately for males and females and compared by analysis of covariance to determine whether they were statistically different. 2. Annul i were verified by recording monthly change in the distance between the otolith margin and the outer edge of the last complete hyaline zone (marginal increment). The term 'marginal increment' in this instance refers -to "the increase in otolith size (by marginal addition) consequent to the completion of the previous year's band" (Thorogood 1987). 3. Back-calculated lengths at each age were compared for various age classes. Method of Back-Calculation of Fish Length Having tested the above prerequisites, back-calculation here defined as the process of determining how large an individual fish was at some previous age (Smith 1983) was determined. The body length of the fish at any previous age in its life history was calculated with the following equation: L n = O R n (Lc - S) + S ORc where L n and L c refer to the lengths (in mm) at age n and capture 19 respectively, O R n and ORc refer to otolith radii (in mm) at age n and capture, and S is the y-intercept (in mm) of the regression of standard length on otolith radius (Ricker 1975; Bagenal & Tesch 1978). One of the pr imary uses of back-calculated lengths is the estimation of growth rates. Annua l length increments for individual Fish are the differences, between successive lengths-at-age. Instantaneous Growth Rates Monthly instantaneous growth rates were calculated for individual year classes. Regression slopes for the length—weight relationship were divided by their correlation coefficients to estimate slope for geometric mean functional regressions of weight versus length (Ricker 1973). Functional regressions are recommended for describing relations between weight and length because both variables are subject to natural variability (Ricker 1973). Functional regression slopes were used to derive y-intercepts from the equation y-int. = y - (b/r)x (Ricker 1973). These slope estimates were employed to calculate instantaneous rates of growth in weight (G) as: G = (b/r) (In L 2 - In L , ) (Ricker 1975), where r=correlation coefficient of the length — weight relationship; b ^ ordinary regression slope of weight on length; b/r = slope of the functional length — weight regression; L ^ m e a n length at time t; and L 2 =mean length at time t+1 . 20 Theoretical Growth Model Attempts were made to fit the length-at-age data to two growth models, von Bertalanffy and Gompertz. A growth model offers a generalized description of the pattern of growth. The von Bertalanffy model assumes that fish grow towards some theoretical asymptotic maximum size, and that the closer the length gets to the max imum the slower the rate of change of size. The Gompertz model also assumes that fish size tends to an asymptote, but the curve is markedly •asymmetrical with an inflexion point situated at a size well below half the asymptotic size (Ricker 1979). The two models differ in the sense that von Bertalanffy model describes only the portion of the curve beyond the inflexion point, i.e., the part having decreasing curvature. Gompertz growth curve describes data on both sides of the inflexion point, including the early years of increasing increments. Prel iminary results indicated that the data fitted well to the Gompertz model, and failed to fit the von Bertalanffy model. Following this attempt, it was considered logical to describe a theoretical growth curve using the Gompertz model which was written in terms of length: L t = L 0 * exp{G(l - exp[ —gt])} where L t is the length of the fish (mm) at age t in years, L 0 is a hypothetical length at t=0 , G is a growth parameter, and g a second growth parameter. The parameters of Gompertz growth equation ( L 0 , G , and g) were derived by a microcomputer program F I S H P A R M (Prager et al. 1987) which implements Marquardt 's algorithm for nonlinear least squares parameter estimation (Conway et al. 1970). 21 2.3 RESULTS 2.3.1 Size Composition A total of 949 specimens of C. globiceps were analysed; out of this total 30 specimens were too small to be sexed since external sex determination is impossible for juveniles less than 18 m m . There were 459 males and 460 females in the entire sample. The length-frequency distributions for males, females, and combined sexes of C. globiceps collected between 1980 and 1988 are illustrated in Figure 3 and Appendices 1, 2 and 3; the frequency distribution is unimodal. Size composition statistics are summarized in Table 1. Statistical differences in size were detected between males and females. Females were larger than males (t=3.77; df=917; p<0.001). The smaller fish (5.0 m m to 49.9 mm) contributed the greatest densities to the population as exemplified from the fact that 56.1% of C. globiceps were smaller than 49.9 m m (Figure 3). 2.3.2 Otolith Description A n otolith is composed primari ly of aragonite (calcium carbonate) and a high molecular weight protein, otolin (Degens et al. 1969). The sagitta, which is the largest of the three otoliths in C. globiceps, is elliptical in shape, and bluntly rounded at the anterior end. The lateral face is concave, while the medial face, traversed by a deep sulcus, is convex. The anterior point tends to be short in Figure 3. Length-frequency distribution of C. globieeps> 23 Table 1. Summary statistics for length and body weight for Q. globiceps population at Helby Island. GROUP Q U A N T I T Y R A N G E M E A N ± S . E . N Female SL (mm) 18.1-139.1 52 .7±0 .972 460 W T (gm) 0.20-98.50 6.70±0.372 Male SL(mm) 18.4-121.9 47 .5±0 .977 459 WT(gm) 0.13-60.50 5.42±0.347 Combined SL(mm) 8.5-139.1 49 .0±0 .702 949 WT(gm) 0.01-98.50 5.87±0.249 24 very young fish and long in the older specimens. Variations in the shape of otoliths was best reflected by the variability of the anterior point. The otoliths of older specimens often showed protrusions on the dorsal margin of this tip. 2.3.3 Otolith Analysis A regular pattern of alternating opaque and hyaline rings resulting from seasonal variations in the chemical components of the otolith can be seen (Figure 4). In many temperate fishes these rings represent seasonal periods of fast and slow growth (Panella 1974). The first summer growth zone appears as a broad opaque band while the first winter zone is usually a narrower hyaline band. The second and subsequent opaque zones are generally narrower than the first one. There was a wide range among otoliths in the clarity of their bands. A total of 480 pairs of saccular otoliths were examined out of which, 55 pairs or 11.5% were considered unreadable, occasionally because they were too opaque and/or marks were too broad and diffused to accurately determine their number and location. Occasionally, some otoliths were eliminated because readings differed by more than one ring. The proportion of unreadable otoliths was slightly higher in older fish. A total of 425 C . globiceps comprising 214 females and 211 males were aged. A total of five age groups were established, although 96.9% of the fish were less than estimated age 5 + (Figure 5). The 1+ to 3+ age groups were reasonably well represented, with the largest number of individuals in the 1 + 25 F i g u r e A . (a) I l l u s t r a t i o n o f the l a b y r i n t h , showing t h e p o s i t i o n o f the o t o l i t h chambers ( a f t e r B l a c k e r 1 9 7 4 ) , (b) o t o l i t h ( s a g i t t a ) o f C. globiceps showing opaque and h y a l i n e z o n e s . 26 20 10 0 20 10 0 >• 20 M 0 o- 20 L U 10 ^ 0 30 20 10 0 20 10 | MALES M=6 Q FEMALES F = 7 M=II F=16 Jim. 1 M=37 1 Ik ~ iJ k M=44 F=55 n 1 M=78 F=74 M=35 F=22 • 1 2 + -0 10 20 30 40 50 60 70 80 90 100 S T A N D A R D L E N G T H ( m m ) Figure 5. Length-frequencies of male and female C. globiceps by estimated age groups. ( F i s h c o l l e c t e d during 1987 and 1988) . T a b l e 2 . Sample s i z e and mean l e n g t h (+ SE) f o r m a l e s , f e m a l e s and combined s e x e s o f C . g l o b i c e p s o f e a c h age g r o u p . Age A g e - l e n g t h d a t a by sex  ( y r ) Sex n Mean l e n g t h + SE (mm) Sex compar ison  t d . f . P n S e x e s combined 0 + M 35 26.9+ 0 .728 F 22 27.9+ 0 .801 1 + M 78 40 .0+ 0.472 F 74 40.5+ 0 .496 2 + M 44 53.1+ 0 .716 F 55 55.1+ 0 .707 3 + M 37 70.4+ 0.654 F 40 69.6+ 0 .555 4 + M 11 80.7+ 0 .699 F 16 80.2+ 0 .677 5 + M 6 91.5+ 2 .848 F 7 90.5+ 1.235 Mean l e n g t h + SE L e n g t h r a n g e (mm) (mm) 0.84 55 0.41 57 2 7 . 3 + 0 .542 0 .71 150 0 .48 152 4 0 . 3 + 0 .341 1 .95 97 0 .06 99 54 .2 + 0 . 5 1 3 0 .88 75 0 .38 77 7 0 . 0 + 0 .426 0 .53 25 0 .60 27 8 0 . 4 + 0 .486 0 .33 11 0 .75 13 9 0 . 9 + 1 .411 1 8 . 4 - 32 .2 32 .2 - 4 7 . 2 4 7 . 0 - 6 6 . 0 6 1 . 6 - 7 6 . 6 7 7 . 0 - 8 5 . 9 8 6 . 1 - 105*1 n=sample s i z e ; t = t - s t a t i s t i c ; d . f . = d e g r e e s o f f r e e d o m ; P = p r o b a b i l i t y 28 and 2 + age groups. Also worth noting is the fact that age classes greatly overlapped in length (Figure 5, Table 2, and Appendix 4). F ish in which the first opaque zone had not yet been delineated were considered to be age 0 + fish. 2.3.4 L e n g t h - F r e q u e n c y A n a l y s i s Six modal classes were identified by the distribution mixture method (MacDonald & Pitcher 1979) (Figure 6 and Table 3) and from the age-at-length data, corresponded to 0 + , 1 + , 2 + , 3 + , 4 + , and 5+ year classes (Figure 7). Length-frequency histograms followed over time indicated that there are three major age groups and progressively smaller numbers of four and five year age group fish existing in the population (Figure 8). These histograms were not separable by the MacDonald—Pitcher analysis because of small sample sizes. Following the modes over the months shown reveals that recruitment began in spring (between Apr i l and May) and peaked in summer (around August). 2.3.5 T i m e o f A n n u l u s F o r m a t i o n The distance from last annulus to the edge of the otolith is an indicator of the time of annulus formation (Thorogood 1987). Capture date is related to the date of annulus formation by this distance. Hence, as this distance decreases, the more recently the annulus was formed. The delineation of the hyaline zone and the subsequent formation of the new opaque zone, evidence of the regular F i g . 6 N = 462 60 50 . 40 . 30 20 . o 10 -0 10 20 30 40 50 60 70 80 90 100 110 120 130 Standard Length (mm) Figure 6. Length-frequency distribution of C. globiceps sampled between Apr i l 1987 and October 1988-Figure 7. Length-frequency distribution of age 0 + through 5 + fish derived from otolithic ageing. Table 3. Parameters of MacDonald-Pitcher length-frequency analysis. p 0.0351 0.2984 0.3482 0.1964 0.0799 0.0420 S.D. 3.76 4.68 5.13 3.88 2.52 5.09 Mean 25.9 38.8 49.4 67.7 78.4 88.9 length (mm) Age 0+ 1 + 2+ 3+ 4+ 5+ p - r e l a t i v e abundance of mode as proportion of t o t a l sample S.D.=standard deviation of mode. 31 Apr i l 1987 October 1988 |—|_j—j N=38 10 20 30 40 50 60 70 80 90 100 110 120 130 140 Standard Length (mm) Figure 8. Length-frequency distributions of C. globiceps showing progression of modes. 32 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Time (Months) Age groups — — Age-1 + — H - Age-2+ — * - Age-3+ - s Combined Figure 9. Otolith marginal increment plotted by month for the three age groups. 33 formation of a single band per year was indicated by a sharp decline in the mean marginal increment in the fish sampled between Apr i l and M a y 1988 (Figure 9). A l l age-1, age-2 and age-3 fish had completed annulus formation by M a y . Rapid growth of the opaque zone, as represented by increases in the marginal increments of fish with 1, 2, and 3 opaque rings occurred throughout the summer and early fall months. Likewise, hyaline zones were formed once annually, predominantly during late winter and earl}' spring. Despite a lack of data for several months, an examination of seasonal annulus formation for individual age classes (1+ to 3 + ) showed little deviation from the pattern displayed by the entire stock (Figure 9). Based on these observations, I conclude that each hyaline—opaque sequence (incremental area) represented one year's growth and that incremental numbers increased in number with an increase in fish size. 2.3.6 G r o w t h i n L e n g t h Back-Calculated Lengths A strong positive linear relationship was found between body length of the fish and otolith radius for the entire size range of fish available (Figure 10 and Table 4). Analysis of covariance indicated no significant differences between male and female fish for standard length and otolith radius relationship (Table 4). 20 r 2 3 Otolith Radius (mm*30) Figure 10. Relationship of saccular otolith radius to fish standard length for C. globiceps. 35 Table 4. Summary of standard length-otolith radius regression statistics (y=a+bx) and analysis of covariance between male and female _C_. globiceps. GROUP N a b r 2 F P Female Male Combined 191 175 366 -22.68 -27.23 -24.97 28.51 29.81 29.17 0.85 0.91 0.88 1096.12 1723.53 2653.89 0.001 0.001 0.001 Analysis of covariance (Males vs Females). Source of variation D.F. SS MS F P Equality of 1 87.44 87.44 2.89 0.09 adj. means Error 363 11000.81 30.31 Equality of 1 40.36 40.36 1.33 0.25 slopes Error 362 10960.45 30.28 Table 5. Back-calculated lengths at each age for combined data of C. globiceps. Age a t c a p t u r e ( y r ) N o . o f f i s h Mean l e n g t h a t c a p t u r e (mm) Back 1 c a l c u l a t e d l e n g t h s (mm) 2 3 a t s u c c e s s i v e a n n u l i 4 5 0 + 57 2 7 . 3 1 + 152 4 0 . 3 32 .0 2 + 99 54.2 31.4 47 .4 3 + 77 70 .0 l 32.2 50 .2 6 3 . 3 4 + 27 8 0 . 4 32 .4 51 .7 6 6 . 3 76 .7 5 + 13 90 .9 33 .8 5 0 . 5 65 .9 78 .3 8 6 . 7 T o t a l 425 Grand mean Growth Increment 32.4 5 0 . 0 65 .2 17 .6 15 .2 1 2 . 77 .5 8 6 . 7 3 9 .2 CO OS Table 6. Back-calculated lengths at each age for male C . globiceps. Age a t No. o f Mean l e n g t h Back c a l c u l a t e d l e n g t h s (mm) a t s u c c e s s i v e a n n u l ! c a p t u r e ( y r ) f i s h a t c a p t u r e (mm) 1 2 3 4 5 0 + 35 26 .9 1 + 78 4 0 . 0 31 .9 2 + 44 53 .1 31 .1 4 6 . 7 3 + 37 70 .4 32 .3 50 .2 63 .2 4 + 11 80 .7 31 .9 5 1 . 6 66 .2 7 6 . 8 5 + 6 91 .5 3 4 . 3 5 1 . 8 66 .7 7 8 . 6 8 7 . 4 T o t a l 211 Grand mean Growth i n c r e m e n t 3 2 . 3 50 .1 65 .4 77 .7 8 7 . 4 1 7 . 8 1 5 . 3 1 2 . 3 9 . 7 Table 7. Back-calculated lengths at each age for female C. globiceps. Age a t c a p t u r e (y r ) N o . o f f i s h Mean l e n g t h a t c a p t u r e (mm) Back c a l c u l a t e d l e n g t h s (mm) 1 2 3 a t s u c c e s s i v e a n n u l i 4 5 0 + 22 27 .9 1 + 74 4 0 . 5 32 .0 2 + 55 55 .1 31 .6 4 8 . 1 3 + 40 69 .6 32.1 50 .1 6 3 . 4 4 + 16 80 .2 I 1 32 .8 5 1 . 7 66 .4 76 .7 5 + 7 9 0 . 5 33 .3 5 1 . 6 6 6 . 8 78 .4 8 6 . 0 T o t a l Grand mean 214 32 .4 5 0 . 4 6 5 . 5 7 7 . 6 8 6 . 0 Growth i n c r e m e n t 18 .0 15 .1 1 2 . 1 8 .4 39 Therefore, a common regression line was calculated and provided the equation S L = - 2 4 . 9 6 + 29.17 O R ( r 2 = 0 . 8 8 , p<0.01). This equation demonstrates that growth in these two parameters (SL and OR) is proportional. However, an increase in the variability between standard length (SL) and otolith radius (OR) was evident with larger fish (Figure 10). The above regression equation formed the basis of back-calculation. A total of 366 specimens (191 females and 175 males) were used for back-calculations. Back-calculated lengths from different age classes are in close agreement. Within 1 + , 2 + , and 3 + age groups, the females, on the average, tend to be slightly larger than the males for back-calculated lengths whereas males are slightly larger than females for age groups 4 + and 5 + . The same trend is reflected in observed lengths (Tables 5, 6 and 7). In the sample of C. globiceps studied, reversed Lee's phenomenon is demonstrated. Stated briefly, Rosa Lee's phenomenon refers to a situation whereby the older the fish whose otolith is used for the calculation, the lower the value obtained (Ricker 1969). Reversal of Lee's phenomenon occurs when back-calculated lengths from older fish are greater than those calculated from younger fish (Ricker 1979). Negative size-selective mortality is the most likely cause of this phenomenon in C. globiceps population. Growth appeared to be faster amongst the age groups 1+ and 2 + after which it decreased slightly, especially after age 3 + . Age 1+ individuals grew 17.6 m m in length by the end of their first year, whereas members of age group 2 + increased in length by 15.2 m m during the second year. Age 3 + fish grew 12.3 m m in length by the end of their third year (Table 5). M e a n back-calculated 40 lengths (Tables 6 and 7) illustrated that both sexes grew in length at approximately the same rate. Theoretical Growth Curves A plot of fish length against age, as determined from otoliths, showed a linear relationship which tended to level off at older ages (Figure 11). In the early part of growth, from age 0 + to 2 +, the increase in size by age was greatest. The rate of acceleration decreased gradually from age 3 + . The Gompertz growth curve resulted in a better fit to the length data than the von Bertalanffy growth curve. The reason for this being that the age data were more complete, i.e., very young age groups and older age groups were included. When these data were plotted, a slight sigmoidal pattern (a feature best described by Gompertz model) was observed. The von Bertalanffy growth model does not conform to the sigmoid length-growth resulting from an inflexion point early in the life of the fishes (Yamaguchi 1975). This limitation is known to occur in short-lived species, which show an inflexion point early in life (Moreau 1987). The von Bertalanffy model is usually applied to the final stanza of life which is taken to start when the fish has recruited to the fishery (in commercial stocks), and has been used extensively in fishery work because of its application in yield computations. A comparison of theoretical and empirical growth curves shows almost complete concordance (Figure 11) indicating that the Gompertz model adequately describes the growth pattern exhibited by this species at Helby Island. Growth parameters estimated from the Gompertz curve fitting are presented in Table 8. 41 The null hypothesis that the three Gompertz growth parameters are equal in males and females was examined using the 95% confidence limits attached to each parameter (Table 8). It can be seen that there is a complete overlap of the limits between sexes, suggesting that there is no significant difference in growth between males and females. The growth curve for combined data was therefore computed and the following Gompertz equation described the growth of C. globiceps: L t = 26.7mm * exp{1.58(l - exp[-0.30t])}. Goodness of fit was tested by an analysis of variance (Table 9) which indicated that the model accounted for a highly significant amount of variation. The fitted model is shown in Figure 11. M e a n observed, back-calculated, and theoretical growth were compared in Tables 10 and 11. B y estimated age 1+ male fish obtained a mean back-calculated length of 32.0 m m and age 5 + males averaged 87.4 m m . Estimated age 1+ females had a mean back-calculated length of 32.3 m m and those of age 5+ averaged 86.5 m m . F r o m Table 11 it can be seen that observed lengths, back-calculated, and theoretical lengths predicted from Gompertz model agreed closely. Differences between sexes are not apparent, therefore, comparisons were made on combined sexes. Predicted lengths from the Gompertz growth model and observed lengths were compared by using t-test for paired comparisons (Sokal & Rohlf 1981). The two lengths compared reasonably well (t=0.23, d f=5 , p = 0.824) adding further support to the accuracy of the model. 42 100 0 1 2 3 4 5 6 Age (years) Curves Gompertz — 1 — Empirical Figure 11. Gompertz growth function for C . globiceps. 43 Table 8. Parameter estimates derived from the Gompertz growth model, for C. globiceps. Group Parameter Estimate Asymp. S .E . 1 .96(S.E.) Lower 1 imit Upper 1 imit Male L0(mm) 26.5 0.6100 1 .1956 25.3 27.7 G 1.64 0.0648 0.1270 1 .51 1 .77 9 0.29 0.0258 0.0506 0.24 0.34 Female L0(mm) 27.1 0.6917 1 .3557 25.7 28.4 G 1 .53 0.0498 0.0976 1 .43 1 .63 9 0.31 0.0268 0.0524 0.26 0.37 Combined LQ(mm) 26.7 0.4569 0.8955 25.8 27.6 G 1 .58 0.0395 0.0773 1 .50 1 .66 9 0.30 0.0185 0.0363 0.27 0.34 Table 9. Analysis of variance for Gompertz model. Source of D.F. SS MS F P variation F E M A L E Model 3 659363.13 219787.71 12129.56 0.001 Error 211 3823.40 18.12 M A L E Model 3 579219.21 193073.07 9962.49 0.001 Error 208 4031.92 19.38 COMBINED SEXES Model 3 1238518.59 412839.53 21994.65 0.001 Error 422 7919.07 18.77 44 Table 10. Summary of mean observed, back-calculated and theoretical growth in length from the present study: Males and Females. Observed Back-calculated Predicted length from length (mm) length (mm) Gompertz model Estimated age Male Female Male Female Male Female 0+ 26.9 27.9 - - 26.5 27.1 1+ 40.0 40.5 32.0 32.3 40.0 " 40.9 2+ 53.1 55.1 49.8 50.1 54.5 55.2 3+ 70.4 69.6 65.2 65.0 68.6 68.8 4+ 80.7 80.2 77.7 77.4 81.6 80.8 5+ 91.5 90.5 87.4 86.5 92.8 90.9 Table 11. Summary of mean observed, back-calculated and theoretical growth in length from the present study: Combined sexes. Estimated Observed Back-calculated Predicted length from age length (mm) length (mm) Gompertz model 0+ 27.3 - 26.7 1+ 40.3 32.2 40.4 2+ 54.2 50.0 54.9 3+ 70.0 65.1 68.7 4+ 80.4 77.6 81.2 5+ 90.9 87.0 91.7 Length—Weight Relationship 45 The length—weight relationship of C. globiceps was calculated using a random sample of 460 females, 459 males, and a pooled sample of 944 fish. The results are plotted in Table 12 and Figure 12. Functional regressions of weight versus length were calculated and the slope estimates were used in the calculation of instantaneous growth rates. The statistics for these regressions are presented in Table 13. The significance of correlations was determined by t-tests. Highly significant correlations are found between length and weight of males, females, and pooled data (Table 12). The regression lines for males and females were compared to test whether the slopes (b values) and intercepts (a values) differed significantly between sexes by the method of analysis of covariance ( B M D P Statistical Software 1983). The slopes and intercepts of males and females are not significantly different from each other (Table 14). Since there was no significant difference between slopes and intercepts of both sexes, a common slope and intercept were calculated (Table 12). The null hypothesis that slopes are not significantly different from 3 (i.e., isometric growth) was tested by t-test which revealed that the b value differs significantly from 3, (i.e., allometric growth) indicating a deviation from the cube law (Table 14). 46 1 <D 1 O -2 COMBINED DflTR 1 2 Log Length (cm) Figure 12. Length-weight relationship of C . globiceps. 47 Table 12. Length-weight relationship in C. globiceps males, females and combined data. Regression equation Group N r 2 p Mali ! W=9.8698 * l f T 6 L 3 2 6 7 4 459 0.981 0.001 Log W=-5.0057+3.2674 Log L Female W=1.1363 * 10~5 I 3 2 3 2 3 460 0.978 0.001 Log W=-4.9445+3.2323 Log L Combined W=1.5913 * 10~5 L 3 1 5 5 2 944 0.975 0.001 Log W=-4.7982-f3.1552 Log L Table 13. Length-weight relationship described as Log W=log a'-f-(b/r)log L (functional regression). Group a' b/r N r Male 1.4058 * 10~5 3.3004 459 0.981 Female 1.3630 * i o - 5 3.2683 460 0.978 Combined 1.9901 * 10"5 3.1968 944 0.975 48 Table 14. Analysis of covariance between sexes for length-weight regression and t-test for comparison of calculated regression coefficient with regression coefficient of the cube law. Source of variation D.F. SS MS F P Equality of 1 36.94 36.94 2.77 0.096 adj. means Error 916 12231.80 13.35 EquaUty of 1 34.10 34.10 2.56 0.110 slopes Error 915 12197.70 13.33 t-test Group b-3 S.E. of b D.F. t P Male 0.2674 0.0236 457 11.34 0.001 Female 0.2323 0.0199 458 11.62 0.001 Combined 0.1552 0.0110 942 14.11 0.001 49 Growth Rates Instantaneous growth rates were calculated for individual age groups from otolithic ageing. Male and female samples were pooled so as to increase sample sizes for the calculations of instantaneous growth rates. The highest monthly instantaneous growth rates were obtained during June and August (Table 15 and Figure 13). The value for age 5 + during June and August is slightly higher than the preceding age 4 + . It is worth noting that there was only one specimen in age 5+ in June and this more than likely confounded the results for the age group in question. Growth rates markedly decreased between August and October. Generally, growth rates were higher for age 0 + and decreased progressively with age. 2.3.7 A g e - L e n g t h R e l a t i o n s h i p Data were analysed initially for difference in length between sexes of each age group (Table 16). It can be seen from analysis of covariance that there is no significant difference (p>0.05) between the mean length of each sex within each age group. Because there appears to be no apparent difference in the age —length data between the sexes these data were combined for regression analysis. The age—length regression calculated for combined sexes (Figure 14) was highly significant; regressions for males and females are also displayed (Table 16 and Figure 14). 50 Table 15. Instantaneous growth rates of C . globiceps in their age 0+ through age 5-)-. Rates were calculated for the intervals between the 1987 and 1988 collections. Instantaneous rate Collection dates 0+ Age 1 + 2+ 3+ 4+ 5+ May 1987 0.793 - 0.115 0.144 0.057 0.019 June 1988 0.712 0.324 0.229 0.166 0.112 0.131 August 1988 - 0.045 0.039 0 - 0.032 October 1988 51 80 i 60 ho D) C CO T J C CO co 20 ( n o d a t a ) 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Time (Months) Figure 13. Instantaneous growth curves for age 0 + through 3+ C . globiceps. 52 2.3.8 T o t a l L e n g t h - S t a n d a r d L e n g t h R e l a t i o n s h i p It is possible to convert data secured with total length to standard length and vice versa (Figure 15). The relationship between the two was defined by the equations: Total length to standard length S L = - 1 . 5 1 8 3 + 0.8327 T L ( r 2 =0.995, n = 941). Standard length to total length T L = 2.1267 + 1.1950 S L ( r 2 =0.995, n = 941). 2.3.9 S e x R a t i o The sex ratio for the entire sample (n = 919) comprising 459 males and 460 females was fairly balanced, having an overall ratio of 0.998:1 in favour of females. Chi-square test revealed that sex ratio did not differ significantly from 1:1 ( X 2 =0.001, p = 0.97). Analysis of sex ratio by age groups was carried out on a sample of 425 individuals consisting of 214 females and 211 males (Table 17). F r o m the table it can seen that sex ratios for age groups 0 + to 5+ and for combined age groups were not significantly different from unity (1:1). 53 Table 16. Summary of age-length regression statistics (y=a+bx) and analysis of covari-ance between male and female C. globiceps. Group N a b r 2 F P Female 214 -1.83 0.07 0.935 3022.23 0.001 Male . 211 -1.71 0.07 0.937 3103.84 0.001 Combined 425 -1.76 0.07 0.936 6176.79 0.001 Analysis of covariance (Males vs Females). Source of variation D.F. SS MS F P Equabty of 1 0.055 0.055 0.55 0.457 adj. means Error 422 42.27 0.10 Equality of 1 0.129 0.129 1.29 0.257 slopes Error 421 42.14 0.10 Table 17. Sex ratios of C. globiceps by age groups. Age group M F Sex ratio X2 P 0+ 35 22 1.60:1 2.965 0.085 1+ 78 74 1.05:1 0.105 0.746 2+ 44 55 0.80:1 1.222 0.269 3+ • 37 40 0.93:1 0.117 0.732 4+ 11 16 0.69:1 0.926 0.336 5+ 6 7 0.86:1 0.077 0.782 Combined 211 214 0.99:1 0.021 0.884 age groups 54 4 3 0 5 co 3 CD D) 2 < 0 5 4 3 2 1 0 Combined N=425 20 40 60 80 Standard Length (mm) 100 Figure 14. Age-length relationship of C. globiceps determined by otoliths-Total Length vs. Standard Length Figure 15. Relationship of total length to standard length and vice versa . 56 2.4 DISCUSSION Clear annual patterns were distinguishable in sagittae of C. globiceps (Figure 4) which made it possible to age these fish. The sagittae of teleost fish are the preferred hard structure for age determination (Bagenal 1974; Wil l iams & Bedford 1974; Panella 1980) and have been suggested to be the most accurate ageing structure (Six & Horton 1977). In the present study it was possible to estimate the age of 88.5% of the specimens with the annual zones clearly evident in whole otoliths. The otoliths of C . globiceps, like those in other teleosts, serve many functions such as hearing and equilibrium (Popper & Coombs 1980). Furthermore, otoliths record a large amount of information about an individual's condition over time and details of its life history. Likewise, otoliths from a collection of individuals provide insights into population dynamics. Hyal ine ring formation in fish occurs during a period of slow growth, generally associated with low water temperatures, reduced food availability, seasonal spawning, and altered photoperiod (e.g. Wil l iams & Bedford 1974), although the exact mechanism remains obscure (e.g. Panella 1974). For C . globiceps at Helby Island the onset of hyaline zone formation in October (Figure 9) coincides with the beginning of a seasonal drop in water temperature to about 1 0 . 1 ° C . The current study did not look into food intake and gonadal activity as they relate to hyaline zone formation, therefore no support is offered for these hypotheses. 57 Chadwick (1976a) speculated that the period from Apr i l to M a y m a y have been the time of annulus formation for C. globiceps at a location on the west coast of Vancouver Island. This study confirms the speculation that annulus formation occurred during the spring (April /May) (Figure 9). Observed rates of otolith marginal increment deposition indicated that growth was rapid from June to September, with a significant . decrease in the rate of growth over the remaining months with the lowest being Apr i l to M a y . The seasonality of increment deposition shown here is consistent with most other temperate fishes in the northern hemisphere (Grossman 1979; Maceina & Betsill 1987). The study of age determination in C. globiceps population of Helby Island, using otoliths and verified by length-frequency analysis, revealed three major age groups (1 + , 2 + , and 3 + ) and two smaller age groups (4+ and 5 + ). Whereas five age classes were observed in this study, Chadwick (1976a) observed as many as six age classes based on counts of vertebral rings, at Port Renfrew, Vancouver Island, B . C . Standard lengths attained by 0 + and 4+ age classes observed by Chadwick (1976a) closely paralleled those displayed in this study, but the lengths for age 1 + , 2+ and 3+ individuals were not comparable with those in the present study (Table 18). Since his analyses were based on a single collection of fish with very small sample sizes (10 and 41 fish at California and B . C . respectively), it is highly possible that age 0 + and 1+ in his study actually correspond to age 0 + in the present study. A major discrepancy between this study and Chadwick's ageing work emanates from the extremely small sample sizes he used. For example, as shown in Table 58 Table 18. Age-length data from Chadwick's (1976a) study and the present study, showing the discrepancy in the former study. Chadwick's study Present study Age groups N Standard N Standard length (mm) length (mm) 0+ 8 26.1 57 27.3 1 + 21 32.2 152 40.3 2+ 3 43.7 99 54.2 3+ 6 54.2 77 70.0 4+ 1 79.0 27 80.4 5+ 1 113.0 13 90.9 6+ 1 106.0 0 0 59 18, age groups 4 + , 5 + , and 6 + in his study are based on a single specimen for each age group. While the results in this study do suggest that C. globiceps lives up to at least 5+ years, some larger specimens in the collections, presumably much older than those whose ages were determined, could not be used for otolith analysis since their otoliths were not readable. Clearly, there remains a need for additional ageing work on these fish. The good agreement between the location of age classes identified by modal analysis and by otolith interpretation supports both methods of age-class identification (Figures 6 and 7). Some discrepancies resulted from the failure of modal analysis to identify mixed age modes at times when length overlap between adjacent age groups was great. Rare age classes may also be missed by modal analysis. Both the difficulty in definition of poorly represented age groups and the possibility of misclassification of highly overlapping distributions (particularly if the distributions deviate from the assumed normality) are recognised problems in modal analysis (McNew & Summerfelt 1978; MacDonald & Pitcher 1979; MacDonald 1987). The otolith interpretations appear to provide the more reliable means of age class identification, but there is no reason to suspect either method as being substantially incorrect. Other studies of age determination in marine cottids suggest lifespan ranging from 1 to 2 years (e.g. Oligocottus snyderi, Freeman et al. 1985) to 13 years (e.g. Scorpaenichthys marmoratus, a subtidal cottid, O'Connell 1953). Oligocottus maculosus, another abundant intertidal cottid, was estimated by Craik (1978) to survive up to four years. Wells (1986) determined the age of Clinocottus analis 60 using otoliths at Point Fermin , California and concluded that max imum lifespan was 6 years for females and approximately 8 years for males. Differences in lifespan between sexes in C. globiceps were not evident in this study since there were no deviations in sex ratios from unity in all collections. There was no discernible pattern which would suggest that either sex lived longer. Growth of C . globiceps was confined to spring and summer (Figure 13), the time when peaks in temperature, nutrients and pr imary productivity (Parsons et al. 1984) occur. Rapid growth in summer has also been reported for other intertidal cottids, Oligocottus maculosus (Moring 1979), O. snyderi (Moring 1981; Freeman et al. 1985), Clinocottus analis (Wells 1986). Growth was reduced as winter approached and probably ceased during winter. It has been suggested (Moring 1979, 1981) that reduced foraging activity caused by increased wave action during winter might partially explain this growth reduction. Gonadal development during the winter months (C. globiceps was found in non-reproductive condition during the months Apr i l through October, therefore, winter reproductive activity is strongly suspected) is also thought to be responsible for reduced growth (Moring 1981). The radius of otolith annuli provided reasonably accurate estimates of length at the time of their formation. The accuracy of back-calculations was confirmed by the agreement between observed lengths and back-calculated lengths. The back-calculated growth of a fish species quantifies the growth attained in each previous age group (Ricker 1969). The back-calculated lengths for the earlier seasons of life were greater in older fish for C . globiceps (Table 5). This 61 relationship appears to be a negative size-selective mortality (Ricker 1969, 1979) and is manifested as a reversal of Rosa Lee's phenomenon. This has been described for another large cottid Leptocottus armatus, by Weiss (1969). Other explanations put forward for Lee's phenomenon include: 1) size-selective mortality of the more rapidly growing individuals (Jones 1958), 2) errors in back-calculation or the use of an inappropriate method (Lee 1920), 3) non-random sampling of the stock (Ricker 1969), 4) the presence of false annuli which increase in frequency as the individuals grow older, and 5) the contraction of the annuli towards the nucleus (Bilton 1974) as a result of passive dissolution or active uptake of calcium from the hard part (Yamada 1956; Ouchi et al. 1972; Mug iya & Watebe 1977). Given the current data there is no way of directly testing the above ideas. However, several indirect evaluations can be made. The concordance of observed lengths and back-calculated lengths strongly suggest that the back-calculated lengths are fairly accurate. The fish collecting technique employed in this study appeared to give representative samples as there was a noticeable lack of skewed length frequency distributions or other evidence suggestive of non-representative sampling. The usage of opaque bands as annuli in this study also appeared to be justified due to their annual formation during the study period. In addition the close agreement of back-calculated lengths between different year classes is a further indication that this assumption was also reasonable. Unfortunately the fifth explanation cannot be satisfied, but upon examination of the data there is a strong feeling that it does not explain the observed phenomenon. The presence of reversed Lee's phenomenon in this species is most likely attributable to the 62 greater survival of faster growing individuals, i.e., negative size-selective mortality of the slower growing individuals. M e a n observed, back-calculated, and theoretical lengths-at-age appeared to be realistic and relatively consistent for the C. globiceps examined (Tables 10 & 11). Growth in length was very rapid (exponential) during the first two years of life but appeared to become asymptotic thereafter (Figure 11). Thus the Gompertz growth model more accurately reflected growth in early and later years. Other growth models except the von Bertalanffy were not used and this is not intended to imply that the Gompertz model was the most appropriate. The use of the Gompertz model was best assessed on its 'goodness of fit' to empirical data. The von Bertalanffy model did not fit the data, especially in the early years when growth displayed a point of inflexion, that is, up to the age corresponding to that point, the growth rate progressively increases with age, whereas subsequently it progressively decreases. This is reflected by the" parameters G and g with values 1.58 and 0.30 representing growth rate before and after the point of inflexion respectively. One of the important aspects of growth is the relationship between length and weight. For many species it has been found that weight increases as the cube of length, but for others it has been shown that weight accrues at a greater or lesser rate than the cube (Le Cren 1951). The length —weight regression analysis showed that combined regression coefficient for both sexes differed significantly from 3 indicating a deviation from the cube law (Table 14). Therefore the fish is showing a positive allometric growth (b = 3.1552) that is, the fish becomes 63 heavier for its length as it grows larger (Bagenal & Tesch 1978). Differences in length — weight relationship may occur as a result of gear selectivity, sex, different seasons, and times of day (Bagenal & Tesch 1978). In the case of C . globiceps population no differences in length—weight relationship were found between sexes. Gear selectivity does not seem to be important considering the method of fish collection, however, different seasons might be important but were not investigated in this study. Sex ratios are one of the most easily obtained of all fishery statistics but they usually remain undiscussed, except in the more atypical fish species (Warner 1975), and lampreys (Hardisty 1961; Potter et al. 1974), and as a consequence are poorly understood (Emlen 1973). The sex ratio of C. globiceps was 0.998:1 (males to females). There is no evidence to suggest a trend towards a preponderance of either males or females with increasing age (Table 17). It is reasonable to assume that species, which mate randomly or form pairs and have little or no size difference between the sexes, would have even sex ratios (Emlen 1973), and such was the case for C. globiceps. Although the overall sex ratio did not deviate statistically from a 1:1 ratio, few individual collections were dominated by one sex. Movements of sexually segregated groups within the intertidal zone and small sample sizes over a restricted period of time could account for a predominance of one sex in some samples. III. F I S H P R O D U C T I O N 3.1 INTRODUCTION The term 'fish production' as used in this thesis and by many authors (Ricker 1946; Al len 1951; Chapman 1978a, 1978b; Weatherley & Gil l 1987) is defined, in the sense of Ivlev (1945) as total weight of body tissue produced by a population of fish during a given interval of time, including growth by fish that died during the time interval. Le Cren (1969) considered production to be "the best epitome of the population dynamics and environmental performance" of a species. Since production estimates account for all growth in a fish population as such they could serve as valid comparative indices for assessing the ecological success of a species in different environments. Production in the present study was estimated by the instantaneous growth rate method (Ricker 1946) and the size-frequency method, formerly called the Hynes method (Hynes & Coleman 1968; Hamil ton 1969; Benke 1979; Hynes 1980; G a r m a n & Waters 1983). Other approaches to studying production involve tracking energy transformations in populations or communities of organisms. L indeman (1942) introduced the major concept that an organism's success in an environment might be a function of its ability to fix and retain energy. This concept underlies many of the energetic approaches to production studies, but few investigators have taken this route in fish production studies. Winberg (1956) first drew attention to the possibility of studying the bioenergetics of fish 64 65 populations. M a n n (1965) described a bioenergetic approach to the study of the quantitative relationship between fish and their food; Na iman (1976) studied productivity of pupfish population from an energetic viewpoint. Recently Bennett (1984) estimated production of an intertidal clinid Clinus superciliosus by developing the population energy budget. The present study attempts to estimate production of two tidepool cottids. Because of the strong homing patterns of Clinocottus globiceps and Oligocottus maculosus, it was considered that the best estimates should follow the procedures for lake and enclosed watershed production estimates. Basic emphases in this study are focused on (i) estimating and comparing annual biomass and production of C. globiceps and O. maculosus by age group in the population, also total biomass and production by the populations as a whole, (ii) evaluating the influence of physical characteristics (depth, shoreline index, perimeter, and surface area) of' tidepools on production of these two species. 66 3.2 MATERIALS AND METHODS 3.2.1 C o l l e c t i o n of F i s h e s The technique used in collecting fish has been described in the earlier sections. Production estimates were carried out on two tidepool cottid species, C . globiceps and 0. maculosus which often occur in sympatry. The latter species was selected on the basis of numbers, biomass, and the ease with which it could be aged. A total of 17 collections made between Ju ly 1986 and Ju ly 1987 yielded C. globiceps (n = 218); out of this total O. maculosus were found in 6 collections (n=169). Between August 1987 and October 1988, a total of 29 collections comprised C. globiceps (n=352) and out of this total, O. maculosus were caught in 16 collections (n=1382). 3.2.2 A b u n d a n c e E s t i m a t e s a n d D e t e r m i n a t i o n of A g e Though it is being advocated that populations should be sampled regularly in fish production studies (Chapman 1978a) so as to account for seasonal variability in growth and abundance, often this is not practical. The classical methods of estimating abundance e.g. mark—recapture and removal methods have been found to be less reliable, moreover they require intensive sampling to be reasonably accurate (Chadwick 1976b; Pot et al. 1984). 67 Since fish were totally removed from each tidepool, censuses of the resident species were considered to be accurate, thus the abundance of the chosen species was based on these censuses. Information on the age structure of C. globiceps has been presented previously, therefore only age data for O. maculosus will be presented here. The procedures for weight and length measurements in O. maculosus were the same as those described earlier for C. globiceps. Age structure of O. maculosus was determined by using the age—length relationship developed by Craik (1978) for populations at Helby Island. She described the age —length relationship by an ordinary predictive regression of the form y = a+bx, using a sample of 87 individuals caught in 1976. 3.2.3 Production Information on density, growth rates and rate of change in numbers (e.g. due to mortality) of a species is fundamental in production calculations. With this information two methods can be used to estimate production of a species. The first, the Al len curve (Allen 1951) is a graphical representation whereby numbers of individuals are plotted against mean individual weight at two or more time intervals. These points are joined by a curve and production during a given time interval is equal to the area under that curve. The second, the instantaneous growth rate method (Ricker 1946), involves integration to determine the mean biomass during a given time period. Production is then calculated by multiplying the mean biomass by the instantaneous growth rate during that time period. Both methods will give exact results for identical data. 68 A more recent method, the size-frequency method which was originally suggested by Hynes (1961) and developed by Hynes & Coleman (1968), Hamilton (1969), and Benke (1979) for estimating secondary production in aquatic invertebrates, has been found to be useful in fish production studies (Garman & Waters 1983; F reeman & Freeman 1985). This method does not require cohort separation (i.e., ageing of individuals), therefore it is useful where ageing of sampled fish poses problems. In this study I chose to use the instantaneous growth rate method. I also attempted to compute production by the size-frequency method and compared the results as a way of checking the accuracy of each method. Instantaneous Growth Rate Method The fundamentals of this method were laid down by Ricker (1946) and Ricker & Foerster (1948) and over the years has become one of the most common models in fish production studies. This method requires ageing of sampled individuals. The method assumes that the ratio of the growth rate to mortality rate (G/Z) is constant for the interval of time for which production is calculated. Furthermore it also assumes, when production estimates are based on a single complete fish collection, that the age structure and numerical abundance of the species in question are constant from year to year, an assumption that is often not valid. The first assumption that G/Z is constant may be considered valid when the time interval is short, usually a fortnight to a month (Chapman 1978a). 69 Keeping track of the mean weight (W) and density (N) of one age-class by continuous monitoring would render the second assumption redundant, but the monitoring process itself might affect growth and mortality. However, several studies (Thomson & Lehner 1976; Grossman 1982; Yosh iyama et al. 1986; N . J . Wil imovsky pers. comm.) have shown that regular sampling of tidepool fish populations does not have significant effects on densities and species richness in the long run. Grossman (1982) demonstrated that tidepools are completely repopulated three months following complete removal of fish. I therefore decided that a series of single complete collections probably were the only practical way to estimate production in this environment. Basic statistics pertinent to this method as outlined in Mahon & Balon (1977) are listed below: • N i —number of fish in each group i, • "Wi - m e a n weight in g of an individual in age group i, • B i —initial biomass in g of age group i which equals N i * W i , • Zi,i—1 --instantaneous mortality rate between ages i and i— 1, this equals - ( I nN i - I n N i - l ) , • Gi,i—1 - instantaneous growth rate between ages i and i—1, this equals InWi - I n W i - l , • Hi, i—1 -instantaneous rate of increase in biomass between i—1 and i, this equals G —Z for this interval, _ pj • Bi , i — 1 - m e a n biomass between i — 1 and i, this equals Bi(e — 1). H Total production as calculated by this method was obtained by multiplying the 70 mean biomass (B) by the instantaneous growth rate (G) on a yearly basis, and summing up for each age class. Production was calculated from the time of collection (and not time of spawning) since this has been shown to be the most accurate method (Halyk & Balon 1983) and since the times of spawning for this species were unknown. The ratios of production to mean biomass (P:B) and production to initial biomass (P:Bi) were calculated for each species. Size-Frequency Method This method requires that individuals be grouped into some length groups, and production is simply calculated by annual summation of individual losses from one size group to the next. Since this method has the potential of being a real solution to the tedious and often expensive task of ageing many individuals as required by both Allen curve and the instantaneous growth rate method, its accuracy was determined by comparing the production estimates with those of the instantaneous growth rate method. This method assumes that animals spend an equal amount of time in each size class (Hamilton 1969). G a r m a n & Waters (1983) suggested "use of nonequal length groups, chosen to reflect age- or size-dependent changes in growth rate" as a way of reducing errors due to this assumption in the production estimates by the size-frequency technique. T h u s , the number and size of length classes were chosen according to the predetermined lengths-at-age, based on size frequency distributions for C. globiceps and O. maculosus. Production was calculated using G a r m a n & Waters (1983) formulation following the notation of Newman & Mart in (1983). The following pertinent information was required: 1. The annual mean number of individuals within the kth length group, which was estimated by = - f v n y where i(i=l, 2,..., a) denotes sampling dates, k(k=l, 2,..., c) denotes length groups, Nik represents the density value within the kth group on date i and Di is a weighting factor representing the interval between the ith date and the first sampling date (in days). 2. The annual mean weight of each length group in the population, which was expressed as where Wik is the average weight of those individuals within the kth size group on date i. Using the above formulation, production is calculated as the sum of the biomass lost from the population through mortality between size classes, multiplied by the number of average cohorts which develop during the period (same as the number of size classes: Hamilton 1969) and was expressed as P-t = [ ^ [ ^ . i ( / V - . i - / v . 2 ) + g W.k(Nk^ - Nk+1) + W.C(NC^ - tfc)]](^) 72 where CPI=cohort production interval, which in fish is represented by the average max imum age attained (in years) by individuals in the population (Garman & Waters 1983). Total production was estimated for each species by summing production between collections made from Ju ly 1986 through Ju ly 1987 (I name this time span 'first year') and from August 1987 to October 1988 (referred to as 'second year'). Production estimates for the second year were corrected to annual values by multiplying by 365/404. Production was described on an area basis, as most aquatic production is known to depend on area (Le Cren 1972). The tidepools were grouped into four vertical zones: upper, middle, lower, and base. The zones follow the levels that were fixed relative to the zero tide level. Production estimates were calculated by zones and statistical tests for significant differences between zones were carried out. Analyses were performed with Statistical Analysis System (SAS) computer programs (SAS Institute 1988). 73 3.3 RESULTS 3.3.1 Distribution within the Study Area Clinocottus globiceps and Oligocottus maculosus, are both bottom-dwelling and often occur sympatrically, although C. globiceps tends to favour the mid to lower intertidal pools. C. globiceps were collected from pools as high as 2.70 m above zero tide level, but occur in great numbers in middle and lower tidepools. They were rarely observed in pools higher than this level (Figure 16). O. maculosus though widely distributed, are concentrated in the mid and higher level tidepools (Figure 17). Distribution within the intertidal area can be influenced by many interacting factors. A great amount of habitat cover, i.e., primarily large rocks and crevices but also dense algal (and/or eel grass) growth, allows C. globiceps to inhabit higher pools. O. maculosus can be found in fair abundance in the more rocky pools with only moderate amounts of macrophytic cover. It may frequently occur in great numbers in open sandy areas of some tidepools. The vertical range inhabited by C. globiceps and O. maculosus reflects the size of the fish. Larger C. globiceps, although found throughout the area, were more concentrated on the lower level than were smaller fish. Young C. globiceps were usually found at tide levels between 0.70 m and 1.50 m (Figure 16). Juveniles less than 25 m m were not common in pools inhabited by adults. Young O. 74 60 40 • 20 • 0 80 L 60 N=12 UPPER N=295 MIDDLE N=241 LOWER N=21 BASE _l L. 0 10 20 30 40 50 60 70 80 90 100 110 120 130 Standard Length (mm) Figure 16. Length-frequencj' distribution of C. globiceps from tidepools at different levels. 75 200 150 100 50 0 200 0 150 1 ioo 3 o- 50 cu L ° 200 150 100 . 50 -0 50 N=412 UPPER N=614 MIDDLE N=323 LOWER N=0 BASE 0 10 20 30 40 50 60 70 80 90 Standard Length (mm) Figure 17. Length-frequency distribution of O. maculosus from tidepools at different levels. 76 maculosus do not occur as low as do older fish. Larger O. maculosus were found inhabiting pools as high as those inhabited by juveniles (Figure 17). The degree of exposure, i.e., how long a pool is actually emerged or submerged, may also affect the distribution of intertidal fishes (Green 1971b). This factor is a function of the vertical position of the pool in relation to tide level. 3.3.2 Age of Oligocottus maculosus The results show that O. maculosus survives up to four years. The data presented in Table 19 are compared with an earlier study by Craik (1978) who used otoliths in ageing this species. Age groups as computed from age —length regression L o g age = - 0 . 7 6 7 7 + 1.447 log total length (Craik 1978) compared well with otolithic ageing as shown in Table 19. Only 15% of the collected O. maculosus were in age group 0, 50% were in group 1, 30% in group 2, and 4% in group 3, with only 0.2% in age group 4. 3.3.3 Density M e a n densities of both species fluctuated considerably in all zones during the two-year period (Table 20). Densities of all ages were usually less in the upper and middle levels than in the lower zone for C . globiceps whereas, O. maculosus had high densities in the upper to middle zones than the lower zone (Figure 18 and Appendix 5). The density of C. globiceps ranged from 1.2 (at the base level) 77 Table 19. Age-structure of 0. maculosus from the present study as determined by age-length regression: Log(age)=-0.7677-fl.447Log(total length). Craik's (1978) data are presented for comparison. Present Study Craik's (1978) Study Age N % total SL T L N %total SL T L 0+ 232 15.0 18.9 23.7 - - - -1 + 796 51.4 29.7 36.6 127 34.8 34.3 42.1 2+ 461 29.7 43.2 52.7 194 53.2 41.0 50.1 3+ 58 3.7 55.2 67.0 43 11.8 53.3 64.9 4+ 3 0.2 61.3 74.3 1 0.3 60.1 73.0 Total 1550 365 78 to 2.7 fish per m 2 of pool area (middle level) in the first year, and from 0.6 (base level) to 7.5 fish per m 2 of pool area (lower level) in the second year (Table 20). Dur ing the first year O. maculosus density ranged from 0.1 (base level) to 5.3 fish per m 2 (lower level) and in the second year the range was 9.8 (lower level) to 15.7 fish per m 2 (middle level) (Table 20). O n the whole, densities of O. maculosus were higher than those of C. globiceps. There was no significant difference in densities among zones and years for either species (Tables 21 and 22). Relative differences among zones were fairly constant among years, as indicated by the lack of significant zone Xy e a r interaction (Tables 21 and 22). Because fewer pools were sampled between Ju ly 1986 and Ju ly 1987, most analyses involving densities were based on collections made between August 1987 and October 1988. The zones with the highest densities (middle and lower for C. globiceps, and upper and middle for O. maculosus) held substantially more age 1+ fish than the lowest density zones. Densities of older age groups were lower than those of younger age groups (Appendix 5). The absolute number of C. globiceps and O. maculosus in each year-class at successive intervals throughout their life history was not estimated for the tidepools at Helby, and consequently no quantitative pattern of mortality and survival for the entire duration of any year-class can be illustrated. However, from the relative frequency of age groups in the population of both species, it was possible to derive the survival curves for the population as a whole (Figure 19). The empirical survival curves shown in Figure 19 indicate a high mortality rate during the first two years of life (note the steepness of the right limb of Table 20. Mean annual density (N/m2), I n i t i a l biomass (Bi, g/m2), mean biomass (5, g/m2), mean instantaneous growth rates (G), annual production (P, g/m 2/year) and annual production to biomass r a t i o s (P:Bi, P:B) of C. globiceps and 0. maculosus. Production and biomass f o r 1987-88 i s based on corrected annual values obtained by m u l t i p l y i n g by 365/404. July 1986 - July 1987 August 1987 - October 1988 Zones Zones Cllnocottus olobiceps Upper Middle Lower Base Upper Middle Lower Base N - 2.7 1.9 1.2 0.7 1.0 7.5 0.6 Bl - 18.6 13.0 13.5 2.4 4.1 29.5 6.7 B - 21.1 11.3 8.3 1.8 2.6 26.0 -G - 0.51 0.78 0.57 0.94 0.77 0.66 -P - 10.8 8.8 4.7 1.7 2.0 17.1 -P:Bi - 0.58 0.68 0.35 0.71 0.49 0.58 -P:I - 0.51 0.78 0.57 0.94 0.77 0.66 -Oligocottus maculosus N _ 2.2 5.3 0.1 15.6 15.7 9.8 Bi - 2.5 5.7 0.03 10.5 14.3 9.4 -B - 1.8 4.9 0.03 9.3 13.1 7.6 -G - 1.00 0.76 1.00 1.20 1.04 1.09 -P - 1.8 3.7 0.03 11.2 13.6 8.3 -P:B1 - 0.72 0.65 1.00 1.07 0.95 0.88 -P:B - 1.00 0.76 1.00 1.20 1.04 1.09 -Table 21. Analysis of variance s t a t i s t i c s for growth, density, biomass and production by years and zones for C. QJJQ Source D.F. SS F P Growth Year 1 0.1070 0 .94 0 .35 Zone 1 0.1709 1 .50 0 .24 Zone*year 1 0.2326 2 .04 0 .17 Error 16 1.8262 Density Year 1 0.4671 0 .04 0 .85 Zone 1 28.7909 2 .40 0 .14 Zone*year 1 35.1182 2 .93 0 .10 Error 19 227.5485 Production Year 1 122.7273 1 .64 0 .22 Zone 1 82.4755 1 .10 0 .31 Zone*year 1 75.8039 1 .01 0 .33 Error 19 1419.168 Biomass Year 1 371.4819 1 .20 0 .29 Zone 1 237.7747 0 .77 0 .39 Zone*year 1 478.5106 1 .54 0 .23 Error 19 5891.303 Table 22. Analysis of variance s t a t i s t i c s for growth, density. biomass and production by year and zones for 0. maculosus. -Source D.F. SS F P Growth Year 1 1.2496 16.82 0 .002* Zone 1 0.0329 0.44 0 .52 Zone*year 1 0.0195 0.26 0 .62 Error 12 0.8917 Density Year 1 210.2645 1.79 0 .21 Zone 1 45.3809 0.39 0 .55 Zone*year 1 23.6696 0.20 0 .66 Error 11 117.6899 Production Year 1 99.2201 1.74 0, .22 Zone 1 34.2435 0.60 0, .46 Zone*year 1 10.1877 0.18 0, .68 Error 10 571.7371 Biomass Year 1 145.6370 1.90 0. .20 Zone 1 33.1877 0.43 0. .53 Zone*year 1 32.4460 0.42 0. 53 Error 10 765.0710 81 o 100 80 60 40 20 0 Upper 1987-1986 Upper Middle I Lower Base Oligocottus maculosus Middle - , Lower Zone Base Age groups tm o+ \zzu i =_B 2 + _ S 3 + C sa 4+ o 5+ Figure 18. Density of C. globiceps and 0 . maculosus by age group, in each zone for the period 1986-87 to 1987-88. 82 60 0 1 2 3 4 5 6 Age (years) Species Q. globiceps —I— Q. maculosus Figure 19. Age-frequency distribution curves for populations of C . globiceps and O. maculosus from Helby Island. Table 23. Summary of regression a n a l y s i s v a r i a b l e s (defined i n t e x t ) . C o r r e l a t i o n s of each set of v a r i a b l e s were tested f or s i g n i f i c a n c e , r-squared values are the proportion of the t o t a l v a r i a t i o n accounted f o r by the independent v a r i a b l e . Oliaocottus maculosus r r ^ P C l i n o c o t t u s alobiceDs P Variables D.F. D.F. r r 2 Densities N with depth 10 -0.68 0.46 0.016 14 0.19 0.04 0.480 N with area 10 -0.25 0.06 0.442 14 0.56 0.31 0.024 Growth G with B 11 -0.13 0.02 0.663 16 -0.06 0.004 0.809 Gl with BI 6 -0.35 0.13 0.391 6 -0.10 0.01 0.809 G2 with B2 11 0.24 0.06 0.438 11 0.02 0.001 0.936 G3 with B3 8 -0.03 0.001 0.945 7 -0.68 0.47 0.040 G with N 11 -0.10 0.01 0.734 16 0.03 0.001 0.913 Gl with NI 6 -0.43 0.18 0.292 6 -0.32 0.10 0.440 G2 with N2 11 0.22 0.05 0.473 11 -0.15 0.02 0.620 G3 with N3 8 -0.04 0.002 0.906 7 0.73 0.53 0.030 Gl with PI 6 -0.01 0.0001 0.985 6 0.04 0.002 0.928 G2 with P2 11 0.42 0.18 0.154 11 0.32 0.10 0.292 G3 with P3 8 0.57 0.32 0.086 7 0.54 0.29 0.130 Production i P with B 11 0.98 0.96 0.0000 15 0.92 0.85 0.0000 P with Bi 11 0.96 0.93 0.0000 15 0.96 0.93 0.0000 P with G 11 -0.14 0.02 0.645 16 0.06 0.004 0.815 P with depth 10 -0.45 0.20 0.139 14 0.23 0.05 0.383 P with area 10 -0.04 0.002 0.902 14 -0.25 0.06 0.357 P with perimeter 10 -0.10 0.02 0.690 14 -0.37 0.14 0.159 P with SI 8 0.21 0.04 0.565 9 -0.24 0.06 0.469 84 the curves). The very youngest fishes (0+ age group), mostly still planktonic were not adequately represented in the area studied. This is depicted by the ascending left limb and the dome of the survival curves in Figure 19. The surface area of the pool was significantly correlated with density of C . globiceps; there was a weak but non-significant negative correlation between density of O. maculosus and surface area of pools (Table 23). Depth of pools correlated negatively with density for O. maculosus; there was no correlation between these two variables for C. globiceps (Table 23). 3.3.4 B i o m a s s There were large differences in biomass by age groups between the two species at different zones (Appendix 5). When biomass is analyzed by zones for both species, O. maculosus exhibits higher biomass than C. globiceps at the upper and middle zones (Table 20). Biomass of all age groups combined for C. globiceps was substantially more in the lower zone than in the middle and upper zones (Table 20). Relative differences in biomass among zones for both species did not change over the two years, as indicated by the lack of significant zone X year interaction (Tables 21 and 22). Age groups 1+ & 2 + , and 2+ & 3+ had the highest biomass for O. maculosus and C. globiceps respectively (Appendix 5). M a x i m u m biomass during the second year was 10.5 g / m 2 , for C . globiceps and 11.4 g / m 2 for O. maculosus for combined zones. 85 3.3.5 G r o w t h Annua l instantaneous growth rates decreased with increasing age for both species (Appendix 5). There were large yearly differences in growth within zones, but no significant differences were found among zones (Tables 21 and 22). The lack of significant differences among zones as indicated by A N O V A was probably attributable to lack of consistent differences among vertical location of pools rather than lack of variability. There was no relationship between population density - and growth rate (Table 23); the amount of variability in growth explained by changes in density was very low ( r 2 of 0.001 and 0.01 for all age groups of C. globiceps and O. maculosus respectively) and insignificant statistically. Correlations with age groups 2+ and 3 + for C . globiceps and 1+ and 2 + for O. maculosus were positive and insignificant (Table 23) indicating that among age groups, growth was usually not negatively correlated with its density. Correlations of mean biomass and growth rates of all age groups for both species were computed (Table 23). Significant negative correlation (r=—0.68, p<0.05) between growth and mean biomass for C . globiceps between age groups 2 + and 3 + indicated that growth was negatively related to population biomass by age groups. The correlations for all age groups were insignificant for both species (Table 23); the correlation for age group 1+ and 2 + for O. maculosus was positive, but weak, suggesting that there are no large density-related differences in growth between these age groups. 86 3.3.6 P r o d u c t i o n Total production for the two species in all zones was 8.1 g / m 2 / y e a r in the first year and 6.9 g / m 2 / y e a r in the second year for C. globiceps, and 1.9 and 11.0 g / m 2 / y e a r over the same period for O. maculosus. Almost 33% of C. globiceps production in the second year occurred between age 1+ and 2 + , whereas 65% of O. maculosus production occurred between age 1+ and 2+ during the same period. Production was less between ages 0 + and 1+ in both species (Appendix 5) because of lower densities of these age groups. Production was also analyzed for those tidepools where these species do not overlap. For example, some pools at the lower level were devoid of O. maculosus whereas C. globiceps were absent from some of the upper level pools. Second year data were used in these analyses. C. globiceps average production value was estimated as 12.3 g / m 2 / y e a r , almost twice the estimate for a sympatric situation. The estimate for O. maculosus was 11.7 g / m 2 / y e a r , a value that is comparable to the one given above for a sympatric case. Speculations on the effect of species interaction on production cannot be made here because production estimates of other tidepool residents were not computed. However, it is worth noting that the high production value for C. globiceps could be attributable to the bigger size of fish in the lower pools since density (7.8 fish per m 2 ) was very similar to the estimate (7.5 fish per m 2 ) when the two species occurred together. Differences in production among zones were not large except for lower zone in 87 the second year for C. globiceps (Figure. 20). Production was greater in the lower zone than in the upper and middle zones for C. globiceps (Figure 20 and Table 20). The lower production in pools of upper and middle zones was probably due to their obviously lower density (Table 20). Production of O. maculosus was greater in the upper and middle zones than in the lower zone; there were no O. maculosus caught from the bottom zone in the second year (Table 20). Annua l production for the two species differed by zones; estimates for O. maculosus in the upper and middle zones were 11.2 and 13.6 g / m 2 / y e a r respectively, with corresponding values for C. globiceps being 1.5 and 2.0 g / m 2 / y e a r respectively. Production estimates for O. maculosus and C. globiceps in the lower zone were 8.3 and 17.1 g / m 2 / y e a r respectively. Though production tended to vary among zones as the above values suggest, analysis of variance by zone and year revealed no significant differences in annual production among both zones and years for the two species (Tables 21 and 22). Aga in , there was a lack of significant interaction of zone with year ' (Tables 21 and 22), indicating that relative differences among zones were fairly constant over time (Figure 20). Production was poorly correlated with instantaneous growth rates for most age groups (Table 23). For all age groups combined, variability in growth accounted for less than 0.4 and 2% of changes in year class production for C. globiceps and O. maculosus respectively. Production was significantly positively correlated with mean biomass for both species. Fluctuations in mean biomass accounted for 85 and 90% of fluctuations in production rates of C. globiceps and O. maculosus 100 80 60 40 >— CO CD 20 E o 3 C 10 o Upper 1987-1988 __L Upper 1986-1987 o T3 O 8 -Upper 60 r-50 -40 -30 -20 -10 -0 - I Upper Middle Middle Middle Lower Base Clinocottus globiceps 5 Lower Base Oligocottus maculosus I Lower Base Oligocottus maculosus 1. Middle Lower Zone Base Age groups !o+-1+ 1Z21+-2+ KS4+-5+ Figure 20. Production of C . globiceps and O. maculosus by age group, each zone for the period 1986-87 to 1987-88. 89 Figure 21. Relationship between production and and initial biomass of 1987 and 1988 year classes of C. globiceps and 0. maculosus. 90 respectively (Table 23). The relationship between production and initial biomass was investigated by regressing the former on the latter for both species. A plot of the production and initial biomass values for C . globiceps and O. maculosus is presented in Figure 21. Variabil i ty in initial biomass accounted for 93% of variation in production for both species. Depth, a physical character of pools, varied substantially among pools with values ranging from 0.1 to 1.9 m. There was a positive but non-significant correlation between tidepool depth and production (r = 0.24, p>0.05) for C. globiceps. However, the correlation between depth and production for O. maculosus was high enough (r= —0.45, p<0.14) to suggest a negative relationship (Table 23). Surface area, perimeter and shoreline index (a measure of the degree of regularity or irregularity of a pool) seemed to be somewhat less important. No correlation was apparent between production and these variables (Table 23). For example, variations in pool surface area accounted for 6 and 0.2% of changes in production for C. globiceps and O. maculosus respectively. 3.3.7 S i z e - F r e q u e n c y E s t i m a t e of P r o d u c t i o n The estimates of production by the size-frequency method for C. globiceps and O. maculosus made for the period spanning August 1987 through October 1988 are presented in Table 24. There are substantial differences in the values obtained by this method and the instantaneous growth rate method. The estimates from the size-frequency method are generally higher for C. globiceps in all zones. The same trend was observed for O. maculosus, but the middle zone had lower -91 estimates. Annua l production estimates ranged from 6.9 g / m 2 / y e a r (instantaneous growth rate method) to 15.8 g / m 2 / y e a r (size-frequency method) for C. globiceps and from 11.0 g / m 2 / y e a r (instantaneous growth rate- method) to 19.5 g / m 2 / y e a r (size-frequency method) for O. maculosus (Table 24). The estimates by the instantaneous growth rate method are lower for both species because growth rates might have been underestimated as a result of nonsynchronous cohort development and prolonged recruitment (usually starting around Apr i l through August). 3.3.8 Annual Production to Biomass Ratios Annua l production to mean biomass ratios (P:B) were lower for the larger C . globiceps and higher for the smaller 0. maculosus (Table 20). Production to initial biomass ratios (P:Bi) followed the same trend as P:B ratios, being lower for the larger species and higher for the smaller species (Table 20). There were annual variations in both ratios (P:B and P:Bi), with the first year having lower values than the second year. These ratios also varied considerably between zones for both species (Table 20). The P:B ratio was higher in the upper followed by middle and lower zones for C. globiceps during the second year. O. maculosus exhibited the same trend with P:Bi ratio. This trend is consistent with the observation that there were more juvenile fish in the higher levels than lower levels (see Figures 16 and 17). 92 Table 24. Annual production, mean biomass and P:B ratios as estimated by instantaneous growth rate (IGRM) and size-frequency methods (SFM) for 1987-88. Mean biomass and production are expressed as g.m~2 and g.m~2.year-1 respectively. C. globiceps IGRM SFM Zone B P P:B B P P:B Upper 1.8 1.7 0.94 4.7 6.2 1.33 Middle 2.6 2.0 0.77 5.7 7.7 1.35 Lower 26.0 17.1 0.66 27.7 33.6 1.21 Combined 10.1 6.9 0.68 12.7 15.8 1.24 zones 0. maculosus Upper 9.3 11.2 1.20 16.8 31.7 1.89 Middle 13.1 13.6 1.04 2.1 8.0 3.81 Lower 7.6 8.3 1.09 11.1 18.7 1.68 Combined 10.0 11.0 1.10 ~ 10.0 19.5 1.95 zones 93 3.3.9 Estimation of Production from Initial Biomass Mathews (1970) noted that the estimation of fish production by either Al len curve or instantaneous growth rate method is very tedious and time consuming, an observation which became apparent in the course of this study. A number of authors (Waters 1969, 1977; Mathews 1970; Al len 1971; Chapman 1978b) have suggested that the ratio of production to mean biomass (P:B) m a y be used to estimate production of fishes and invertebrates by multiplying the mean biomass at a given site by a predetermined P:B ratio. However, mean biomass is not easy to determine since it requires either continuous (and regular) sampling throughout the year or ageing of individuals in a sample to estimate growth and mortality rates. Initial biomass, on the other hand, can easily be determined from a single estimate of abundance and mean weight. Moreover, its estimation does riot require that the individuals collected be aged. The suitability of initial biomass in predicting production was tested by regressing production on mean and initial biomass. The results presented in Table 23 show that changes in initial biomass accounted for 93% of the variation in production, while variability in mean biomass explained 85% of fluctuations in production. 94 3.4 DISCUSSION Production rate is regulated by the numbers of fish present and their biomass which in turn are regulated by growth and density. It follows therefore that the accuracy of calculating annual production depends on how accurately growth and density are determined. Estimates of population size in particular are sometimes subject to large sampling errors and are often the principal source of inaccuracy in production calculations (Chapman 1978b). Dur ing this study, fish sampling involved complete collection, therefore it can be assumed that numerical abundance estimates are accurate. However, it was not possible to differentiate between fluctuations in abundance attributable to mortality or to other factors e.g. predation. One could argue that since the tidepools are not entirely closed systems, movements into or out of the study pools could affect accuracy of density estimates, but the strong homing tendencies of these species (Green 1971b, 1973) make the movement factor less important in causing density fluctuations. While the potential lifespan is at least 4 and 5 years for O. maculosus and C . globiceps respectively, only a minute fraction of the tidepool population approach such an age (Tables 18 and 19). Beverton & Holt (1957) believe that predation is the most general cause of natural mortality in fish populations. I have no direct evidence to suggest that C. globiceps and 0. maculosus are eaten by other tidepool fishes. In this context Gibson (1982) notes that predation has rarely been observed in intertidal fishes and therefore does not constitute an important factor controlling the populations 95 of intertidal fishes. Apar t from predation and other extrinsic factors causing death (Paling 1968), senescence has been demonstrated in some wild fish populations (Gerking 1957; Cra ig 1985). The presumed increase in mortality rate suffered by both species after the second year of life (Figure 19) would qualify as age-dependent, and may thus resemble senescent mortality described in other fishes. However, the onset of senescence in the wild populations would be difficult to demonstrate, even with absolute estimates of population size unless the presumed predation factor of mortality could be measured. Oligocottus maculosus had the highest production: 11.0 g / m 2 / y e a r by the instantaneous growth rate method and 19.5 g / m 2 / y e a r by the size-frequency method. C. globiceps produced 6.9 and 15.8 g / m 2 / y e a r by instantaneous growth rate and size-frequency methods respectively. Mean biomass for the two species was remarkably similar: 10.2 g / m 2 (instantaneous growth rate) and 14.05 g / m 2 (size-frequency) for C . globiceps and 10.0 g / m 2 (instantaneous growth rate) and 11.0 g / m 2 (size-frequency) for O. maculosus (Table 24). Young age groups (1 + and 2 + ) made a significant contribution to production in both species, most obviously in O. maculosus; age groups 1+ and 2+ contributed 65.4% to total production for the species. In contrast, in C. globiceps age groups 1+ and 2 + contributed 32.8% to production. This observation is consistent with several studies (Mathews 1971; M a n n 1971; Chadwick 1976b) which have shown that young fish are the most productive and m a y contribute up to 95% of the total production for all ages. Mor ing (1976) reported average densities of 1.64, 2.66 and 2.86 fish per m 2 of 96 "effective sampling area" (same as tidepool area) for 0. maculosus at three sites in Tr inidad B a y , California. Although these densities compared with densities of 0.1 to 5.3 fish per m 2 in the first year, they are substantially lower than densities of 9.8 to 15.7 fish per m 2 reported for the second year in the present study (Appendix 5). This difference is consistent with the observation that densities of 0. maculosus increase as one moves to higher latitudes northward away from the southern limits of their range (Yoshiyama et al. 1986). The present study is a first attempt to determine densities of C. globiceps in tidepools. A study by Wells (1986) estimated the average density of Clinocottus analis from southern California to range from 8.5 to 27.0 fish per m 2 of tidepool surface. Estimates from the present study of 0.6 to 7.5 fish per m 2 of tidepool area for C . globiceps appear reasonable considering the area that was sampled (460 m 2 ) . The above densities represent total initial biomass of 11.4 g / m 2 for O. maculosus and 10.5 g / m 2 for C . globiceps during 1987 — 88. Annua l variations in biomass and production of C. globiceps and O. maculosus were closely and positively linked to changes in population numbers in the tidepools at Helby Island (Figure 22). Annua l variation in growth was little, and for the most part did not depend on density ( r 2 =0.001 and 0.01 for C. globiceps and O. maculosus respectively). Whatever controlled density therefore directly affected C. globiceps and O. maculosus production. Population age structure of both species varied from year to year, but population structure within zones remained relatively constant (Figure 18). These results are consistent with suggestions that carrying capacity may be determined by available 97 6 8 Density (N/m2) 14 10 15 20 Density (N/m2j Figure 22. Relationship between production and density of 1987 and 1988 year classes of C. globiceps and O. maculosus. 98 habitat and possibly foraging sites with migration to and retention within pools having available suitable habitat (Chapman 1966; Bachman 1984). Furthermore, Bohlin (1978) suggested that amount of suitable habitat available may be size or age dependent, resulting in different size or age structure among zones. The present study has shown that C . globiceps populations exhibited a tendency of larger individuals to favour lower rather than higher level pools (Figure 16). These observations are consistent with the above suggestion. Production as examined by zones for the two species reflects the distribution patterns of these species. One feature of the production estimate for C. globiceps is apparent: the similarity of the estimate at the upper and middle zones (1.7 and 2.2 g / m 2 / y e a r respectively). The regressions of log P on log B for the two zones were computed and compared by analysis of covariance as a way of determining interzone variation (Portt et al. 1986). The slopes of regressions did not differ significantly between sites (p>0.52). These results support the null hypothesis that the rates of fish production will be similar in physically similar habitats. There was high production for O. maculosus in the upper and middle zones than at the lower zone (Table 20). Aga in , differences in production among zones were primari ly related to the numerical component of production. The significant correlation between pool area and density of fish (C. globiceps) indicated that surface area of pools could partly explain the variability in fish density. Depth has long been known to influence the size and age of fishes present. F o r example, Heincke (1913) (cited in Cushing 1981) stated that larger plaice live in deeper water. Green (1971b) observed that 0-year O. maculosus 99 were concentrated in shallow upper tidepools whereas larger fish inhabited deeper pools in the intertidal zone. Similar observations as pointed out earlier were made for C. globiceps. However, production was not significantly correlated with depth of pools (p>0.05, r = 0.24) for C. globiceps. Production of O. maculosus on the other hand was negatively correlated with depth of pools (p<0.05, r = —0.45). This species is known to prefer shallow water (Nakamura 1976b) which leads to the suggestion that the negative correlation is consistent with the distribution patterns of this species. Furthermore, these differences between species for some correlations seem to suggest that no two species are identical with respect to the factors that govern their intertidal distribution. The ratio of production to biomass indicates the turnover rate of production of the habitat. The tendency is for the value to decrease with increasing trophic level and greater longevity (Emlen 1973). B y virtue of their larger . size, C. globiceps had a lower production to mean biomass (P:B) ratio of 0.62 and the smaller O. maculosus had a higher P:B ratio of 0.98. General ly the zones with the higher ratios (upper and middle for C. globiceps and O. maculosus) are those with large populations of juvenile fish. Other researchers have pointed out that the P:B ratio, which varies in a trend similar to production to initial biomass (P:Bi) ratio, is higher for the younger age classes (Chapman 1965; Hunt 1966; Hopkins 1971) and similar observations were made in this study (Appendix 5). This suggested in the present study that the P:B ratio might be related to the mean biomass of individuals present at any given zone. This idea was tested by regressing the P:B ratio on log mean biomass for C . globiceps (Figure 23). The resultant regression was significant and takes the form 100 Figure 23. Relationship between P:B ratio and mean biomass (B) of C. globiceps. 101 P:B = 1.37 - 0.30 In B ( r = - 0 . 5 5 ; n = 1 7 ; p<0.05). The log of mean biomass accounted for 30% of the variation in the P:B ratio. It can be seen from Figure 23 that P:B ratio is negatively correlated with mean biomass that is, turnover rate of production declines with increasing biomass indicating some density dependent changes probably on growth and survival. In this study, a higher proportion of variability in production of C. globiceps was explained by variation in initial biomass (Bi) ( r 2 =0.93) than was explained by the variation in mean biomass (B) ( r 2 =0.85). The standard error of the estimate was also lower when production (P) was regressed on initial biomass (Figure 21). Since statistically speaking, B i is as good a predictor of P as B and given the relative ease with which B i can be estimated makes it a more useful predictor of production. This is further corroborated by the range and variance of P:Bi for C. globiceps (0.22 to 0.89, s 2 =0.033) which were less than the range and variance of P:B (0.46 to 2.83, s 2 =0.400). This implies that P:Bi is a better ratio than P:B to predict production from biomass data (Portt et al. 1986). Production estimates from the size-frequency method did not on the whole, compare well with the instantaneous growth rate method (Table 24). A number of things might have contributed to this discrepancy: the size-frequency method is not affected by nonsynchronous cohort development, thus it gave higher estimates of production. The instantaneous growth rate method is subject to effects of nonsynchronous development and protracted recruitment, both of which depress production estimates. Based on these differences, higher estimates by the size-frequency method could be considered more accurate. IV. G E N E R A L D I S C U S S I O N A N D C O N C L U S I O N S The objectives of this thesis have been to investigate the age and growth of Clinocottus globiceps and to evaluate production dynamics of this species in comparison with Oligocottus maculosus. The present study has demonstrated the usefulness of otoliths in providing age and growth information for Clinocottus globiceps. This work has found that C. globiceps lives at least 5 years, the largest individual measured 168 m m total length. In B . C . waters, a total length of 190 m m has been recorded from the west coast of Vancouver Island (Hart 1973), but no maximum age was reported. The lack of any deviations in sex ratio from unity as shown in Table 17, suggests that males and females have about equal lifespans at Helby Island. The population of C. globiceps at Helby Is. consisted of three major and two smaller year-classes. Some age groups are dominant during certain times of the year. For example, in August 1988 newly recruited juveniles dominated the length-frequency distributions (Figure 8). Similar changes in the proportion of recruits to adults were found in Oligocottus maculosus at a location on west coast of Vancouver Island (Craik 1978), although recruits dominated the population in Ju ly . Length-frequency distribution data for winter months are lacking. It would be instructive to conduct winter sampling so that the progression of modes over time from length-frequency distributions could be followed for the entire year. 102 103 Age determinations based on the interpretations of otolith annuli are not perfect and unavoidably contain errors, since otolith growth is known to change as fish age (Beamish 1979). Furthermore the interpretation of the zonation in otoliths is still a subjective process relying to a great extent on the experience of the observer. The otoliths of C. globiceps being thin and flat with clearly visible zones could be useful in developing new techniques of reading whole otoliths. Exploited fish populations where most ageing work has been conducted may be more prone to error because of the disrupted age classes, and hence it would be advantageous to use unexploited fish populations such as cottids to improve otolithic ageing. In the present study otolithic ageing was corroborated by a length-frequency distribution technique (MacDonald & Pitcher 1979), and the two ageing methods agreed well. Growth was confined to spring and summer and was rapid mainly during the first two years of life after which it slowed, which is probably the time when C. globiceps reach maturity. B y plotting the regression lines for body length on otolith radius, and measuring the distance from the nucleus of the otolith to a particular point on the otolith, the body length of the fish at any previous age in its life history was easily calculated. F r o m the back-calculations it was possible to gather some information on mortality through the demonstration of reversed Lee's phenomenon. C. globiceps demonstrated negative size-selective mortality (Ricker 1969) probably arising from natural mortality factors. The marine tidepool environment where the present study was undertaken provided an opportunity to explore production dynamics of two cottid species. 104 Production in the ecological sense is a synthesis of growth and biomass of each cohort of the population,. therefore links density, growth, mortality and survival rates to give an overall picture of the ecological success of a species. The present study has shown that Oligocottus maculosus was a very successful species in the study site and its ecological success was best seen in its production estimate which was over 1.5 times that of Clinocottus globiceps. A high biomass (due to high population numbers) as well as high growth rate of O. maculosus over the whole of their lifespan contributed to the high production figure. Th is study attempted a quantitative estimate of tidepool fish production, and the estimates indicate that tidepools may support substantial secondary production. However, it must be stressed that m y estimate of production in the tidepool habitat is very preliminary and probably an underestimate for the following reasons. F i rst , the younger age groups were not adequately represented in the samples (Figure 19). This discrepancy has been shown in other tidepool cottids (Grossman 1982; Yoshiyama et al. 1986) and is probably attributable to the passive dispersal of the planktonic larval phase of these species which in turn influences recruitment into tidepools. Production between ages 0 + and 1+ can be up to 95% of total production (Mathews 1971). In this study, however, production was only 11.3 and 12.2% between age 0+ and 1+ C. globiceps and O. maculosus respectively (Appendix 5). Second, the study pools were not optimal habitats for older fish. C. globiceps and O. maculosus older than age 4+ were fewer, hence their contribution to production was not substantial. Perhaps large individuals, especially of C. globiceps are found in sub tidal pools where waters are more turbulent and deeper, thus difficult to sample. 105 Production estimates are not of much value in themselves, only when they are compared one with another (as was the case in the present study) and with other factors yield a better understanding of the production processes in the ecosystem. M a n y studies cited herein treated production as a consequence of fish growth and population dynamics, in isolation from the physical (e.g. habitat dimensions and temperature) and biotic (e.g. competition and predation) environment. The present study attempted to relate production estimates to physical characteristics of tidepools (Table 23). Production of the entire tidepool fish taxocene and the interactions between species as they relate to production could be a possible avenue for future research in this area. Le Cren (1972) and Weatherley & Gil l (1987) correctly pointed out that it would be desirable to view fish production in relation to the complex of production processes of the aquatic ecosystem of which fish form an integral part. Another aspect which needs to be looked into is the annual seasonal effects on production. The present study did not take into consideration seasonal variation in production. It is worth suggesting that as one incorporates the physical parameters of the environment into production dynamics of a species, a particular season could become an important environmental variable affecting production. F o r example, reduction (or cessation) of growth in winter would undoubtedly result in little or no production of the fish. A full understanding of the production ecology of tidepool fishes requires a greater emphasis towards understanding the population dynamics of the species in question. This can come only as a result of a long-term basic research coupled with empirical experimentation. L I T E R A T U R E C I T E D Al len , K .R . 1951. The Horokiwi Stream: a study of a trout population. F ish . Bul l . N .Z . 10:238p. Al len, K .R . 1971. Relation between production and biomass. J . F i s h . Res. Board C a n . 28:1573-1581. Atkinson, C . E . 1939. Notes on the life history of the tidepool johnny (Oligocottus maculosus). Copeia 1939: 23-30. Bachman, R .A . 1984. Foraging behaviour of free-ranging wild and hatchery brown trout in a stream. Trans. A m . F i s h . Soc. 113:1-32. Bagenal , T . B . (editor) 1974. The Ageing of F i s h . Unwin Brothers, L td . , Surrey. 234p. Bagenal , T . B . and F .W. Tesch. 1978. Age and growth, p. 101-136. In T . B . Bagenal (ed.) Methods for Assessment of F i s h Production in Freshwaters. IBP Handbook No. 3. 3rd ed. Blackwell Scientific Publications, Oxford, England. Balon, E . K . 1974. F ish production of a tropical ecosystem, p.249-676. In E . K . Balon and A . G . Coche (ed.) Lake Kar iba: A Man-made Tropical Ecosystem in Central Afr ica . Monogr. Biol. 24. Dr . W . Junk Publishers, The Hague. Beamish, R . J . 1979. New information on the longevity of the Pacific Ocean perch (Sebastes alutus). J . F i s h . Res. Board Can . 36:1395-1400. 106 107 Benke, A . C . 1979. A modification of the Hynes method for estimating secondary production with particular significance for multivoltine populations. L imnol . Oceanogr. 24:168-171. Bennett, B . A . 1984. A population energy budget for Clinus superciliosus L . , with an assessment of the role of resident fish as predators in the intertidal zone. Marine Biology Letters 5:323-334. Bennett, B . A . and C . L . Griffiths. 1984. Factors affecting the distribution, abundance and diversity of rock-pool fishes on the Cape Peninsula, South Afr ica. S. A f r . J . Zool. 19:97-104. Beverton, R . J . and S . J . Holt. 1957. O n the dynamics of exploited fish populations. U . K . M in . Agric. F i s h . , F i s h . Invest. (Ser.2) 19:533p. Bilton, H . T . 1974. Effects of starvation and feeding on circulus formation on scales of young sockeye salmon of four racial origins, and of one race of young kokanee, coho, and chinook salmon, p.40-63. In T . B . Bagenal (ed.) The Ageing of F ish . Unwin Brothers, Surrey, England. Blacker, R.W. 1974. Recent advances in otolith studies, p.67-90. In F .R. Harden Jones (ed.) Sea Fisheries Research. Paul Elek (Scientific Books) L t d . , London. B M D P Statistical Software, Inc. 1983. B M D P IV: One-way analysis of variance and covariance. Regents of Un iv . of California, California. Bohlin, T . 1978. Temporal changes in the spatial distribution of juvenile sea trout Salmo trutta in a small stream. Oikos 30:114-120. Brothers, E . B . 1987. Methodological approaches to the examination of otoliths in aging studies, p.319-330. In R .C . Summerfelt and G . E . Ha l l (ed.) The Age and Growth of F ish . Iowa State University Press, A m e s , Iowa. 108 Campana , S . E . 1983a. Calcium deposition and otolith check formation during periods of stress in coho salmon, Oncorhynchus kisutch. Comp. Biochem. Physiol . 75A:215-220. Campana , S . E . 1983b. Factors affecting the production of daily growth increments in the otoliths of fishes. P h . D . Thesis, University of Brit ish Columbia, Vancouver, Canada. 150p. Chadwick, E . M . P . 1976a. A comparison of growth and abundance for tidal pool fishes in California and Brit ish Columbia. J . F ish Biol. 8:27-34. Chadwick, E . M . P . 1976b. Ecological fish production in a small Precambrian shield lake. E n v . Biol. F ish . 1:13-60. Chapman, D.W. 1965. Net production of juvenile coho salmon in three Oregon streams. T rans . A m . F i s h . Soc. 94:40-52. Chapman, D .W. 1966. Food and space as regulators of salmonid populations. A m . Nat . 100:345-357. Chapman, D .W. 1978a. Production, p.202-217. In T . B . Bagenal (ed.) Methods for Assessment of F ish Production in Freshwaters. IBP Handbook No. 3. 3rd ed. Blackwell, London. Chapman, D .W. 1978b. Production in fish populations, p.5-25. In S .D . Gerking (ed.) Ecology of Freshwater F i s h Production. Blackwell, London. Clemens, W . A . and G . V . Wilby. 1961. Fishes of the Pacific coast of Canada. F i s h . Res. Board Can . Bul l . 68:443p. Conway, G.R. , N.R. Glass, and J . C . Wilcox. 1970. Fitt ing nonlinear models to biological data by Marquardt 's algorithm. Ecology 51:503-507. Cra ig , J . F . 1985. Ag ing in fish. C a n . J . Zool. 63:1-8. Craik, G . J . S . 1978. A further investigation of the homing behavior of the intertidal cottid, Oligocottus maculosus Girard. Ph .D . Thesis, University of Brit ish Columbia, Vancouver, Canada. 278p. Cushing, D . H . 1981. Fisheries Biology: A Study in Population Dynamics. 2nd ed. The University of Wisconsin Press, Madison, WI, U S A . 295p. Degens, E . T . , W . G . Deuser, and R . L . Haedrich. 1969. Molecular structure and composition of fish otoliths. Mar . Biol. 2:105-113. DeMart in i , E . 1978. Spatial aspects of reproduction in buffalo sculpin, Enophrys bison. E n v . Biol. F i s h . 3:331-336. DeMart in i , E . and B. Patten. 1979. E g g guarding and reproductive biology of the red Irish lord, Hemilepidotus hemilepidotus (Tilesius). Syesis 12:41-55. Emlen , J . M . 1973. Ecology: A n Evolutionary Approach. Addison Wesley Publ. Co l . , Don Mil ls , Ontario, Canada. 493p. Freeman, B . J . and M . C . Freeman. 1985. Production of fishes in a subtropical blackwater ecosystem: The Okefenokee swamp. L imnol . Oceanogr. 30:686-692. Freeman, M . C , N . Neally, and G . Grossman. 1985. Aspects of the life history of the fluffy sculpin, Oligocottus snyderi. F i s h . Bul l . (U.S.) 83:645-655. G a r m a n , G . C . and T . F . Waters. 1983. Use of size-frequency (Hynes) method to estimate annual production of a stream fish population. C a n . J . F ish . Aquat. Sci . 40:2030-2034. Gerking, S .D . 1957. Evidence of aging in natural populations of fishes. Gerontologia (Basel) 1:287-305. 110 Gibson, R .N . 1967. Studies on the movements of littoral f ish. J . A n i m . Ecol . 36:215-234. Green, J . M . 1971a. High tide movements and homing behaviour of the tidepool sculpin Oligocottus maculosus . J . F i s h . Res. B d . C a n . 28:383-389. Green, J . M . 1971b. Local distribution of Oligocottus maculosus and other tidepool cottids of the west coast of Vancouver Island, Brit ish Columbia. C a n . J . Zool. 49:1111-1128. Green, J . M . 1971c. Field and laboratory activity patterns of the tide pool cottid Oligocottus maculosus. Can . J . Zool. 49:255-264. Green, J . M . 1973. Evidence for homing in the mosshead sculpin (Clinocottus globiceps) J . F ish . Res. Bd . Can . 30:129-130. Grossman, G . D . 1979. Demographic characteristics of an intertidal bay goby (Lepidogobius lepidus). E n v . Biol. F i s h . 4:207-218. Grossman, G . D . 1982. Dynamics and organization of a rocky intertidal fish assemblage: the persistence and resilience of taxocene structure. Amer . Natur . 119:611-637. Grossman, G . D . and V . deVlaming. 1984. Reproductive ecology of female Oligocottus snyderi Greeley, a Nor th Amer ican intertidal sculpin. J . F i s h Biol. 25:231-240. Ha lyk , L . C . and E . K . Balon. 1983. Structure and ecological production of the fish taxocene of a small floodplain system. Can . J . Zool. 61:2446-2464. Hamil ton, A . L . 1969. O n estimating annual production. L imnol . Oceanogr. 14:771-782. I l l Hardisty, M . W . 1961. Studies on an isolated spawning population of brook lampreys {Lampetra planeri). J . A n i m . Ecol . 30:339-355. Har t , J . L . 1973. Pacific fishes of Canada. F i s h . Res. Board C a n . Bul l . 180:740p. Hopkins, C . L . 1971. Production of fish in two small streams in the North Island of New Zealand. N.Z. J . M a r . Freshw. Res. 5:280-290. Hubbs, C . L . and K . F . Lagler. 1958. Fishes of the Great Lakes Region. Univ . of Michigan Press, A n n Arbor. 213p. Hunt , R .L . 1966. Production and angler harvest of wild brook trout in Lawrence Creek, Wisconsin. Wise. Conserv. Dept. Tech. Bul l . 35:52p. Hynes , H . B . 1961. The invertebrate fauna of a Welsh mountain stream. A r c h . Hydrobiol. 57:344-388. Hynes , H . B . 1980. A name change in the secondary production business. L imnol . Oceanogr. 25:778. Hynes , H . B . and M . J . Coleman. 1968. A simple method of assessing the annual production of stream benthos. L imnol . Oceanogr. 13:569-575. Ivlev, V . S . 1945. The biological productivity of waters. Engl ish version 1966. J . F i s h . Res. B d . Can . 23:1727-1759. Jones, R. 1958. Lee's phenomenon of "apparent change in growth rate" with particular reference to haddock and plaice, p.229-242. In Some Problems for Biological F ishery Survey and Techniques for their Solution. Int. C o m m . N.W. Atlantic F i s h . Spec. Publ. N o . l . 112 L e Cren, E . D . 1951. The length-weight relationship and seasonal cycle in gonad weight and condition in the perch (Perca fluviatilis). J . A n i m . Ecol . 20:201-219. L e Cren 1969. Estimates of Fish populations and production in small streams in England, p.269-280. In T . G . Northcote (ed.) Symposium on Salmon and Trout in Streams. H.R. Macmil lan Lectures in Fisheries, Universi ty of Brit ish Columbia, Vancouver, Canada. L e Cren, E . D . 1972. F ish production in freshwaters. Symp. Zool. Soc. Lond. 29:115-133. Lee, R . M . 1920. A review of the methods of age and growth determination in fishes by means of scales. F ishery Invest. Lond. Ser. 2, 4:1-32. L indeman, R . L . 1942. The tropho-dynamic aspect of ecology. Ecology 23:399-418. Lowenstein, O. 1971. The Labyr inth , p.207-240. In W . S . Hoar and D . J . Randall (ed.) F ish Physiology. V o l . V . Academic Press, N . Y . MacDonald, P . D . M . 1980. A F O R T R A N program for analyzing distribution mixtures. McMaster Universi ty , Dept. of Mathematical Sciences Statistical Technical Report 80-ST- l :73p. MacDonald, P . D . M . 1987. Analysis of length-frequency distributions, p.371-384. In R . C . Summerfelt and G . E . Ha l l (ed.) The Age and Growth of F i s h . The Iowa State Un iv . Press, A m e s , Iowa. MacDonald, P . D . M . and T . J . Pitcher. 1979. Age groups from size-frequency data: a versatile and efficient method of analyzing distribution mixtures. J . F i s h . Res. Bd. Can . 36:987-1001. 113 Maceina, M . J . and R.K. Betsill 1987. Verification and use of whole otoliths to age white crappie, p.267-278. In R . C . Summerfelt and G . E . Hal l (ed.) The Age and Growth of F ish . The Iowa State Univ . Press, Ames , Iowa. Mahon, R. and E . K . Balon 1977. Ecological fish production in Long Pond, a lakeshore lagoon on Long Point, Lake Er ie . E n v . Biol. F i s h . 2:261-284. M a n n , K . H . 1965. Energy transformations by a population of fish in the River Thames. J . A n i m . Ecol . 34:253-275. M a n n , R . H . K . 1971. The populations, growth and production of fish in four small streams in southern England. J . A n i m . Ecol . 40:155-190. Mathews, C P . 1970. Estimates of production with reference to general surveys. Oikos 21:129-133. Mathews, C P . 1971. Contribution of young fish to total production in the River Thames near Reading. J . F ish Biol. 3:157-180. McNew, R.W. and R .C . Summerfelt. 1978. Evaluat ion of a max imum likelihood estimator for analysis of length-frequency distributions. Trans . A m . F i s h . Soc. 107:730-736. Miller, D . J . and R .N . L e a . 1972. Guide to the coastal marine fishes of California. Calif. Dep. F ish Game, F i s h Bul l . 157:235p. Mitchell , D . F . 1953. A n analysis of stomach contents of California tidepool fishes. Amer . Midi . Nat . 49:862-871. 114 Moreau, J . 1987. Mathematical and biological expression of growth in fishes: recent trends and further developments, p.81-113. In R .C . Summerfelt and G . E . Ha l l (ed.) The Age and Growth of F i s h . The Iowa State University Press, Iowa. Mor ing, J . R . 1976. Estimates of population size for tidepool sculpins, Oligocottus maculosus, and other intertidal fishes at Trinidad B a y , Humboldt County, California. Calif. F ish Game 62:65-72. Moring, J . R . 1979. Age structure of a tidepool sculpin, Oligocottus maculosus, population in northern California. Calif. F ish Game 65:111-113. Moring, J . R . 1981. Seasonal changes in a population of the fluffy sculpin Oligocottus snyderi, from Trinidad B a y , California. Calif. F i s h Game 67:250-253. Morr is , R.W. 1960. Temperature, salinity and southern limits of three species of Pacific cottid fishes. L imnol . Oceanogr. 5:175-179. Morr is , R.W. 1961. Distribution and temperature sensitivity of some eastern Pacific cottid fishes. Physiol . Zool. 34:217-227. Mug iya , Y . and N . Watabe. 1977. Studies on fish scale formation and resorption. II. Effect of estradiol on calcium homeostasis and skeletal tissue resorption in the goldfish, Carassius auratus, and the killifish Fundulus heteroclitus. Comp. Biochem. Physiol . 57A: 197-202. Na iman , R . J . 1976. Productivity of a herbivorous pupfish population (Cyprinodon nevadensi's) in a warm desert stream. J . F ish Biol. 9:125-137. N a k a m u r a , R. 1971. Food of two cohabiting tidepool Cottidae. J . F ish . Res. Board C a n . 28:928-932. 115 N a k a m u r a , R. 1976a. Temperature and vertical distribution of two tidepool fishes (Oligocottus maculosus and O. snyderi). Copeia 1976:143-152. N a k a m u r a , R. 1976b. Experimental assessment of factors influencing microhabitat selection by the two tidepool fishes Oligocottus maculosus and O. snyderi. M a r . Biol. 37:97-104. Newman , R . M . and F . B . Mart in . 1983. Estimation of fish production rates and associated variances. Can . J . F i s h . Aquat. Sci . 40:1729-1736. O'Connel l , C P . 1953. Life history of the cabezon Scorpaenichys marmoratus (Ayres). Calif. Dep. F ish Game, F i s h Bul l . 93:76p. Ouchi , K . , J . Y a m a d a , and S. Kosaka . 1972. On the resorption of scales and associated cells in precocious male parr of the masu salmon (Oncorhynchus masou) [in Japanese, English summary] . Bul l . J a p . Soc. Scient. F i s h . 38:423-430. Pal ing, J . E . 1968. Causes of mortality, p.226-235. In W . E . Ricker (ed.) Methods for Assessment of F ish Production in Freshwaters. IBP Handbook No.3. Maxwel l Publishers, London. Panella, G . 1974. Otolith growth patterns: an aid in age determination in temperate and tropical fishes, p. 28-39. In T . B . Bagenal (ed.) The Ageing of F ish . Unwin Brothers, Surrey. Panella, G . 1980. Growth patterns in fish sagitta, p.519-560. In D . C . Rhoads and R .A . Lutz (ed.) Skeletal Growth of Aquatic Organisms: Biological Records of Environmental Change. Plenum Press, N . Y . Parker, R.R. 1963. Effects of formalin on length and weight of fishes. J . F ish . Res. Board Can . 20:1441-1455. 116 Parsons, T .R . , M . Takahashi , and B. Hargrave. 1984. Biological Oceanographic Processes. 3rd ed. Pergamon Press, Oxford. 330p. Pienaar, L . V . and J . A . Thomson. 1969. Allometric weight-length regression model. J . F i s h . Res. Board Can . 26:123-131. Popper, A . N . and S. Coombs 1980. Auditory mechanisms in teleost fishes. A m . Sci . 68:429-440. Portt, C . B . , E . K . Balon, and D . L . G . Noakes. 1986. Biomass and production of fishes in natural and channelized streams. Can . J . F i s h . Aquat . Sci . 43:1926-1934. Pot, W. , D . L . G . Noakes, M . M . Ferguson, and G . Coker. 1984. Quantitative sampling of fishes in a simple system: failure of conventional methods. Hydrobiologia 114:249-254. Potter, I.C., F . W . H . Beamish, and B . G . H . Johnson. 1974. Sex ratios and lengths of adult sea lampreys (Petromyzon marinus) from a Lake Ontario tributary. J . F ish . Res. Board C a n . 31:122-124. Prager, M . H . , S . B . Sai la, and C . W . Recksiek. 1987. F I S H P A R M : a microcomputer program for parameter estimation of nonlinear models in fishery science. Old Dominion Univ . Tech. Rep. 87-10:37p. Richkus, W . A . 1978. A quantitative study of intertidepool movement of the wooly sculpin Clinocottus analis. M a r . Biol . 49:277-284. Ricker, W . E . 1946. Production and utilization of fish populations. Ecol . Monogr. 16:373-391. 117 Ricker, W . E . 1969. Effects of size-selective mortality and sampling bias on estimates of growth, mortality, production, and yield. J . F ish . Res. Board C a n . 26:479-541. Ricker, W . E . 1973. Linear regressions in fishery research. J . F i s h . Res. B d . C a n . 30:409-434. Ricker, W . E . 1975. Handbook of computations for biological statistics of fish populations. F i s h . Res. Board C a n . Bul l . 191:382p. Ricker, W . E . 1979. Growth rates and models, p.677-743. In W . S . Hoar , D . J . Randall , and J . R . Brett (ed.) F ish Physiology. Vol . VIII. Academic Press, N .Y . Ricker, W . E . and R . E . Foerster. 1948. The computation of fish production. Bul l . B ingham Oceanogr. Coll. 11(4):173-211. S A S Institute Inc. 1988. Statistical Analysis System. Cary , North Carol ina. Shetter, D .S. 1936. Shrinkage of trout at death and on preservation. Copeia 1936:60-61. Six, L . D . and H . F . Horton 1977. Analys is of age determination methods for yellowtail rockfish, canary rockfish, and black rockfish off Oregon. F ish . Bul l . 75:405-414. Smith , C . L . 1983. Summary of round table discussion on back calculation. U . S . Dep. C o m m . , N O A A Tech. Rep. N M F S 8:45-47. Sokal, R.R. and F . J . Rohlf. 1981. Biometry: The Principles and Practice of Statistics in Biological Research. 2nd ed. W . H . Freeman & Company, N .Y . 859p. 118 Thomson, D . A . and C . E . Lehner. 1976. Resilience of a rocky fish community in a physically unstable environment. J . Exp . Mar . Biol. Ecol . 22:1-29. Thorogood, J . 1987. Age and growth rate determination of southern bluefin tuna, Thunnus maccoyii, using otolith banding. J . F i s h Biol. 30:7-14. V a n Oosten, J . 1929. The life history of the lake herring (Leucichthys artedii LeSueur) of Lake Huron, as revealed by its scales, with a critique of the scale method. Bul l . U . S . Bur . F ish . 44:265-448. Warner, R.R. 1975. The adaptive significance of sequential hermaphroditism in animals. Amer . Natur. 109:61-82. Waters, T . F . 1969. The turnover ratio in production ecology of freshwater invertebrates. Amer . Natur . 103:173-185. Waters, T . F . 1977. Secondary production in inland waters. A d v . Ecol . Res. 10:91-164. Weatherley, A . H . and H.S . Gil l . 1987. The Biology of F ish Growth. Academic Press, London. 443p. Weiss, E . F . 1969. The age and growth of the marine cottid Leptocottus armatus. Proc. Montana Acad . Sci . 29:63-71 Wells, A . W . 1986. Aspects of ecology and life history of the woolly sculpin, Clinocottus analis, from southern California. Calif. F i s h Game 72:213-226. Wil l iams, R. and B . C . Bedford 1974. The use of otoliths for age determinations, p. 114-124. In T .Bagenal (ed.) The Ageing of F i s h . U n w i n Brothers, Surrey. 119 Winberg, G . G . 1956. Rate of metabolism and food requirements of fish. F ish . Res. Board C a n . , Trans l . Ser. , No. 194. Y a m a d a , J . 1956. O n the mechanism of the appearance of the scale structure. VI . Some observations associating with the absorption of scale in the goldfish (in Japanese, Engl ish summary) . Bul l . F a c . F i s h . Hokkaido Univ . 7:202-207. Yamaguchi , M . 1975. Estimating growth parameters from growth rate data. Oecologia (Berlin) 20:321-332. Yoshiyama, R . M . 1980. Food habits of three species of rocky intertidal sculpins (Cottidae) in central California. Copeia 1980:515-525. Yoshiyama, R . M . , C . Sassaman, and R . N . L e a . 1986. Rocky intertidal fish communities of California: temporal and spatial variation. E n v . Biol. F i s h . 17:23-40. 120 A P P E N D I C E S Appendix 1. Length-frequency distribution for C. globiceps: Females. Length interval Frequency Percent 0-9.9 0 0.0 10-19.9 8 1.74 20-29.9 64 13.92 30-39.9 79 17.17 40-49.9 72 15.65 50-59.9 56 12.17 60-69.9 69 15.00 70-79.9 61 13.26 80-89.9 41 8.91 90-99.9 8 1.74 100-109.9 1 0.22 110-119.9 0 0.0 120-129.9 0 0.0 130-139.9 1 0.22 Total 460 ~ 100 121 Appendix 2. Length-frequency distribution for C . globiceps: Males. Length interval Frequency Percent 0-9.9 0 0.0 10-19.9 16 3.48 20-29.9 95 20.70 30-39.9 93 20.26 40-49.9 75 16.34 50-59.9 44 9.59 60-69.9 47 10.24 70-79.9 53 11.55 80-89.9 23 5.01 90-99.9 10 2.18 100-109.9 1 0.21 110-119.9 1 0.21 120-129.9 1 0.21 130-139.9 0 0.0 Total 459 100 Appendix 3. Length-frequency distribution for C. globiceps: Combined data. Length interval Frequency Percent 0-9.9 5 0.53 10-19.9 49 5.16 20-29.9 159 16.75 30-39.9 172 18.12 40-49.9 147 15.49 50-59.9 100 10.54 60-69.9 116 12.22 70-79.9 114 12.01 80-89.9 64 6.74 90-99.9 18 1.90 100-109.9 2 0.21 110-119.9 1 0.11 120-129.9 1 0.11 130-139.9 1 0.11 Total 949 100 123 Appendix 4. Age-length key for Helby Island Clinocottus globiceps. showing the number of f i s h per age class by sex. Length Length d i s t r i b u t i o n of age groups Interval (mm) 0+ 1+ 2+ 3+ 4+ 5+ M F M F M F M F M F M F 15-19. 9 4 2 20-24. 9 6 1 25-29. 9 11 10 30-34. 9 14 9 8 7 35-39. 9 29 24 40-44. 9 26 29 45-49. 9 15 14 13 11 50-54. 9 19 19 55-59. 9 7 12 60-64. 9 4 13 3 2 65-69. 9 1 0 12 21 70-74. 9 15 13 75-79. 9 7 4 4 9 80-84. 9 6 7 85-89. 9 1 0 3 3 90-94. 9 2 4 95-99. 9 0 0 100-104 .9 0 0 105-109 i.9 1 0 TOTAL 35 22 78 74 44 55 37 40 11 16 6 7 Appendix 5. Mean annual density (N/m2), i n i t i a l biomass (Bi, g/m2), mean biomass (B, g/m2), mean instantaneous growth rates (G), annual production (P, g/m2/year) and production to biomass r a t i o s (P:B) of C. globiceps and 0. maculosus by age groups. Age groups were designated 0+ to 5+. Biomass and production values for 1987-88 are based on corrected annual values by multiplying by 365/404. Clinocottus globiceps. Zone: Middle 0+ 1+ July 1986 2 + . to July 1987 3+ 4 + 5+ N 0. 7 5 .4 3.9 2.7 1.6 1.8 Bi 0. 3 9 .9 16.9 25.4 23.3 35.7 B 1.0 10.3 14 .1 50.9 29.1 G 1.76 0 .72 0.89 0.45 0.32 P 1.8 7.4 12.5 22.9 9.3 P:B 1.8 0.72 0.89 0.45 0.32 Lower N 2. 2 2 .4 3.4 1.7 1.3 0.7 Bi 0. 6 5 .1 12 .7 16.6 21.7 21.4 B 2.5 7.8 12 .8 19.0 14.6 P 6.8 5.7 10.0 10.0 11.2 P:B 2.72 0.73 0.78 0.53 0.77 Base N 1. 0 0 .5 _ 0.5 0.8 3.1 Bi 0. 3 0 .4 - 6.2 12.3 48.1 B 2.2 - 6.3 16.3 G 2.64 - 0.05 0.5 P 5.8 - 0.3 8.1 P:B 2.64 - 0.05 0.5 A l l zones P 2.9 6.3 7.5 11.1 9.5 Percent 7.7 16.9 20.1 29.7 25.6 Appendix 5 —Continue. Clinocottus globiceps August 1987 to October 1988 Zone: Upper 0 + 1+ 2+ 3+ 4 + 5+ N 1 .3 1. 5 0.2 0.3 0.2 -Bi 0 .5 2. 8 0.8 4.2 3.5 -5 0.9 0.6 2.5 3.2 G 1.44 0.67 1.4 0.16 P 1.6 0.4 3.5 0.5 P:B 1.44 0.67 1.4 0.16 Middle N 1 .6 1. 8 1.7 0.8 0.3 0.1 Bi 0 .9 3. 0 6.4 8.1 4.5 2.0 1 1 5.1 3.3 3.2 0.5 G 1.06 0.98 0.82 0.31 0.6 P 1.1 5 2.7 1 0.3 P :B 1.1 0.98 0.82 0.31 0.6 Lower N 8 .2 21 .7 7.4 5.4 1.6 0.6 Bi 4 .7 37 .9 34 .2 55.3 25. 5 18.4 1 7.7 31.7 35.9 41.6 13.2 G 1.14 0.88 0. 65 0.41 0.63 P 8.8 27.9 23.3 17.2 8.3 P :B 1.14 0.88 0.65 0.41 0.63 A l l zones P 3. 83 11.1 9.83 6.23 2.87 Percent 11.3 32.8 29 18.4 8.5 Appendix 5 —Continue. Oligocottus maculosus July 1986 to July 1987 Zone: Middle 0+ 1 + 2+ 3+ 4 + N 1.6 4.6 4.2 0.4 0,2 Bi 0.3 2.0 7.5 1.3 1.2 B 0.8 3.8 3.3 1 G 0.75 1.47 0.73 0. ,5 P 0.6 5.6 2.4 0. .5 P:B 0.75 1.47 0.73 0. .5 Lower N 10.2 5.3 0.5 -Bi 7.5 7.4 2.3 -I 7 2.8 G 0.69 0.93 P 4.8 2.6 P:B 0.69 0.93 Base N 0.1 0.1 - - -Bi 0.01 0.1 - - -B 0.03 G 1 P 0.03 P :B 1 A l l zones P 0.21 3.47 1.67 Percent 3.9 64.9 31.2 Appendix 5 --Continue. Oligocottus maculosus August 1987 to October 1988 Zone: Upper 0+ 1+ 2 + 3+ 4 + N 11.0 37.0 13.8 0.5 Bi 1.8 14.1 24 .0 2.3 -5 2.9 17.4 7.4 G 0.83 1.45 0.8 P 2.4 25.3 5.9 P:B 0.83 1.45 0.8 Middle N 9.3 40.4 26.4 2.1 0.2 Bi 1.1 18.6 42.3 8.1 1.4 B 5.7 25.6 19.8 1 .3 G 1.23 1.22 0.81 0. 11 P 7 31.3 16 0. 14 P:B 1.23 1.22 0.81 0. 11 Lower N 2.7 27.9 6.7 2.0 _ Bi 0.5 15.2 14 .4 7.6 -B 3.8 11.4 7.6 G 1.13 1.51 0.46 P 4.3 17.2 3.5 P:B 1.13 1.51 0.46 A l l zones P 4.57 24.6 8.47 Percent 12.2 65.4 22.5 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items