The Significance of Coloration in Fishes of 5

182 BULLETIN SOUTHERN CALIFORNIA ACADEMY OF SCIENCES VOLUME 75 1976 ORIENTATION TO DEPTH AND SEASONAL VERTICAL TEMPERATURES 183 (August 1974-July 1975) at King Harbor, Cali- was contributed by Occidental College Research Fund, Radovich, J. 1961. Relationships of some marine Terry, C. B. 1975. The role of depth, tenveoSttre, fornia, using diver transects. We also recognize the support of the City of Redondo organisms of the northeast Pacific to water tem- diet and feeding periodicity i - ition- 2. Beach, Harbor Department and particularly Harbor- peratures, particularly during 1957 through 1959. ing among the surfp otocidae) The bathymetric distributions of two of the Dept. five species master Donald Campeau. California Dept. Fish and Game, Fish Bull., 112: in King Har .A. Thesis, (M. minimus and E. jacksoni) differed lege. significantly from the remaining three with M. 1-62. Biolo minimus occupying the shallow intertidal and sub- . P. III, and C. R. Feldmeth. 1971. tidal fringe, E. jack.soni the middle subtidal, and LITERATURE CITRO Simpson, G. G., A. Roe, and R. C. Lewontin. 1960 Quantitative Zoology. Harcourt, Brace, an quential mortality of the fish fauna impounded the R. vacca, H. earn, and P. furcatus the deeper Alevizon, W. S. 1975. Spatial 'erlap and competi- New York, 440 pp. in construction of a marina at Dana Point, rocks and rock-sand interface. tion in congeneric California. California Fish and Game, 57:167- 3. Sur hey (Embiotocidae) Depth distribution varied seasonally (except off Santa Barbara, Cal nia. Copeia, 1975(2): 1952. A revisi y Embio- 176. for Al. naninuts) Tarp, F. H. with adults of each species mov- 352-356. tocidae (the surfperc • orma Dept. Fish ing deeper in the late summer and fall and shal- and Game, Fish tre1-99. Accepted for publication May 1,1976. ý t lower in the winter and spring. Bray, R. N., and A. Wing. 1975. Food, activ- 4. The fall descent corresponded to periods of ity, and habij of three "picker-type" micro. increased surface water temperature and deepening carnivorous es in the kelp forests off Santa of the thermocline. The three deeper species Barbara, •. fornia. Fish. Bull., 73(4):815-829. chose the coolest available water during this period. Briggs, J. 1974. Marine zoogeography. Mc- THE SIGNIFICANCE OF COLORATION IN FISHES OF 5. Seasonal abundance decreased during the fall Graf II Book Co., 475 pp. HYPSOBLENNIUS and may represent some emigration in response THE GENUS GILL to warm thermal conditions. Gra J. B. 1970. Temperature sensitivity of 6. Each species appears to occupy a preferred o species of intertidal fishes. Copeia, 1970: GEORGE S. LOSEY, JR.' thermal range. This range is especially obvious 19-56. in the deeper water species where the distributions AnsmAcr: The social behavior and associated colorations of five Pacific species of tend to center at temperatures at or below I6°C edgepeth, J. W. (Ed.). 1957. Marine Biogeog- Hypsoblennius were observed under field and laboratory conditions. Behavior patterns were Ernbiatoca jacksoni appears to be more thermal raphy. In Treatise on ?Amine Ecology and Falco- classified as to their functional relationship to types of behavior such as aggression and tolerant. Micrometrus ',animus appears to c ecology, Vol. I, Ecology. Geol. Soc. Amer. courtship. The occurrence of various color patterns was then tested for correlation to these select the warmest water available or more Mem., 67:1-1296. types of behavior. This analysis indicates that coloration is of potential value as a communi- ably to tolerate whatever temperature is a able cation signal for differentiatng between submission, aggression and courtship. Coloration was in its preferred habitat. Hubbs, C. L. 1948. Changes on the fish fauna of also correlated with habitat type and showed striking differences between sympatric species. 7. western North America correlated with changes Juveniles prefer warmer water t 1 adults. Young are born from spring to late - imer and in ocean temperatures. J. Mar. Res., 7:459-482. generally orient to shallower, warm ater. Sub- adults are less depth or temperatur stricted than . 1960. The marine vertebrates of the outer Species of the genus Hypsoblennius inhabit tidal Chaetodon Iunula. In this paper, I will describe either adults or juveniles. coast. In Biogeography of Baja California and pools, subtidal reefs and algae beds along the the potential signal value of coloration in five Adjacent Seas, Part II. Syst. Zool., 9:134-148. 8. Temperature orientation ears to play a Pacific coast of North, Central and South America. species of Hypsoblennius and explore the relation- role in controlling depth di ution and, there- Like many of the Blenniidae, they are territorial ships of these colorations to their taxonomic affin- 1974. Review of Marine Zoogeography by ities and their behavioral ecology. fore. influences spatial ut lion by these surf- and show frequent social interactions. The be- J. C. Briggs. Copeia, I974(4):1002-1005. been thoroughly perches. havior of some blenniid fishes has Wickler, 1957, 1961, 1963, 1965; deecribed (e.g., METHODS Morris, R. W. 1961. Distributions and temperature Wilson, 1969; Thompson and Bennet, 1969; Gib- sensitivity of sonic eastern Pacific cottid fishes. ACKN EDGMENTS Physiol. ?Ail., 34:217-227. son, 1968; Phillips, 1971; Fishelson, 1963; Robins from et al., 1959; Tavolga, 1960) but fishes of the genus All of the data presented below were drawn have largely escaped attention laboratory observations. Field observations made The following Oc• ntal College students participated Nakamura, R. 1976. Temperature and the vertical Hypsoblennius in California, Mexico, and South America pro- in sonic aspects this study: John Hypsoblennius species have the Helly, Jo Ellen distribution of two tidepool fishes (Oligocottat (Losey, 1968). vided a subjective confirmation of the laboratory Hose. Shuichi hikawa, Susan Jelley, Marc Lytle, ability to display a variety of color patterns and inacrdosas, G. snyderi). Copeia, 1976:143-152. 1, 300 I, and 550 1 Howard Mc erry, Laurie Post, Steve Subber, and can alter many features of their coloration in a results. Fish were held in 150 Bill Westpt We are indebted to Karl Ehrlich and matter of seconds. The signal value of coloration aquariums at the Scripps Institution of Ocean- Myron II of Occidental College for consultation Norris, K. S. 1963. The functions of temperature (e.g., ography and supplied with running sea water. in the ecology of the percoid fish Girella nigricans changes in fishes is well known in blenniids were used only for the tropical and dcsi of the thermal preference gradient and 1963) and other fishes (e.g., Aquarium heaters statistic %time. We also thank Mark Terry for his (Ayres). Ecol. Monogr., 33:23-62. Wickler, 19572 species. Fish were held in densities similar to Barlow, 1974; Baerends and Bacrends van-Roon, help with many aspects of this study. We gratefully those encounterd in the field and offered a variety acknowledge the support of Southern California Quast, J. C. 1968. Fish fauna of the rocky inshore 1950). Hamilton and Peterman (1971) have Edison's Research and Development Grant to the zone. In North, W. J., and C. L. Hobbs. Utiliza- drawn attention to the compromise between Junior author and especially Kevin Muench, who tion of kelp-bed resomces in Southern California. countershading as a camouflage coloration and 'Dept. 7.00logy and Hawaii Institute of Marine supetNities this research program. Additional support California Fish and Game, Fish Bull. 139:35-56. contrasting colorations for communication in Biology, Univ. Hawaii, Honolulu, Hawaii 96744. •

184 BULLETIN SOUTHERN CALIFORNIA ACADEMY OF SCIENCES VOLUME 75 1976 COLORATION IN HYPSOBLENNIUS 185 'ratio: I. The species of Hypsoblettnius studied with 100 notes on zoogeography and ecology. Habitat sympatry is indicated by connecting vertical bars. an

11. jenkinsi (Jordan and Everman)—California and northel a Baja California. Found in barnacle tests and bore holes on shallow, subtidal reefs with dis- --•- ruptive background. 50 [ //. genti/i, (Girard I—California and north-central Baja California. Found in a variety of habitats front grass beds to sea walls and buoys front barely subtidal depths to at least 30 m at the southein extent of its range. It has limited habitat 2 3 0X overlap with //. lentinsL RIPENESS INDEX

FLEE TENDENCY ... II. brevip' 's (Ounther)—Southern Baja Califor- Figure 2. The distribution of the male action patterns nia to Central America and northern Peru. It over the ripeness of the male, expressed as cumulative 5 10 20 30 40 >40 inhabits barnacle tests on shallow, wave-washed percent. DISTANCE TO (CM) rocks in barren habitats. [ II. robustus (Hildebrand).—Central Peru. Habitat Figure 3. The distribution of the male action patterns identical to /I. jenkin. over the proximity of a courting female to the male, Figure I. The agonistic action patterns of Hypso- si For sexual behavior, the females exhibited little expressed as cumulative percent. blennins spp. arranged as to functional motivational II. sordidus (llennet)—Central Peru to central display other than coloration change and a sub- context. Chile. It inhabits the same areas as 11. robsestus missive posture so attention was focussed on male See text for explanation. as well as the habitat type of 11. brevipinnis. behavior. Sexual ripeness of males was indicated spawned repeatedly with her partner and ignored the courtship efforts of neighboring males. The by the occurrence of nest preparation behavior: of substrata: gravel bottom with cobble and Not ripe, no rubbing, spitting or circle-spitting; typical situation, however, is for the female to 1) approach a courting male, deposit on the order of broken shells, opaque plastic tubes, and Balanus of attacking and fleeing. Morris (1958) intro- rub and spit occur but circle-spitting is lacking; 2) 1000 eggs in his refuge and leave. Males continue barnacle tests resting on the bottom and mounted duced a diagrammatic hypothesis of the causation 3) rub, spit and circle-spit occur but circle-spit on vertical, artificial rock surfaces. Fish were of agonistic behavior based on an orthogonal inter- is less common than spit; 4) same as No. 2 but to court so long as they have space in their refuge for egg deposition. Courting males release a fed an overabundance of cut frozen shrimp and action between the tendencies to flee and attack. circle-spit is more common than spit. pheromone which is an attractant to other sexually living invertebrates such as brine shrimp, Emerita This has proven to be a useful tool for initial The frequency of courtship action patterns were ripe males (Loscy, 1969). They are all carnivores sand crabs and assorted benthic amphipods in order studies but Hindc (1966) has drawn attention to then plotted over both the ripeness of the male to encourage normal foraging behavior. Fish were its limitations as an explanation of causation. I (Fig. 2) and the proximity to an approaching fe- and, in the field, may wander a meter or more from held under reversed photoperiod and observed used the framework of the "attack/flee" (A/F) male in order to indicate relative intensity of their refuge while foraging. Feeding patterns in an aquarium may be highly variable depending on from a darkened chamber through slits in a blind. hypothesis in order to indicate the functional ex- courtship movements. Tape recorded notes were made of essentially all pression of agonistic action patterns according to the feeding regimen. The behavior groups and of the activities of groups of fish during two to whether they were followed by attack, flee or patterns that were observed and used in the five-hour observation periods. neutral activities. Sexually ripe individuals were BEHAVIORAL DESCRIPTION analysis of coloration are briefly described below Two methods were used in order to approximate omitted from this analysis. Several values were with an indication as to which species exhibit them the motivational state of an individual as evidenced calculated for each pattern: I) Percent followed The five species of Hypsoblennius treated below (Table 2). by its agonistic and sexual behavior. These Exploration.—Locomotion about the habitat by attack (move0 toward opponent or physical show a wide range of geographic distribution, dif- methods are only intended to indicate the junc- combat) times 90 X 10 2 described the angle from fering degrees of sympatry with other Hypso- without overt social interaction. tional expression of the motivational status of a the flee axis for the attack vector (48° for head-up blenttcus species and a diversity of habitat types Submission.—Any action patterns with a high "signal sender:" that is, information that would be flee to attack ratio (Fig. 1): 1) 13ackwalk, the display, Fig. I); 2) percent followed by retreating0 (Table 1). Their general behavior is, however, of major importance to another individual, the (move away from opponent) times 90 X 10-2 quite similar. All individuals are territorial except, fish moves posteriorly propelled by moving the signal receiver, in forming an impression of the described the angle from the attack axis for the perhaps, the females of H. gentilis. All except paired fins against the substrate with the head held the behavioral tendencies of the signal sender. flee vector (12° for head-up display, Fig. 1); 3) the possibly H. genii/is show only limited movement down (Fig. 413); 2) Flee, the actor swims away In other words, we are concerned with analysis at percent preceded by an aggressive act by the op- about the habitat and may occupy the same from its opponent; 3) Side-roll, the fish rolls on the level of communication and prediction of an ponent (move toward, physical combat, head-up, territory for years (Stephens, et al., 1970). Hypso- its side, dorsum toward the opponent with the individual's subsequent behavior rather than at tail-up or bile) described the distance from the blennius pen/ills appears to he far more vagile. dorsal fin erect but inclined laterally up to 90° the level of a causal analysis of motivation. A origin (33% for head-up display, Fig. 1). Pelagic prejuveniles settle out of the plankton toward the substrate and remains motionless (Fig. signal receiver can profitably form an impression These three measures were used to describe an and immediately show simple agonistic behavior 4A). This is an apparent appeasement behavior of the probability that a signal sender will attack, area in the A/F framework for each action pat- and begin territorial defense within several days. that arrests the opponent's aggression. In only one flee or behave sexually without differentiating tern (Fig. I). The area reflects both the proba- They threaten nearly all congeneric intruders. of 34 occurrences was it answered by an attack. between the many possible causal factors of social bility of a following attack or retreat and the Both males and females are usually polygamous In H. sirloins, however, it appears to serve as a behavior. intensity of the behavior as indicated by the pre- but pair bonding has been observed on occasion. threat display. The first method was used to differentiate ago- ceding behavior of the opponent which probably One female H. jenkinsi was observed within the Mild Aggression.—Movement toward another nistic behavior according to the relative probability contributed to the causation of the display. territory of a male for nine days in the field. She individual without any overt agonistic display. 186 BULLETIN SOUTI1ERN CALIFORNIA ACADEMY OF SCIENCES VOLUME 75 1976 COLORATION IN HYPSOBLENNIUS 187 TABLE 2. The occurrence of action patterns in spe- A cies of Hypsoblennius. 1, indicates present but rare; 2, indicates that it is common; Dash indicates that it was not observed. No courtship was seen in sordidus.

Act ion brest- GALORE* Pattern leakinst salaams sardhlas plants ernijhts Backwalk 2 2 2 2 2 Flee 2 2 2 2 2 circle Side Roll 2 2 2 2 2 Head-up 2 2 1 2 2 cave Nod 1 1 1 2 Tail-up 1 2 1 Run 2 2 2 2 1 Carousel 2 2 2 2 1 courting Tail-beat 1 1 1 1 female Open-mouth 1 1 2 1 2 Bite 2 2 2 2 2 Head-jerk 2 2 7 2 2 Quiver 2 2 2 2 Circle-cave 1 2 1 Rear 2 Rub 2 2 2 2 Spit 2 2 2 2 Circle-spit 2 Number of individuals circle sampled 46 It 2 11 19 spit Figure 4. Display postures of Hypsobiennius spp. A) side-roll; B) backwalk; C) low intensity head-up; D) high intensity head-up; low intensity tail-up; E) F) (Fig. 4E and F): 2) Run, this is a reciprocal high intensity tail-up. "fighting" movement in which the opponents alter- nate between charging and withdrawing a few cm in quick succession. The body is in a high intensity Aggression.—Any action pattern with a low flee head-up or a tail-up posture; 3) Carousel, the fish to attack ratio and low intensity: I) Fled up, pivot around each other in head to tail position. the body is arched to raise the head which is This type of combat is usually fast and short lived enlarged by expansion of the oro-branchial in Hypsohlennius species with little tail-beating or chamber. The dorsal fin is usually erected. The other ritualized combat actions. It is a strong actor faces the opponent at low threat intensities attack but there was insufficient data for A/F (Fig. 4C) but turns the head 15° to 45° to one analysis; 4) Tail-beat, a rare behavior consisting of side at higher intensities (Fig. 4D). It is usually an caudal beats toward the opponent's flank; 5) Bite, initial threat or challenge; 2) Nod, the head is biting opponent. slowly raised and lowered in repeated resumption Male courtship L—Actions that were common of the head-up display. Between the species, it in response to a distant female in fully ripened ranges from a rare display of irregular form to a males or to close females in partially ripened common threat behavior with highly ritualized nudes: I) Head-up without quivering or head form in H. breripinnis; 3) Open mouth, the mouth jerking (see below). Figure 5. Some display movements of Hypsoblennius spp. is opened, usually to the maximum extent, and Male courtship H.—Actions common in re- held open while facing the opponent (Fig. 4F). sponse to a female that is close to a fully ripened Strong Aggression.—Actions having a low flee male. These are usually associated with a head-up tremely high frequency and low amplitude. It out spawning (Fig. 5); 4) Rearing, during the to attack ratio and high intensity: I) Tail-up, the or tail-up posture: 1) Head-jerk, the head is occurs during courtship and throughout spawning; head-up display, the head is raised even further body is arched with the head held next to the sub- twitched rapidly toward the female, frequently ac- 3) Circle-cave; the male circles its refuge with by swimming movements of the pectoral fins. strate, the tail is raised and the anal and soft dorsal companied by a momentary opening of the jaws. jerky swimming movements. In some species, it is Female courtship I.—Approaching the male's fin erected. The head-up display slowly changes to It frequently precedes a quiver bout; 2) Quiver, rare and of highly variable form while in others it refuge without agonistic display but not contacting a tail-tip as the intensity of an encounter builds lateral flexures of the body are made at ex- is common during intense courtship and through- or entering the refuge. 188 BULLETIN SOUTHERN CALIFORNIA ACADEMY OF SCIENCES VOLUME 75 1976 COLORATION IN HYPSOBLENNIUS 189 H. robustus and ienkinsi H. sordidus H. jenkinsi and H. robustus H. sordidus .1% • ■ *''-'' ..c; 1- - •:,..,,.'.. • ■._ _ ,i • t . 4:* • .1i , z - •- .. ..e. r•':

'','s . .•. ....7..;2;i.,,==2-s saddle --'..- _ . = ' saddle ' ------,-- - /

!. - i - C' , 0 • 0 'AS lit;. ,, ' ''' . ‘ ''' • . . I' , • ' 4... ': T A : B C D E F 4. ' ?Jr .; . A ..■ ' mottle. - •„,„, - H. gentilis H. brevipinnis mottle '---:■‘-=.-"'--. -' , ..- ritP ' • • • .. , ,., ---.„ .- Stripe ...... • --- bland 7. chromatophore Hypsoble tttt ittt spp. melanophores 4 Figure The major areas for Areas are defined by groups of -././../74,0 - _ that act more or less simultaneously to produce changes in pattern. H. genttlus :. .1:.... • A ' . H.—As C D, brevipinnis adults, a s _ ,.....„... .-a,N ,...■:.- FF':1'' Female courtship above but contacting and if present. In H. bland or entering the male's refuge (Fig. 5). true saddled coloration is never seen but the -..- Male nesting behavior—Behavior usually seen probable homology of these colorations is sup- saddle only in males: I) Rub, rubbing the walls of the ported by saddle-shaped pigment areas found in H. brevipinnis , refuge with the anal fin pads; 2) Spit, taking gravel prejuveniles. This is the "resting" color of all of or detritus into the mouth and spitting it out. This the species except H. brevipinnis; 2) Mottle is a i• .. , , li ir g ; .. display has a range of intensities (Fig. 5) and disruptive pattern seen in all species that lacks • ' Ofbtt '. . r- t i grades into; 3) Circle-spit, similar to spit except prominent saddles. This pattern was common in H. brevipinnis; 3) •• I: • , - ' the male swims in a circle over the refuge before Bland is a homogeneous colora- ,. , ‘,... - sadd lestripe mottle spilling (Fig. 5). This appears to be appetitive to tion that lacks mottling and is seen in all species; . , courtship behavior and rarely if ever occurs when 4) Reticulate pattern extends over the head and is 410 f' gentilis and is females are visible, it has not been observed in seen only in sexually ripe male H. I 5) . field observations. acquired and lost only over a period of days; ,.• ,._...... Spawning.—The female deposits eggs on the Stripe is a mid-lateral darkening that lacks promi- t . . refuge wall. The male may quiver, head-jerk, nent saddles or mottling. It is the "resting" color bland female stripe head-up and tail-up inside or outside of the refuge of H. brevipinnis and is also prominent in H. sordidus when it is over a bland background. It ,.'.••••., '., ad/or remain inactive outside of the refuge. ,} 1 Sperm are deposited on the wall of the refuge dur- is somewhat evident in H. gentilis in similar situ- . 4 -- • r.. , • I Itt"1.4 AI 1.-P-a..0_301,-1111Pa'.u.c.11 ing courtship that are sufficient for fertilization of ations; 6) Darkness is a variable of all of these ,. , N Iff A 7!'h:.1.1b,..0-... . . 4 the eggs. patterns which may range from a light tan to * - - 7) lntergrades such as sad- r oculate , ,..e'r stripe/blandel - - almost totally black; --.., dle/bland may be found in most species where only COLORATION DESCRIPTIONS about half of the body is altered. In all species et. , ,....• ■ , except H. sordidus the anterior saddles or stripe ntergrades ,.. 4,5: V The major color patterns and some i are the first to be replaced by bland or reticulate " - J are depicted in figure 6. Their method of forma- coloration. In H. sordidus the opposite is found: I 2 3 4 tion can be compared between species by referring Ventral View of Chin bland ‘j--' the posterior saddles and stripe are first to fade. to fields of chromatophores that act as more or Head pattern.—Aside from general mottling, less separate units to produce various features of several special markings occur: 1) Chin bars are Figure 6. The major color patterns of Hypsoblennius species. The views of the chin of H. gentilis range from 7). formed in all species by darkening pigment area normal for all of the species (I.) to full reproductive ripeness (4.). the coloration (Fig. Trunk pattern.-1) Saddle is a disruptive mot- A; 2) Red chin is found only in sexually ripe male tled coloration formed by darkening pigment areas H. gentilis in which the chin bars are slowly re- ../1.11171111.1.

/90 BULLETIN SOUTHERN CALIFORNIA ACADEMY OF SCIENCES VOLUME 75 1976 COLORATION IN HYI'SOBLENNI US 191

H. jenkinsi H. robustus COLOR PATTERN H. gentilis 100 "6-8'"'ki egg. - _ 100 s9b e - - 44, • • COLOR PATTERN

1 100 R ct. . expbr. \\acog 50 50 os9. d 50 0 — saddle saddle'/bland bland saddle saddle/bland bland BODY DARKNESS c, 2 9 et. 2 100 100 0 saddle saddle/bland bland 50 50 BODY DARKNESS (-) 100 2 9 ctsP- >- 0 o light dark light dark LU 9 c t 2 SPOT DARKNESS 09 100 100 Lit

s, ogg. 50 50 0 /occ'• light 1/2 dark dark 0 0 SPOT DARKNESS light 1/2 dark dark rght dark 51 100 Film,. 8. The occurrence of color patterns in various motivational states in //yp.mbienm'us jenkinxi and 11. mild agg. explor. robust'', Abbreviations are sub., submission; explor., exploration; mild agg., mild aggression; agg., aggression; st. agg., strong aggression; ,t ct., male courtship; 9 Cl.. female courtship. Due to an error in data collection, the half-dark category is lacking for 1/. robust/is. >- u 50

placed by a bright red to red/ black field traversed Fit; patiern.—The fins generally track the trunk LIJ by a few delicate silver streaks. This chin colora- coloration and may lie transparent in bland-light tion appears and fades with the reticulate pattern: or stripe colorations. The dorsal spot is a small 3 I Upper postorbital spot appears in all species dark area in many species that does not show any ce 0 by darkening pigment area 13. It ranges from a differentiated function except in female H. light 1/2 dark dark shade that matches the rest of the head to an genii/is in which it was large and ranged from intense blue/black shade; 4) Lower postorbital D. SPOT DARKNESS white to a metallic blue/black. spot ;ippears only in //. brevipinnis in area F and 100 ranges from a solid black to intense yellow out- lined by paired black lines. Its color changes ap- ANALYSIS OF COLORATION peared to occur in less than one second. A hint of this spot occurs in IL surdidtts but it is largely un- The frequency at which each element of the colora- differentiated. tion occurred during Ow various categories of be-

4 FiVire 9. The occurrence of color patterns in various motivational states in 11 yosobtrunius geoid's. For alibi eviations see figure 8.

4;4k t 192 BULLETIN SOUTIIERN CALIFORNIA ACADEMY OF SCIENCES VOLUME 75 1976 COLORATION IN HYPSOBLENNIUS 193 TAM 3. The relative probability of difference between the frequency distribution curves that describe the TABLE 4. The relative probability of difference be- H. sordidus coloration in various motivational states (see Fig. 8). Due to multiple testing, the actual p level is 32 X p, tween the frequency distribution curves that describe COLOR PATTERN so these probabilities should be used only to indicate the relative divergence of the curves. Probabilities cal- the coloration in various motivational states (Fig. 8). 100 culated by the Kolmogorov-Smirnov method. Abbreviations are: submis. = submission, agg. = aggression, Due to multiple testing, the actual p level is 14 X P, ctsp. = courtship. Bold face type indicates broader divergence. so these probabilities should be used only to indicate the relative divergence of the curves. Abbreviations HYPSOBLENNIUS JENKINS! are as in Table 3. 50 eler MOTIVATIONAL STATE OF ANIMAL behavioral HYPSOBLENNIUS ROBUSTUS category explore submis. mild agg. agg. strong agg. ctsp. I cr/9 MOTIVATIONAL STATE OF ANIMAL 0 Probability of difference between frequency behavioral stripe mot. saddle bland category submit tnild agg. agg. strong agg. distributions of trunk color pattern %) BODY DARKNESS 100 submis. 0.2 Probability of difference between frequency mild agg. 0.2 0.2 distributions of trunk color pattern -- agg. 0.2 0.05 0.1 submis. e;plar• strong agg. 0.02 0.01 0.01 0.2 mild agg. 0.2 50 ctsp. I. a 0.01 0.01 0.01 0.2 0.2 agg. 0.05 0.2 ctsp. I . 9 0.2 0.02 0.02 0.2 0.2 strong agg. 0.01 0.2 0.2 (CUMULATIVE ctsp. 2, 6 0.01 0.01 0.01 0.2 0.2 0.2 ctsp. d 0.01 0.2 0.2 0.2 ctsp. 2, 9 0.05 0.05 0.05 0.2 0.2 0.2 ctsp. 9 0.01 0.2 0.2 0.2 light dark Probability of difference between frequency Probability of difference between frequency - SPOT DARKNESS distributions of body darkness distributions of body darkness 100 submis. 0.2 captor. mild agg. 0.2 FREQUENCY mild agg. 0.1 0.01 sub, mild egg. agg. 0.01 0.01 0.2 agg. 0.2 0.2 •••••• strong age. 0.01 strong agg. 0.2 0.2 0.2 50 0.01 0.2 0.2 0.2 ctsp. I. 8 0.01 0.01 0.2 0.2 0.2 ctsp. 1 0.2 0.2 0.2 wag' ctsp. I. 9 0.2 0.2 0.01 0.01 0.1 ctsp. 9 0.2 0.05 0.01 0.01 ctsp. 2. 8 0.01 0.01 0.2 0.2 0.2 0.2 0 ctsp. 2. 9 0.2 0.2 0.01 0.01 0.05 0.2 light dark or dark stripe, mottle or more rarely saddle with Probability of difference between frequency distributions a light upper postorbital spot; 2) Aggression-strong Figure 10. The occurrence of color patterns in various of darkness of postorbital spot aggression, light saddle/bland or bland with dark motivational states in Hypsoblennius sordidus. For submis. 0.2 abbreviations see figure 8. mild agg. 0.01 0.01 postorbital spot. Note the reversal of the dark- age. 0.01 0.01 0.2 ness of the trunk! strong agg. 0.01 0.01 0.2 0.2 H. brevipinnis.-1) Exploration-submission, light postorbital spot. Note that a parental male cisp. I. d 0.01 0.01 0.2 0.2 0.2 light mottle, mottle/stripe or stripe with light may be nonreceptive to females if its refuge is full ctsp. I. 9 0.2 0.2 0.01 0.01 0.01 upper and lower postorbital spot; 2) Mild ag- of eggs. ctsp. 2. 6 0.01 0.01 0.2 0.2 0.2 0.2 gression, same as exploration except a darkened ctsp. 2. 9 0.2 0.2 0.01 0.01 0.01 0.2 upper postorbital spot and light to dark trunk; 3) Aggressivatrong aggression, dark stripe or stripe/ DISCUSSION bland with dark upper postorbital spot. The lower havior was obviously nonrandom. In figures 8, 9, dark upper postorbital spot; 2) Exploration-sub- frostorbital spot darkens only during intense ag- Cryptic Colorations.-Species of Hypsoblennius 10, and II the color patterns and common inter- mission, light saddled trunk with light upper post- gression; 4) Male courtship, dark stripe/bland show three types of coloration that appear to serve grades are arranged on the abscissa in order of orbital spot; 3) Female courtship, light saddle/ with both postorbital spots dark; 5) Female court- the function of crypticity: saddle, mottle and decreasing erypticity, which was subjectively deter- bland to bland trunk with light postorbital spot. ship, light mottle with postorbital spots dark. stripe. All of these colorations may he adopted by mined and coincides with the order in which II. gentilis.-I) All aggression-all male court- Subjective examination of the families of cumu- isolated fish and in nonsocial contexts but bland coloration changes occurred. The cumulative per- ship, dark saddle/bland, bland or reticulate with lative percent curves indicates that H. gentilis has was only seen during social interactions. Hypso- cent of the occurrence of each pattern is plotted saddle and bland intergrades and dark upper the least divergence between curves whereas H. blennius jenkinsi and H. robustus are nearly always for each category of behavior. The probability of postorbital spot. Note that reticulate can only brevipinnis has the most clear separation. found in saddle coloration and are restricted to difference between curves is estimated by their serve to indicate an increased probability of court- Some data are available for the response of male disruptive habitats. Hypsoblennius gentilis and degree of divergence (Tables 3-6). Examination ship since it is a long term reflection of gonadal H. robustus to approaching females in different particularly H. sordidus show both saddle, stripe of the figures indicates that coloration alone is ripeness; 2) Exploration-submission, as in H. colorations. The scanty observations are not and mottle coloration. Saddle is usually seen when sufficient to distinguish the following behavioral jenkinsi but dorsal spot is dark. amenable to statistical analysis but examination of they are found over a disruptive background while categories with a high degree of accuracy: H. sordidus.-Courtship was not observed but table 7 indicates its nonrandom nature. Males mottle and stripe appear to be more common over jenkinsi and H. robustus.-1) All aggression- the following categories could be distinguished: which had courted did not attack a female in barren backgrounds. These species have appar- all male courtship, dark saddle/bland trunk with 1) Exploration-submission-mild aggression, light courtship coloration with a light bland trunk and ently analogous methods of forming the stripe

P, t, 194 BULLETIN SOUTHERN CALIFORNIA ACADEMY OF SCIENCES VOLUME 75 1976 COLORATION IN HYPSOBLENNIUS /95 Tom 5. The relative probability of difference between the freqttency distribution curves that describe thc H. brevipinnis Tome 6. The relative probability of difference be- coloration in various motivational states (see Fig. 9). Refer to Table 3 for explanation. tween the frequency distribution curves that describe COLOR PATTERN the coloration in various motivational states (see Fig. HYPSOBLENNIUS GENTILIS 100 9). Refer to Table 3 for explanation. MOTIVATIONAI. STATE OF ANIMAL HYPSOBLENNIUS BREVIPINNIS behavioral category explore submis. agg. age. strong agg. ctsp. I dv9 MOTIVATIONAL STATE OF ANIMAL Probability of difference between frequency behavioral mild strong distributions of trunk color pattern category explore submis. egg. agg. BIlg• submis. 0.2 Probability of difference between frequency mild agg. 0.2 0.2 mottle stripe bland distributions of trunk color pattern agg. 0.05 0.2 0.2 strong agg. 0.01 0.2 0.2 0.2 HEAD DARKNESS submis. 0.2 mild agg. 0.2 0.2 ctsp. I. 1 0.01 0.05 0.05 0.2 0.2 100 sub. & 9 cusp. ctsp. I. 9 0.01 0.05 0.05 0.2 0.2 088. 0.2 0.01 0.01 strong agg. 0.01 0.01 0.01 0.2 ctsp. 2, 1 0.01 0.1 0.1 0.2 0.2 0.2 0. 0 0.01 0.2 ctsp. 2. 9 0.05 0.2 0.2 0.2 0.2 0.2 ctsp., 0.01 0.01 0.01 50- ctsp., 9 0.2 0.2 0.2 0.05 0.05 Probability of difference between frequency distributions of body darkness Probability of difference between frequency 17- distributions of bead darkness submis. 0.2 <- 0 mild agg. 0.2 0.2 5' I ght dc,rk mild agg. 0.2 agg. 0.02 0.02 0.02 UPPER SPOT agg. 0.05 0.2 strong agg. 0.2 0.2 0.2 0.2 100 strong agg. 0.05 0.2 0.2 ctsp. I. 8 0.2 0.2 0.2 0.2 0.2 ctsp., 8 0.05 0.2 0.2 0.2 ctsp. I. 9 0.1 0.2 0.2 0.01 0.2 ctsp., 9 insufficient data cisp. 2. 8 0.2 0.02 0.2 0.2 0.2 0.2 >- ctsp. 2. 9 0.2 0.2 0.2 0.05 0.2 0.2 u 50 Probability of difference between frequency distributions of upper postorbital spot darkness Probability of difference between frequency submis. 0.2 distributions of dorsal spot darkness ce 0 mild agg. 0.1 0.1 submis. 0.2 ight dark agg, 0.01 0.05 0.2 mild agg. 0.2 0.2 strong agg. 0.01 0.05 0.2 0.2 agg. 0.1 0.2 0.2 100 LOWER SPOT ctsp., 8 0.01 0.05 0.2 0.2 0.2 strong agg. 0.2 0.2 0.2 0.2 ctsp., 9 insufficient data ctsp. I. 9 0.05 0.2 0.1 0.2 0.2 ctsp. 2. 9 0.01 0.2 0.1 0.2 0.2 0.2 Probability of difference between frequency distribu- 50--' tions of lower postorbital spot darkness submis. 0.2 mild agg. 0.2 0.2 pattern. Hypsohlennins sordidus includes the low- stripes. Hypsoblennius species, however, show () agg. 0.2 0.2 0.2 er segment of each saddle in the stripe while in both types of patterns in apparent response to their light dark strong agg. 0.2 0.2 0.2 0.1 genii/is, the stripe is clearly separated from the cryptic value in different habitat types. ctsp., 0.01 0.01 0.01 0.01 0.2 pigment areas that form the saddles. Both species Hypsoblennius brevipinnis appears to be addi- Filure II. The occurrence of color patterns in various ctsp., 9 insufficient data may he found in disruptive and barren habitats in tionally adapted to avoiding predation in its barren motivational states in Hypsoblennius breripinnis. For the field. Hypsoblennius brevipinnis was only habitat through not showing displays that take it abbreviations see figure 8. found in barren habitats and has stripe and, less far from the substratum such as circle-spit and my ability to distinguish more discrete types of commonly, mottle as its cryptic coloration. Only by showing vigorous but short duration agonistic of coloration: 1) Submission and exploration, trunk coloration and intergrades. This should //. brevipinnis is found in the stripe or stripe/ interactions. Fighting in this species develops usually a light, cryptic coloration; 2) Aggression enable H. brevipinnis to have a more efficient mottle coloration after death; all other species are quickly with shorter periods of threat display and male courtship, usually a dark, increasingly communication system and thus spend less time saddled. (Losey, 1968). noncryptic coloration; 3) Female courtship, a out of its refuge on the barren substratum settling Stripe and mottle appear to function as cryptic Communication.-The pattern and darkness of light bland, and probably noncryptic pattern. territorial disputes and engaging in courtship. colorations in barren habitats. They probably the trunk and the darkness of special "signal spots" Hypsoblennius brevipinnis showed the greatest It is also striking that the two pairs of sym- serve to reduce the number of light to dark inter- had nonrandom relationships to the behavioral degree of divergence between the curves which patric species have divergent coloration codes faces that can serve as a clue of movement for state of the individual in all of the species studied. describe the occurrence of coloration elements and whereas allopatric species from the same habitat visually hunting predators. Barlow (1972) has As such, they serve as an expression of the motiva- thus the highest potential for communication. The have similar codes. Hypsoblennius gentilis shows noted a correlation between body depth and color- tional status of the individual which may be per- larger number of behavioral categories that can limited sympatry with H. jenkinsi since they are ation: He found that deep bodied fishes tended ceived by other individuals. Three behavioral be distinguished is partially due to the addition of usually found in different habitats. And yet the to be barred whereas long thin fishes displayed types are commonly discernable through patterns the lower postorbital spot as a signal organ and courtship colorations are radically different due

/96 BULLETIN SOUTHERN CALIFORNIA ACADEMY OF SCIENCES VOLUME 75 1976 COLORATION IN HYPSOBLENNIUS 197 TABLE 7. The reaction of H. robumus males to an approaching female in various color patterns. The previ- strongly favored due to the demands of its habitat Hinde, R. A. 1966. "Animal Behaviour." McGraw- ous behavior of the male is categorized as to "parental" (guarding eggs), "quiver courtship" (having courted for fast, efficient communication to avoid exposure Hill Book Co., N.Y.: 1-534. with quivering). "head-up courtship" (having courted with head-up but not quiver), and "no courtship." The to predation. response of the male to the female is categorized as to A = "attack" (bite or chase), N = no overt reaction, Krejsa, R. J. 1960. The eastern tropical Pacific Q = quiver, M. "mate." fishes of the genus Blenniolus, including a new CONCLUSIONS island endemic. Copeia, 1960:322-336. FEMALE COLOR PATTERN RODY PATTERN saddle saddle/bland bland A study of the social behavior, habitat type and Losey, G. S. 1968. The comparative behavior of RODY IIARKNESS dark light dark light dark light light some Pacific fishes of the genus Hypsoblennius POSTORRITAL SPOT no data dark light coloration of five species of Hypsoblennius re- DARKNESS vealed that: 1) Saddled and striped color patterns Gill (Blenniidae). Doctoral Dissertation, Univ. California, San Diego: 1-276. PREVIOUS MAI.E appear to be used as cryptic colorations in disruptive and barren habitats, respectively; 2) BEHAVIOR Sexual pheromone in some fishes Species that inhabit both types of habitat display 1969. parental AA AAA of the genus Hypsoblennius Gill. Science, 163: both types of color pattern; 3) Color pattern 181-183. quiver AAA forms a potentially informative signal of the courtship NNN QQQQQ behavioral status of the individual in terms of Morris, D. 1958. The reproductive behaviour of the QQ QQQQ 4) Sympatric Supp., No. 6: mmm agonistic and sexual tendencies; and ten-spined stickleback. Behav. species appear to have divergent courtship color- I-152. head-up AA ations. courtship A NN N NNN Robins, C. R., C. Phillips, and F. Phillips. 1959. Some aspects of the behavior of the blennioid fish ACKNOWLEDGMENTS ocellala Poey. Zoologica, 44:77-84. no courtship A AA A A A Chaenopsis This study was a portion of my doctoral dissertation Phillips, R. 1971. The relationship between social at the Scripps Institution of Oceanography. I thank behavior and the use of space in the benthic fish my major professor, Carl I.. Hubbs, and all of the Chasmodes bosquianus Lacepede (Teleostei, to the red chin and reticulate trunk of ripe male placement in order to reduce interspecific ag- members of my dissertation committee for their Blenniidae) I. Ethogram. Zeit. fur Tierpsychol., II. genii/is and the unique dorsal spot in courting gression. They show interspecific agonism that guidance and support. 29:11-28. females of this species. is very similar to intraspecific interactions despite Hypsohlennius rob:emus and H. .sordidus show having slightly different threat codes in terms of Springer, V. G. 1967. The Pacific South American LITERATURE CITED complete sympatry in Peru and have strikingly dif- action patterns (Loscy, 1968). blenniid fish, Hypsoblennius sordidus. Copeia, ferent coloration changes. Both have similar Hypsoblennitts robustus and IL jenkinsi form a 1967:461-464. saddle patterns in this disruptive habitat but the likely species pair. They probably in Baerends, G. P., and J. M. Baerends-van Roon. 1950. speciated fishes. posterior most saddle in H. sordidus can he dark- allopatry An introduction to the study of cichlid 1968. Osteology and classification of the after their ancestral species was split Behav. Suppl., No. 1:1-242. ened considerably. This saddle may or may not into two populations separated by a warm water fishes of the family Blenniidae. Bull. U. S. Nat. be darkened in preserved specimens (Victor G. harrier. They show little divergence in mor- Mus., No. 284:1-85. Springer, pers. comm.). The possession of a stripe phology, habitat, behavior or coloration. Barlow, G. W. 1972. The attitude of fish eye lines in relation to body shape and to stripes and bars. pattern in H. sordidus can probably be ascribed to Hypsoblennius shows highly diver- Stephens, .I. S., R. K. Johnson, G. S. Key, and 1. E. brevipinnk Cypeia, 1972:4-12. its ability to occupy barren habitats. But the gent morphology and coloration. Krejsa (1960) McCosker. 1970. The comparative ecology of blennies of divergency between aggressive colorations is strik- placed this species and H. proteus in the genus three sympatric species of California the genus Hypsoblennius Gill (Teleostorni, ing: Ilypsoblennius sordidus lightens and loses Bknohdus. Three of the major characters that -. 1974. Contrasts in behavior between Cen- tral American fishes and coral reef surgeon Blenniidae). Ecol. Monogr., 40:213-233. its posterior saddles first while 11. robustus darkens led to this separation were an additional pelvic cichlid fishes. Amer. Zool., 14:9-34. and loses its anterior saddles when becoming ray, foreshortened body and unique coloration. bland. It is frustrating that courtship was not seen Tavolga. W. N. 1960. Underwater communication Springer (1967, 1968) concluded that pelvic ray Communication" in IL .cordidus since this is a likely reason for the Fishelson, L. 1963. Observations on littoral fishes in fishes. "Animal Sounds and counts in this group are too variable to be used to Risso ed. W. E. Lanyon and W. H. Tavolga. Publ. No. divergence. In all of the other species the color- generic of Israel I. Behavior of Blennius pa as a character. The foreshortened body (Teleostei, Blenniidae). Israeli. Zool., 12:67-91. 7 Amer. Inst. Biol. Sci., 93-136. ation of male courtship is nearly the same as that and several related meristic counts arc an obvious for aggression. This suggests a link between the adaptation to living in a barren environment where Thompson, J. M., and A. E. Bennet. 1969. The physiological mechanisms for producing this ex- the only available cover is empty barnacle tests. Gibson, R. N. 1968. The agonistic behavior of L. (Teleostei). Behav., oyster blenny, Omaha:itch:a anotius (Valenci- pression of sexual activity and that of aggression. This study has shown that its coloration has also juvenile Blennius pholis 30:192-217. ennes) (Fllenniidae). Australian 1. Mar. Freshw. Thus. if IL sordidus acquired a light courtship most probably arisen as a response to its unique Res., 4:227-233. coloration due to selective pressure for reproduc- habitat with similar coloration responses in H. tive isolation, the change in aggressive coloration sordidus. Its unique lower postorbital spot is a Hamilton, W. 1. III, and R. M. Peterman. 1971. Countershading in the colorful reef fish Chaeio- Wickler, W. 1957. Vergleichende Verhaltensstudien might he a necessary result. It is doubtful that this highly ritualized signal organ that obviously serves don lunuln: concealment, communication or an Grundfischen I. Beitrage zur Biologic, change is a response to pressure for character dis- a communicative function. This again may be both? Anim. Behav., 192:357-364. Besonders Zur Ethologie von Blennius 198 BULLETIN SOUTHERN CALIFOR NIA ACADEMY OF SCIENCES VOLUME 75 1976 INFLUENCE OF TIDES AND WAVES ON GRUNION 199 A.sso im Vergleich zu einigen anderen Boden. Ecsenius bicolor (Pisces. Teleostei, Blenniidae) earlier workers to be controlled by tides (Thomp- fischen. Zeit. fur Tierpsychol., 14:393-428. Zeit. fur Tierpsychol., 22:36-49. son, 1919; Clark, 1925). After analyzing nu- merous grunion runs, Walker (1949) hypothesized . 1961. Uber das Verhalten der Blenniiden . 1967. Specialization of organs having a that the timing of runs is set by the initiation of sex Runula rind Aspidontus (Pisces. Blenniidae). Zeit. signal function in some marine fish. Stud. Tiop. product maturation at the second preceding moon fur Tierpsychol.. 18:421-440. Oceanog., 5:539-548. phase 17.9 days from the mid-point of a run series. Walker's thorough study, although never published --. 1963. Zion Problem der Signal hitching. Wilson, D. P. 1969. Notes from the Plymouth in its entirety, led to successful predictions of Cali- am Beispiel der Verhaltenmimikry zwischen Aquarium. J. Mar. Biol. Assoc. U. K., 28: fornia grunion runs (Walker, 1952). Since these Aspidontus und Lahroides (Pisces, Acanthop- 345-351. terygii). ?tit. fur Tierpsychol., 20:657-679. studies, nothing has been published on the periodicity of grunion runs. Our studies on the Gulf of sardina, began in 1968. We . 1965. Zur Biologic und Ethologie von Accepted for publication May California grunion, L. 1,1976. have attempted to elucidate some of the factors in- fluencing the periodicity of grunion runs by comparing runs of another species in a different tidal and wave regime. Fig 1. Night spawning of the Gulf grunion thes sardina) on the beach of El Golfo de INFLUENCE OF TIDES AND WAVES ON THE SPAWNING METHODS Clara, Sonora, Mexico. Erect female in fore- BEHAVIOR OF THE GULF OF CALIFORNIA GRUNION, und is attended by five males. (Photo by D. A. LEU RESTII ES SA RD/NA (JENKINS AND EVERMANN) From 1968(0 1973, L. sardina runs were observed omson.) along the beaches of El Golfo de Santa Clar 2 Sonora, Mexico, in the northern Gulf of C DONALD A. THOMSON' AND KEVIN A. Muglicti fornia. Data on 53 runs were compiled and ti Santa Clara, Sonora, Mexico, in the northern Gulf of each run was compared with the moon an of California. Spawning runs are also known to Anstancr: A preliminary analysis of 53 spawning runs of Gulf grunion, Leuresthes and the time and height of the tides before ring occur on several suitable beaches south along both sardina (Jenkins and Evermann) from 1968 to 1973 in the nor n Gulf of California suggests and after each run. [These data may be ained coasts of the upper Gulf to Bahia Concepcion, that the initiation of a spawning run is strongly dependent tidal amplitude. Gulf grunion from the senior author upon request] many Baja California Sur and Guaymas, Sonora. How- inns occur fortnightly from January to May following a s ding series of higher high tides instances, observers watched for runs r to and ever, such runs are sporadic and do not con- after the new and full moon phases. The runs begin abo days after the time of full moon after a predicted run series to deter the num- sistently occur following each new and full moon and about four days after the new moon. The day • runs of L. sardina result from a ber and intensity of runs in a seri II runs in a phase. The occurrence of California grunion mid-season shift in the time of the higher high spring from early morning (0300 to 0500) mean values. runs are likewise less predictable in the northern to late afternoon (150(1 to 1700). series were used in the calculation ble for the El and southern margins of their range. The accelerated spawning act of the Gulf grun ppears to be an adaptation to the short Since tide information was una period, low amplitude sea waves typical of the p ted coasts of the ripper Gulf of California Golfo de Santa Clara region, r were compared The most striking difference between the spawn- rather than increased daytime predation of spr g fish by sea birds. with tide predictions for Pu Peftasco, Sonora ing runs of the two species of grunion is the day- (Thomson, 1968-73). 0 ations show that time run of the Gulf grunion. However, the the Puerto Peiiasco tides approximately syn- reason for this anomaly became clearer when the chronous with those of olfo although slightly pattern of spring tides during a grunion breeding season was examined. The tides of the northern Carl L. Hobbs was the first to publish an a nt 1974: lis's in antplitude. Moffatt and Thomson, 1975). Both ulf grunion during night Gulf of California are of the mixed sernidiurnal of grunion spawning behavior when he re d a species exhibit the unique behavior of spawning Spawning behavior • ved, filmed with 16 mm type. The higher high water occurs during early fairly accurate description by J. P. lopl ow a out of water, depositing their eggs under the sand and day runs was and Super-8 mm c as and compared with that morning (0300-0500) in January and February Idler to David Starr Jordan (Hubbs. 19 Much of beaches following spawning . a high tide. Their The duration of the which is the beginning of the Gulf grunion spawn- later, one of Hobbs graduate stride (Walker, runs are highly predictable and coincide with the of the Californ union. e period of sea waves washing ing season. Only night runs occur early in the sea- 1949) slid the most thorough analys' the spawn- and full moon spawning act a new phases which cause the fort- son, closely following the early morning high tide. ing periodicity on the beach timed with stopwatches. of the Calif a grunion, nightly series of spring tides. The tinting of the During mid-season there is a shift in tidal ampli- Leuresthes tennis (Ayres). As 'kite to Carl runs of both species following a moon phase is tude so that the two daily high tides are about Hobbs' longtime interest in the 'on, this paper similar but the California grunion is known to TP AND GRUNION RUNS equal. Daytime runs begin when the afternoon reports a preliminary analysis c effect of tides spawn only at night whereas the Gulf grunion the height of the night high and waves on the spawning high tide approaches avior and period- spawns in the daytime as well as at night. Spaw runs of Gulf grunion occur from Jan- tide. When the afternoon high tide exceeds the icity of another species of nion, the Gulf of Spawning runs of L. tennis were believed by uar ugh May, whereas California grunion early morning high tide only daytime runs occur. California grunion, sardina Lear es (Jenkins begin in early March and end in After comparing grunion runs with tide heights it and Evermann ) . ru sually was timing its run with The only two species of grunion in the world 'Dept. Ecology and Evolutionary Biology, Univ. August. Fortnightly Gulf grunion runs occur appeared that L. sardina Arizona, Tucson, Arizona ocean. L. tennis and L. sardina, are geographically 85721. consistently on the beaches in the vicinity of San the higher high tide, regardless of when this tide 'Southern California Edison Co., Rosemead, occurred. During the transition of the higher high isolated and morphologically distinct ( Moffatt, California 91770. Felipe, Baja California Norte and El Golfo de