Bulletin 7>^ / ILLINOIS NATURAL HISTORY SURVEY JtStlIIQtttl Pn«/e o ^ O £ SB Ih ns .„ C <0 5«H 3 5 32 U 16 8 24 33 9 10 7 1 34 40 6 9 10 1 2 13 3 5 2 1 4 1 2 2 1 4 2 1 a.H 0.0 5.6 4.5 0.3 0.5 1.1 44.9 3.3 4.7 22.8 0.2 tr. 0.1 0.4 3.0 4.2 0.0 0.4 1.0 2.9 he trend was downward through the sea- ons to the lowest figure of the year in pring. Of some interest is the large number of mpty stomachs taken in the winter from he warmouths of Park Pond. Thirty-four ler cent of the warmouth stomachs col- ected there in winter (44 per cent of the tomachs collected in January) contained 10 food. These high percentages may re- lect the influence of partial ice cover and old water on the feeding activities of war- nouths. A large percentage of stomachs mpty at a time in which digestion was low indicated that the fish were going ong periods between feedings. A large iroportion of stomachs empty in summer vas due to rapid digestion and reduced eeding activity after midday. Tables 6—11 show seasonal changes in :inds of food eaten by warmouths in Park J'ond and Venard Lake, as indicated by inalysis of stomach contents. The following discussion and figs. 6 and 7 emphasize the highs and lows of the sea- sonal trends in foods most commonly found in the warmouth stomachs from the two study areas. Winter.—In the warmouth collections from Park Pond, fish (Pisces) were found in a larger percentage of stomachs in win- ter than in any other season, fig. 6. Cray- fish (Decapoda), which ranked second to fish in percentage of total volume in win- ter, were present in only 9 per cent of the winter-collected stomachs examined (all of them from large fish) and did not com- prise so large a percentage of the total vol- ume in winter as during the spring and summer. Dipteran and caddisfly (Tri- choptera) larvae, cladocerans, and amphi- pods were found in a smaller percentage of stomachs in winter than in any other season of the year. In the warmouth collections from Ven- ard Lake in winter, the animal groups 22 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 Table 12.—Stomach contents of 99 largemouth bass collected from Venard Food Item Stomachs From 23 Largemouth Bass Collected During March-Mav, 1949 Percentage of Stomachs Containing Organism Average Number of Organisms in Stomachs Containing Them Average of Volume Percentages Percentage of Total Volume Cestoda Gastropoda Cladocera Copepoda Isopoda Decapoda Araneae Hydrachnellae. . . . Ephemeroptera. . . Zygoptera Anisoptera Hemiptera Homoptera Hymenoptera. . . . Coleoptera larvae. Coleoptera adults. Trichoptera Diptera Pisces Filamentous algae Higher plants. . . . Organic debris 22 39 4 83 57 61 9 39 13 26 70 4 2 17 6 3 2 5 0.0 0.0 5.0 0.0 22 1 2'9 0.0 0.0 10.3 16.6 36.9 tr. 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.3 5.0 0.0 0.0 1.9 0.0 13.3 1.6 0.0 0.0 7.4 26.3 37.8 tr. 0.0 0.0 0.0 0.0 0.0 0.5 0.0 0.0 0.2 11.0 found in the largest percentages of stom- achs were cladocerans, mayfly nymphs, dip- teran larvae, copepods, and isopods; each of these groups was in winter at or near its peak for the yea.T in the percentage of stomachs in which it was represented. Fish and crayfish, which in winter led all other food groups in percentage of Park Pond stomachs in which they were found, were not found in any of the Venard Lake stomachs during the winter. Spring.—In the warmouth collections from both Park Pond and Venard Lake, the nymphs of damselflies were found in a larger percentage of stomachs, and com- prised a somewhat larger percentage of total volume of food, in spring than at any other season. Mayfly nymphs were present in a larger percentage of the stom- achs from Venard Lake in spring than at any other season ; at this season, they were present in a large proportion of the stom- achs from Park Pond, also. In the Park Pond collections, the per- centage of stomachs containing fish and the percentage containing dragonfly nymphs were lower in spring than at any other sea- son. The fragments of so-called higher plants, mostly rootlets or parts of leaves, that were found in 23 per cent of the stomachs probably were taken accidentally with other organisms. About two-thirds of the stomachs that contained plant frag- ments also contained crayfish. In the Venard Lake collections, isopods occurred in a larger percentage (57 per cent) of stomachs in spring than at any other season, fig. 7. They did not com- prise so large a percentage of total volume in spring as in winter, but their average of volume percentages (29 per cent) was greater and it was greater than that of any other food item taken during the spring. Dragonfly nymphs were present in nearly half of the Venard Lake stomachs collected in spring ; the percentage of stom- achs containing these nymphs, the average of volume percentages, and percentage oi the total volume were much greater dur- ing the spring than at any other season, Annelida, CoUembola, Neuroptera, and Bryozoa were represented as Venard Lake August, 1957 Larimore: Life History of the Warmouth 23 Lake during the perioc 24 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 FREQUENCY OF OCCURRENCE PER CENT CADDISFLIES 1 DRAGONFLIES DIPTERA MAYFLIES DAMSELFLIE3 FISH 1 I r 10 20 30 PER CENT AMPHIPODS CRAYFISH CADDISFLIES DIPTERA MAYFLIES DAMSELFLIES I r 20 30 PER CENT 10 20 30 PER CENT CLADOCERA DAMSELFLIES AMPHIPODS MAYFLIES CADDISFLIES DIPTERA 1 I r 10 20 30 PER CENT I 40 40 WINTER SPRING CRAYFISH August, 1957 Larimore: Life History of the Warmouth 25 FREQUENCY OF OCCURRENCE PER CENT DRAGONFLIES COPEPODS ISOPODS DIPTERA WINTER MAYFLIES CLADOCERA 1 \ T" 10 20 30 PER CENT 100 DAMSELFLIES DRAGONFLIES DIPTERA SPRING CLADOCERA ISOPODS MAYFLIES 1 T" 20 30 PER CENT OSTRACODS CRAYFISH CLADOCERA MAYFLIES SUMMER CADDISFLIES DIPTERA 10 20 30 PER CENT ORGANIC DEBRIS OSTRACODS MAYFLIES CLADOCERA COPEPODS FALL DIPTERA "1 \ r 10 20 30 PER CENT PER CENT OF TOTAL VOLUME Fig. 7.—For each of the most important foods taken from the stomachs of warmouths col- lected from Venard Lake in each of the four seasons of 1948 and 1949, the percentage of stom- achs containing these foods (frequency of occurrence) and the percentage of the total volume of food represented by each of these important foods. 26 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 the winter months, fig. 6. Although fish comprised more than one-half the bulk of food in winter, they were taken from only 15 per cent of the stomachs collected in the autumn. Dipteran larvae (largely chi- ronomids) comprised less than 1 per cent of the total volume of food each season but occurred in a large percentage (33 per cent or more) of stomachs each season — the largest percentage in the fall. In the fall collections, dipteran larvae were pres- ent in 46 per cent of the stomachs; the av- erage of volume percentages for these lar- vae was 9.2. Caddisfly larvae were pres- ent in a large proportion of stomachs and ranked first among all food groups in av- erage of volume percentages. Mayfly and damselfiy nymphs were found in a smaller proportion of stomachs collected during the fall than during the summer, whereas dragonfly nymphs were present in about the same percentages of stomachs in these two seasons. In the warmouth collections from Ven- ard Lake, mayfly and damselfly nymphs were present in about the same percentages of stomachs during the summer as during the fall. In November, nymphs of the mayfly, Siphlonurus sp., showed a sudden pulse of occurrence that extended into De- cember. Caddisfly larvae were present in a smaller percentage of stomachs collected in fall than in the summer months. Through- out the four seasons, dipteran larvae were found in consistently high percentages of stomachs—38 to 49 per cent—the highest percentages in the fall and winter, fig. 7. Chaoborus sp., a dipteran larva, was found only occasionally during the spring and summer ; but, during the fall months and in December, it was found in large num- bers of stomachs. Gastropods (snails) and ostracods were found in higher percentages of Venard Lake warmouth stomachs in the fall than at other seasons. Daily Changes During the summer, when water tem- peratures were high, digestion in the stom- achs of fish was rapid, and food remained in these stomachs for only a few hours. At this season, it was possible to determine daily feeding periods of warmouths at Park Pond by comparing the percentages of empty stomachs taken in morning col- lections with those taken in afternoon col- lections, table 13. The influence of indi- vidual fish size, and of size groups repre- sented by few individuals, was reduced by eliminating from the calculations all war- mouths less than 2.0 inches or more than 6.4 inches total length. Only 4 per cent of the warmouth stom- achs collected in the morning were empty, whereas 50 per cent of those collected in the afternoon contained no food. On July 8 and August 1, 1949, collec- tions were taken soon after sunrise and as late in the evening of the two days as fish could be taken with the shocking apparatus without the use of artificial lights. The daily feeding pattern was quite evident in these collections, table 13. Stomachs re- moved from fish collected between 6:00 and 7:15 a.m. on July 8 were "relatively full of very dark material" ; in contrast, the stomachs collected between 6:15 and 7:45 P.M. "seemed rather empty, the up- per intestine completely empty, with only the last three-fourths inch of the lower in- testine containing heavy black material representing the early morning feeding." In stomachs of fish of selected size groups (excluding fish with empty stom- achs), total volumes of the food masses averaged consistently lower (by 24 per cent or more) for individuals taken in the afternoon than for those taken in the morning. Monthly collections of stomachs from Venard Lake in the warm period of the year (1949) indicated for the warmouths of this body of water a daily feeding pat- tern somewhat similar to that of the war- mouths of Park Pond. Of four collec- tions from Venard Lake in the warm months, when the daily feeding pattern might be evident, two were taken in the morning and two in the evening; 13 per cent of the stomachs in the two morning collections and 26 per cent of the stom- achs in the two afternoon collections were empty. The differences between morning and afternoon collections were less evi- dent at Venard than at Park Pond, prob- ably because the collections were taken from Venard late in the morning and early in the afternoon. There was evidence of some feeding ac- tivity by warmouths in late afternoon. Although most of the food materials found August, 1957 Larimore: Life History of the Warmouth 27 Table 13.—Number of warmouths in morning and afternoon collections from Park Pond, June 2-September 2, 1949, and number of those warmouths with empty stomachs. Figures in- clude only those for fish between 2.0 and 6.4 inches total length. 28 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 mostly by warmouths between 2.0 and 3.4 inches, whereas the nymphs of Siphlonurus, which were larger, were eaten generally by larger fish, up to 5.2 inches in length. Caddisfly larvae were eaten by a larger percentage of small warmouths than of large ones ; the seasonal trend in consump- tion was somewhat similiar for fish of all sizes. Damselfly and dragonfly nymphs were utilized as food by warmouths of all sizes except those less than 2.0 inches long. Fish were eaten by a greater percentage of large warmouths (over 5.0 inches in length) than of small ones. The average volume of food found in the stomachs of warmouths of various sizes was not directly proportional to the length or weight of the fish ; the larger the war- mouth the greater was the volume of food taken in proportion to its size. The stom- achs of very small warmouths occasionally contained relatively great amounts of food, but small w^armouths feeding on many small items seldom experienced the ex- treme distention of the stomachs that oc- curred in many large warmouths when feeding on comparatively large fish or crayfish. The percentage of stomachs that were empt\^ was smaller among small war- mouths than among large ones. The per- centages of Park Pond warmouths with empty stomachs were as follows: fish of 1.9 inches or less total length, 3 per cent; fish of 2.0-3.4 inches, 18 per cent; fish of 3.5-4.9 inches, 24 per cent; fish of 5.0-6.4 inches, 32 per cent; and fish of 6,5 inches or larger, 28 per cent. Small warmouths feeding on many small items apparently had a more certain food supply than had large warmouths, which had relatively fewer large organisms on which to feed. Interspecific Competition Largemouth bass stomachs were col- lected from Venard Lake at the same time the warmouth stomachs were obtained. Be- cause only four largemouths were taken during the winter, their food habits for this season were omitted from consider- ation here. The majority of the bass taken (92 per cent) were between 5.5 and 9.0 inches total length. Of 107 bass stomachs collected, 99 contained food materials, table 12. Cladocerans were found in surprisingly large numbers in the stomachs of large- mouths up to 7.5 inches in length—791 of them in the stomach of one 6.4-inch bass. Of the largemouth stomachs collected in the fall months, cladocerans (almost ex- clusively Simocephalus sp.) were found in 59 per cent. Very few were found in the bass stomachs collected in summer. Cladoc- erans were found in relatively high per- centages of the warmouth stomachs col- lected from Venard Lake throughout the year ; they did not show an increase in utilization by warmouths during the fall months comparable to the increase in uti- lization by largemouth bass. Cladocerans were found in 32 and 37 per cent of the warmouth stomachs collected, respectively, in the summer and fall and in 96 per cent of the warmouth stomachs collected in De- cember. Copepods were found in very few stomachs of the largemouth bass, and ostracods were found in none. Isopods were found in greater percentages of both largemouth and warmouth stomachs col- lected in spring than at any other season. Fish and crayfish together comprised 64 per cent (36 and 28 per cent, respective- ly) of the total volume of largemouth bass food. The percentages of bass stomachs containing fish or crayfish were low in the spring, high during the summer, and low again in the fall. The seasonal trend was somewhat similar to that for warmouths at Venard Lake. Bass stomachs collected in August contained very few items except fish and crayfish. That mayfly nymphs and midge (Dip- tera) larvae were important bass foods at Venard Lake was shown by the consist- ently large percentage of stomachs in which they occurred. Percentages were larger in spring and fall than in summer. Mayfly nymphs were found in at least as large a percentage of bass stomachs each season as was any other kind of food organism ; they were found in a larger percentage of bass than of warmouth stomachs. The fall increase in utilization of mayfly nymphs by bass was not followed by a similar in- crease by warmouths. Nymphs of Siphlo- nurus sp. accounted for the fall increase in consumption of mayfly nymphs by bass; J as many as 125 were found in each of sev- " eral bass stomachs taken during Novem- ber. Only in the spring collections did the II August, 1957 Larimore: Life History of the Warmouth 29 nymphs of Caenis sp. occur in bass stom- achs as frequently as those of Siphlonurus; Caenis was consistently the species of may- fly most abundantly taken by warmouths. The variation in utilization of these may- flies may have come from differences in their habitats: Siphlonurus is generally concentrated in deeper water than is Caenis and would be available to large- mouths feeding in open areas of the lake. Caenis is a shallow-water mayfly and would be taken by warmouths feeding along the banks and in shallow weed beds. The percentage of bass stomachs that contained damselflies (Zygoptera) and/or dragonflies (Anisoptera) decreased from spring to fall. A smaller percentage of warmouth than of largemouth stomachs contained nymphs of the Odonata. Large- mouth bass stomachs contained more adults and subimagoes of damselflies than did warmouth stomachs; in the June col- lection of bass stomachs, these forms out- numbered the nymphs taken. Larval and adult beetles (Coleoptera), bugs (Hemiptera), and bees and ants (Hymenoptera) occurred at peak abun- dance in bass stomachs during the summer, especially in June. The incidence of these insects was much greater in largemouth bass than in warmouths at Venard Lake. Larvae of the aquatic beetle Peltodytes sp. were eaten in large numbers by a few largemouths ; in June, 131 of the larvae were found in the stomach of one individ- ual and 115 in the stomach of another. Certainly the foods and feeding areas of warmouths and largemouth bass over- lapped in Venard Lake. However, even though largemouth bass and warmouths fed on the same kinds of organisms, and even though several of these organisms fol- lowed similar seasonal patterns of occur- rence in the stomachs of the two fishes, the competition was somewhat reduced by dif- ferences in feeding habits. Warmouths tended to consume the organisms on the soft bottoms, in shallow waters, and along the banks; largemouths fed more on the surface organisms and free-swimming forms in deeper or more open parts of the lake. General Conclusions on Food Habits Considerable differences have been ob- served in the contents of the stomachs of warmouths taken in small numbers at dif- ferent seasons or from widely separated localities, table 14. Forbes (1903:48-9) analyzed the stom- ach contents of warmouths collected at scattered localities in Illinois and neigh- boring states and considered the foods uti- Table 14.—Average of volume percentages for the food of warmouths studied by Forbes (1903), McCormick (1940), and Rice (1941) and for the food of warmouths of approximately the same sizes, and collected at about the same times of year, from Park Pond and Venard Lake. 30 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 lized by fish of different sizes. The small warmouths in his study had eaten large numbers of Entomostraca, as had the small warmouths at Venard Lake and at Park Pond. In his six adult warmouths, table 14, crayfish were not represented, and fish made up a larger percentage of the food (47 per cent) than has been reported in other studies. Forbes related the espe- cially piscivorous habit of this species to the large size of its mouth. Dragonflies were noticeably absent in all the warmouth stomachs he examined. Data in studies made by McCormick (1940:73) and Rice (1941:26) at Reel- foot Lake, Tennessee, table 14, empha- sized the differences in foods utilized by a species in successive years, even at the same location and during the same season of the year. McCormick examined 69 warmouth stomachs which contained food and found that the average of volume per- centages of insects was 38.14 per cent. The comparable figure at Park Pond was 40.9 per cent. Crayfish were higher and fish were lower in the average of volume per- centages of warmouth food at Reelfoot Lake than at Park Pond. A year after Mc- Cormick's study. Rice examined another series of warmouth stomachs from Reelfoot Lake. Of 45 stomachs which contained food, only 1 had food other than crayfish, table 14. For the periods of time covered by the studies cited in table 14, warmouths at both Venard Lake and Park Pond utilized a greater variety of food items than did warmouths in the other places listed, as is indicated by the percentages of "miscella- neous insects" and "miscellaneous items." Lewis & English (1949:321) examined the stomachs of 29 warmouths from Red Haw Hill Lake, Iowa. The fish were collected from April through July and ranged from 40 to 177 mm. in total length. In the 17 stomachs that contained food, "food items occurred as follows: 2- to 4-inch fish, 7; crayfish, 4; vegetable de- bris, 2 ; unidentified insect larvae, 4 ; leech, 1 ; dragon-fly naiad, 1 ; unidentified in- sects, 2; and snails, 1." These figures probably refer to the number of stomachs in which each kind of food was found and not to the numbers of individual food items. The conclusion from this study was that "On a volumetric basis, fish and cray- fish were the most important food items." In a rather general statement involving five fishes, Black (1945:463) mentioned that the warmouth in Shiner Lake, Indi- ana, hunts the northern mimic shiner, No- tropis volucellus volucellus (Cope), to the exclusion of almost all other food. At Park Pond, where many species of min- nows were present, small sunfish were more commonly taken by warmouths than were minnows. Hunt (1953:29) examined 25 small warmouths from the Tamiami Canal west of Miami, Florida. Twelve of these fish, ranging from 1.4 to 3.5 inches in total length, contained food material composed exclusively of animals, including the fol- lowing organisms: dragonfly, damselfly, and mayfly nymphs ; dipteran larvae of va- rious kinds ; a few scuds and large ostra- cods ; and a large number of small shrimps, Palaemonetes paludosa. Huish (1947:15-6) examined 17 war- mouths from Lake Glendale (southern Illinois) during the summer of 1946. These fish, caught on artificial flies, ranged from 5.0 to 7.5 inches in total length. Of the 14 fish whose stomachs contained food, there were 3 with small fish, 1 with a tadpole, 3 with dragonfly nymphs, and 7 with crayfish. Fish, crayfish, and immature forms of aquatic insects comprised the important foods for most of the warmouths involved in the present study. Diets of the war- mouths in Park Pond were found to differ from the diets of the warmouths in Ven- ard Lake both as to the kinds and to the amounts of certain organisms eaten in va- rious seasons and by fish of different sizes. This study and others, some of which have been cited above, make it seem very un- likely that there is any specific diet or highly restricted food preference for this species. Food items of many kinds are ac- ceptable to the warmouth; this fish may feed upon any of those items that are read- ily available. REPRODUCTION Whether a fish population overcrowds its habitat is determined in part by its rate of reproduction—the development of sex products and the subsequent growth and survival of young fish. August, 1957 Larimore: Life History of the Warmouth 31 Development of Sex Products In interpreting the stages of develop- ment of sex products in a fish such as the warmouth, which spawns over a period of several months, one must keep in mind that all of the germ cells do not go through the cycle of maturation simultaneously ; instead, small groups of these cells mature at intervals and are spawned. This proc- ess is accompanied by a continuous recruit- ment of additional cells from the primor- dial stock. Thus, in a gonad in spawning condition, in addition to the fully matured sex products, there are other groups of cells representing earlier stages in the maturation process. Annual Sexual Cycle.—Terms de- scriptive of the appearance of the fish gonad are useful for designating stages of devel- opment associated with seasons, table 15. Some of the terms used here are the same as those used by Bennett, Thompson, & Parr (1940:17). Partly because the gradual process of growth of germ cells varies among indi- viduals and partly because the develop- mental process is influenced by climatolog- ical conditions, the periods during which the various designated stages may be found overlap and are not exactly the same from year to year for either individual fish or for populations. Overlapping of develop- mental stages was evident in warmouths Table 15.—Appearance and significance of each stage in the development of warmouth gonads, related here to the seasonal occurrence of each stage in warmouths from Park Pond, 1948 and 1949. Stage of Development OF Gonads Appearance of Gonads Male Female Significance of Stage IN Developmental Process Season of Stage Latent Poorly developed Well developed Spawning condition Partly spent Spent Clear pinkish white to colorless; a narrow translucent strand Pinkish white, opaque, becoming ribbon-like White, opaque; ribbon-like, with wavy margins Appears as in preceding stage but flowing milt when gently pressed White to gray; more nearly flat than in spawning condition Muddy white, becoming smaller and less distended Light amber, often with small red dots; lobelike in form and somewhat trans- lucent Pinkish orange to light yellow; slightly granular and somewhat enlarged Bright yellow; very granular and fully distended, with opaque eggs Appears as in preceding stage but flowing eggs when gently pressed Yellow to orange, with congested blood vessels; less distended than in spawning condition Reddish-orange, flaccid, with congested blood vessels A quiescent period; gonad containing only primor- dial germ cells Initial maturation of sex products Advanced development of germ cells; heavy yolk accumulation in ova Completed germ cells free in gonad, ready to be discharged No germ cells free in gonad, but a considerable stock of well-developed ova and sperm remain- ing Remaining matured germ cells resorbed; gonad reorganizing, but effect of spawning still evident All year for fish under 3.5 inches; July 1-April 15 For fish over 3.5 inches March 1-May 1 April 15-June 1 May 15-August 15 June 1-August 20 June 15-Septem- ber 1 32 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 collected from Park Pond during the spawning season of 1949. These fish were divided into two size groups established on the basis of differences in maturation: in the first group were "large" warmouths, those of more than 5.4 inches total length ; in the second group were "small" warmouths, those between 3.5 and 5.4 2. All the "large" individuals took part in spawning activities, whereas a consid- erable proportion of the specimens be- tween 3.5 and 5.4 inches remained imma- ture during the entire nesting season. 3. The "small" warmouths recovered more quickly from the effects of spawning than did the "large" ones; all of the MAY JUNE JULY MAY JUNE JULY MALES 3.5-5.4 IN MALES OVER 5.4 IN. MAY JUNE JULY MAY JUNE JULY FEMALES LATENT CH SPAWNING CONDITION POORLY TO WELL DEVELOPED PARTLY SPENT SPENT Fig. 8.—Percentage of warmouth males and females (two size groups) in each of five stages of sexual development during May, June, and July. These warmouths were collected from Park Pond, May 13 to August 1, 1949. inches total length. Warmouths shorter than 3.5 inches were eliminated from this part of the study because they would not mature in the then current season. The following generalizations may be made re- garding the season of greatest sexual ac- tivity of the warmouth, fig. 8 : 1. The "large" warmouths (over 5.4 inches total length) attained spawning condition sooner and spawned over a longer period than did the "small" fish (3.5-5.4 inches total length). "small" warmouths were in a latent condi- tion by August 1, whereas among the larger fish only 14 per cent of the males and 33 per cent of the females possessed gonads that had become reorganized by that date. 4. The males ripened slightly earlier in the season than did the females and re- mained sexually active somewhat longer. In the smaller size group (3.5-5.4 inches total length), a considerable propor- tion of both males and females had latent August, 1957 Larimore: Life History of the Warmouth 33 gonads after the first week in May, fig. 8. The continuous rise in percentage of latent gonads found in this size group during the nesting season may have resulted from ( 1 ) rapid reorganization of the gonads of early spawners, (2) failure of some small indi- viduals to spawn and prompt return of these individuals to the latent condition, (3) growth of fish during May, June, and July, which resulted in the recruitment of some sexually undeveloped fish into the 3.5-5.4-inch group and the loss of some sexually mature fish from this group to the size group beyond 5.4 inches, and/or (4) inadequacy of collections and their failure to represent true proportions of individuals in the various developmental stages. Field observations showed that sexually mature females do not have free ova (which can be forced out by gentle pres- sure on their sides) except immediately before and during the spawning act, whereas sexually mature males may be in- duced to extrude milt during much of the spawning season. These observations may explain the fact that many more males than females were classed as ripe. The scarcity of small males that were classed as completely spent was due probably to the difiiculty of separating partly spent fiom completely spent individuals in the small sizes. By expressing the weight of gonads at regular intervals through the year as per cent of body weight, one can follow the increase and diminution in size of the sex glands and fit the observed spawning con- dition of fish into the annual sexual cycle (James 1946 and others). Statistics on monthly gonad weight-body weight rela- tionships for warmouths in Illinois were based on 222 females and 260 males col- lected in Park Pond from earlv October, 1948, to early September, 1949, fig. 9. The fish were divided into three size groups: (1) less than 3.5 inches total length, (2) 3.5-5.4 inches total length, (3) more than 5.4 inches total length. Most of the warmouths of 3.5 inches and longer total length were sexually mature ; most below this length were sexually im- mature. During the period beginning with Sep- tember and ending with March, there was no appreciable change in gonad weights among warmouths. Soon after the initial rise of water temperatures in March, the gonads in warmouths longer than 3.5 inches began to enlarge, and they increased rapidly in weight during April and May. The ratio of gonad weight to body weight increased most rapidly in the large war- mouths (over 5.4 inches) ; males in this group showed their greatest average gonad weight in May and females showed their greatest average gonad weight early in June. The ratio of gonad weight to body weight for both male and female war- mouths between 3.5 and 5.4 inches in length averaged highest the first week in June. Soon afterwards, however, the ratio of gonad weight to body weight declined rapidly for females in the two larger size groups. Among males of both groups, there was a drop in weight of testes, but the males remained sexually developed later in the season than did females. The ratio of gonad weight to body weight may be at a minimum immediately following cessation of sexual activity, but it may in- crease slightly with reorganization of the gonads. Evidence from the cycle of changes in the ratio of gonad weight to body weight lends support to the conclusion previously drawn that large warmouths mature ear- lier in the season than do small ones and also remain active reproductively over a longer period. In this series of specimens, initial ripening of the sex products oc- curred during the second week in May, 1949, fig. 9. At the time other warmouths were spawning, warmouths under 3.5 inches total length showed no increase in the rela- tive weight of the gonads; in fact, an apparent decrease took place. This decline was probably associated with an improve- ment in condition (an increase in the body weight) of the smaller fish during May, June, and July. The irregularities that appeared during winter months in gonad weight-body weight relationships of fe- ir^ales in the two smallest length groups, fig. 9, were attributed to changes in body weight rather than to changes in gonad weight; these irregularities corresponded to changes in the coefficient of condition, fig. 17. Increasing length of day and rising tem- peratures associated with spring are known to stimulate gonad development in fishes. 34 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 In 1949, the ratio of average gonad weight to average body weight of Park Pond warmouths began increasing early in March and roughly corresponded to an up- swing in mean monthly air temperatures above 40 degrees F. (United States Weather Bureau 1948-1949, Danville Station). Spawning was actually initiated at water temperatures (12 inches deep) of about 70 degrees F. Low ratios of gonad weight to body weight for the summer months of July and August, 1949, indi- 4.0 3.5 UJ August, 1957 Larimore: Life History of the Warmouth 35 cated that most of the spawning was com- pleted by early July. Information obtained from collections of warmouths made the previous summer suggested that the spawn- ing season in 1949 may have been shorter than in most years. Fecundity.—Probably one of the more difficult problems in studying the repro- duction of a fish is how to estimate the total number of ova developed in a season by an individual female. The problem is especially complex in a fish that, like the warmouth, spawns an indefinite number of times over a rather long period and that contains large numbers of eggs too small to be counted without some magnification. Usually an extended period of sexual ac- tivity in fish is associated with a continu- ous maturation of eggs and sperm, so that sex products may ripen and be discharged at frequent intervals during the spawning season. Thus, at almost any time during the spawning season, in the ovary are sev- eral sizes of ova representing various stages of development from primordial germ cells to yolk-laden eggs ready for spawning. The egg-counting problem presented by the warmouth was not solved by any of the methods previously reported for de- termining the number of eggs in fish ova- ries. The method used—based on the dry weight of the egg mass in an ovary—was developed as an efficient way to estimate the numbers of ova of various sizes. The first step involved measuring several hun- dred ova and sorting them into seven size groups. A low power binocular microscope (Xl8) equipped with an ocular microm- eter was used to measure each of the sev- eral hundred ova to the nearest 0.05 mm. The seven size groups were as follows: I, 0.15-0.30 mm.; II, 0.35-0.45 mm.; Ill, 0.50-0.60 mm.; IV, 0.65-0.75 mm.; V, 0.80-0.90 mm.; VI, 0.95-1.05 mm.; and VII, 1.10 mm. and over. Group I in- cluded most small ova except the undiffer- entiated germ cells in the ovigerous lam- ellae. All ova diameters were measured as the ova appeared at random on the horizontal scale of the ocular micrometer. Clark (1925:5), in using a similar system of measurement, proved this was a reliable method of measuring eggs that were not spherical. The ova were separated into the various size groups as they were measured. Twenty ova belonging to each size group were then placed in a platinum crucible of known weight and put in a drying oven at 45 degrees C. After remaining in the dry- ing oven for 48 hours, the eggs, in the crucible, were moved to a desiccator, where they were left until repeated weighings showed no changes in weight. (It is now believed that the use of a desiccator was not necessary.) Each sample was weighed to the nearest 0.01 mg. ; through calcula- tions, the tentative average dried weight for ova in each of the various samples was determined. For each size group, five ad- ditional samples of 20 eggs each were dried and weighed before a final average dried weight was determined. The average dried weight determined for eggs in each size group was assumed to be the same as the average dried weight for eggs in a similar size group in other ovaries. Steps in processing each ovary for which a calculation of egg numbers was desired were as follows : ( 1 ) The connective tissue sheath surrounding the ovary was removed, and the eggs were teased apart; (2) a ran- dom sample of several hundred eggs was taken from the total mass of eggs in the ovary ; the eggs in the sample were meas- ured individually and separated into size groups, and the percentage of eggs in each size group was determined; (3) the mass of eggs remaining was washed, placed in a drying oven at 45 degrees C. for 48 hours, moved to a desiccator, and kept there until repeated weighings showed no changes of weight. Steps in calculating the total number of eggs in an ovary were as follows: 1. The average weight (dry) deter- mined for eggs of each size group was mul- tiplied by the percentage of eggs of the random sample in that size group. The products from the calculations for all the size groups were added and the sum mul- tiplied by 100 to give the calculated dry weight of 100 representative eggs of the sample. 2. The weight of the dried eggs (ex- clusive of eggs in the random sample) was divided by the calculated dry weight of 100 representative eggs, as determined from the random sample, and the result- ing quotient was multiplied by 100; to this product was added the number of eggs 36 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 in the random sample. The sum of these numbers was the calculated total number of eggs in the ovary. I'he calculated number of eggs in each size group in the ovary was determined by multiplying the calculated total number of eggs in the ovary by the percentage of the random sample represented by the size group. A modification of the dry-weight method described above, and one that did not re- quire the initial work of determining the average dried weight of eggs of each size group was as follows : After the eggs were teased apart and the connective tissue removed, ( 1 ) eggs in a random sample were counted (and measured if there was interest in size groupings), dried, and weighed, (2) the eggs remaining were dried and weighed, and (3) the total num- ber of eggs was then calculated on the basis of the total dried weight of all eggs in the ovary including those in the ran- dom sample. This method was similar to that used by Katz & Erickson ( 1 950 r 1 76 ) for estimating fecundity of herrings, in which only one size group of eggs was involved. When the numbers of ova in only a few ovaries are to be calculated, this modified dry-weight procedure is faster than the method first described. However, once the average dried weight of ova of each size group has been determined, the first method requires less work and is faster be- cause the sample ova that are measured, sorted, and counted do not have to be dried and weighed. Ovaries for which estimates of num- bers of eggs were made—38 ovaries from Park Pond and 10 from Venard Lake — had been divided into four groups: those taken from Park Pond warmouths in ( 1 ) January and March, (2) April, May, and June, (3) July and August, and (4) those taken from Venard Lake warmouths. May 25, 1949. In these ovaries, there was a pos- itive correlation between estimated num- bers of eggs in individual fish and total length of fish. Coefficient of correlation values ranged from 0.64 to 0.98. Seasonal variations in the number of eggs in warmouth ovaries were considera- ble. There was a marked increase in num- ber of eggs per ovary from late winter to the peak of the spawning period, fig. 10. Immediately after the peak of spawning ac- tivities, the number of eggs per ovary was considerably reduced. At that time, the correlation between number of eggs and size of fish was lowest ; this low correla- tion was due to the depleted condition of the ovaries, some of them being partly spent, some entirely spent, and others part- ly recovered. Females showing recovery from spawning contained more eggs than did the spent fish. During the fall and winter months, the number of eggs per ovary increased gradually; the final and greatest increases took place in the spring when groups of small eggs were adding yolk and undergoing final maturation. Fish of comparable sizes in different bodies of water did not produce comparable numbers of eggs. For example, a Venard Lake female, 5.3 inches in length, con- tained 40,400 eggs, whereas a female of this length from Park Pond contained only 12,500 eggs, table 16. This large differ- ence may be explained in part by dififer- ences in environmental stresses upon these warmouth populations resulting from ( 1 ) a rapidly expanding population in Venard Lake in 1947 and 1948 in contrast to an older and more stable population in Park Pond, (2) a greater concentration of fish in Park Pond than in Venard Lake, and (3) a higher incidence of parasitic infesta- tion in the more concentrated population of fish. Venard Lake was stocked in the spring of 1947, and the fish population expanded rapidly during the 1947 and 1948 growing seasons. Although the population probably had attained its maximum size by the time ovaries were collected for egg counts in 1949, the rapid expansion of the popula- tion in the preceding two seasons may have been at least partly responsible for the fact that the number of ova per female war- mouth was greater for fish collected from Venard Lake than for those from Park Pond. Park Pond had a higher population den- sity of fish other than warmouths than had Venard Lake, which had only warmouths and largemouth bass. Where many fish are concentrated within a limited volume of water, there may be severe competition for food and space among members of this population. The smaller number of eggs per warmouth in Park Pond may have been ugust, 1957 Larimore: Life History of the Warmouth 37 64 56 VENARD LAKE O MAY, r=.88 PARK POND • JAN -MAR r = .98 ^ APR-JUN r = .97 A JUL-AUG r = .64 , 48 I : 40 t > l32 c u D 5 24 J 5,6 8 - 2.0 3.0 4.0 5.0 6.0 7.0 TOTAL LENGTH OF FISH, INCHES Fig. 10.—Estimated number of ova from warmouths of various total lengths in collections from Park Pond (January-March, April-June, and July-August) and Venard LakeJMay, duced more ova than did females of similar sizes from Park Pond. Each graphic symbol repre- sents one female from which the number of ova was estimated. 38 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 associated with interspecific competition among several species rather than intra- specific competition among warmouths. Of the warmouths for vvhicli ova counts were made, the females from Venard Lake were in much better body condition (aver- age of coefficient of condition 82.2) than were Park Pond females of similar sizes taken at about the same time (average of coefficient of condition 73.8). This differ- Table 16.—Estimated numbers of ova in 38 warmouths from Park Pond and 10 from Venard Lake, 1949, arranged according to seasons and in order of increasing total length of fish. August, 1957 Larimore: Life History of the Warmouth 39 2nce in body condition indicated that the Menard Lake females had a greater amount Df reserve energy, which was available for the production of eggs. Functions of the gonads of warmouths nay be disturbed by internal parasites. Two ovaries in the series examined were parasitized by plerocercoid tapeworms and rontained fewer ova than expected, table 16. However, in both Park Pond and Ven- ird Lake there was such a high correlation )etween egg production and length of fish, egardless of differences in parasitic infes- :ations, that it does not seem possible that parasites greatly influenced egg production n these populations. Fecundity in warmouths may be re- iuced by lack of suitable nesting space, by )vercrowding of the population, unfavor- ible weather conditions, or other circum- tances which limit spawning opportunities md result in large numbers of mature eggs )eing retained and resorbed in the ovaries, llxamination of ovaries from warmouths aken from Park Pond during the middle ind last of June, 1949, revealed that only mall percentages of mature eggs were )resent in these fish and that some of these ggs were being resorbed, indicating that lot all the ova produced were actually dis- :harged. It was found that the spawning )eriod for warmouths was generally horter in this water area than in other vater areas under observation in Illinois. Seasonal Development of Ova.— Development of warmouth ova through he seasons was observed in the fish col- ected from Park Pond, figs. 11 and 12. rhe development was followed by assign- ng the ova collected at various times to he size groups defined on page 35. The easonal occurrence of eggs in these size [roups was as follows: 1. From January to early April, ova- ies contained only the smallest eggs (size ;roup I, 0.15-0.30 mm. diameter). 2. During the second week in April, the iva began to increase in size ; some ova v^ere in group II. 3. By the middle of May, in some fish wo-thirds of the eggs were of the smallest ize, or size group I, and small numbers of ggs were in size groups II, III, IV, and J . In other fish only one-third of the eggs vere in size group I, and group VI, as veil as groups II, III, IV, and V, was represented. In still other fish, there were relatively more eggs in groups V and VI than in groups II, III, and IV. 4. Toward the end of May, all of the size groups of ova were well represented, and some of the eggs appeared to be ripe. 5. By the end of the first 2 weeks of June, the ovaries of all the mature fish had discharged most of the largest eggs. Some of the fish appeared to be preparing for a final spawn ; about one-fourth of their eggs were in size group V and only low per- centages in groups II, III, and IV. Other fish appeared to have completed spawning, and their ovaries appeared to have begun reorganizing, as small eggs again com- prised three-fourths of the ova present. 6. On completion of spawning in July or early August, ovaries contained many pulpy eggs that were undergoing rapid re- sorption ; the only other eggs present were those of group I. The ovigerous lamellae were poorlv organized, showing no recov- ery from the production and crowding of large volumes of mature eggs. 7. By mid-August, ovaries were fully reorganized ; the well-arranged ovigerous lamellae contained many small ova which comprised a part of the egg-stock for the next season. Prespawning Activities Nest building and spawning activities of the centrarchids probably have been stud- ied more thoroughly than the spawning be- havior of any other family of fishes (Bre- der 1936 and others). Although the repro- ductive behavior of the warmouth is much like that of other sunfishes, it may differ in the location and construction of nests, recognition of sex, courtship, care of eggs and larvae, and spawning schedule. Location and Construction of Nests.—Warmouths appear to exercise selection in their choice of nesting sites. Both available bottom materials and cover influence this selection. In Venard Lake, where nesting was easily observed, the fol- lowing types of bottom were available: loose silt, silt containing sticks and leaves, rubble, rubble covered with a thin layer of silt, sand with loose silt, and clean sand. No nests were found on clean sand (such as is often selected by bluegills and pump- kinseeds), and the only nests seen on loose 40 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 silt were closely associated with tree roots or mats of submerged plants. Even though the Venard Lake warmouths used each of the bottom types (except sand) as nesting sites, they showed some preference for rub- ble lightly covered with silt and detritus. The warmouths at Venard Lake were not so consistent in nesting on a particular kind of bottom as they were in selecting a spot near a stump, root, rock, clump of vegetation, or some similar object. This habitual preference for a location adjacent to a stationary object may account for the lack of nests on the clean sand bottom in Venard Lake, Nests were never found on an area of bottom completely exposed, such as the bluegill usually selects. In labora- tor}^ aquariums, the locations most often used by warmouths for nesting were near the vertical drain pipes. <> O U. o a: UJ m < o I- li. o z LiJ o q: LlJ a. 100 August, 1957 Larimore: Life History of the Warmouth 41 In Park Pond, warmouths nested among weed masses, stumps, roots, and brush ; they nested in areas where the water was less than 4 feet deep and were most fre- quently^ seen where the shale and rubble spread out at the ends of the old spoil banks or had filled in the back portions of the flooded strip channels. They were not seen nesting where the banks were steep and sloped off quickly into deep water. Warmouths build nests within a wide range of water depths, and consequently nest locations vary in their distances from shore. Earlier observers of warmouths re- corded a variety of water depths selected : 6 to 10 inches (Richardson 1913:412); mostly 2 to 10 inches (Carr 1940:109) ; 3 feet (Hubbs 1919:144); 2 or 3 feet (Toole 1946:33). In Venard Lake, the depths of water over warmouth nests that 100 42 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 were found ranged from 6 to 60 inches. Most of the nests were covered with 2 to 2.5 feet of water ; nests in deeper water were more difficult to locate and may have been more common than counts showed. The limited areas of shallow water or the high transparency of the water in Venard Lake might account for the comparatively great depths at which nests were found in this body of water. Reports of some observers indicate that the warmouth is gregarious in its nesting habits, forming colonies of nests (Richard- son 1913:412; Carr 1940:110). It seems probable, however, that colony formation is a result of restricted nesting habitat. Observations of Carr (1940:109) in Flor- ida and of the author in Illinois support the assumption that the reason warmouth nests are sometimes found close to each other is that the fish are tolerant of each other rather than gregarious in their habits. The warmouth colony studied by Carr was formed on what may have been the only bottom of suitable depth not cov- ered with ooze. In Venard Lake, in May and June, 1947, nests were scattered in shallow water along the shore line with no indication of colonization. Three nests within a linear distance of 4 yards were the only ones that were found very close together. Nest construction by warmouths in lab- oratory aquariums was observed many times in the course of the study reported here. As in other sunfishes, the male ex- cavates the nest. Violent sweeping motions of the tail clear the loose debris away from the selected spot and produce a shallow, irregular concavity. The male begins each sweeping movement by approaching the nesting spot with his nose low and close to the bottom. As he enters the nest site, he turns abruptly upward, giving three or four violent sweeps with his tail while bal- ancing in an almost vertical position and checking his forward motion with his pec- toral fins. The loose material in the nest area is stirred up, and much of it settles outside the spot being cleared. The size and neatness of the nest depend to some extent on the amount of time the male spends in its construction. Many nests in natural waters are rather shapeless oval de- pressions of 4 by 8 inches from which loose silt has been cleared. The male may con- tinue to improve the nest if a female is not immediately available for spawning. One warmouth male nesting in an aquarium worked on his nest until it was a beauti- fully symmetrical depression 18 inches in diameter and 5 inches deep. Though most warmouth males under natural conditions spend no more than a few hours in clear- ing small nesting spots, the male in the aquarium spent a week working on his nest while waiting for the female that was present to become ripe. A mature female warmouth isolated in an aquarium during the breeding season constructed a shallow nest; her attach- ment to the nest was much weaker than that characteristic for a male. Preliminary Courtship.—The pre- liminary courtship phase of warmouth nesting was observed by the author only in laboratory aquariums. Normally it ap- peared as an aggressive threat to other males, serving to drive them away, and as a persuasive gesture to females in spawn- ing condition. During the first week in May, 1947, three unripe warmouths were placed in each of four aquariums previously filled with water and supplied with a layer of sand and gravel. By the second day, some of these fish had selected favorite corners and were accepting food. Two of the three fish in each tank soon began constructing nests and making advances at the third fish. On the basis of behavior, only the nest-building warmouths appeared to be males; the third fish in each of the tanks showed no interest in nesting and appar- ently was being courted by the aggressive males. Courtship in each tank progressed to the act of driving the nonnest-building warmouth into the nest depression and go- ing through motions of spawning. Vivid spawning colors (discussed in the follow- ing section) were displayed by only the ag- gressive males ; no color changes were shown by the fish that were being courted. After these courtship activities had con- tinued for several days, the fish were ex- amined. Several had become ripe since be- ing put in the aquariums and were flowing milt. Dissection of the nonaggressive, non- nest-building fish revealed that they, too, were males, although not in advanced stages of development, as were the aggres- sive males. August, 1957 Larimore: Life History of the Warmouth 43 These initial observations of spawning suggest that ( 1 ) warmouth males begin construction of the nests in the absence of females and well before their testes are ripe; (2) sex recognition among war- mouths is based on behavior and response to courting; and (3) in a small group of warmouths, during the breeding season some individuals assume dominance over less aggressive fish. When a female that is not yet ready to spawn is placed in a tank with a nesting male, she is charged, nipped, and driven to the surface. She remains quiet and retir- ing, ignoring as much as possible the male's advances. Being unable to escape the male in an aquarium, she may finally be killed by his continued aggression. Under nat- ural conditions, the female does not become exposed to the unavoidable advances of the nesting male before she is ready to spawn. Spawning In the warmouth, the mating act, which includes the deposition and fertilization of eggs, requires the simultaneous ripening ot sex products and synchronization of be- havioral attitudes in a male and female. Many environmental conditions, as well as the state of maturity of the fish, afFect the spawning process. Size and Age at Sexual Maturity. —The attainment of sexual maturity in fishes is influenced by both age and size. In the warmouth, size seems to be more important than age in determining when a fish attains maturity. However, there is considerable variation among warmouths in the size (and age) at which maturity is reached. As might be expected, this vari- ation is greater between fish of different populations than among fish within a sin- gle population. Sexual maturity is attained by war- mouths when the fish are between 3 and 4 inches in length. In Venard Lake, both males and females matured at 1 year of age and at lengths between 3.1 and 3.4 inches. In Park Pond, warmouths did not mature until they were 2 years old and at least 3.5 inches in length. Thus, at the time the fish became sexually mature in Park Pond, they were somewhat larger than the sexuallv mature 1-vear-old fish in Venard Lake. Hile (1941 :319) found for the rock bass that rapid growth appears to be correlated with an early attainment of sexual maturity. Nesting Season.—The observed nest- ing season for the warmouth in central Illinois begins during the second week in May, reaches its peak early in June, starts to decline after the first of July, but often extends well into August. The length of the nesting season differs among different populations of warmouths in different lakes and probably varies considerably from year to year. The length of the season varies also with the size of the fish ; large war- mouths spawn over a longer period than do small ones, fig. 8. Although the exact length of the spawn- ing season is difficult to determine for an individual fish, studies of gonads have shown that a fish may spawn several times during a summer. In Texas, Toole (1946:33) reported, "One pair of these fish was observed to spawn three different times during one year from April to Octo- ber." At the Natural History Survey lab- oratory, two females that spawned early in June, 1948, were examined 2 weeks later and were found to have well-devel- oped ova. These fish were not spent after the one spawning period and would un- doubtedly have produced more ripe eggs during the same season. Warmouths that were collected in July, 1948, and that pre- sumably had spawned at least once, were placed in a small pond on the following August 10; they produced a good brood of young in the pond that season. Deposition and Fertilization of Eggs.—Only when a female is ready to lay her eggs will she allow a male to guide her to the nest for spawning. In getting the female to the nest, the male assumes a very aggressive attitude, approaching her Vv'ith his opercles widely spread and his mouth open. The body of such a courting male becomes bright yellow in color and his eyes become blood red. The adjust- ment to these colors is very rapid, requir- ing only 5 to 10 seconds. If the female is ready to spawn, she is easily directed to- ward the nest, and spawning soon follows. The number of females contributing to the complement of eggs in a nest may de- pend upon how many females are ripe and available to the male. It is probably not 44 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 uncommon for more than one female to spawn in a single nest, as has been observed for other centrarchids. That such polyg- amy seldom occurs after the male assumes close guardianship of the eggs is indicated by the fact that freshly laid eggs are not commonly found in nests containing eggs in advanced stages of development. In a laboratory aquarium, however, a male guarding yolk-sac fry brought a female to the nest and proceeded with spawning ac- tivities. One female observed in a large aquarium in which two males had nests only 10 inches apart alternated between the two nests during a continuous spawning se- quence. During an hour of spawning ac- tivity, the female spawned with both males. When she was in one nest, the male in the other nest showed no concern for her; he would inspect the newly depos- ited eggs, stir the nest with his tail, and wait for the female to again approach his nest. The female spawned almost contin- uously for 40 minutes and for another 20 minutes at brief intervals and less vig- orously. When spawning was finished, the female was temporarily removed from the aquarium ; gentle pressure on the sides of this female did not cause the discharge of any eggs. On entering the nest site, both male and female begin to circle, the female being nearer the center of the nest, slightly on her side and somewhat beneath the male, fig. 13. As they circle inside the nest, the female works her jaws three or four times and suddenly jerks her body violently, giv- ing the male a sharp thump on the side. Each time the female jerks, she extrudes about 20 eggs. The thump she gives the male probably stimulates a discharge of sperm, although no milt was ever seen coming from the genital pore. After cir- cling the nest several times, the female interrupts the activities and leaves the nest site. The male usually follows her a short distance but returns quickly to the nest to assume guardianship. At this point in the spawning activity, males often have been observed to fan the nest with sweep- ing motions of the tail in a manner similar to that exhibited when nest building. Spawning activities like those observed in aquariums were carefully watched in Venard Lake. In nature, when a female Fig. 13.—Warmouths spawning in an aquarium. The light-colored fish is the male; the darker fish below and slightly on her side is the female. August, 1957 Larimore: Life History of the Warmouth 45 is ready to spawn, she makes her appear- ance near a nest and accepts the advances of the male. After a few spawning turns inside the nest, she retires, usually to a clump of weeds several yards away. The male remains over the nest for a few min- utes before again making advances toward the female. This procedure is repeated un- til the female has discharged her ripe eggs. With spawning completed, the female swims away, and the male settles down quietly to protect and fan the eggs. History of Embryos and Larvae When the warmouth starts its life and development, it is confronted by greater stresses of physical and biological factors than it will face at any other time during its life. Temperature changes or tempera- ture extremes, disease and predation, and dependence on the constant protection of a parent fish, which may at the same time be exposed to many dangers, result in high losses in the period of early development of the warmouth. Development of Embryos.—Four groups of warmouth eggs, 10 in each group, were artificially inseminated in or- der to observe the gross development of the embryos and the exact length of the incu- bation period. The following account in- cludes the time sequence of certain easily discerned stages of development at tem- peratures between 25.0 and 26.4 degrees C. Within this temperature range, the average developmental period of the 40 eggs was 34 hours and 30 minutes and the interval between hatching of the first egg and hatching of the last was 2 hours and 40 minutes. When eggs and sperm were mixed in a petri dish and then immediately flooded with water, a high percentage of the eggs became fertilized. The inseminated eggsi measured 0.95 to 1.03 mm. in their great- est diameters during their first and second minutes in water. These measurements are slightly below those of the largest ova taken from preserved ovaries. Differences in size measurements between preserved and live ova may have been due to differ- ences in shapes : irregular shapes of the ova preserved intact in the ovaries and almost spherical forms of the live ova in water. Each of the live eggs, translucent and light amber in color, contained a single, dark amber oil droplet 0.35 mm. in diameter. Within 3 minutes after an egg was im- pregnated by a sperm cell, a thin perivitel- line space could be seen between the outer membrane (chorion) and the egg cell proper. Thirty minutes later, a blastodisc was evident as a slightly raised cap, giving the egg a somewhat oval appearance. The first division of the blastodisc occurred 43 minutes after impregnation. Each blasto- mere then measured 0.4 mm. across. The second, third, and fourth divisions took place at 60, 75, and 90 minutes, respec- tively. The resulting group of blastomeres appeared whitish, the yolk was very pale yellow, and the oil droplet remained dark amber in color. After 2 hours and 15 minutes, the blas- tomeres formed an oval-shaped mass at one end of the yolk. The segmentation cavity was formed beneath this mass, and at 2 hours and 30 minutes the blastoderm be- gan growing down over the yolk mass. The blastoderm had grown over two- thirds of the yolk mass within about 11 hours after impregnation, and a thickened band of cells at the margin of the blasto- derm had appeared as the germ ring. About an hour later (12 hours and 15 minutes after impregnation), the blastoderm cov- ered all but a small plug of yolk, which contained the oil droplet. The first differ- entiation among the dividing cells was visi- ble in the live egg after 14 hours and 15 minutes. A groove extended around the egg from a patch of opaque cells near where the yolk plug and oil droplet en- tered the blastoderm. This groove, formed by the neural plate and neural folds, be- came quite distinct during the following hour (15 hours and 10 minutes after im- pregnation). After 16 hours and 30 minutes, the primordial form of the embryo was de- fined. The ensuing process of organ for- mation, however, could not be discerned. Movement of the embryo was first ob- served 25 hours after impregnation. The first egg hatched 33 hours and 20 minutes after impregnation. All eggs had hatched by the end of the following 2 hours and 40 minutes (36 hours after im- pregnation). Fry emerging from the eggs early during the hatching period were smaller (2.30-2.60 mm. in total length) 46 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 «5 August, 1957 Larimore: Life History of the Warmouth 47 than those emerging later (2.65-2.85 mm. in total length). The oil droplet in the newly hatched fish was the same size (0.35 mm. in diameter) and color as it was in the newly fertilized egg. The greatest depth of the fry was 0.80 mm. across the yolk sac. It is interesting to note here that war- mouth eggs fertilized with sperm from a green sunfish showed no difference in rate of development from the rate described above. There was a high percentage of survival of both embryos and fry of the warmouth and green sunfish cross. Development and Growth of Lar- vae.—Specimens for a study of growth of larvae were collected from a nest in a lab- oratory aquarium. The eggs were laid dur- ing the morning of June 24, 1947. Daily collections were made until the postlarvae left the nest. Then the free-swimming lar- vae were transferred to an outdoor tank, where observation and sampling were con- tinued. The specimens, preserved in alco- hol, served as materials for the following descriptions of developmental stages. Ob- servations indicated that total length of a larva is a better indicator of the stage of its development than is actual age, which was known for each specimen. Measure- ments made with an ocular micrometer to the nearest 0.01 mm. were used in the de- scription of body form. In general, de- scriptions follow the procedure used by Fish (1932) ; terms for growth stages are those suggested by Hubbs (1943:260). Prolarva, 3.4 mm., soon after hatching, fig. 14^^ : Total length 3.4 mm.; length to anus 1.7 mm.; length of yolk sac 1.0 mm. Large oval yolk mass containing one oil droplet 0.3 mm. in diameter. Head de- flected sharply downward in front of yolk sac, making the midbrain the most forward part of the body and the forebrain lying directly beneath it. Head from front of forebrain to end of hindbrain 0.65 mm. Optic capsule faint, 0.23 mm. in diam- eter. Notochord straight. Myomeres in- complete anteriorly, numbering about 8 in front of anus and 17 behind anus. Embry- onic marginal fin fold complete except for faint break where intestine penetrates fin to outside. Fin extending forward on back to point 1.0 mm. from front of body. No visible ray development nor pigmentation. Prolarva, 4.6 mm,, 48 hours old, fig. 145: Total length 4.6 mm.; length to vent 2.0 mm. ; length of head 0.63 mm. ; greatest depth of body in front of vent 0.70 mm.; greatest depth of body behind vent (excluding fin fold) 0.35 mm. Forebrain still somewhat deflected, with globular cerebellum extending high above anterior part of medulla. Eyes well pigmented, each 0.35 mm. in diameter; optic fissure still quite apparent. Mouth indistinct. Fin fold extending forward on back to point 1.2 mm. from front of body, entire except for break at anus. Very weak indications of fin rays below and above end of straight notochord, giving effect of diphycercal tail. Pectoral lobes present. Myomeres indis- tinct forward, about 10 to anus, 19 caudad from anus. Branchial elements forming. Postlarva, 5.3 mm., 4 days old, fig. 14C: Total length 5.3 mm. ; length to anus 2.2 mm. ; greatest depth of body anterior to anus 0.6 mm. ; greatest depth of body posterior to anus 0.35 mm.; diameter of eye 0.41 mm.; length of head 0.90 mm. Cranial flexures almost straightened, but cerebellum high and bulblike. Optic cavity distinct. Fin fold beginning on back 1.5 mm. from front of body, becoming slightly narrow on caudal peduncle but wide again at tail. Faint indication of rays forming in areas of the anal and soft dorsal fins. Distinct fin rays on either side at end of the straight notochord. Kidney apparent through body wall. Pectoral fin lobes well developed but with no rays. Myomeres 10 anterior to anus, 19 posterior to anus. Branchial arches well formed and with de- veloping gills. Mouth gape extending obliquely forward from point below mid- dle of eye. Dark row of spots on either side of ventral fin fold ; two large chro- matophores between bases of pectoral fins. Postlarva, 7.6 mm., fig. 14D: Total length 7.6 mm.; length to anus 3.4 mm.; length of head 1.5 mm.; diameter of eye 0.55 mm. ; greatest depth of body anterior to anus 1.15 mm.; greatest depth of body posterior to anus 0.65 mm. Myomeres 1] before anus, 19 behind anus. Fin fold still complete except for break at anus ; high at soft dorsal fin region, quite low above and below caudal peduncle. End of notochord bent upward at 40-degree angle, giving ap- pearance of heterocercal tail. Caudal fin rays well developed on lower side of noto- chord, with middle rays longest. Rays weak but distinct in unformed anal fin; 48 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 rays very vi'eak in soft dorsal. Rays visible in pectoral fins. Mouth only moderately oblique. Pigmentation much more devel- oped. Row of spots along ventral fin fold spreading as stellate chromatophores over ventral surface of body. Series of dark dashes indicating lateral line. Heavy chro- matophore lying above anus. Some color apparent at base of caudal fin rays. Six stellate chromatophores between bases of pectoral fins and a row of five chromato- phores on each side across branchiostegals. Distinct, dark chromatophores scattered over top of head. Postlarva, 8.8 mm., fig. 14£: Total length 8.8 mm. ; length to anus 3.9 mm. ; length of head 2.0 mm.; length of snout 0.4 mm. ; diameter of eye 0.75 mm. ; depth of caudal peduncle 0.65 mm. Caudal pe- duncle long and narrow. Notochord with upturned end but tail appearing essentially homocercal. Fin fold present immediately anterior to anus. Anal and soft dorsal fins separate from caudal fin but each broad at base, due to some remaining parts of the embryonic marginal fin fold. Rays distinct in all fins present. Pelvic fins not devel- oped. Otic region large and clear. Anus protruding from ventral line of body. Lat- eral line chromatophores quite distinct. Ventral spots larger, with more chromato- phores scattered over head region. Postlarva, 12.0 mm., fig. 14F: Total length 12.0 mm.; length to anus 5.4 mm.; length of head 2.8 mm. ; diameter of eye 1.0 mm.; length of snout 0.7 mm.; great- est depth of body 2.45 mm. ; length of cau- dal peduncle 2.4 mm. ; depth of caudal peduncle 1.15 mm. Fin fold remaining only as short keel in front of anus. Pelvic fins present but weak and with indistinct rays. Spinous dorsal developed only as a row of short stubs. Distribution of pig- ment about the same as in 8.8 mm. stage, except spots appearing more distinct. More dark chromatophores around mouth ; a ver- tical row present at base of caudal rays. Juvenile, 15.7 mm. (not photographed). Body form essentially like that of adult fish. Total length 15.7 mm.; length to anus 7.0 mm.; length of head 4.35 mm.; diameter of eye 1.4 mm.; length of snout 0.85 mm. ; greatest depth of body (at about anterior insertion of spinous dorsal) 3.7 mm. ; length of caudal peduncle 2.9 mm., depth 1.5 mm. Anus protruding only slightly from abdomen. No trace of em- bryonic marginal fin fold. Pelvic and spi- nous dorsal fins well formed. Pigmentation much heavier than in earlier stages. More color apparent over head and caudal pe- duncle. Belly rather free of pigment. Many large chromatophores scattered over back. Heavy row of spots forming circle behind eye and distributed over top of head. Chro- matophores noticeable on soft dorsal, anal, and caudal fins. Behavior of Larvae.—Activities of the warmouth larvae during their early life in the nest were limited to a few feeble movements. There was definite sequence, however, in the behavioral development of these small fish. The following description of behavior was based on laboratory ob- servations in aquariums with water of 2-1—25 degrees C. (75-77 degrees F.). Immediately upon hatching, the delicate prolarvae dropped down onto the sand and silt between coarse gravel particles of the nest. As the heavy yolk sac restricted movement, the prolarvae were difficult to see in the nest. When the prolarvae were between 36 and 48 hours old, fig. 145^ they began making feeble jumps an inch or so above the bottom of the nest. Most of these prolarvae were between gravel par- ticles in the nest, but a few could be seen resting on the flat surfaces of the largest particles. Although the yolk supply was about ex- hausted by the fourth day, the young fish still limited their movements to poorly di- rected jumps above the nest. They did not begin active swimming until the end of the fifth day, when they appeared as in fig. 14-D. At this time, they swam about the nest in rather compact groups. Their movements were well controlled, and they showed remarkable ability to avoid a dip net. In the aquariums, these small fish had no food supply and starved in 10 or 11 days after hatching, but in outdoor tanks they began feeding at least by the seventh day after hatching. School formation among postlarval war- mouths in natural habitats was not so ob- vious as in postlarvae of certain other sunfishes, for the warmouths remained either among dense submerged vegetation or else in small pockets of open water closely surrounded by plants. The indi- vidual shown in fig. 14F was taken from a 1 August, 1957 Larimore: Life History of the Warmouth 49 school near the nest in which it had hatched. The schools gradually dissolved as individuals began independent searches for food. No juvenile warmouths were observed in large groups. Factors Affecting Survival.—Rate of survival of warmouth eggs and young is influenced greatly by many physical and biological factors. Incubating eggs are readily affected by adverse weather condi- tions. Sudden drops in water tempera- tures promote the rapid growth of fungi infecting the eggs; often, entire nests of eggs are destroyed early in the spawning season as a result of low temperatures and fungi. For example, many warmouth nests in Venard Lake contained eggs during the last week in May, 1947, but, after several days of cold weather, the eggs in every nest observed were covered with fungi. Although heavy rains and high turbidity were not seen to affect nesting of adults or survival of fry, rapidly falling water levels might disturb nesting. In several Illinois lakes, minnows and sunfishes were observed destroying eggs and larvae in unprotected warmouth nests. In laboratory aquariums, warmouths were seen to rob poorly guarded nests; they charged in to snap up eggs or larvae. Postlarval and juvenile warmouths which have left the nest are eaten in great numbers by larger fish. Venard Lake sup- ported a heavy spawn of both largemouth bass and warmouths in the summer of 1947. On June 30 of that year, bass 1.75 inches long were voraciously feeding upon warmouths 0.75 inch long, which had been eating large numbers of postlarval warmouths. In the laboratory, a 0.75-inch warmouth ate 11 postlarvae (4 days old) in 5 minutes; another ate 12 in the same length of time. Survival of small warmouths is closely related to the density and composition of the fish population, the time of year, and the character of the habitat in which they are produced. Fry hatched late in the spawning season are in a population with a larger number of potential predators (fish only slightly larger than themselves) than are the fry produced earlier. How- ever, because the density of aquatic vegeta- tion increases during June and July, sur- vival in the late summer broods is fre- quently higher than in early broods. GROWTH Whether one is studying a single spe- cies of fish or the entire fish population of a body of water, it may become necessary to consider the growth of individuals in the one or more species involved. An analysis of growth is not always the objective of such a study, nor is determination of the morphological relationships which must be known before an analysis of growth can be made. The ultimate value of a growth study may come from its use in deter- mining the factors that govern or influence growth of fish under particular conditions. Relative Growth Various parts of a fish's body grow at differential rather than uniform rates. The differential rates are not necessarily the same even for closely related species in the same habitat, nor for fish of the same spe- cies in different habitats. Consequently, when making a growth analysis of a se- lected species in a given habitat, one must determine several morphological relation- ships, namely, those of body growth to scale growth, body growth to growth of tail fin, and growth in length to growth in weight. Lewis & English (1949) and Hennemuth (1955) have plotted some of these relationships for two populations of Iowa warmouths, and Jenkins, Elkin, & Finnell (1955) for several populations of Oklahoma warmouths. Relation of Body Growth to Scale Growth.—A regression line to show the relationship between length of anterior radii of scales and length of body was con- structed from measurements of 1,068 war- mouths from Venard Lake and Park Pond, fig. 15. Data were obtained from collec- tions made approximately monthly begin- ning in June, 1948, and ending in Novem- ber, 1949. Regression lines constructed for warmouths from Venard Lake and for those from Park Pond proved to be so similar that data from the two lakes were combined. Key or representative scales were taken from the side of each of the fish near a point where the tip of the pectoral fin laid backward touched the third row of scales below the lateral line. Fish were sep- arated into total length classes at 0.5-inch 50 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 TOTAL LENGTH OF FISH IN INCHES Fig. 15.—Regression line expressing the re- lationship between total length of fish and radius of scale for 1,068 warmouths collected from Park Pond and Venard Lake, early June, 1948, through middle November, 1949. The dots show the anterior radius of scale X41 for fish of various total lengths (one-half-inch intervals). intervals, and the average of the total lengths for the fish in each of these classes was determined. Scales of these fish were placed in a scale-reading machine that magnified 41 times, and images of the an- terior radii of the scales were measured and averaged for the fish in each length class. Average total lengths of fish and corresponding average anterior radii of selected scales were used in developing the regression line shown in fig. 15 and the following equation : L—0.5278+ 1.048 S where L^total length of fish in inches and S=41 X anterior radius of scale in inches Relation of Body Growth to Tail Growth.—Growth of the body of a fish in relation to growth of its tail, or caudal fin, may be ascertained by comparing the stand- ard length of the fish with its total length. As defined by Hubbs & Lagler ( 1947 : 13 ) , total length includes the caudal fin, where- as standard length does not. The relationship between growth of body to growth of tail was calculated from measurements of 264 warmouths taken from Park Pond, summer, 1948, and No- vember, 1949, and Venard Lake, October, 1949. As the average body length of the warmouths increased, the average tail length became relatively less, table 17. In fish less than 4.0 inches total length, the average total length was 1.259 times the average standard length ; in fish of 4.0 to 6.9 inches, the average total length was 1.240 times the average standard length; and, in fish longer than 6.9 inches, the average total length was 1.211 times the average standard length. Relation of Growth in Length to Growth in Weight.—The relationship between growth in length and growth in weight was calculated from data on 866 warmouths collected from Park Pond be- tween early June, 1948, and early No- vember, 1949. Size groups were estab- lished at 0.1 -inch intervals. Average weights were determined for each group within the size range beginning with 3.3 and ending with 8.2 inches total length. The length-weight relationship for each of Table 17.—Factors derived from measurements of warmouths from Venard Lake, October, 1949, and Park Pond, summer, 1948, and November, 1949, for converting standard length (S.L.) to total length (T.L.), and the reverse, with the same and with different units of measurement. Total Length, I.NCHES August, 1957 Larimore: Life History of the Warmouth 51 the specimens (866) was expressed by the equation : log W= -4.49867+3.04902 log L where W=weight in grams and L^'standard length in millime- ters Length-weight relationships were calcu- lated for the especialh' heavy and especially light warmouths in the Park Pond popula- tion. For warmouths heavier than aver- age, the equation was as follows: log W- -4.36191 +3.01387 log L For those lighter than average, the equa- tion was as follows : logW= -4.35603+2.95352 log L From these equations, it may be seen that fish either heavier in relation to length, or lighter in relation to length, than the av- erage bore a systematic relationship of length to weight roughly paralleling that of the average. Such divergence from the average as was discernible was found par- ticularly among the fish of greater lengths and weights, fig. 16. Coefficient of Condition The coefficient of condition (Cj, based on total length in inches and weight in pounds, was computed for the 866 Park Pond warmouths used in the analysis of length-weight relationships. Fish smaller than 3.3 inches total length were not con- sidered because individuals were weighed only to the nearest 0.01 pound, and greater weighing preciseness would have been nec- essary if smaller fish had been used ; war- mouths of more than 8.2 inches total length were not used because few were available. Condition (C) increased progressively with increased size of fish. The average C (weighted to compensate for differences in numbers of individuals in groups) for each of several size groups, 3.3—4.2, 4.3-5.2, 5.3-6.2, 6.3-7.2, and 7.3-8.2 inches total length, was 72.6, 74.8, 78.6, 80.9, and 82.6, respectively. Warmouths within the size range 3.3-4.2 inches total length showed a wide seasonal variation in condition (C). This variation among small warmouths may have been caused by their dependence upon food items that fluc- tuated widely in abundance from month to month, such as cladocerans and certain in- sects and their larvae. Larger warmouths, STANDARD LENGTH IN MILLIMETERS 50 100 150 200 0.80 0.70 0.60 - 9 0.50 2 0.40 - 0.30 0.20 - 0.10 0.00 1 .0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 100 total LENGTH IN INCHES Fig. 16.—Curves illustrating the relationship between length and weight of warmouths col- lected from Park Pond, early June, 1948, through middle November, 1949. feeding more upon crayfish and fish, varied less in body condition from one season to the next than did small warmouths. The condition (C) of warmouths for each of three size groups, 4.3-5.2, 5.3-6.2, 6.3-7.2 inches total length, was similar in seasonal fluctuations. A sudden, severe drop in condition occurred in September, 1948; a low level in condition lasted through October, and was followed by a rapid recovery by mid-November. Con- dition declined gradually during the winter and spring, but then began an increase that continued through May and June. Condi- tion remained relatively high and constant throughout the summer of 1949 and then declined during the fall months. Fig. 17 suggests that for the warmouths of Park Pond a definite cycle of condition associ- ated with seasons could not be established. Since food habits were studied for these warmouths collected from early October, 1948, to early November, 1949, it was pos- sible to associate the foods eaten with the seasonal variations in condition of the fish. The low level of condition in the winter (1948—1-9) was coincident with a compar- atively low consumption of crayfish, dip- 52 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 teran larvae, amphipods, and mayfly the warmouths used in the present study nymphs and with a comparatively high are as follows: consumption of fish and dragonHy nymphs, 1. There was a regular increase in the fig. 6. During the spring of 1949, when number of annuli accompanying an in- OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV 1948 1949 Fig. 17.—Coefficient of condition of 392 warmouths of three size groups (6.3-7.2 inches, 5.3-6.2 inches, 4.3-5.2 inches) collected from Park Pond, early October, 1948, through early November, 1949. Warmouths collected and weighed in the summer of 1948 were so few in num- ber that data on them were not included in the graph. the warmouths were eating high percent- ages of crayfish (volume) and damselfly nymphs (frequency), the condition of these fish was steadily improving. The rela- tively good condition of warmouths during the summer of 1949 was associated with the extensive use of mayfly nymphs, cad- disfly larvae, and crayfish. There were no consistent differences in condition between male and female war- mouths in Park Pond. Scale Method of Calculating Growth The method of calculating warmouth growth from fish lengths and scale meas- urements is a composite of methods devel- oped by several authors but generally fol- lows the procedures suggested bv Hile (1941). Validity of the Annulus as a Year- Mark.—Age and growth studies made from the scales of warmouths in Venard Lake and Park Pond indicate that the an- nulus is a reliable year-mark in this spe- cies. Hile (1941:201-4) outlined the most important features of a valid annu- lus. Four points used to test the validit}' of the scale method of age determination for crease in size of fish, and fish assigned to any single age group were within a cer- tain length range, table 18. 2. Lengths calculated from scale meas- urements agreed reasonably well with ac- tual lengths of fish of corresponding ages, tables 18 and 21. 3. Calculated lengths were similar for the same age groups of fish collected in different years and consistent for different age groups of fish collected in the same or different years. Because the calculated lengths were verv similar for warmouths in the 1948 and 1949 Park Pond collec- tions, data for these two collections were Table 18.—Averages of total lengths of war- mouths of various ages collected from Park Pond, June, 1949. August, 1957 Larimore: Life History of the Warmouth 53 combined and are not shown separately; however, consistency in calculated lengths of fish of the same and of different age groups is shown in tables 21 and 26. 4. There was similarity among war- mouths of different year classes with re- spect to growth rates in certain calendar years, tables 23 and 24. Characteristics of the Annulus.— The true annulus on the warmouth scale appears as a result of resumption of growth of bodv and scales after the cessa- tion of growth during winter months. True annuli or year-marks usually show several rather definite characteristics, fig. 18. Across the anterior field of the scale the annulus appears as a break in the arrange- ment of circuli; it is bordered on the in- side by closely spaced, incomplete circuli and on the outside by complete, widely spaced circuli. The radii in the anterior field are slightly distorted in the region of the annulus. On the lateral fields of the scale, the annulus and newly formed circuli Fig. 18.—Warmouth scale: A, wide spacing of circuli especially evident following resurnp- tion of growth in spring; B, first annulus; C, second annulus; D, crowding of circuli during period of slow growth resulting from habitat disturbance ; E, new circuli cutting across ends of circuli laid down in previous season. 54 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 "cut over" the circuli laid down near the end of the previous season. The annulus extends only part way across the posterior field of the scale. Slight differences in spacing of circuli and in the lengths of ctenii (small surface spines on scale) may aid in recognizing the location of the year- mark in the posterior field. The first annulus formed on the war- mouth scale does not show as much "cut- ting over" in the lateral fields as do later annuli, nor are the differences in spacing of circuli so apparent in the first as in later annuli. The differences which may exist in length and distribution of ctenii are most useful in recognizing the first an- nulus, Annuli formed later are generally more difficult to recognize with certainty, mainly because they are closer together than those earlier marks laid down when the fish was increasing rapidly in length. Time of Annulus Completion. — In 1949, the time of annulus completion in warmouths at Park Pond was deter- mined from 129 specimens collected be- tween April 7 and May 20. Sixty-eight per cent of the warmouths collected May 13 and 14 had laid down the year's annuli on their scales, so that the average time of annulus completion appeared to be about the first week of May, table 19, at which time the water temperature 1 foot below the surface of the water was about 70 de- grees F., fig. 9. Small fish usually lay down the year-mark earlier in the spring than do the larger ones, because the former begin to grow at an earlier date. Fifteen per cent of 39 fish collected from Park Pond on April 7 and 9 showed new an- nuli on their scales; all but one of these fish were less than 4 inches total length. As annulus formation is associated with the resumption of growth of the fish in spring, the annuli on the scales of fish may not be completed at the same time each year, and the time may vary from one pop- ulation to another. Warmouths in Venard Lake showed a period of annulus forma- tion in 1949 that differed from the pe- riod shown by the warmouths in Park Pond. In the fish collected from Park Pond, some year-marks appeared before April 7, but only 68 per cent of the fish in the May 13-14 collections had completed annulus formation; the period in which annuli were being completed covered more than 6 weeks. In Venard Lake, on the other hand, none of the specimens col- lected on April 5 had formed an annulus; in the collection of May 12, about 5 weeks later, 30 of 31 fish (nearly 97 per cent) had completed annulus formation. Ecological conditions in Venard Lake, where warmouths were confined to a small area having only minor fluctuations in Table 19.—Percentages of warmouths with and without new annuli, Park Pond collec- tions, 1949. Date August, 1957 Larimore: Life History of the Warmouth 55 true annuli on the scales in later collec- tions, these marks gave little trouble in age determinations. The formation of the marks coincided with a 4-week period of shoreline dredging with a dragline. Such a severe disturbance of the habitat must have reduced the food supply or its avail- ability and caused a temporary stoppage of growth that produced the false marks. Growth in Park Pond Park Pond, an 18-acre lake in a flooded stripmine area, in 1948 and 1949 sup- ported an old (60 years at least), rather large fish population of about 36 species native to the region. Most of these spe- cies had been introduced into the lake from the Salt Fork in times of flood, page 4. Collection and Preparation of Ma- terials.—In the period beginning June 7, 1948, and ending November 12, 1949, 1,420 warmouths were collected from Park Pond. Hoop nets, fig. 19, were used in collecting 367 warmouths ; of these fish, 298 were collected between early June and mid-September, 1948, and 69 between June 27 and July 5, 1949. An electric shocker rigged for operation from a row- boat (Larimore, Durham, & Bennett 1950) was used to collect specimens for growth analyses and food studies. A total of 788 fish were taken by this method in collections made at monthly intervals from early October, 1948, to early November, 1949, except that no collections were made in February and October, 1949. Even though Park Pond supported the largest naturally established warmouth population that had then been examined in Illinois, warmouths were never taken there in great numbers; a good day's take might consist of 2 dozen warmouths from the usual set of six hoop nets or 50 warmouths from the operation of the electric shocker. Large specimens predominated in the hoop net catches ; fish of all sizes were present in collections made by shocking. On August 22, 1949, rotenone was applied to a shal- low, isolated, 0.47-acre slough in Park Pond, and a census was made of the fish population. Scales for growth studies were taken from 265 of the 504 war- mouths collected from this area. Fig. 19.—Collecting hsh with a hoop net in Park Pond. 56 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 Ages were determined for 1,328 of the 1,420 warniouths collected (265 from the shallow, isolated slough and 1,063 from other parts of Park Pond). Scales from 84 fish were regenerated and unreadable ; scales of 8 other specimens were so difficult to read that ages could not be determined with certainty. Fish from the slough that was treated with rotenone were considered separately because they showed growth rates significantly different from those taken in the other collections. Impressions of the warmouth scales were made on cellu- lose acetate slides, and the images of these impressions were projected for study at a magnification of 41 diameters. Measure- ments along the median, anterior radius of a magnified scale image were marked on a manila paper strip, and the strip was used for calculating past growth on a nomograph, as described by Carlander & Smith (1944). Calculated lengths, based on a straight-line relation between scale length and body length, were corrected for an intercept of 0.53 inch in body length (fig. 15 and equation on page 50). The following sections, "Growth Dif- ferences Between Sexes," "History of Successive Year Classes," "Fluctuations in Annual Growth," and "Seasonal Growth," refer to Park Pond warmouths other than those from the slough. Growth Differences Between Sexes.—Sex was determined by dissection or by visible discharge of sex products for 600 specimens in year classes 1944 through 1948. Calculated lengths of males and fe- males of each of the year classes were aver- aged and compared, table 20. Fish in both the 1943 and 1949 year classes were rep- resented by so few specimens for which sex was determined that they were omit- ted from the calculations. Very consistent, although rather small, differences existed between the growth rates of males and females. Males were larger than females at the end of each year of life in the five year classes considered. The greatest differences occurred in the 1944 year class, but the small number of specimens (only seven females) made this growth comparison less reliable than that for other j^ear classes. The next oldest brood, the 1945 year class, was represented by 187 specimens. Males in this group av- eraged only a little longer than females. Schoffman (1940:32) observed that the lengths and weights of male and female warmouths of the same ages in Reelfoot Lake, Tennessee, were either the same or only slightly different. Since the actual dififerences in lengths of male and female warmouths were rather small in those year classes represented by substantial numbers of specimens, data for the two sexes were not separated in the growth analyses discussed in the following paragraphs. History of Successive Year Glasses. —All the specimens were assigned to year Table 20.—Average calculated total lengths in inches for male and female warmouths, representing five year classes, collected from Park Pond, October, 1948, through November, 1949. Year Class August, 1957 Larimore: Life History of the Warmouth 57 Table 2L—Average calculated total lengths in inches for 1,063 warmouths, representing nine year classes, collected from Park Pond, June, 1948, through November, 1949. Year Class 58 Illinois Natural History Survey Bulletin V^ol. 27, Art. 1 lUU ao August, 1957 Larimore: Life History of the Warmouth 59 each year of life and for each calendar year, table 23. Percentages of expected growth during selected calendar years may be read from table 23 in diagonal rows from lower left to upper right. The percentages of ex- pected growth in each calendar year, when averaged, show clearly the fluctuations in annual growth, table 24. Actual length increments exceeded the expected increments in only 1945 and 1946, table 24. Poor growth in 1942 and 1943 may have been due to heavy floods, which caused the water to remain muddy for 6 or 7 weeks during the early summer of each of these years. The exceptionally good growth of war- mouths during 1945 and 1946 may have re- sulted from an artificial thinning of the fish population. On Tune 26, 1945, May 15, 1946, and July 29, 1946, Dr. George W. Bennett and other members of the Illi- nois Natural History Survey staff sprayed most of the shallow waters of Park Pond with rotenone to reduce the numbers of small fish in the population. Although no estimate could be made of the percentage of the total fish population killed by these par- tial poisoning operations, the great num- ber of small fish destroyed may well have allowed a substantial increase in growth rates of the surviving fish. Seasonal Growth.—Growth patterns of the 1946, 1947, and 1948 year classes of warmouths in Park Pond during the summer of 1949 are shown in fig. 22. The length increment for each fish was calcu- lated from scales. The growth increment on each scale used was measured on the Table 24.—Average percentages of expected annual length increment attained in each cal- endar year by warmouths collected from Park Pond, June, 1948-November, 1949; year classes combined. 60 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 grew rapidly through July and showed some growth in each of the months through October. The growth pattern for the 1947 year class was intermediate between the patterns for the 1946 and 1948 year classes. Seasonal growth rates of three year classes of warmouths in Venard Lake are shown in figs. 23 and 24. Measurements of warmouths from Park Pond, fig. 22, from Venard Lake, figs. 23 long. Most of these warmouths were less than 3.5 inches in length. Age determinations for 265 of the 504 warmouths revealed a growth rate consid- erably less than was found in other areas of Park Pond, table 25. Of the 265 speci- mens aged, very good growth was found in 23 large fish belonging to the 1944 and 1945 year classes. These fish did not seem to be representative of the population of Table 25.—Average of calculated lengths of 265 warmouths from vphich scale samples were taken, after being collected from Park Pond Slough, shown with similar calculated lengths of warmouths from other parts of Park Pond. Group of Fish August, 1957 Larimore: Life History of the Warmouth 61 length increment for each year of life was then calculated for each size group as a means of determining the growth rate — whether fast, intermediate, or slow. Of the 313 warmouths considered, 94 had been collected with hoop nets during the summer of 1948; these 94 were fish of the 1944 year class and were faster-grow- ing individuals than the specimens (219) that had been taken from the slough. War- mouths taken from the slough had been collected after being poisoned with rote- none; they belonged to the 1946, 1947, and 1948 year classes, table 26. For the fish taken in hoop nets, the difference in average calculated lengths be- tween the largest and smallest size groups was 0.55 inch for the first year of life, 0.74 inch for the second, 0.55 inch for the third, and 0.42 inch for the fourth. The decline in differences between these two size groups in the third and fourth years of life may indicate compensatory growth among individuals of the smallest group in these years. However, the compensatory growth that occurred was slight and it did not overcome the length advantage held by the fish that grew most rapidly during the first _vear of life. In the slow-growing population from the slough, the maximum differences in length between the two extreme size groups of the various year classes declined little or not at all after the second year. In the 1946 year class, after an increase in length difference at the end of the second year, the differences were about the same at the end of the third and fourth growing seasons, table 26. A study of compensatory growth in these four year classes of warmouths from Park Pond suggested the following con- clusions: 1. Warmouths that were largest at the end of the first year of life increased this length advantage in the second year of life. Table 26.—Compensatory growth, in inches, in four year classes of warmouths collected from Park Pond. Fish in the 1944 year class were collected in hoop nets from several parts of Park Pond in the summer of 1948; those in 1946-1948 year classes were taken from a slough to which poison was applied on August 22, 1949. Year Class AND Place OF Collection 62 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 2. Warmouths that were smallest at the end of the first year of life showed no com- pensatory growth in the second year but showed a slight compensatory growth in the third year. 3. Although warmouths that grew fast the first year of life underwent a decline in annual length increment after the sec- ond growing season, they retained their length advantage over warmouths that grew slowly the first year. 4. Warmouths that grew slowly the first year of life showed more compensa- tory growth in later years if they were members of fast-growing populations than if they were members of slow-growing populations. These conclusions are in fair agreement with those from similar studies done on several other sunfishes. Hubbs & Coop- er (1935:678) found no compensatory growth during the second year of life in the longear sunfish, pumpkinseed, or blue- gill, or in bluegillX pumpkinseed hybrids. Their data did not include growth rates beyond the second vear. For the rock bass, Hile (1941:332) stated: "First-year ad- vantage in size may be retained over 1 or 2 additional years, but more probably it will be increased in the second and/or third year of life. Compensatory growth occurs in the later years." Sizes and Longevity.—A 9.6-inch male was the largest warmouth collected from Park Pond. This fish weighed 1.0 pound and was 6 years of age. The ma- jority of the large fish were males. Al- though the males grew slightly faster than the females, table 20, it did not necessarily follow that the males reached greater max- imum sizes than did the females. The oc- currence of more large males than large females in the collections may have indi- cated only that the former were more read- ily taken than were the latter—a logical hypothesis in view of the differences in be- havior during the nesting season. The sed- entary nest-guarding habits of the males would have made them very vulnerable to collection by shocking. Schoffman (1940:36) mentioned spawn- ing habits to explain the greater percentage of females than males in the groups of large warmouths he collected from Reel- foot Lake, Tennessee. As his collections were taken with traps operated during the breeding season, nest-guarding males were not caught so readily as females. The larg- est warmouth handled by Schofifman (1940:34) was a 9.29-inch female. Growth in Venard Lake In 1948 and 1949, warmouths in Ven- ard Lake, an artificial lake of 3.2 acres, were associated with only one other species, the largemouth bass. Both species had been introduced in April, 1947, page 5. The 1,102 Venard Lake fish used in this study were from collections made with an electric shocker each month (ex- cept January and February) in a period beginning September, 1948, and ending October, 1949. Methods used for scale preparation and age determination were similar to those described for the coUec- a tions from Park Pond. I Since Venard was a recently stocked lake, it contained only a small number of year classes of warmouths: 1947, 1948, and 1949. A comparison of growth rates was made between warmouths of the first, fast-growing year class (1947) and those of the two following year classes (1948 and 1949). The following points seem ap- parent, figs. 23 and 24: 1. Both actual and calculated lengths of warmouths of the first year class to be spawned in the lake (1947) averaged more at the end of the first year and of each suc- ceeding j'ear of life than did those of later year classes. 2. The actual length range for mem- bers of the first year class was greater than that for members of each succeeding year class. 3. The average calculated lengths of warmouths of 1947 and 1948 year classes collected in successive months of 1949 showed a decline. 4. The average annual growth of war- mouths in Venard Lake was very similar to that of warmouths in their first 3 years of life in Park Pond, table 27, in spite of large ecological differences in the two habi- tats. Growth in Other Water Areas The rate of growth of warmouths may be influenced by various environmental fac- tors or combinations of them. This fact is August, 1957 Larimore: Life History of the Warmouth 63 illustrated by the differences observed in the growth rates of warmouths taken from 12 Illinois water areas, table 27. The most rapid growth recorded in Illi- nois warmouths was in Enright Pond in McLean County in which some members of the first brood produced in the lake at- tained 6 inches in total length during their first 13 months. Thinning the total fish population by intensive angling resulted in an increase in the growth rate of war- mouths in Onized Lake, a 2-acre body of water in central Illinois (Bennett 1945: 396-7). Exceptionally rapid growth of war- mouths usually accompanies the expansion of fish populations in new reservoirs. Dur- ing the first 6 years after impoundment of 7.0 if) LjJ I o X I- 6.0 - 5.0 - 4.0 - LjJ _l 3.0 O 2.0 LO RANGE OF ACTUAL LENGTHS AVERAGE OF ACTUAL LENGTHS AVERAGE OF CALCULATED LENGTHS AT FIRST ANNULUS AVERAGE OF CALCULATED LENGTHS AT SECOND ANNULUS SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT 1948 1949 Fig. 23.—Averages of actual total lengths and averages of calculated total lengths at time of formation of first annulus and at time of formation of second annulus for warmouths of the 1947 year class taken in 12 collections from Venard Lake, late September, 1948, through middle October, 1949; also range of actual total lengths in each collection. 64 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 5.0 4.0 CO X o X3.0 o z _J -J 2.0 \- 1.0 0.0 RANGE OF ACTUAL LENGTHS AVERAGE OF CALCULATED LENGTHS AT FIRST ANNULUS AVERAGE OF ACTUAL LENGTHS ^ 1948 YEAR CLASS 1949 YEAR CLASS X J. SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT 1948 1949 Fig. 24.—Averages of actual total lengths and averages of calculated total lengths at time of formation of first annulus for warmouths of the 1948 year class taken in 10 collections from Venard Lake, early October, 1948, through middle October, 1949; averages of actual total lengths of warmouths of the 1949 year class taken in 4 collections, late June through late October, 1949; also range of actual total lengths in each collection. Lake Glendale, an 82-acre lake in Pope County, Illinois, warmouths showed a growth rate that was exceptionally fast for the species (Dr. Donald F. Hansen of the Illinois Natural Historj^ Survey, unpub- lished studies of Lake Glendale). Another example of improved growth rate in a new impoundment is given by Hall & Jenkins (1953:34) ; they found that, in Tenkiller Reservoir in Oklahoma, the growth rate of warmouths was rapid during the first year of impoundment. Jenkins (1953:79) found that in Grand Lake, Oklahoma, the growth rate of warmouths gradually de- clined during the years of impoundment. Data from several studies of warmouth growth in other states are summarized by Carlander (1950:191-2; 1953:370-1). An inspection of these data and those given in table 27 reveals a wide range of differ- ences in warmouth growth rates. PARASITISM No attempt will be made here to survey all published records concerning parasites of the warmouth. Reference should be be made, however, to several important studies involving autopsies of compara- tively large numbers of warmouths. Holl (1932:99-100) examined 90 warmouths from North Carolina and discovered an in- August, 1957 Larimore: Life History of the Warmouth 65 teresting seasonal fluctuation in relative numbers of parasites and in percentages of fish infested. In a study of centrarchids from southern Florida, Bangham (1939: 265) examined 143 warmouths and found all of them infested with parasites of one the genus PhysOj any one of several spe- cies of fish, and the great blue heron, Ardes herodias L. As little age immunity has been demonstrated in intermediate hosts of most flukes, the numbers of metacer- cariae of this strigeid probably continue Table 27.—Growth rates of warmouths in 13 water areas in Illinois. Water Area 66 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 latius worms were seen handling as a red tuft from the anus of a fish taken from the water. Camallanus attaches to the inner intestinal wall. The warmouth serves as the final host, and infestation may take place at any time. The only parasite which infested no- ticeable numbers of warmouths from Ven- ard Lake was a leech, Illinobdella moorei Meyer. During the autumn and winter of 1948, leeches of this species were present in such large numbers that they appeared as compact fringes on the fins of the war- mouths. The caudal fins of the fish were severely damaged, frequently suffering ex- tensive destruction of the rays. Although the infestation was still heavy during the spring of 1949, only relatively few of the leeches were seen during the following summer months. An examination of 25 ovaries from war- mouths taken from Park Pond on No- vember 12, 1949, revealed the following: (1) 14 ovaries contained one or more par- asites; (2) ovaries of all sexually mature females were parasitized; (3) 24 plero- cercoids of Proteocephalus ambloplites oc- curred in 12 ovaries, as many as 4 in 1 ovary; (4) 27 metacercariae of Posthodip- lostomum minimurn occurred in 10 ova- ries, 12 m 1 ovary; and (5) 4 sexually undeveloped adult acanthocephalans, Lep- torhynchoides thecatus (Linton), were found in 4 ovaries (1 in each ovary). In spite of these parasites, no sterile fish were found and no primary damage to the ovaries was evident. Warmouths were collected in sufficient numbers from Venard Lake and Park Pond to permit tentative conclusions to be drawn relative to the influence of para- sites on the general physical condition of these fish. Warmouths from Venard Lake had fewer internal parasites and a consist- ently higher coefficient of condition (C) than had warmouths of similar sizes taken from Park Pond. At the same time, among the heavily infested warmouths of Park Pond, no positive relationship could be shown between a fish's coefficient of con- dition and the number of parasites present. Therefore, it is believed that the difference in condition of warmouths from the two lakes was more directly a result of differ- ence in densities of the total fish popula- tions than of parasitism. Even though no harmful effects of para- sites on the condition (C) of warmouths in V^enard Lake or Park Pond could be demonstrated, possibilities of some other harmful effects must be recognized. Fe- male warmouths from Park Pond pro- duced much smaller numbers of eggs than did Venard Lake females, table 16, which were less heavily infested. BEHAVIOR Observations on the behavior of the warmouth are scattered through many sec- tions of this publication. For example, the aggressive behavior of the nesting male is described (page 43) under "Spawning" in the section on reproduction. It seems de- sirable to bring together here certain as- pects of observed behavior of the war- mouth, although to do so will mean some repetition. General Activity and Disposition The warmouth has a quiet disposition ; it moves around relatively little and dis- plays no showy activity except during the nesting season. It seeks the cover of weed masses, stumps, or rocky banks (pages 6 and 8), and avoids intense light. Reproductive Behavior Tinbergen (1953:23) describes syn- chronization, persuasion, orientation, and reproductive isolation as functions of mat- ing behavior in animals. These functions, along with defense of the nest area, the spawning act, and parental care, are con- sidered here as reproductive behavior in the warmouth. Defense of the Nest Area.—The nesting warmouth male displays an aggres- sive threat toward other fish that approach his nest area (page 42) . He assumes a bel- ligerent attitude by swimming toward the intruder with his mouth open and his opercles spread ; at the same time, his eyes become red and his body becomes light yel- low in color. As the nesting male nears the intruder, he usually turns abruptly to one side or upward and, with vigorous movements of his tail fin, forces small pulses of water toward the intruder. He may also nip the intruder. The entire August, 1957 Larimore: Life History of the Warmouth 67 threat attitude associated with defense of the nest area is similar to the persuasive behavior employed by the nesting male in courting a female (page 43). Synchronization.—The spawning pe- riod for warmouths extends over several months (page 43). Male warmouths be- come ready to spawn earlier and remain capable of spawning later in the season than do females. Thus, a ripening female generally encounters many males ready to spawn. More precise synchronization for the actual discharge of sex products is brought about by the preliminary court- ship and persuasive gestures of the male and finally by the thump given the male as the female extrudes a group of eggs (page^44). ^ Orientation.—The special orientation for mating in warmouths consists simply of the male having an established nest, the female with ripening eggs wandering into the vicinity of the nest, and the male initi- ating the persuasive actions. A signal — such as sound, odor, or color display, used by many animals to attract a mate from considerable distances— is not known to be given by the male warmouth. However, the female probably receives some internal physiological stimulus to wander as her eggs ripen and they become free for dis- charge. Persuasion.—The male warmouth's threat attitude, described above, serves to initiate the action for persuading the fe- male to spawn. A female that is not ready to spawn responds to the threat as any other intruder would and is driven from the nesting area (page 43). On the other hand, a female that is ready to spawn quietly submits to the aggressive male. The threat, in which the male spreads his opercles and shows some display of color, is followed by attempts to guide the female to the nest depression. With only mild re- sistance and casual reluctance, the ripe fe- male accepts more and more of the male's actions and soon enters the nest to remain with him for periods of time that become increasingly longer until spawning actu- ally takes place. The Spawning Act.—The series of signals and responses described above cul- minates when the male and female come together to deposit their sex products si- multaneously. Although the aggressive at- titude of the male makes it seem that he is controlling the spawning activities, the fe- male enters the nest only when she is ready, she gives the final signal ( thumping the male's side) for extrusion of eggs and milt, and she leaves the nest depression for short intervals between egg laying. The spawning signals and responses follow a definite sequence ; it is interesting to recall that a female warmouth in a laboratory spawned alternately during one continu- ous spawning sequence with two male warmouths (page 44). After having been brought to a spawning attitude by one male, the female then responded to either of the two nesting males. Reproductive Isolation.— Hybrids are produced between the warmouth and a great number of the species of Lepomis, and yet such hybrids seldom occur in large numbers in natural populations. What forms the reproductive isolation that pre- vents greater hybridization was not de- termined in this study. There is little spa- tial separation of the various sunfishes. The warmouth is usually found living and even nesting with several species with which it could genetically hybridize, and yet few hybrid individuals are formed. In the absence of any other observable isolat- ing barrier, the isolation appears to result from a lack of the specific signals and re- sponses necessary to bring a warmouth to successful spawning with an individual of another species. In the laboratory, male warmouths have courted green sunfish and bluegill females but have not succeeded in spawning with them and seldom are able to guide them to the nest depressions. Ap- parently, the series of specific signals and responses is not followed through to suc- cessful spawning. Parental Care.—After the warmouth fry leave the nest area, they receive no parental care. In ponds and lakes, the fry scatter into dense weed masses (page 48), and thus it becomes impossible for the male parent to keep the young together for close care. Protection afforded by the dense weed masses eliminates most of the needs for parental care. The male warmouth seems to lack the drive to care for his free- swimming young; even in a laboratory aquarium without vegetation in which the young may hide, he shows little interest in his frv after thev leave the nest. 68 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 Feeding BehaviorGroup Behavior The warmouth is not a gregarious fish, even though large numbers of individuals may be concentrated in a comparatively small area. I'he following observations concern the social relations among war- mouths. Aggregations.—There is no school formation among warmouths except that immediately associated with the nest (page 48). Aggregations form around desirable cover, such as the riprapping along a dam (page 8), but little social structure can be detected in such groups. Even during the winter, when many fishes form groups, warmouths show no tendency to gather together except in response to choice hab- itats. The nesting colonies that have been reported (page 42) are probably due to restricted nesting habitat rather than to a gregarious nature of the species. Hierarchy.—To what extent an order of dominance occurs in a natural war- mouth population has not been observed. Attempts at observations on dominance are hampered by the difficulty of identify- ing individual fish in a natural setting; also, the order of dominance becomes com- plicated by nesting behavior, mating ag- gression, feeding activities, and local move- ments. A hierarchy is quickly established among warmouths in a restricted group, such as that in an aquarium. The aggressiveness of a fish, as for food or space, and the dom- inance of the fish relative to other mem- bers of the group, determine its position in the hierarchy. The smaller the group the more stable and definite the order of dominance appears. In groups of more than three or four, the order may change frequently. Nesting studies in the labora- tory revealed that a male in spawning con- dition tends to assume dominance over one not so sexually advanced (page 42). The attitude of aggression which initiates the breeding behavior temporarily affects any existing hierarchy. Witt (1949:34) discovered a definite hierarchy among five warmouths in an aquarium. He found no correlation be- tween the order of dominance and the errors the fish made in learning to distin- guish a worm on a hook from a worm that is free. I Warmouths have a simple pattern of taking food. When a food item is sighted, the fish turns toward it, judges its accepta- bility as food, and then may move in quickly to snap it up. An unacceptable food item may hold the warmouth's atten- tion for several minutes. Seldom is a mo- tionless object picked up by a warmouth. Suction created as the warmouth quickly opens its wide mouth aids in the capture of food. This suction causes a loud noise when the fish gulps an item of food from the water's surface and may be responsible for taking a considerable amount of detri- tus with the food. Learning Witt (1949:27) found that warmouths could learn to distinguish a worm on a hook from a free worm. As isolated indi- viduals, warmouths learned about as quick- ly as did bluegills and more quickly than did largemouth bass, but in groups the warmouths made more errors than did either largemouths or bluegills. Individ- uals of all three species exhibited a fair de- gree of learning, making the majority of their errors in the first two of the seven trial periods. After being penalized for making an error, the warmouth was not so cautious as the bluegill in its approach to a hooked worm. Warmouths do not seem so cautious in taking fishing lures or so quick in recogniz- ing artificial situations as most other sun- fishes. In ponds and in laboratory aquar- iums, warmouths were seen to strike re- peatedly at artificial lures without, ap- parently, becoming suspicious that the lures were unnatural. In an aquarium, a rest- ] ing warmouth, molested by a succession of i lures dangled before its face, apparently was so undisturbed by the experience that ! it turned to snap at a lure more attractive I to it than the others. On several days at Ridge Lake, Coles County, Illinois, a fish- erman repeatedly hooked and released! what appeared to be the same large war- mouth by dangling a worm in front of am old piece of tile. This warmouth may have learned, but, if so, its memory did not last: from one fishing trip to the next. The warmouth's gullibility toward baits August, 1957 Larimore: Life History of the Warmouth 69 may be a desirable trait for a warm-water sport fish. ECONOMIC RELATIONS Warmouths attain their greatest im- portance as food and sport fish in the lower Mississippi River valley and states bor- dering the Gulf of Mexico. There they are commonly taken with live bait by cane- pole fishermen. In the midwestern and eastern states, warmouths usually are not taken in large numbers but are caught on a wide variety of baits and lures. Because of their gamyness and plumpness, they are attractive to most anglers. The Warmouth as a Food Fish The warmouth is now of little commer- cial value, partly because in most states its sale is illegal ; where it can be legally sold, the warmouth is not an important food fish in comparison to the larger species now be- ing marketed. In North Carolina during the early part of this century the warmouth was taken in gill nets and other nets and sold throughout the year (Smith 1907:235). At Reelfoot Lake, Tennessee, in 1937, it was one of the seven sunfishes that as a group comprised approximately 10 per cent of the weight of the commercial catch (Kuhne 1939Z':58). Most people consider the warmouth an excellent table fish. At times, however, this fish may have a "muddy" flavor, which is generally blamed on its association with silt bottoms and muddy waters, but which is caused at least partly by the food organ- isms comprising its diet. Warmouths taken off silt-covered bottoms of Park Pond usu- ally had an excellent flavor; they were in- termediate between the bluegill and the largemouth bass in both flavor and texture of flesh. The Warmouth as a Sport Fish An early angling critic, Henshall (1903:59), was very enthusiastic about the warmouth; he wrote, 'Tor its size, it is the gamest member of the family except the black-bass." In a discussion following a paper by Lovejoy (1903:120), Henshall pointed out that this sunfish takes a fly well, responds to almost any kind of bait, and is an excellent table fish. Evermann & Clark (1920:393), Baker (1937:44), Curtis (1949:266), and others have praised the fighting qualities of the war- mouth or have termed it "an excellent small game fish." The value of the warmouth as a sport fish is enhanced by the wide variety of nat- ural and artificial lures that are effective in catching it. Through most of its range, the warmouth is taken more commonly on natural baits (earthworms, minnows, grasshoppers, crickets, or grubs) than on artificial baits. The yield to the warmouth fisherman, using either natural or artificial lure, is often restricted by the difficulty of work- ing the lure in close enough to weed masses, brush, and other dense cover to present it properly to the fish without get- ting the hook snagged. This difficulty in- creases during the summer as aquatic vege- tation grows rank. Floating lures, such as poppers, are efifective during the summer, because they can be dropped in pockets of open water among water weeds—where the warmouths may be hiding, feeding or nesting—and then be lifted out without becoming entangled. Worm fishing with a long pole offers similar advantages in fishing for warmouths around dense vege- tation and heavy brush. Most Illinois fishermen believe that warmouths may be taken in greater num- bers during the spring and early summer than at other seasons. At the Pollv^vog Association property and at the flooded limestone quarries (Fairmount Quarries) near Fairmount, Vermilion County, good catches of warmouths are usually made in May and June but seldom later in the sum- mer—at least not on the artificial lures that are relatively effective during the ear- lier months. Most of the warmouths caught at Ridge Lake (Dr. George W. Bennett, unpublished creel records from Ridge Lake, Coles County, Illinois) have been taken during the first month of the summer fishing season. Although the catch of warmouths at Lake Glendale (Dr. Donald F. Hansen, unpublished creel rec- ords from Lake Glendale, Pope County, Illinois) was distributed rather evenly in the period May through August in 1945, the catch of warmouths in 1946 was much 70 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 higher in April and IVIay than during the summer months. No warmouths were taken at Lake Glendale during September in either year. These records form an in- teresting contrast with the records of Ricker (1945:330) for Muskellunge Lake, Indiana, where a striking increase in the catch of warmouths occurred during September. Censuses of sport fishing reported by Ricker (1945) for three Indiana lakes show that warmouths were taken regu- larly bv anglers but not in abundance. Lewis & English (1949:317) recorded only four warmouths taken during 6,513 man-hours of fishing in Red Haw Hill Reservoir, Iowa, even though warmouths were fairly common in the lake. They sug- gested that the low catch was due to the difficulties of angling among the dense marginal vegetation of this lake. Kuhne (1939rt:51) calculated a take of war- mouths at Reelfoot Lake, Tennessee, that amounted to 1.02 per cent by weight of the anglers' catch for 1937. The combined catches of resident and non-resident fisher- men amounted to only 0.02 warmouth per hour (Kuhne 1939rt:48). In a creel cen- sus for the period March 1 through Sep- tember 30, 1952 (Cobb 1953:21), war- mouths comprised 2.05 per cent of the weight of all fish taken by sport fishermen at Reelfoot Lake. Creel records for Illinois lakes show that the warmouth usually is not abundant in the anglers' catches. In Onized Lake, Illinois, Bennett (1945:380-3) reported only 105 war- mouths caught during 7,526.9 hours of fishing in a period beginning in 1938 and ending in 1941. This catch represented about 0.01 warmouth per hour. Even though this species comprised 18 per cent of the total number of fish (6 per cent by weight) in the final census of 1941, it made up only 2.6 per cent by numbers (2.9 per cent by weight) of all fish caught in the period of study. Since only a few war- mouths were caught during a period when the other sport fish were being severely cropped in this 2-acre lake, one might have expected warmouths to replace the other fish of desirable sizes removed by angling. However, only 13 warmouths of 6 inches or more in length were recorded in the final census. At Horseshoe Lake, Alexander County, in southern Illinois, 2 per cent of the fish- ermen's catch during the summer of 19563 was composed of warmouths (Brucel Muench, 1956, report to the Illinois De- partment of Conservation and Southern, Illinois University). At Venard Lake, 20 of the 101 war- mouths planted in this lake early in 1947 were caught by anglers later in the same year. This take represented about 20 per cent of the number of warmouths planted] but only 14 per cent of the total catch. About 52 per cent of the 240 largemouthj bass that had been planted with the war- mouths were taken by anglers in 1947. In 1946 at Lake Glendale, in southerni Illinois, the percentage of warmouths} in the anglers' creel was not far be- low the percentage of warmouths in the total fish population. Warmouths were first caught in Lake Glendale the thirc summer after impoundment of the lake ir 1940 and they increased in the anglers* catches during each of three summers fol- lowing their first appearance (Dr. DonaU F. Hansen, unpublished creel records fror Lake Glendale, Pope County, Illinois), In the last year of the 3-year sequence^ warmouths comprised 4.5 per cent of the total number of fish taken. When the lake was drained and the fish populatior censused, warmouths made up 6 per cent of the total number of fish and 5 per cent of the total weight; 57 per cent of the warmouths were over 6 inches in total length. Several central Illinois ponds that, as part of the life history study reported herej had been stocked with warmouths pro- duced hook-and-line yields that were lo\ in proportion to the populations of thes^ fish. The exploitation rate from angling was proportionally lower than for most other centrarchids inhabiting these waters. Fly and plug fishermen caught relatively few warmouths ; most large catches of war- mouths from these stocked ponds were taken on live baits. Warmouth populations in the creeks and rivers in most parts of Illinois contribute very little to the creels of anglers. How- ever, anglers who fish a few of the streams of southern Illinois report the common oc- currence of warmouths in their creels. The warmouth is probably not abundant August, 1957 Larimore: Life History of the Warmouth 71 enough in the Mississippi River from Ca- ruthersville, Missouri, to Dubuque, Iowa, to be considered of much importance in the sport fishery (Barnickol & Starrett 1951: 319). The Warmouth as a Laboratory Fish The warmouth is a desirable fish for laboratory experimentation. It is rela- tively easy to transport from the field and to keep alive in the laboratory. It is large enough to be easily handled and yet small enough to be accommodated in most aquar- iums. It has a quiet disposition, quickly be- comes adapted to laboratory conditions, and readily feeds on a wide variety of foods. In the laboratory, the warmouth will nest and spawn, apparently undisturbed by the presence of an observer. The wide va- riety of foods acceptable to it simplifies the task of keeping this fish for long periods in the laboratory. Such characteristics as its tolerance for low concentrations of dis- solved oxygen, its rapid color responses to excitement, and its unusual individual and group behavior present interesting prob- lems for study. The warmouth has been used in Natural History Survey laborato- ries in studies of food conversion, learning, group behavior, and marking techniques, as well as in studies reported in the pres- ent paper. The Warmouth in Artificially Established Populations Several combinations of species have been used by fisheries biologists in seeking to establish fish populations that will pro- duce and maintain good sport fishing in lakes and ponds. In some waters, the largemouth bass and bluegill have seemed to be suitable companion species (Swingle & Smith 1941:271). In many Illinois lakes, however, this combination has not proved satisfactory, as bluegills have tended to overpopulate the water (Bennett 1944:186). Lovejoy (1903:116-7) considered the warmouth one of the three best species to be used in stocking small ponds in the south. He wrote, "It grows to much larger size than the bream, thick and fleshy, with large mouth, and is to some extent cannibalistic, but not enough so to make it objectionable. It will eat a few of its own young, but not enough to miss them —just enough to make the balance grow well." The stocking of inland waters with war- mouths for sport fishing was begun before the turn of the century. Records indicate that the distribution of warmouths by state and federal agencies has been sporadic and probably never on a large scale. For ex- ample, an Oklahoma state agency distrib- uted 36,300 iingerlings in the calendar year 1935, and the United States Bureau of Fisheries distributed 53,160 fingerlings in the fiscal vear ending June 30, 1936 (Earle 1937:16, 23). In the 12-month period beginning September 1, 1946, a Texas state agency distributed 134,345 warmouth fingerlings, and in 1947 the United States Fish and Wildlife Service distributed 20,348 warmouth fingerlings and 20 warmouths at least 6 inches in length (Tunison, Mullin, & Meehean 1949:55,58). The Fish and Wildlife Service distributed 64,040 warmouth fin- gerlings in 1949 and 710 in 1950 (Dun- can & Meehean 1953:5-6); 4,600 war- mouth fingerlings and 610 warmouths at least 6 inches long in 1951 and none in 1952 (Duncan & Meehean 1954:4-5). In Alabama, Swingle (1950:49-73) stocked 10 of 34 experimental ponds with warmouths in combination with large- mouth bass, bluegills, and other fishes. Seven of the 10 ponds containing war- mouths produced populations that were considered balanced and 3 produced popu- lations that were considered unbalanced. Warmouths comprised less than 6 per cent of the total weight of fish in all but 1 of the 10 ponds, a pond with a population judged to be unbalanced ; in this pond warmouths made up 11.3 per cent of the weight. Bluegills far outnumbered the warmouths in each population. The relatively low proportions of war- mouths encountered (usually less than 10 per cent by weight, table 4) indicate that these fish have no tendency to become dominant at the expense of other kinds of fishes. However, even these low propor- tions may represent overcrowding for the warmouths themselves, as indicated by slow growth and the occurrence of a high percentage of small individuals reported in 72 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 several censuses. Growth studies in Illi- nois indicate that as low a proportion of warmouths as 10.4 per cent by weight, found in Park Pond Slough (a weed- choked channel in Park Pond) may repre- sent overcrowding for these fish. Growth of warmouths in this channel was consid- erably slower than was that of warmouths in Onized Lake, just preceding 1941, when warmouths made up 6.5 per cent of the total weight of the fish population (Bennett 1945:382, 397), and slower than that of warmouths in Lake Glendale just preceding 1946, when warmouths made up 5.0 per cent of the total weight (unpublished information from Dr. Don- ald F. Hansen of the Illinois Natural His- tory Survey). In Onized Lake, the fish population had been thinned by excessive fishing, and in Lake Glendale the popula- tion had been expanding during the 6 years following impoundment of the water. Experimental Species Combina- tions.—As part of a series of manage- ment experiments by the author, 17 ponds in central Illinois were stocked with war- mouths in various combinations that in- cluded largemouth bass, smallmouth bass, several pan fishes, and minnows. Because these experiments have not yet been com- pleted and because they are not an integral part of the life history study reported here, the stocking combinations are listed below with consideration given princi- pally to the early development of the pop- ulations and such factors as directly relate to the life history of the warmouth. Warmouths (Adults); Largemouth Bass (Fingerlings and Yearlings).—This combination of species and sizes was first tested in 3-acre Enright Pond over a pe- riod of 15 months. Sixteen adult war- mouths, 4 yearling largemouths, and 60 fingerling largemouths per acre were re- leased in May, 1947. Warmouths spawned the first summer, and both species pro- duced broods of young the second sum- mer. Growth of all fish was rapid ; some of the first-brood warmouths attained lengths as great as 6 inches in a little more than a year. There was a desirable distri- bution of numbers in size groups of both species. Warmouths (Fingerlings and Adults); Largemouth Bass (Fingerlings and Adults).—Both species were introduced in numbers and sizes simulating a "pyramid of numbers." This combination was tried in Enright Pond after termination of the experiment described above; the popula- tion was established during the early fall months of 1948 with 90 fish of each spe- cies per acre. Moderate-sized broods of both species were produced the next sum- mer, and in each of the seven following summers the population was studied. The striking difference between what oc- curred in this warmouth-largemouth pop- ulation and what usually occurs in a blue- gill-largemouth population was that in the Enright Pond population the bass success- fully produced a brood each year and fish of the companion species (warmouths in Enright) never produced such large num- bers of young that they dominated the pop- ulation. In 187 hours of recorded fishing during the sixth summer (there were fewer records for other years), 98 large- mouths and 16 warmouths were caught at a rate that averaged 0.6 fish per hour. The number and sizes of fish of each spe- cies in this population were more nearly constant from year to year than in popu- lations started with fish of one size. Warmouths (Adults); Largemouth Bass (Yearlings and Adults).—Venard Lake was stocked in 1947 with 32 adult warmouths and 70 yearling and 5 adult largemouths per acre. The growth and competition for food in this population have been discussed previously in this pa- per. The bass gained an early dominance! over the warmouths; by the end of the] third growing season, the lake was be- coming overcrowded with bass. Warmouths (Adults); Largemouth*^ Bass (Adults).— Fifteen adult war- mouths and 22 adult largemouths per acre were released in Reece Pond in May, 1949. This 2.5-acre pond was character- ized by a large proportion of shallow water and dense masses of aquatic vegetation (Potamogeton foUosus). Both species of fish spawned the first summer and they produced broods in each of the 7 succeed- ing years. The extensive vegetation per- mitted the survival of more young fish than could grow well in this pond. Warmouths (Adults); Largemouth Bass (Adults) Added 1 Year Later.—In April and May, 1948, approximately 20 August, 1957 Larimore: Life History of the Warmouth 73 adult warmouths were released in a 1-acre pool above Venard Lake. They produced a large brood in the summer of 1948. The following spring about 20 adult large- mouths were added to the pool. A small brood of bass was spawned, and the young grew well ; by the end of the summer they were feeding on small warmouths. This combination and sequence of setting up the population allowed the warmouths to become well established, may have limited of sport fishes—was investigated in three populations containing warmouths and largemouths. Three adult warmouths and 30 finger- ling largemouths per acre were released in June, 1952, in Parkhill Pond, a 3-acre pond which contained a large established population of the bullhead minnow, Pime- phales vigilax (Baird & Girard). When the study was terminated at the end of about a year, which included parts of Fig. 25.—Central part of Kearney Pond, McLean County, stocked with warmouths and largeraouth bass. the size of the first bass spawn, and pro- vided small forage fish for the bass. Warmouths (Established Population) ; Largemouth Bass (Adults).—In June, 1951, Kearney Pond (2.5 acres), fig. 25, containing a small population of war- mouths, principally yearlings, was stocked with 5 adult largemouths per acre. In the following October, 20 more adult bass per acre were added, along with 40 finger- ling and adult warmouths per acre. The warmouths produced a moderate-sized brood in 1951. In 1952, the largemouths produced a large brood, the warmouths a relatively small one. This relative spawn- ing success of the two species was main- tained in each of the 3 following years, or until the study was terminated. Warmouths (Adults); Largemouth Bass (Fingerlings) ; Minnows.—The in- fluence of minnows—both as a forage item and as a predator on the eggs and fry two breeding seasons, there was an abun- dance of minnows, the bass and warmouths had grown exceptionally fast, and the warmouths had produced broods the first and second summers. The largemouths, which were 10 to 12 inches in total length early in the second season, did not produce a brood. Since the warmouths spawned successfully even though an abundant min- now population was present, it seems likely that the largemouths would have produced a brood the next year. In July, 1948, Lutz Pond contained a large population of several species of min- nows, the most abundant of which was the bullhead minnow. This 1.5-acre pond was then stocked with 20 adult warmouths and 60 fingerling largemouths per acre. War- mouths spawned the first summer (1948) and produced a large brood; these young fish grew riapidly. The warmouths pro- duced another brood (1949) before the 74 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 2-year-old largemouths spawned in 1950. The minnow population declined rapidly during!; the third summer. Kearney Pond (2.5 acres) contained minnows (species unidentified) and dart- ers, Etheostorna ni(jritm Rafinesque, when it was stocked in June, 1949, with 10 adult warmouths and 4b fingerling largemouths per acre. The warmouths, minnows, and darters reproduced well in the first sum- mer. Growth of the sport fishes was good. The numbers of minnows and darters de- clined during the summer. In Parkhill Pond and Lutz Pond, two broods of warmouths were produced before the first spawn of bass. In all three ponds, the abundance of small forage fish provided food for the warmouths and largemouths, which grew rapidly; large- mouths provided some fishing the second summer; and a bass brood of moderate size was spawned the third summer in the presence of two broods of warmouths. Warmouths (Fingerlings and Adults) ; Minnows.—In August, 1948, 34 finger- ling and 14 adult warmouths per acre were introduced into Longworth Pond (2 acres), which contained a large population of fathead minnows, Pimephales promelas Rafinesque. The warmouths spawned suc- cessfully during the remainder of the 1948 season and again the following summer. The broods in both years were small ; evi- dently the minnows had a depressive effect on the warmouth population. Warmouths (Adults); an Established Sunfish Population.—Three experiments were conducted to see if a small number of warmouths could successfully reproduce and survive in an established population consisting of several species of sunfishes. Seven large adult warmouths were planted in a one-half-acre pond, Green Gravel Pit, which at the time (June, 1947) contained a population of bluegills, redear sunfish, green sunfish, and large- mouth bass. Only one warmouth (orig- inal stock) was recovered when poison was applied to the pond in August, 1948. The warmouths had failed to establish a brood during the two intervening spawning sea- sons. In November, 1949, Taylor Pond (2 acres) was stocked with 69 adult war- mouths per acre. A few weeks before, it had been stocked with 100 bluegill fin- gerlings, 100 largemouth fingerlings, and 15 largemouth adults per acre; a few adult green sunfish, longear sunfish, and bluegills also were added. The population was killed during the second spawning season (June, 1951); two adults, each a half pound in weight, were the only warmouths recovered from a rather large sunfish pop- ulation (337 pounds per acre). Twenty-three adult warmouths were re- leased in June, 1952, in a 3-acre pond, McCarty, which contained a new but large population of bluegills and largemouth bass. No young warmouths (definitely identified) were taken from this pond dur- ing the following 4 years. Warmouths (Adults); Redear Sunfish (Adults); Smallmouth Bass (Adults and Fingerlings).—This combination of spe- cies was tested in two ponds. In July, 1951, 2-acre Taylor Pond (mentioned in connection with another ex- periment) was stocked with 21 adult war- mouths per acre. These fish produced a small brood in the same summer. In the following fall and spring, 10 fingerling and 7 adult smallmouth bass and 17 adult redears per acre were added. A small brood of smallmouths, a moderate-sized brood of warmouths, and a relatively large brood of redears were produced in the summer of 1952. Observations the next 2 years revealed the following: The smallmouths produced a very small brood in 1953 and no brood in 1954. Growth of the original stock of bass was good, but growth of both the 1952 and 1953 year classes was very poor. The warmouths and redears reproduced successfully each year and at first grew at satisfactory rates ; however, by the spring of 1955 there were relatively few over 6.5 inches in length. Sparks Pond (3 acres), fig. 26, was stocked with 22 adult smallmouth bass in November, 1949. The following spring the smallmouths spawned very successfully. In June, 25 adult warmouths and 34 adult redear sunfish were added to the popula- tion of this pond. Both of these species re- produced, although the brood of war- mouths was quite small in numbers. During the following 7 years, these ob- servations were made: Smallmouth bass of the first brood (1950) did not grow well after the first summer. The bass spawned August, 1957 Larimore: Life History of the Warmouth 75 Fig. 26.—South side of Sparks Pond, Woodford County, stocked with warmouths, redear sunfish, and smallmouth bass. each year, but the fingerlings disappeared before attaining 1.5 inches in length. The only successful brood of bass after the first was that produced in 1954, which came after the redear sunfish population had been reduced in numbers by poison ap- plied to part of the pond, a tremendous number of small sunfish had been lost over the spillway during a severe flood, and bass of the 1950 brood had become less nu- merous. Growth in this 1954 brood of bass was poor. The redear sunfish spawned very suc- cessfully each year. The original stock and the first brood grew very well. Broods produced later showed much slower growth. After the fourth year, there were very few redears over 6 inches in length, although redears of smaller sizes were nu- merous. The warmouth population was slow to develop. Warmouths spawned successfully each year, but the broods produced were small. However, by the fourth year war- mouths were numerous and had become large enough to be attractive to anglers. Warmouths ; Largemouths ; Bluegills (Adults of One Sex).—Four experiments were set up in attempts to produce war- mouthXbluegill hybrids. Kearney Pond (mentioned in connec- tion with other experiments) contained a 4-year-old warmouth-largemouth popula- tion when 8 adult female bluegills per acre were added, July, 1955, in an effort to produce hybrids with the warmouths. No hybrids were found in the two spawn- ing seasons after the bluegills were added. Dunmire Pond (4.5 acres), fig. 27, was stocked in May, 1950, with 16 adult war- mouths, 19 adult male bluegills, and 100 fingerling largemouth bass per acre. In July, 1955, 10 more adult male bluegills per acre were added. The warmouths and largemouths grew well, spawned success- fully each year, and produced good fish- ing. The male bluegills grew exception- ally large (1.2 pounds), but no hybrids were observed in the first 6 years after the pond was stocked. A shallow 3-acre pond on the Univer- sity of Illinois Golf Course near Savoy was stocked with 45 adult warmouths, 185 adult female bluegills, and 907 largemouth bass fingerlings. These fish were placed in the pond in two groups, one group in each of the summers of 1949 and 1950. In the third summer following the original stock- ing, a large brood of bluegills was pro- duced; one or more male bluegills must have been accidentally introduced in 1950. No warmouth Xbluegill hybrids were col- lected from this pond. During the summers of 1949 and 1950, 19 adult warmouths and 59 adult male bluegills were released in Green Gravel 76 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 Pit, mentioned in connection with another experiment. Broods of warmouths were produced in each of these two summers and the two followinji summers that the study was continued. The male hluegills built nests, but no warmouthXbluegill minnow population and grow well when they become large enough to utilize the minnows as food. Largemouth bass in a pond with war- mouths apparently grow faster and pro- duce better fishing than do smallmouth Fig. 27.—North arm of Dunmire Pond, Woodford County, stocked with warmouths, large- mouth bass, and male bluegills. hybrids were collected. This half-acre pond, in which a substantial warmouth population had been developed, should have offered a desirable situation for hy- brid production. General Conclusions About Spe- cies Combinations.—Several general conclusions may be drawn from the pre- liminary observations on these experimen- tal populations. Usually when sexually mature warmouths are released in a pond before the middle of August, they will pro- duce a brood the same summer. In estab- lished warmouth populations, a high pro- portion of each new brood is spawned so late in the season that the fish are too small in their second summer of life to reproduce then. Small numbers of war- mouths when introduced into a pond over- crowded with other sunfish seem unable to establish a population. Warmouths repro- duce successfully in the presence of a large bass in a pond with warmouths. There is little difference in growth rates be- tween warmouths that develop in a pond with largemouths and those that develop in a pond with smallmouths. Warmouths will not establish a large enough popula- tion to support angling and materially reduce the survival of young bass unless adult warmouths are introduced a year be- fore adult bass are added, or unless fin- gerling bass, instead of adults, are intro- duced with the adult warmouths. There is no assurance that hybrids will be produced when bluegills of only one sex are introduced into a warmouth popu- lation. SUMMARY 1. The ecological life history of the warmouth, Chaenobryttus gulosiis (Cuv- ier), was studied intensively in two habi- August, 1957 Larimore: Life History of the Warmouth 77 tats of central Illinois: Venard Lake, a 3.2-acre artificial impoundment stocked only with warmouths and largemouth bass, and Park Pond, an 18-acre flooded strip- mine area containing a fish population of 36 species. The intensive investigations in these two areas were supplemented by ob- servations in other habitats and by pub- lished records on warmouth habitats and populations. 2. Field observations and published rec- ords indicated that the warmouth is usu- ally associated with habitats characterized by soft bottoms and dense stands of aquatic vegetation. 3. In the water areas under observa- tion, small and medium-sized (less than 5 inches total length) warmouths remained in protected areas of shallow water throughout the year, whereas larger indi- viduals spent more time in deep, open wa- ters. 4. Laboratory experiments supported field observations demonstrating that war- mouths are able to survive in water having very low concentrations of dissolved oxy- gen. The critical oxygen tension observed was 2.5 cc. per liter at 20 degrees C. Tol- erance for low oxygen concentrations al- lows the warmouth to survive and grow in a wide range of habitats and to survive during periods of water conditions that are generally considered unfavorable to fish. 5. The food habits of warmouths from Park Pond and Venard Lake were studied through a 12-month period. In volume and frequency of occurrence, the various food items identified in warmouth stom- achs showed little similarity in the two areas, although crayfish and nymphs of mayflies, dragonflies, and damselflies were abundantly utilized at both places. Dur- ing the summer months, feeding activity was at a peak early in the morning; it practically ceased in the afternoon. 6. Postlarval warmouths observed in the laboratory fed first on protozoa and bacteria. There was a general increase in size of food items taken by warmouths of Park Pond and Venard Lake as the fish increased in size; the percentage of stom- aches that were empty was higher among large fish than among smaller ones. 7. In Venard Lake, warmouths and largemouth bass consumed about the same kinds of foods, but differences in their feed- ing habits may have prevented extensive competition between these species. 8. Seasonal changes in appearance and weight of gonads indicated that the war- mouths collected from Park Pond and Venard Lake attained sexual maturity when between 3.1 and 3.5 inches total length and that fast-growing fish matured earlier in life than did slow-growing ones. Warmouths over 5.4 inches total length attained spawning condition earlier in the nesting season, and spawned over a longer period, than did fish of smaller sizes. Males matured slightly earlier in the season than did females. In central Illinois, the spawn- ing season for warmouths generally extends from mid-May through mid-August. 9. An estimation was made of the total number of eggs in ovaries of warmouths of different sizes, from different water areas, and taken at different times of year. Total egg counts ranged from 4,500 to 63,200 per ovary. Females from Park Pond consistently produced fewer eggs than did those from Venard Lake. 10. A month before the beginning of the spawning season, groups of developing eggs began moving away from the primor- dial egg-stock in sexually mature war- mouth females. There was a gradual with- drawal of eggs from the egg-stock through- out the spawning season. Ova in advanced maturation were resorbed if not spawned before the cessation of nesting. 11. In all instances of warmouth nest- ing observed in the field, the male con- structed the nest, usually near some pro- jecting object and on a bottom of loose rubble containing some silt and detritus. No colony formation was observed. 12. Sex recognition among warmouths observed in the laboratory was based ap- parently on behavior and response to court- ing. Males displayed temporary color changes during courtship and spawning. There was evidence that many males and females spawned two or more times dur- ing a summer; in some instances, more than one female contributed to the com- plement of eggs in a nest. 13. In the laboratory, incubation of eggs lasted about 34.5 hours at tempera- tures between 25.0 and 26.4 degrees C. Immediately after hatching, the prolarvae dropped to the bottom of the nest. The 78 Illinois Natural History Survey Bulletin Vol. 27, Art. 1 i volk supply was exhausted in 4 days, and the larvae attempted feeble, poorly di- rected jumps. By the fifth day, they swam actively. They began feeding by the sev- enth day; considerable pigmentation had developed and the caudal fin appeared homocercal. The 15.7-mm. young were essentially like an adult in body form. 14. The mathematical relationship (in inches) between the anterior radius of a warmouth scale magnified 41 times (S) and the total length of the fish (L) was expressed bv the equation : L—0.5278+ 1.048 S 15. In the populations studied, as the body length of the warmouth increased, the tail became relatively shorter; differ- ent mathematical relationships between standard length and total length were found for fish of various sizes. 16. The relationship of standard length in millimeters (L) to weight in grams (W) was expressed for 866 Park Pond warmouths by the equation : log W= -4.49867+3.04902 log L 17. The coefficient of condition (C) for 866 warmouths from Park Pond showed no consistent seasonal cycle. Sea- sonal variations in condition were greater in warmouths between 3.3 and 4.2 inches than in larger fish. Coefficient of condition increased progressively with increase in size of fish. 18. The annulus was found to be a re- liable year-mark in the warmouth. War- mouths in Park Pond completed the 1949 annulus between April 7 and May 20. Warmouths in Venard Lake completed the 1949 annulus over a shorter period than did those in Park Pond, where ecological conditions in the habitat varied greatly. Dredging of the shore of Venard Lake during August, 1948, is believed to have caused the formation of a false annulus. 19. Females from Park Pond were con- sistently smaller than males of the same ages. The difference was small, however. 20. Age was determined for 1,063 war- mouths from Park Pond ; it was found that fish of certain year classes had con- sistently grown more rapidly than others. Growth for all year classes was better in certain years than in others. 21. The 1946, 1947, and 1948 year classes in Park Pond showed different growth patterns for the summer of 1949. The fish in each year class grew rapidly during May and June. Although growth continued through the summer for the younger fish, it declined rapidly after June for the 1946 year class. The 1947 year class showed a growth pattern inter- mediate between the earlier and later year classes. Growth rates were different for warmouths in different parts of Park Pond. 22. A comparison of length increments for the first and for later years of life showed that warmouths in Park Pond with the greatest length increment for the first year added to this length advantage the second growing season. Fish that grew slowly the first year showed a slight growth compensation during the third year, although they did not overcome the length advantage held by the larger fish. 23. Three year classes, represented by 1,102 warmouths, were studied in Venard Lake. Fish of the first year class spawned in the lake grew faster than did those of succeeding year classes. The length range in a single year class was greater during the first summer than in succeeding years. The average growth in length for war- mouths in their first 3 years in Venard Lake was similar to that for warmouths of comparable ages in Park Pond. 24. Warmouths in Park Pond were heavily infested with Posthodiplostomum minimum, Proteocephalus ambloplites, and Camallanus oxycephalus. Except for an infestation of the leech, Illinobdella moorei, warmouths in Venard Lake were relatively free of parasites. No direct harm- ful effect of parasites was established. 25. Laboratory and field observations showed that the warmouth has a quiet dis- position. In its reproductive and group behavior, it is similar to other centrarchids, but it displays certain behavioral char- acteristics peculiar to the species. 26. Reports and field observations dem- onstrated that the warmouth is caught on a wide variety of baits and lures, and that warmouth fishing is best during the spring and early summer months. The warmouth has been praised by sport fishermen for its fighting qualities. It is a useful fish for laboratory experimentation. 27. That warmouths have no tendency to become dominant at the expense of other kinds of fishes was indicated by the rela- August, 1957 Larimore: Life History of the Warmouth 79 tively low proportions of warmouths re- cies—largemouth bass, smallmouth bass, ported in fish populations of Illinois and and several pan fishes—warmouths tended other states. In 17 ponds in central Illinois to establish small broods each year without stocked with 11 different fish combinations seriously restricting the reproduction or that included warmouths with other spe- growth of companion species. LITERATURE CITED American Fisheries Society 1948. A list of common and scientific names of the better known fishes of the United States and Canada. Am. Fish. Soc. Spec. Pub. 1. 45 pp. Bailey, Reeve M. 1956. A revised list of the fishes of Iowa, with keys for identification. Iowa Ag. Exp. Sta. Jour. Paper J-29I4: 325-77 ; a reprint of pp. 325-77 of Iowa fish and fishing, 3rd ed., by James R. Harlan and Everett B. Speaker, published 1956 for Iowa State Conser- vation Commission, [Des Moines], 377 pp. Bailev, Reeve M. (Chairman) 1952. [Report of] Committee on Names of Fishes. Am. Fish. Soc. Trans. 81 (1951) : 324-7. 1953. [Report of] Committee on Names of Fishes. Am. Fish. Soc. Trans. 82(1952) :326-8. Baker, C. L. 1937. The commercial, game, and rough fishes of Reelfoot Lake. Tenn. Acad. Sci. Jour. 12(l):9-54. Bangham, Ralph V. 1939. Parasites of Centrarchidae from southern Florida. Am. Fish. Soc. Trans. 68(1938): 263-8. Bangham. Ralph \ ., and Carl E. Venard 1942. Studies on parasites of Reelfoot Lake fish. IV. Distribution studies and checklist of parasites. Tenn. Acad. Sci. Jour. 17(l):22-38. Barnickol. Paul G., and William C. Starrett 1951. Commercial and sport fishes of the Mississippi River between Caruthersville, Missouri, and Dubuque, Iowa. 111. Nat. Hist. Surv. Bui. 25(5) :267-350. Beck, John R. 1952. A suggested food rank index. Jour. Wildlife Mgt. 16(3):398-9. Bennett, George W. 1943. Management of small artificial lakes: a summary of fisheries investigations, 1938-1942. 111. Nat. Hist. Surv. Bui. 22(3) :357-76. 1944. The effect of species combinations on fish production. N. Am. Wildlife Conf. Trans, 9:184-90. 1945. Overfishing in a small artificial lake: Onized Lake near Alton, Illinois. 111. Nat. Hist. Surv. Bui. 23(3):373-406. Bennett, George W., David H. Thompson, and Sam A. Parr 1940. Lake management reports. 4. A second year of fisheries investigations at Fork Lake, 1939. 111. Nat. Hist. Surv. Biol. Notes 14. 24 pp. Black. John D. 1945. Natural history of the northern mimic shiner, Notropis volucellus