Fish respiration in the natural oxygen gradient of an anaerobic spring in Florida

MISSING IMAGE

Material Information

Title:
Fish respiration in the natural oxygen gradient of an anaerobic spring in Florida
Series Title:
Copeia
Physical Description:
Article
Language:
English
Creator:
Odum, Howard T.
Caldwell, David K.
Publication Date:

Subjects

Subjects / Keywords:
springs
oxygen gradient
fisheries
fish respiration
Spatial Coverage:
United States -- Florida -- Putnam -- Welaka -- Beecher Springs
Coordinates:
29.48 x -81.67

Notes

General Note:
Pages: 104-106

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
System ID:
AA00004062:00001


This item is only available as the following downloads:


Full Text


Reprinted from COPEIA, 1955, No. 2, pp. 104-106, May 20
Printed in U.S.A.














Fish Respiration in the Natural Oxygen Gradient of an
Anaerobic Spring in Florida

HOWARD T. ODUM AND DAVID K. CALDWELL


ON rare occasions one finds in nature an
ecological circumstance so perfectly com-
posed that it constitutes.a natural experiment
superior to any man-made construction that
could be built at reasonable cost in the labora-
tory. Definite inferences can then be obtained
with a small effort. Such natural experimental
situations have additional value if they are per-
manent enough for data to be cumulative. In
this communication some observations made in
one of the remarkable Florida springs are used
to draw conclusions about the respiration of
fishes. It is hoped that others will be attracted
to these permanent chemostatic gradients.
Beecher Springs, adjacent to the University
of Florida Conservation Reserve, Welaka, Put-
nam County, emerges from a limestone aquifer
without oxygen. As it flows downstream in its
run (Fig. 1) this small stream picks up oxygen
by diffusion and by photosynthesis of the algae.
At a constant temperature diurnally and sea-
sonally of 22 to 230C., the physical and biologi-
cal characteristics of this spring seem constant
as surveyed 6 times over a period of 3 years.
Extensive work on trace elements in another
spring, Silver Springs, Marion County, indi-
cates that a chemostatic condition is to be
expected in the springs of the region. Thus
Beecher Springs seems to constitute a perma-
nent oxygen gradient in which organisms have
become adjusted. White sulfur bacteria and
blue-green algae dominate the boil with other
algal types appearing downstream, and finally
higher plants appear 300 meters from the boil.
The algae have been studied by Larry A. Whit-


ford, North Carolina State College, and will be
the subject of a forthcoming paper by him.
Midges (Tendipedidae and Heleidae) are found
throughout the run and in part are the basis of
the food chain leading to some of the fishes.
The oxygen values on the map (Fig. 1) were
obtained with the Rideal-Stewart-Winkler
method. Essentially similar values were re-
corded on June 6, 1952 and a year later on June
22, 1953. There are clearly eddies of water in
the stream so that oxygen conditions at any one
point fluctuate slightly with a range of about
1 ppm. Some idea of this variation from eddy
to eddy can be obtained from the data in
Table I. The analytical accuracy is about 0.1
ppm plus or minus.
The water is cear and the boil area smells
strongly of sulfurous gases. Other properties
are as follows: carbon dioxide, 10 ppm; pH,
7.2-7.7; chlorides, 166 ppm; phosphorus, 0.14
ppm; nitrate nitrogen, 0.02 ppm; alkalinity,
105 ppm (CaCOa); hydrogen sulfide, 0.02 ppm.
The natural distribution of fishes in this
natural gradient was observed with the aid of
a face mask by swimming underwater where
clear visibility is 30 meters or more. The ob-
served species distribution is indicated in Fig. 1
and is listed below. The eddies of water permit
local temporary microhabitats to acquire more
oxygen so that fishes can move into the higher
oxygen spots to replenish oxygen and thus can
live in parts of the stream where the average
oxygen may be below their physiological toler-
ances. Fig. 1 shows ecological tolerances under
stream conditions by indicating the farthest





H. T. ODUM AND D. K. CALDWELL-ANAEROBIC SPRING


N


100 METERS
Fig. 1. Map of Beecher Springs, Welaka, Florida. Dissolved oxygen values are given in parts per million.
The nearest that the fish were observed to come to the boil is indicated for 12 species in the middle of a
sunny day (June 24, 1953).


point upstream that each species was observed
to go. It is possible that the decrease of the small
hydrogen sulfide concentration downstream is
also involved. In order of distribution down
the spring run the fishes represented are: eastern
gambusia, Gambusia affinis holbrooki (Girard);
sailfin molly, Mollienesia latipinna LeSueur;
Least killifish, Heterandriaformosa Agassiz; red-
fin killifish, Lucania goodei Jordan; shellcracker,
Lepomis microlophus (Gunther); redfin pickerel,
Esox americanus Gmelin; redbreast sunfish,
Lepomis auritus (Linnaeus); largemouth bass,
Micropterus salmoides (Lac6pede); chubsucker,
Erimyon succetta (Lac6pede); bluegill, Lepomis
macrochirus Rafinesque; golden shiner, Notemi-
gonus crysoleucas (Mitchill); and bowfin, Amia
calva Linnaeus.

EXPERIMENTS IN SUBMERGED CAGES
Although Habs and Mueller (1948) found
that the eastern gambusia lived readily in 0.3
ppm oxygen, the abundance of three cyprino-
dont species in the churning outflow headwaters
(boil) where the oxygen is less than 0.3 ppm
day and night seemed remarkable. These fishes
are observed to break the surface continually.
-vollys feeding on the bottom swim to the sur-
face frequently before returning to the bottom.
Although gulping by these fishes has long been
observed, the writers do not know of a cear
demonstration of the.thresholds at which gulp-
ing is necessary for these widespread fishes.
The following experiments were conducted
within the springs.


TABLE I
DISSOLVED OXYGEN, IN PARTS PER MILLION, IN
BEECHER SPRINGS, FLORIDA
For 1953, three analyses for each station are preceded by the
mean of these values. For stations, see Fig. I
Station VI:6:1952 VI:24:1953

Boil A 0.05 0.3 (0.27, 0.30, 0.50)
B 0.26 0.6 (0.50, 0.59, 0.65)
C 0.9 1.1 (0.40, 0.65, 2.16)
BoilD 1.3 0.2 (0.21, 0.13, 0.29)
E 1.7 1.3(1.26, 1.31, 0.99)
F 1.1 2.0 (2.30, 1.80, 2.01)
G ... 2.3 (3.51, 1.15, 2.32)
H 3.6


Gambusia and Mollienesia taken from the
boil were placed in the following situations:
1. Out of water in a dry box on land.
2. In cages in the boil at the surface so that
fish could gulp at the surface.
3. In 4 cages submerged in the boil so that
fishes could not get closer than 1 inch to
the surface. One cage was near the surface
and one was at a depth of 2 feet. A third
cage included only Gambusia and a fourth
only Mollienesia.
4. Submerged below the surface downstream
where the oxygen was 1.3 ppm. (See Fig. 1.)
The amounts of time required for the fishes
to die are summarized in Table II. Where
oxygen was inadequate, distress was noticed
within three minutes and most of the fishes
were dead in about half the time necessary for
all to succumb. Although Gambusia and
MoUienesia are readily observed to pump the
surface film over their gills in stagnant waters








COPEIA, 1955, NO. 2


TABLE II
FISH SURVIVAL EXPERIMENTS IN BEECHER SPRINGS,
JANUARY 11, 1953


Experimental situation Nu tbe


Cage in the boil at the surface
of the water, fish gulping
air*................... 37
Completely out of water*..... 30
Cages completely submerged
in boil:
Gambusia, avg. weight: .23
gm ................... 27
Mollieesia, avg. weight:
0.37 gms............. 27
Gambusia and Mollienesia:
Below the surface 1 inch.. 24
Below the surface 2 feet... 48
Cage completely submerged
downstream*............. 20


* Gambsia and Mollienesia.


where oxygen is low, this is not the means of
respiration in this turbulent but anaerobic
stream. Definite gulping of air is observed. That
gulping is necessary at these low tensions is
demonstrated by the survival of those fishes
with access to the surface. Since there was a
rapid turbulent current cascading upward
through the experimental cages an oxygenated
surface film could not develop. When pre-
vented from gulping, the fish drowned quickly.
Those fishes kept out of water were quite alive
at the end of 22 minutes since they demon-
strated normal behavior when returned to the
water.
Whereas the respiration of air by fishes has
been much studied, the mechanism of aerial
respiration in small cyprinodont fishes seems to
have been neglected. Our dissections indicate
that the Beecher Springs individuals do not


differ from Gambusia affinis holbrooki elsewhere
in having a closed swim bladder, a short intes.
tine opening directly into a short oesophagus,
and pseudobranchs (Vialli, 1926). These latter
vascularized structures lining the pharynx may
be the structures used as a respiratory surface
for gulped air.
Another anaerobic spring in which obligate
gulping by fish may be directly inferred is the
large Warm Salt Spring in Sarasota County,
Florida. Here, tarpon, Tarpon atlanticus (Valen-
ciennes); gar, Lepisosteus sp.; southern varie-
gated cyprinodon, Cyprinodon variegatus varie-
gatus (Lac6pede); sailfin mollys, and eastern
gambusia are found in about half sea water
salinity (17,800 ppm) and a temperature of
31C. The oxygen content is between 0.00 and
0.81 ppm day and night. The tarpon were ob-
served to roll and the smaller fishes to break
the water continually. That tarpon and gars
breathe with a vascularized air bladder is
known from observations and dissections sum-
marized by Babcock (1951) for the tarpon and
Jones and Marshall (1953) for the gar.
These studies were aided by a contract be-
tween the Office of Naval Research, Depart-
ment of the Navy, and the University of Flor-
ida, NR 163-106.

LITERATURE CITED
BABCOCK, Louis L. 1951. The Tarpon. 5th ed.
Privately published, 157 pp., illus.
HABS, H., AND G. MUELLER. 1948. Uber das
Verhalten der Gambusien gegenuber dem Sauer-
stoffgehalt des Wassers. Zeitschr. Hgy. u. Infek-
tienskrankh. 128: 371-8. (Abstract)
JONES, F. R., AND N. B. MARSHALL. 1953. The
structure and function of the teleostean swim
bladder. Biol. Rev. 28: 16-83.
VIALLI, M. 1926. Le pseudobranchie dei Pesci
Arch. Ital. Anat. Embriol., 23: 49-117 (Abstract)

DEPARTMENT OF BIOLOGY, UNIVERSITY OF
FLORIDA, GAINESVILLE, FLORIDA.




Full Text

PAGE 1

Reprinted f r om COPEIA, 1955, No.2, pp 104-106, .. y 20 Printed in U.S.A. Fish Respiration in the Natural Oxygen Gradient of an Anaerobic Spring in Florida HOWARD T. 0DUM AND DAHD K. CALDWELL ON rare occasions one finds in nature an ecological circumstance so perfectly com posed that it constitutes a natural experiment superior to any man-made construction that could be built at reasonable cost in the labora tory Definite inferences can then be obtained with a small effort. Such natural experimental situations have additional value if they are per manent enough for data to be cumulati ve In this communication some obser v ations made in one of the remarkable Florida springs are used to draw conclusions about the respiration of fishes. It is hoped that others will be attracted to these permanent chemostatic gradients. Beecher Springs adjacent to the University of Florida Conservation Reserve, Welaka, Put nam County, emerges from a limestone aquifer without oxygen. As it flows downstream in its run (Fig. 1) this small stream picks up oxygen by diffusion and by photosynthesis of the algae. At a constant temperature diurnally and sea sonally of 22 to 23C., the physical and biological characteristics of this spring seem constant as surveyed 6 times over a period of 3 years. Extensive work on trace elements in another spring, Silver Springs, Marion County, indi cates that a chemostatic condition is to be expected in the springs of the region. Thus Beecher Springs seems to constitute a perma nent oxygen gradient in which organisms have become adjusted White sulfur bacteria and blue-green algae dominate the boil with other algal types appearing downstream, and finally high e r plants appear 300 meters from the boil. The algae have been studied by Larry A Whitford, North Carolina State College, and will be the subject of a forthcoming paper by him. Midges (Tendipedidae and Heleidae) are found throughout the run and in part are the basis of the food chain leading to some of the fishes. The o}"yrgen values on the map (Fig. 1) were obtained with the Rideal-Stewart-Winkler method Essentially similar values were recorded on June 6,1952 and a year later on June 22, 1953. There are clearly eddies of water in the stream so that oxygen conditions at anyone point fluctuate slightly with a range of about 1 ppm Some idea of this variation from eddy to eddy can be obtained from the data in Table r. The analytical accuracy is about 0.1 ppm plus or minus. The w ater is clear and the boil area smells strongly of sulfurous gases. Other properties are as follows : carbon dioxide, 10 ppm; pH, 7.2-7 .7; chlorides, 166 ppm; phosphorus, 0.14 ppm; nitrate nitrogen, 0 .02 ppm; alkalinity, 105 ppm (CaCOa); hydrogen sulfide, 0 .02 ppm. The natural distribution of fishes in this natural gradient was observed with the aid of a face mask by swimming underwater where clear visibility is 30 meters or more. The observed species distribution is indicated in Fig. 1 and is listed below. The eddies of water permit local temporary microhabitats to acquire more oxygen so that fishes can move into the higher oxygen spots to replenish oxygen and thus can live in parts of the stream where the average oxygen may be below their physiological toler ances Fig. 1 shows ecological tolerances under stream conditions by indicating the farthest

PAGE 2

. be im. tnd ; o f ere ler reo ne in ne ut i y in '.1 Is es 1, 4 ( 1. is I i H. T. ODUM AND D. K. CALDWELL-ANAEROBIC SPRING 105 1.3 MICROPTERUS L. AURITua sox GAMIIUSIA UOlLiENESIA HETERANDRIA 100 .. lnRS Fig 1. Map of Beecher Springs, Welaka, Florida. Dissolved oxygen values are given in parts per million. The that the fish were observed to come to the boil is indicated for 12 species in the middle of a sunny day Gune 24, 1953). point upstream that each species was observed to go. It is possible that the decrease of the small hydrogen sulfide concentration downstream is also involved. In order of distribution down the spring run the fishes represented are: eastern gambusia, Gambusia affinis kolbrooki (Girard); sailfin molly, Mollienesia latipinna LeSueur; least killifish, Heterandriajormosa Agassiz; redfin killifish, Lucania goodei Jordan; shellcracker, Lepomismicrolophus (GUnther); redfin pickerel, Esox amel'icanus Gmelin; redbreast sunfish, Lepomis (Linnaeus); largemouth bass, ,"\IIicropterus salmoides (Lacepooe); chubsucker, Erimyzon succetta (Lacepooe); bluegill, Lepomis macrochirus Rafinesque; golden shiner, Notemigonus crysoleucas (Mitchill); and bowfin, Amia calva Linnaeus EXPERIMENTS IN SUBMERGED CAGES Although Habs and Mueller (1948) found that the eastern gambusia lived readily in 0.3 ppm oxygen, the abundance of three cyprino dont species in the churning outflow headwaters (boil) where the oxygen is less than 0.3 ppm day and night seemed remarkable. These fishes are observed to break the surface continually. follys feeding on the bottom swim to the surface frequently before returning to the bottom. :\lthough gulping by these fishes has long been observed, the writers do not know of a clear of the thresholds at which gulp III T g IS necessary for these widespread fishes. he following experiments were conducted within the springs. TABLE I DISSOLVED OXYGEN, IN PAlI.TS PER MILLION, IN BEECHER SPRINGS, FLORIDA For 1953, three analyses for each station are preceded by the mean of these values. For stations, see Fig. 1 Station VI:6: 1952 VI: 24:1953 Boil A 0.05 0.3 (0.27, 0.30,0.50) B 0.26 0.6 (0.50, 0.59, 0.65) C 0.9 1.1 (0.40,0.65,2.16) Boil D 1.3 0.2 (0. 21,0.13,0.29) E 1.7 1.3 (1.26, 1.31, 0.99) F 1.1 2.0 (2.30, 1.80, 2 .01) G 2 3 (3.51, 1.15,2.32) H 3.6 Gambusia and M ollienesw taken from the boil were placed in the following situations : 1. Out of water in a dry box on land. 2. In cages in the boil at the surface so that fish could gulp at the surface. 3. In 4 cages submerged in the boil so that fishes could not get closer than 1 inch to the surface. One cage was near the surface and one was at a depth of 2 feet A third cage included only Gambusia and a fourth only M ollienesia. 4. Submerged below the surface downstream where the oxygen was 1.3 ppm. (See Fig. 1.) The amounts of time required for the fishes to die are summarized in Table II. Where oxygen was inadequate, distress was noticed within three minutes and most of the fishes were dead in about half the time necessary for all to succumb. Although Gambusw and M oUienesia are readily observed to pump the surface film over their gills in stagnant waters

PAGE 3

106 COPEIA, 1955, NO. 2 TABLE II FISH SURVIVAL EXPERIMENTS IN BEECHER S PRINGS, JANUARY 11, 1953 --DuraOxyPer-Experimental situation Number tiOD in gen "Cent of fishes min-ten-sur-utes sian, vival ppm ._------Cage in the boil at the surface i of the water, fish gulping air .... .... 37 160 I 0.28 92 Completely out of water' ... 30 22 In air 93 Cages completely submerged in boil : Gambusill, avg. weight: .23 gIn ....... .... 27 II 0.28 0 M ollientsia, avg weight : 0.37 gros ........... 27 22 0.28 0 Gambusia and M ollienesia : Below the surface 1 inch .. 24 19 0.28 0 Below the surface 2 feet 48 20 0.28 0 Cage completely submerged downstream ..... ...... 20 25 1.3 100 Gambusia and M olli,n.sia. where oxygen is low, this is not the means of respiration in this turbulent but anaerobic stream Definite gulping of air is observed. That gulping is necessary at these low tensions is demonstrated by the survival of those fishes with access to the surface. Since there was a rapid turbulent current cascading upward through the experimental cages an oxygenated surface film could not develop. When pre vented from gulping, the fish drowned quickly Those fishes kept out of water were quite alive at the end of 22 minutes since they demon strated normal behavior when returned to the water. Whereas the respiration of air by fishes has been much studied, the mechanism of aerial respiration in small cyprinodont .shes seems to have been neglected. Our dissections indicate that the Beecher Springs individuals do not differ from Gambusia affinis holbrooki elsewhere in having a closed swim bladder, a short intes tine opening directly into a short oesophagus, and pseudobranchs (Vialli, 1926). These latter vascularized structures lining the pharynx may be the structures used as a respiratory surface for gulped air. Another anaerobic spring in which obligate gulping by fish may be directly illferred is the large Warm Salt Spring in Sarasota County, Florida. Here, tarpon, Tarpon atlanticus (Valen ciennes); gar, Lepisosteus sp ; sou them varie gated cyprinodon, Cyprinodm variegatus varie gatus (Lacepede); sailfin mollys, and eastern gambusia are found in about half sea water salinity (17,800 ppm) and a temperature of 31C. The oxygen content is between 0.00 and 0.81 ppm day and night. The tarpon were ob served to roll and the smaller fishes to break the water continually That tarpon and gars breathe with a vascularized air bladder is known from observations and dissections sum marized by Babcock (1951) for the tarpon and Jones and Marshall (1953) for the gar. These studies were aided by a contract be tween the Office of Naval Research, Depart ment of the Navy, and the University of Flor ida, NR 163-106. LITERATURE CITED BABCOCK, LOUIS L. 1951. The Tarpon. 5th ed. Privately published, 157 pp. illus. HABS, H., AND G. MUELLER. 1948. Uber das Verhalten der Gambusien gegenuber dem Sauer stoffgehalt des Wassers. Zeitschr Hgy. U lnlek tienskrankh. 128: 371-8. (Abstract) JONES, F. R., AND N. B. MARSHALL. 1953. The structure and function of the teleostean swim bladder Bioi. Re:-u. 28: 16-83. VIALLI, M. 1926. Le pseudobranchie dei Pesci Arch Ital. Anat. Embriol., 23: 49-117 (Abstract) DEPARTMENT OF BIOLOGY, UNIVERSITY OF FLORIDA, GAINESVILLE, FLORIDA.





PAGE 1

Reprinted f r om COPEIA, 1955, No.2, pp 104-106, .. y 20 Printed in U.S.A. Fish Respiration in the Natural Oxygen Gradient of an Anaerobic Spring in Florida HOWARD T. 0DUM AND DAHD K. CALDWELL ON rare occasions one finds in nature an ecological circumstance so perfectly com posed that it constitutes a natural experiment superior to any man-made construction that could be built at reasonable cost in the labora tory Definite inferences can then be obtained with a small effort. Such natural experimental situations have additional value if they are per manent enough for data to be cumulati ve In this communication some obser v ations made in one of the remarkable Florida springs are used to draw conclusions about the respiration of fishes. It is hoped that others will be attracted to these permanent chemostatic gradients. Beecher Springs adjacent to the University of Florida Conservation Reserve, Welaka, Put nam County, emerges from a limestone aquifer without oxygen. As it flows downstream in its run (Fig. 1) this small stream picks up oxygen by diffusion and by photosynthesis of the algae. At a constant temperature diurnally and sea sonally of 22 to 23C., the physical and biological characteristics of this spring seem constant as surveyed 6 times over a period of 3 years. Extensive work on trace elements in another spring, Silver Springs, Marion County, indi cates that a chemostatic condition is to be expected in the springs of the region. Thus Beecher Springs seems to constitute a perma nent oxygen gradient in which organisms have become adjusted White sulfur bacteria and blue-green algae dominate the boil with other algal types appearing downstream, and finally high e r plants appear 300 meters from the boil. The algae have been studied by Larry A Whitford, North Carolina State College, and will be the subject of a forthcoming paper by him. Midges (Tendipedidae and Heleidae) are found throughout the run and in part are the basis of the food chain leading to some of the fishes. The o}"yrgen values on the map (Fig. 1) were obtained with the Rideal-Stewart-Winkler method Essentially similar values were recorded on June 6,1952 and a year later on June 22, 1953. There are clearly eddies of water in the stream so that oxygen conditions at anyone point fluctuate slightly with a range of about 1 ppm Some idea of this variation from eddy to eddy can be obtained from the data in Table r. The analytical accuracy is about 0.1 ppm plus or minus. The w ater is clear and the boil area smells strongly of sulfurous gases. Other properties are as follows : carbon dioxide, 10 ppm; pH, 7.2-7 .7; chlorides, 166 ppm; phosphorus, 0.14 ppm; nitrate nitrogen, 0 .02 ppm; alkalinity, 105 ppm (CaCOa); hydrogen sulfide, 0 .02 ppm. The natural distribution of fishes in this natural gradient was observed with the aid of a face mask by swimming underwater where clear visibility is 30 meters or more. The observed species distribution is indicated in Fig. 1 and is listed below. The eddies of water permit local temporary microhabitats to acquire more oxygen so that fishes can move into the higher oxygen spots to replenish oxygen and thus can live in parts of the stream where the average oxygen may be below their physiological toler ances Fig. 1 shows ecological tolerances under stream conditions by indicating the farthest

PAGE 2

. be im. tnd ; o f ere ler reo ne in ne ut i y in '.1 Is es 1, 4 ( 1. is I i H. T. ODUM AND D. K. CALDWELL-ANAEROBIC SPRING 105 1.3 MICROPTERUS L. AURITua sox GAMIIUSIA UOlLiENESIA HETERANDRIA 100 .. lnRS Fig 1. Map of Beecher Springs, Welaka, Florida. Dissolved oxygen values are given in parts per million. The that the fish were observed to come to the boil is indicated for 12 species in the middle of a sunny day Gune 24, 1953). point upstream that each species was observed to go. It is possible that the decrease of the small hydrogen sulfide concentration downstream is also involved. In order of distribution down the spring run the fishes represented are: eastern gambusia, Gambusia affinis kolbrooki (Girard); sailfin molly, Mollienesia latipinna LeSueur; least killifish, Heterandriajormosa Agassiz; redfin killifish, Lucania goodei Jordan; shellcracker, Lepomismicrolophus (GUnther); redfin pickerel, Esox amel'icanus Gmelin; redbreast sunfish, Lepomis (Linnaeus); largemouth bass, ,"\IIicropterus salmoides (Lacepooe); chubsucker, Erimyzon succetta (Lacepooe); bluegill, Lepomis macrochirus Rafinesque; golden shiner, Notemigonus crysoleucas (Mitchill); and bowfin, Amia calva Linnaeus EXPERIMENTS IN SUBMERGED CAGES Although Habs and Mueller (1948) found that the eastern gambusia lived readily in 0.3 ppm oxygen, the abundance of three cyprino dont species in the churning outflow headwaters (boil) where the oxygen is less than 0.3 ppm day and night seemed remarkable. These fishes are observed to break the surface continually. follys feeding on the bottom swim to the surface frequently before returning to the bottom. :\lthough gulping by these fishes has long been observed, the writers do not know of a clear of the thresholds at which gulp III T g IS necessary for these widespread fishes. he following experiments were conducted within the springs. TABLE I DISSOLVED OXYGEN, IN PAlI.TS PER MILLION, IN BEECHER SPRINGS, FLORIDA For 1953, three analyses for each station are preceded by the mean of these values. For stations, see Fig. 1 Station VI:6: 1952 VI: 24:1953 Boil A 0.05 0.3 (0.27, 0.30,0.50) B 0.26 0.6 (0.50, 0.59, 0.65) C 0.9 1.1 (0.40,0.65,2.16) Boil D 1.3 0.2 (0. 21,0.13,0.29) E 1.7 1.3 (1.26, 1.31, 0.99) F 1.1 2.0 (2.30, 1.80, 2 .01) G 2 3 (3.51, 1.15,2.32) H 3.6 Gambusia and M ollienesw taken from the boil were placed in the following situations : 1. Out of water in a dry box on land. 2. In cages in the boil at the surface so that fish could gulp at the surface. 3. In 4 cages submerged in the boil so that fishes could not get closer than 1 inch to the surface. One cage was near the surface and one was at a depth of 2 feet A third cage included only Gambusia and a fourth only M ollienesia. 4. Submerged below the surface downstream where the oxygen was 1.3 ppm. (See Fig. 1.) The amounts of time required for the fishes to die are summarized in Table II. Where oxygen was inadequate, distress was noticed within three minutes and most of the fishes were dead in about half the time necessary for all to succumb. Although Gambusw and M oUienesia are readily observed to pump the surface film over their gills in stagnant waters

PAGE 3

106 COPEIA, 1955, NO. 2 TABLE II FISH SURVIVAL EXPERIMENTS IN BEECHER S PRINGS, JANUARY 11, 1953 --DuraOxyPer-Experimental situation Number tiOD in gen "Cent of fishes min-ten-sur-utes sian, vival ppm ._------Cage in the boil at the surface i of the water, fish gulping air .... .... 37 160 I 0.28 92 Completely out of water' ... 30 22 In air 93 Cages completely submerged in boil : Gambusill, avg. weight: .23 gIn ....... .... 27 II 0.28 0 M ollientsia, avg weight : 0.37 gros ........... 27 22 0.28 0 Gambusia and M ollienesia : Below the surface 1 inch .. 24 19 0.28 0 Below the surface 2 feet 48 20 0.28 0 Cage completely submerged downstream ..... ...... 20 25 1.3 100 Gambusia and M olli,n.sia. where oxygen is low, this is not the means of respiration in this turbulent but anaerobic stream Definite gulping of air is observed. That gulping is necessary at these low tensions is demonstrated by the survival of those fishes with access to the surface. Since there was a rapid turbulent current cascading upward through the experimental cages an oxygenated surface film could not develop. When pre vented from gulping, the fish drowned quickly Those fishes kept out of water were quite alive at the end of 22 minutes since they demon strated normal behavior when returned to the water. Whereas the respiration of air by fishes has been much studied, the mechanism of aerial respiration in small cyprinodont .shes seems to have been neglected. Our dissections indicate that the Beecher Springs individuals do not differ from Gambusia affinis holbrooki elsewhere in having a closed swim bladder, a short intes tine opening directly into a short oesophagus, and pseudobranchs (Vialli, 1926). These latter vascularized structures lining the pharynx may be the structures used as a respiratory surface for gulped air. Another anaerobic spring in which obligate gulping by fish may be directly illferred is the large Warm Salt Spring in Sarasota County, Florida. Here, tarpon, Tarpon atlanticus (Valen ciennes); gar, Lepisosteus sp ; sou them varie gated cyprinodon, Cyprinodm variegatus varie gatus (Lacepede); sailfin mollys, and eastern gambusia are found in about half sea water salinity (17,800 ppm) and a temperature of 31C. The oxygen content is between 0.00 and 0.81 ppm day and night. The tarpon were ob served to roll and the smaller fishes to break the water continually That tarpon and gars breathe with a vascularized air bladder is known from observations and dissections sum marized by Babcock (1951) for the tarpon and Jones and Marshall (1953) for the gar. These studies were aided by a contract be tween the Office of Naval Research, Depart ment of the Navy, and the University of Flor ida, NR 163-106. LITERATURE CITED BABCOCK, LOUIS L. 1951. The Tarpon. 5th ed. Privately published, 157 pp. illus. HABS, H., AND G. MUELLER. 1948. Uber das Verhalten der Gambusien gegenuber dem Sauer stoffgehalt des Wassers. Zeitschr Hgy. U lnlek tienskrankh. 128: 371-8. (Abstract) JONES, F. R., AND N. B. MARSHALL. 1953. The structure and function of the teleostean swim bladder Bioi. Re:-u. 28: 16-83. VIALLI, M. 1926. Le pseudobranchie dei Pesci Arch Ital. Anat. Embriol., 23: 49-117 (Abstract) DEPARTMENT OF BIOLOGY, UNIVERSITY OF FLORIDA, GAINESVILLE, FLORIDA.