2‘ a... u o ... V . ”I.” , . u A 1‘. n . —.. a - I Q -\~ an . . v. at Z. .5. C. u- .c‘ a . 1.. . g o . a - .am . . p‘. ”v a . I :s ”I v“ I L. .3 pg st .M. . ¢ ' .QA up” 5.. .u. a p . .. nu C. u . s . .2 .3 u. . 6— a. r . .u. . . v~ .4 .3 . 4 . c . . L . .. ~ . . . u. .—. .q . c s. .... ~ . ‘n : . v . ~I. . A Z. 3... ~_‘ 2.. NP. —' M. n. .N a. 5 N v ‘ n u .. ' .1 ABSTRACT PRODUCTION AND TROPHIC ECOLOGY OF TWO CRAYFISH SPECIES COHABITING AN INDIANA CAVE BY David Lawrence Weingartner Two species of crayfish inhabit the stream of Shiloh Cave located in Lawrence County, Indiana. One species, Orconectes inermis Cope, is classified as a troglobite and, as such, is an obligate cavernicole highly adapted to the cave environment. The other crayfish, Cambarus laeVis Faxon, maintains a population not only in the cave, but also in the surface stream issuing from the cave; it is considered a facultative cavernicole and is ecologically classified as a troglophile. Major aims of the study, which was conducted primarily during 1969, were the delineations of the life histories, population parameters, productivities and ecological strategies of both species. The avail- ability, procurement and utilization of energy were important themes of the investigation. Comparisons were drawn not only between the troglophile and the troglobite, but also between the epigean and hypogean crayfish. The two crayfish species competed for highly restricted food resources, which were composed primarily of vegetative detritus and benthos. Particulate organic matter, which was mostly wood and leaf fragments and associated microflora, was quite limited quantitatively and was of a highly seasonal nature. The organic drift was in a highly decomposed state with the refractive compound lignin constituting almost half of its content. The benthic standing crop, which was primarily . . . .O a. 1 .... l .1 . F. u ‘ 'h c. .u a . v. *- 0 .c v o. . . w. .n V. AC .. p . .o. . Z. . . Va. I .‘ . y 5 . . , 1 O . . . ;_ .1 . . a r. .. ~. 0. § . . an no o o I no ‘ ._ .. a. q . v. A. .. o a u. .. _ . .“ . g ‘3 . .. .o . n ~ :~ ‘ . :3 A: u . .. ; : V. I 1 . ... .n. r. . a. u .A u. . q . .. w . :a p .4; f s ‘1. . .~‘ ~. . v. , . . £4 .. .. . . . . . v. .. p. .. .. .. S. 4.. ... .J. .1 h; .44 .y .3 L. .3 . J. . A: n . a. o . L. .4. z. :x I. r .. . .~, .~ .c. u. a r . . . . 3. ‘L o. 1* . . 3- . I . I . . : vi . . rt...» . ,. . ~ David Lawrence Weingartner composed of oligochaetes, isopods and amphipods, was quite meager in comparison to that of the surface stream. Crayfish ingestion was determined by chemical analysis of fecal matter. A trophic trend was observed in comparing the epigean C. laevis, the hypogean C. laevis, and the O. inermis crayfish; there was a sequential decrease in both the ingestion rate and the utilization of animal material as a dietary component. Crayfish were individually tagged by a new technique. Both cave species were found to display home range behavior. Crayfish usually resided in a home pool, but wandered into adjacent riffle areas on occasion. Wandering was more frequent and more extensive in C. laevis crayfish. In addition, there was evidence that this species migrated between surface and cave habitats. Various population attributes were studied. In comparison to the epigean crayfish, the hypogean pattern for both species was a small population with low growth, mortality, reproductive and production rates, and with extended longevities. Basal metabolism differed in that it was much lower in the O. inermis crayfish than in the equivalent rates found for the surface and cave C. laevis populations. Energy budgets were constructed, and changes in energy partitioning associated with epigeal-hypogeal and troglophilic-troglobitic transforma- tions were discussed. PRODUCTION AND TROPHIC ECOLOGY OF TWO CRAYFISH SPECIES COHABITING AN INDIANA CAVE BY David Lawrence Weingartner A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1977 .- \-..7.' -‘v “‘ .... a. I. I a . . 5. “ ‘Hd . $~ . «1 s.. a. —.. I 44 . s. ‘2 h hm FM .< w« v. «J Vs l. . . . _ I. V. I .n. I : .~a V. \ a u.“ A. \u. .. _ .. a: . p: ‘ . C. r. >u an . a \. “I r. C. a. : . . a‘ .2 . U: «_ L. . L. .3 r. r. an .I . .o . .N I s .: x u.. . . é. .1 . 2. K. . \ o l. .. . W. —.. .- ... ... a. p. .... o . a. v. .. A. L. . . w...- ._. .s. N... M? I. :- —.. u~ s. w! .a. J. ‘ - 2 L ACKNOWLEDGMENTS In conducting this study, I spent many solitary hours in Shiloh Cave, but it would be misleading to give the impression that I alone should receive credit for the results; many people, through their contributions of time, effort, knowledge or interest, have made this study possible. In particular, I would like to express grateful acknowledgment to the following: Dr. William E. Cooper, my major professor, who has surpassed all requirements of his position to guide this project to a successful conclusion; the members of my guidance committee, Drs. T. Wayne Porter, Niles R. Kevern, Richard Merritt and Kenneth W. Cummins, who offered valuable suggestions; Dr. Thomas L. Poulson of the University of Illinois at Chicago Circle, who shared his expertise in cave ecology; Dr. Charles M. Stine of the Food Science Department, who provided essential research facilities; CDR John M. Hoffmann, Chairman of the Chemistry Department of the U. S. Naval Academy, who has been actively supportive of my efforts to complete the manuscript; Mrs. Linda Zelinski, also of the Naval Academy, who gave technical assist- ance; Mr. John Day, the owner of Shiloh Cave, who gave me permission to carry out this project; Miss Mona Pitman of Bedford, Indiana, who gave me shelter and friendship during the period of field work; my wife, Carol, who has given active support and assistance in all phases of this study; and Dr. E. C. Williams of Wabash College, who instilled in me an interest in cave ecology. ii ‘I.’I"_ . ~.,-——— -‘. n u“... .~. v-‘--.-._ 9 -_- I- -.--._.~ .- -‘y-..’ -. <_. .. —u~¢~n .. ‘v-‘--... -._ -'Q' ~- '~*-- a... .g u. . ~' u, °' A. \ -— -'..1- w._ “ ‘4. D‘ ‘ : 1 .a‘,‘ TABLE OF CONTENTS PAGE CHAPTER I - INTRODUCTION . . . . . . . . . . . . . . . . . . . 1 CHAPTER II - STUDY AIMS . . . . . . . . . . . . . . . . . . . 7 CHAPTER III - STUDY SITE . . . . . . . . . . . . . . . . . . 8 CHAPTER IV - MATERIALS AND METHODS . . . . . . . . . . . . . . l9 Hydrology . . . . . . . . . . . . . . . . . . . . . . . l9 Non-organic Water and Substrate Analyses . . . . . . . . 2O Alkalinity . . . . . . . . . . . . . . . . . . . . 20 Temperature . . . . . . . . . . . . . . . . . . . 20 Carbonate . . . . . . . . . . . . . . . . . . . . 20 Organic Analyses . . . . . . . . . . . . . . . . . . . . 20 Ashing . . . . . . . . . . . . . . . . . . . . . . 21 Energy . . . . . . . . . . . . . . . . . . . . . 21 Cellulose . . . . . . . . . . . . . . . . . . . . 21 Lignin . . . . . . . . . . . . . . . . . . . . . . 22 Crude Protein . . . . . . . . . . . . . . . . . . 22 Food Resources . . . . . . . . . . . . . . . . . . . . . 22 Macroseston . . . . . . . . . . . . . . . . . . . 22 Microseston . . . . . . . . . . . . . . . . . . . 24 Substrate . . . . . . . . . . . . . . . . . . . . 24 Benthos . . . . . . . . . . . . . . . . . . . . . 25 Trophic Studies . . . . . . . . . . . . . . . . . . . . 25 Diet Analysis . . . . . . . . . . . . . . . . . . 28 Predator-Prey Relationships . . . . . . . . . . . 31 Crayfish Census . . . . . . . . . . . . . . . . . . . . 32 Length-Weight Relationships . . . . . . . . . . . . . . 40 Population Attributes . . . . . . . . . . . . . . . . . 41 Tissue Growth . . . . . . . . . . . . . . . . . . 41 Molting . . . . . . . . . . . . . . . . . . . . . 42 Reproduction . . . . . . . . . . . . . . . . . . . 43 Mortality . . . . . . . . . . . . . . . . . . . . 43 Population Size and Structure . . . . . . . . . . 44 Biomass and Productivity . . . . . . . . . . . . . . . . 44 Respiration . . . . . . . . . . . . . . . . . . . . . . 45 Time Frame . . . . . . . . . . . . . . . . . . . . . . 47 CHAPTER V - THE ENVIRONMENT . . . . . . . . . . . . . . . . . 48 Results . . . . . . . . . . . . . . . . . . . . . . . . 48 iii 1" ‘ '7)" (61‘3“.4M‘ in]! .— g Q‘- -- ~ .. '--. .. --~-.‘ — .- or? -~. 7‘ o r. _ -.u. _‘ ‘- 31:5.“ ».. tn if"; J._ .‘_ \ u-__ u 4‘ ~‘ g i‘,. TABLE OF CONTENTS--C0ntinued The Stream . . . . . Hydrology . . . Alkalinity . . . Temperature . . Food Resources . . . . Leaf Input . . . Macroseston . . Microseston . . Substrate . . . Benthos . . . . Discussion . . . . . . . . . CHAPTER VI - THE CRAYFISH . . . . Initial Findings . . . . . . Results and Discussion The Census . . . POpulation Estimates Assumptions Accuracy . Correction for The Population/Capture Capture Rates 0 C O 0 Displacement Trap versus Visual . Temporal Differences Length-Weight Relationships Activity . . . . . . . . . . Introduction . . . . . Results . . . . . . . Circadian Rhythm . Longitudinal Movement Patterns in Individuals . Trolobite Troglophile Effect at Population Discussion . . . . . . Trophic Ecology . . . . . . Results . . . . . . . Predator-Prey Interactions Diet Analysis . Troglobite Troglophile Discussion . . . . . . Population Attributes . . . Results . . . . . . . Tissue Growth . Introduction Troglobite Troglophile iv 0 Level Ratio PAGE 48 48 56 61 61 61 7O 81 86 9O 94 108 108 108 108 108 108 112 112 112 117 117 117 120 122 122 124 124 124 124 124 133 144 149 154 154 154 157 157 159 164 169 169 169 169 172 182 T. I. . . — u“ \A a) Wm. nu, \. .\.. .: .2 D... . s.. —. TABLE OF CONTENTS--Continued Enclosure Growth Study . . . . Epigean Population . . . . . . Molting . . . . . . . . . . . . . . . . . Troglobite . . . . . . . . . . . . . Troglophile . . . . . . . . . . . . Reproduction . . . . . . . . . . . . . . . Troglobite . . . . . . . . . . . . . Troglophile . . . . . . . . . . . . Mortality . . . . . . . . . . . . . . . . Troglobite . . . . . . . . . . . . . Troglophile . . . . . . . . . . . . Longevity . . . . . . . . . . . . . . . . Population Size and Structure . . . . . . Troglobite . . . . . . . . . . . . . Troglophile . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . Tissue Growth . . . . . . . . . . . . . . Hypogean Populations . . . . . . . . Comparison with Epigean Populations Molting . . . . . . . . . . . . . . . . . Reproduction . . . . . . . . . . . . . . . Mortality . . . . . . . . . . . . . . . . General Observations . . . . . . . . Mortality Factors . . . . . . . . . Physiological Aging . . . . . Environmental Stress . . . . . Molting . . . . . . . . . . . Pollution . . . . . . . . . . Predation . . . . . . . . . . Human Interference . . . . . . Disease . . . . . . . . . . . Starvation . . . . . . . . . . Longevity . . . . . . . . . . . . . . . . Population Size and Structure . . . . . . Biomass and Productivity . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . Troglobite . . . . . . . . . . . . . . . . Troglophile . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . Respiration . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . Troglobite . . . . . . . . . . . . . . . . Troglophile . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . PAGE 185 185 190 190 197 198 198 199 200 200 204 208 209 209 219 237 237 237 239 243 246 252 252 254 254 255 255 255 256 259 259 260 261 263 269 269 269 270 276 280 280 281 281 285 287 .— -‘_-—~ _ _ n - C - ,......~ -‘ ~.‘———-—r ..-v —_. bu . -‘-_._- a- .\ ’ ~_. -- .. 47-- .. n I. " ' ‘ u 5..“ ..'.' .n.‘~ ..... _..._ _\ .— Itv. ‘- '— -.-...._..- I TABLE OF CONTENTS--Continued CHAPTER VII - ENERGY AND PROTEIN BUDGETS —— AN INTEGRATION Description of Inter—relationships . Discussion of the Pattern Budget Changes Associated with Adaptation to the Cave Habitat CHAPTER VIII - GENERAL DISCUSSION LITERATURE CITED vi PAGE 292 292 299 305 311 313 . . . A . ... a W. .n . l . .a‘ o. I. ;.~ . . ht ”4. . A .u.. 34 .2 ,. . . a g . . Y. O . .. .fia w.. L. n... ... .44 r~ .4 vv. VI .L .n. —.. r. r. .3 o. .A L. .. a. T —. L. .1 NJ ~.. 7. 0‘. .nq .N‘ ‘w u .u ‘ A .. r j 3 :. E : .3 r. Lv .4‘ : fi« 4 :. nur-on. . . P.‘ am“ I i I 18 designated the "census area". This area comprised that reach of the main stream from 0 meters to the waterfall at 492 meters, and also included a short downstream section of the Black Damp Tributary. A sub-area of 364 meters, designated the "quantitative study area", terminated up- stream at the weir and downstream at the Black Damp Tributary. This reach of the main stream was relatively free of feeder streams and was used for the study of energy budgets. In the evaluation of food re- sources the entire main stream and most of its tributaries were studied. jig?“ a .J, ,1... o .1. . CHAPTER IV MATERIALS AND METHODS Hydrology To measure the discharge of the cave stream, a permanent 90° V- notch weir, meeting specifications of the Water Measurement Manual (1967) published by the U. S. Bureau of Reclamation, was constructed; the single modification consisted of building the base of the notch flush with the substrate, in order to minimize pooling and interference with the normal movement of drift (Figure 8). This type of weir is an ac- curate flow measuring device particularly suited for small flows. A float-actuated pen continuously recorded the head on a kymograph. After an initial calibration against actual flow, the head was there- after converted to discharge by the formula, discharge (l/sec) = 2.48. 0.0147 (head in cm) , this formula was derived from a formula in the Water Measurement Manual by conversion to the metric scale. The drainage areas of the main cave stream and each of its tributaries were estimated from levels of base flow and surface drain- age patterns. The subterranean watersheds so determined are judged to be fairly accurate, but do not have the absolute accuracy of surface watersheds. Precipitation records were obtained from the Oolitic Weather Bureau Station, which is located 2.3 kilometers northeast of the cave entrance, but only an estimated 0.5 kilometers from the watershed of 20 the cave. Precipitation measurements were converted to the metric scale. Non-organic Water and Substrate Analyses Alkalinity The alkalinity of water samples was determined by the titration method described by Welch (1948). Samples were taken quarterly at 7 tributaries, 5 sites in the main cave stream, and 2 sites in the surface stream issuing from the cave. Duplicate samples were taken and tested in all cases. Temperature A maximum-minimum thermometer was placed in the cave stream at the weir. A temperature record was also kept on other mainstream sites and tributaries in conjunction with microseston and alkalinity studies. Carbonate The carbonate content of the mud was determined by the method of Kozlovskii, as described by Il'kovskaya (1965). This method is based on the decomposition of carbonates with HCl solution, followed by absorption of the evolved CO with NaOH and subsequent titration of 2 the excess alkali against acid in the presence of BaClz. Organic Analyses All samples to be chemically analyzed were first oven—dried at 60° C. Chemical analyses were performed in duplicate in all cases. 21 Ashing Ashing was carried out in a muffle furnace at 500° C. Macro- seston, leaves, mud, feces, and detritus from rocky substrates were ashed, and the results were expressed as ash-free dry weight or percentage ignition loss; these expressions are considered to be closely equivalent to the absolute and relative organic content. Energy Energy contents were determined for macroseston, microseston, benthos, leaves, mud, and feces. These determinations were based on quantitative dichromate oxidations as described by Maciolek (1962). A reagent strength of 0.25 N potassium dichromate was used in the analysis of microseston which had been filtered through filter pads, whereas 0.50 N reagent was used for all other determinations. The results, expressed as mg of oxygen consumed, were calibrated against a cellulose standard, adjusted for the incomplete oxidation of certain proximate groups in natural organic matter, and transformed to gram calories by the 3.4 conversion factor suggested by Maciolek (1962). Cellulose Cellulose determinations were made on feces, mud, leaves, and macroseston. The method of Crampton and Maynard (1938), which removes non-cellulolytic organic constituents by digestion with an acetic acid- nitric acid reagent, was employed. 22 Lignin Lignin content was determined in leaves and macroseston by the method of Crampton and Maynard (1938); however, this method, which is based on the use of sulfuric acid, was modified by correcting for protein content (determined by the Kjeldahl method), instead of digest- ing protein with pepsin. Crude Protein Organic nitrogen was determined by the semi-micro Kjeldahl method. Results were expressed as crude protein by multiplying the nitrogen content by the conversion factor, 6.25. Macroseston, leaves, microseston, feces, and mud were analyzed for crude protein content. Food Resources Macroseston Except for the small tributary entering the cave stream at 284 meters, the weir intercepted all water—borne food input into the 364 meter section of stream between the weir and the Black Damp Tributary; this section, designated the "quantitative study area", was utilized for the measurement of energy flow. A drift net was employed in conjunction with the weir to sample macroseston entering the quantitative study area (Figure 8). The net was made of Nitex with 0.23 mm apertures, and, because of the fine mesh, was two meters in length to prevent back-flow. The mouth of the net formed a 22.6° arc, which, when positioned in the middle of the 90° V— notch, filtered 20% of the water at all flow rates. Tests employing 23 marked drift material confirmed that 20% of the drift was also inter- cepted. The drift net collected 24-hour samples and was used for three consecutive days in each 20 day period throughout the year. A temporary weir and drift net were placed just upstream of the Black Damp Tributary at 440 meters. They were employed during a period of low flow in August to measure the macroseston leaving the quantitative study area. The drift collected at this drift station was adjusted by correction factors to compensate for the removal of 20% of the drift at the 79 meter station. The values of the correction factors are based on certain assumptions. Vegetative drift was assumed to entirely origi- nate upstream of the 79 meter station, and a correction factor of 1.25 was applied to the results from the 440 meter station. Although it may not have been true of the few planktonic forms, such as the cyclopoids, animals of aquatic origin were generally assumed to have undergone only short range displacement and no correction was considered necessary. Terrestrial and aerial fauna and exuviae, on the other hand, underwent passive transport in the stream and could have been displaced great distances. Unlike the vegetative drift, however, their point of origin could have been either upstream or downstream of the 79 meter station, and an intermediate correction factor of 1.12 was applied to these drift categories. Collected drift was preserved with 4% formaldehyde and sorted into the components, leaves, wood, animals, exuviae, and debris; after drying in a 60° C oven, their weights were determined. The wood category in- cluded everything of a "woody" nature -even leaf petioles. January and July drift samples, which were collected over four day intervals, were sorted, dried and analyzed for energy, protein, cellulose, lignin, and ash content. 24 Leaves, which made up a major portion of the drift, were studied further to determine the seasonal species ratios entering the cave and to monitor the changes in their composition (energy, protein, cellulose, lignin and ash) in passing from the living state to trypton in the cave stream. Microseston Microseston was considered to be that portion of the particulate matter that was not retained by the drift net. Water samples, usually of a one liter volume, were vacuum filtered through pre-fired glass fiber pads (Reeve Angel 934-AH, Arthur H. Thomas Co.); these pads are reported to retain at least 95% of the particles retained by membrane filters of 0.45 micron pore size. After filtration the pads were oven- dried at 60° C and were later analyzed for energy or protein content. The filtered water was tested for alkalinity as described by Welch (1948). A series of 8 water samples was taken at the weir during the month of February in conjunction with the ebb of the largest flood, in order to relate microseston content to discharge. Other samples were taken quarterly at 7 tributaries, 5 sites in the main cave stream, and 2 sites in the surface stream issuing from the cave. Duplicate samples were taken and tested in all cases. Substrate Mud, consisting of sand, silt, clay, and organic fractions, was collected with a core sampler with a cross-sectional area of 11.6 sq cm and a depth of 1.0 cm. These shallow cores permitted measuring the organic material in the mud layer most accessible to benthic foraging. Mud samples were taken in January and July from 10 pools in the cave ___l "'11:“ -‘. .1; w . SEE! 21.12% a 25 stream and from 2 pools in the surface stream; two cores were taken from each pool site, and these were arbitrarily assigned to one of two longitudinal sample series (designated series "A" or ”B"). These samples were oven-dried at 60° C and energy contents determined. In addition, energy, ash, cellulose, protein and carbonate analyses were performed on the composite of cave mud from each longitudinal sample series. Benthos Quantitative benthic samples were taken in June and December. Sampling followed a stratified random design with 15 sample sites, of which 13 were located within the quantitative study area; each site consisted of a 15 meter longitudinal stream section. The sample sites were restricted to rocky substrates, as mud contained an extreme paucity of fauna. A 0.05 m2 cylinder sampler with a saw-tooth edge was employed to delimit the substrate to be sampled (Figure 9). The contained substrate was removed to a depth of approximately 3.6 cm and washed through a sieve series with a minimum aperture of 0.18 mm. The water remaining within the cylinder was agitated and strained through a net of similar aperture size. The samples were preserved in 4% formaldehyde, and later sorted into taxon and detrital components and oven-dried at 60° C. Trophic Studies Studies were carried out to determine the trophic ecologies of the crayfish populations. These studies focused on determining the relative importance of mud, plant and animal sources in the field diet, but also examined feeding patterns and predator-prey relationships under laboratory conditions. 11:11.... i. . . . p r. .. ‘7‘1 I p, ..... .-_—.- Wu)?" m. w’ In» Figure 8. Figure 9. 26 Weir constructed 79 meters downstream from collapse sinkhole. Visible are the drift net, which captured macroseston, and, at the right, the automatic head recorder. Equipment used for sampling benthos, with a typical pebble-cobble riffle substrate also shown. - «’4. ‘_—— 2'7 ,lapse :pturei head l- \ 1"..3 5:1 '9;— Ir -‘- a. .h- _--J ‘<> . c: v . .d .J a... ”Q ‘2 :u l a.“ k .T ..... ..h e I t: ”.1 .2 .3 .3 7. a T E a. S 2. r. d t E l .2 O .3 .3 S r .. J a .3 t . u 3. S .3 S a a. w. ~ml .l C a p.“ ... kw e .4; MN MI C. 21 t O S L. £4 C § 2 — r c s sq .n a “fl .5 . a r : ..-. sd .3. cc 3 w. a; a v u ‘ ”‘1“ a: ‘3 a” M v “A” .84 h.“ o “.1“ bum-4 .s s :5” M... ... .I u. ”c. .6; :0. : .~ 3. a . .1 . . . . 'de... 4.. u. . 28 Diet Analysis The traditional gut analysis technique is ineffective for quan- titative measurement of dietary components, on account of their vary— ing assimilation rates. Because of the highly macerated condition of their gut contents, this technique is especially unsuited to the study of crayfish, although it has been employed by many investigators. In the present study a method was developed to determine the quan- titative and qualitative nature of ingestion by chemical analysis of feces; this involved comparison of the chemical composition of field- collected feces with that of crayfish maintained on controlled diets of mud, animal or plant material. Feces were collected in the field by placing a rubber bladder over the anus and tail of a freshly captured crayfish (Figure 10). The opening of the bladder was sealed against the abdomen with a rubber band. This arrangement interfered with escape behavior, but not with normal feeding activity. The experimental crayfiSh was then placed in an undisturbed 40 m2 portion of the stream that was enclosed by quarter inch hardware cloth. Two such foraging pens were set up, one in the surface stream for studying epigean C. laevis, and another in the cave stream for studying both 0. inermis and hypogean C. laevis. No more than three crayfish were placed in a pen at any one time, and they were removed after feeding for a period of two days. Neither the bladder nor the pen was observed to interfere significantly with normal behavior or food availability and, because of the delay of approximately 15 hours between ingestion and egestion, a large portion of the feces collected was derived from food eaten before the crayfish was captured. .c. .C .1: I a. ‘1 I T. .n. J. .t t .J 3. . . ... Q. . n. .. ,. - —~1.'L.-:{' 29 In the laboratory 0. inermis and C. laevis were maintained in individual containers under cave conditions of temperature and darkness, and fed various quantitative diets. One group of crayfish was fed a complete diet of filtered mud, isopods, amphipods and three kinds of leaves. The other group was fed a ration consisting of one of the following: (1) mud, composed of sand, silt, clay and organic fractions, that was obtained from the cave stream and filtered through a 0.5 mm aperature sieve; (2) live amphipods; (3) one of three species of shed leaves conditioned for several weeks in water from the cave stream. The three kinds of leaves, Fagus grandifolia, Acer saccharum, and Quercus velutina, are the ones most commonly found in the cave stream. Leaf rations were punched from intact leaves in species-specific shapes. The feces resulting from the various diets were collected in a manner similar to that of the field studies. Feces from both the field and the laboratory studies were weighed and chemically analyzed. The fecal constituents quantitatively determined were ash, crude protein, cellulose and energy. The ratio in the field feces of the dietary sources, mud, animal and plant, was determined from the following simultaneous linear equations with three unknowns: D1 = alx + bly + clz D2 = a2x + bzy + 022 D3 = a3x + b3y + C32 D4 = a4x + b4g + c4z I.O=x+y+z, 1'31... .41»... . hi.— I ‘ :c 1 'V- ~ . ‘~ 30 where the fraction of field feces from x = animal source y = plant source 2 = mud source, where the constituent fractional content in feces derived from a = animal diet b = plant diet 0 = mud diet D field diet, and where the constituent, l = ash 2 = crude protein 3 = cellulose 4 = energy. Solving for three unknownsrequired simultaneous solving of three or four of the above equations; the last equation, which expresses the relationship that the summation of x, y and 2 equals unity, was always included. The ten possible combinations gave up to ten estimates of the dietary nature of the feces, although some combina- tions were not solvable because of inconsistencies of the data. The proportions of the mud, animal and plant components of the feces were converted, by assimilation efficiencies determined in the laboratory studies, to the ingestion proportions of these dietary components. The assimilation of the mud diet was too low to be determined; based on an organic content in mud of 4%, the assimilation efficiency was assumed to be approximately 1%. The field egestion rate and the assimilation data also provided an estimate of the ingestion rate in the field. The accuracy of this method of trophic ”1“ ll 1 n v — ‘ I .4 C h. r a. r. C r. r . . .c .. : l L” n“. e p . «a a C b: p . I; p . A , . .. Cu “O . . . . r. c. w. . S .P . t C. r. L f e U 9 O 3 L. T .7 C A. . 1 . . . S 9. kc . 1 71 5,. W T. .1 I I . .3 a : ... t a .d c. 9 a .1 3. .t. 3 l n...» m... 1 .... C. a o 3 r . . : . S O C: 11 6 an 1.1. 11. _ . I s. . .2 .3 .3 C. .1! w. u: ”a r. o. .1 mu 3 .. a 15 7.. . 4: .V .u. . . I J. .3 .2 a. my .. .1. a 4.. J. :u .t. t. x. ‘5. 2. ... .3. .a. f.‘ x, u... l. L r . . . v. .3 5 :m a: r. .1 .1“ c: .. .. n... w... 1. .: a. . . .i . . . L. ”a I. I I. r. l .. 1. C .3 « I» I II 11~,,.ut Jflw 31 determination was checked by applying it to the complete laboratory diet, in which the rate of ingestion of the various dietary components was known. Predator-Prey Relationships A study was undertaken to investigate three aspects of predator- prey interactions: (1) the effect of bare versus rocky substrate on the predation rate; (2) species-specific predation rates of the two predators, O. inermis and C. laevis; and (3) species-specific predation rates on two of the most common cave stream prey taxa, Asellus sp. (an isopod) and Crangonyx sp. (an amphipod). This study was carried out in the cave in order to best simulate natural conditions. Predator (25-30 mm carapace length) and prey (adult-sized) subjects were taken fresh from the cave stream, with the crayfish confined with mud food for a period of one day in order to adjust them to confinement. The crayfish were then tested individually over a period of 24 hours. The rocky substrate consisted of 6 pebbles that were grouped in the center of the tank and occupied approximately 10% of the total substrate; the purpose of the rocks was to serve as a refuge for the prey species. The experimental design attempted to simulate a low prey density situation typical of cave ecosystems; only 3 Asellus or 3 Crangonyx were introduced into a 0.1 m2 tank containing a single crayfish —-thus exposing the predator to an initial density of 30 prey per square meter. The experimental prey density was lower than those that existed in the cave stream; means of 268 isopods/m2 and 138 amphipods/m2 were obtained in benthic samples (see Table 9). This low prey density was used because it was thought that higher numbers of prey might have a tendency to test satiation levels instead of predation efficiencies. A..‘ ”A -...o-- - t .3 ... r S .2 y. r. .7. 4. ._. I. ,2. i ..._ .3 .J ..-.. t a n .J r O. O T 3 D L F. T. H l t 3 Q. a S .l r .l C .L .l ‘. t a .l C .5 kn u .t r .l .wu : .d .3 n n yfl a s A C. Q. Q .3 ..~ .5 a. S .3 i a. O 3 a. .3 E. t c . J. r e a. d 9 n r E a .: .n. r w. a. C a ..~ 7. 3 .l 3 a. t L 3 a. .. .1 a. I .3 L .4. .1 .. Q. «u z. .. r. u. a. . a v. L. T. .3 .x. u...‘ C. m. .. 3. ..- :. a. T, Z. i .Z i .. i ; . ‘ ., i. 5.1!: SJ ,.l 32 Crayfish Census The census of crayfish was accomplished by visual and trap capture. Traps were constructed of quart jars with plastic funnels inserted in the jar mouth and held in place with a rubber band (Figure 11). The inner surface of the funnel was scored to allow additional purchase for entering crayfish. The traps were of two types: in one type the funnel opening was enlarged to a diameter of 16 mm,and the trap was intended to capture 0. inermis and small C. laevis; in the other type, which was intended to capture larger C. laevis, the funnel opening was enlarged to a diameter of 32 mm, and the opening was covered by a hardware cloth gate.which opened only in the inward direction. The traps were baited with bologna,which was wrapped in bolting silt to prevent the crayfish from eating the meat. Alternating the two types, traps were placed every 5 meters, beginning at the upstream entrance to the cave at 0 meters and continuing to the waterfall at 490 meters. Traps were also placed in the first 10 meters of the Black Damp passage. This census area included not only the quantitative study section, which was located between the weir and the Black Damp Tributary, but also both upstream and downstream buffer zones. The visual captures were carried out using a gasoline lantern and electric head lamp for illumination, and a long-handled dip net for capture. A survey consisted of a two way transit of the delineated area and required an average time of 3.3 hours, although individual surveys varied from 1.8 to 7.3 hours. A census consisted of 25 surveys conducted in a period of approximately 25 consecutive days. The actual length of time to complete a census depended upon weather conditions, since intensive rainfall increased the turbidity of the water and decreased Figure 10. Figure 11. 33 O. inermis rigged for diet study inside stream enclosure. Trap used in crayfish surveys. F‘QEPE'I-Ir——’ f? “" '3" r- 34 mm c n e n V. C I. w. 3, 3 m. I"é‘* ‘A u... a so ‘ ‘ “u-‘, :19. 113:! 3 . . c. .. AU 3 v. 13 a... .C 5.. r S a. . a 3. S v . p" .3 «a . A : n all...“ .1..n'..‘.? (94.x; .h» I. , I V “JV— 5’): Z"! ..‘ H'.‘ JUU'lS‘S H a... In” nus 1”; w 35 the efficiency of visual capture. A census was conducted every four months during the period from March,l969,to March,l970. If a crayfish was captured for the first time, it was individually tagged. Tagging consisted of injecting a color-coded flexible rod into the space between the abdominal muscles and the sternum (Figures 12 and 13). Injection was accomplished by pushing the rod through a hypodermic needle with a wire plunger. To prevent interference with the ventral nerve cord, the rod was positioned parallel to, but to one side of, the midline. The rod consisted of flexible nylon and was made as short as possible to prevent interference with flexure of the abdomen. For the larger crayfish with a carapace length of 18 mm or larger, the rod consisted of a 3.5 mm section of monofilament fishing line with a diam- eter small enough to be injected with a 23 gauge hypodermic needle. For the smaller crayfish nylon sewing thread, 1.5 mm in length and fine enough to be injected with a 26 gauge hypodermic needle, was used for the rod. Crayfish with a carapace length of as little as 7.5 mm were successfully coded and recaptured. The rods were coated with three bands of color, using Testor's pla enamel paint (commonly used by hobbyists for model construction). Because the implanted color-coded rods maintained a stable position, it was possible to read the code unidirectionally. By employing a total of seven colors in a triplicate code, 210 permutations were possible, and, by repeating the same color in non-adjacent positions, 245 individual codes were available. This coding system was expanded by using duplicate codes for different species, sexes, and rod sizes. The code was visible through the trans- parent sternites and was retained after molting. This new tagging technique is a refinement of a liquid injection method developed by Slack (1955a) for group-tagging crayfish, and of a Figure 12. Figure 13. 36 O. inermis with identification tag. The code was injected while the crayfish was secured to the mounting board. C. laevis with identification tag. 37 38 ferromagnetic group-tagging technique described by Stewart and Squires (1968) that is employed in lobster studies. Hobbs (1973) has also developed a marking technique for the identification of individual cray- fish, but the presently employed method appears to cause less trauma and to result in a higher certainty of long-term identification. Each coded crayfish was assigned a number which was marked on its carapace with waterproof ink. In most cases this allowed noting recaptures without handling the crayfish. A paper punch was used to clip the margin of the outer ramus of a uropod. The clipped margin partially repaired itself with the next molt; thus, the appearance of the margin provided evidence as to whether molting had occurred during the inter-census period. A crayfish was clipped every time it was captured in a different census, and the particular uropod clipped depended on the census. The length of the cephalothorax was measured with dividers to the nearest 0.5 mm. The species and sex were recorded, and the sexual form of the male and breeding condition of the female were noted. The linear location of the crayfish in the stream and the type of capture were also recorded. For recaptured crayfish, the loca— tion and type of capture were recorded; if they had molted, they were remeasured, examined for sexual condition, and their carapace number replaced. After collection of data, all crayfish were returned to the stream at the point of capture. Crayfish living in surface streams were also censused for compar- ative purposes. One site was a stream issuing from Sullivan‘s Cave that was sampled with modified (enlarged entrances) minnow traps. The primary surface site, however, was at the base of the cascade in the stream issuing from Shiloh Cave, and it was sampled with modified minnow traps and jar traps. The crayfish population of a 50 meter 39 section of this stream was isolated by placing hardware cloth barriers across the stream. The barrier edges, at both the top and sides, were recurved on both sides to guide crayfish back to their original section of stream. Based on census data, the standing crops of the various crayfish populations were determined by the Schumacher and Eschmeyer (1943) mark~recapture method, which is based on a weighted least squares fit- ting. The formula for estimating the population size is N = [n2(m+u)] (DWUJN‘ (M?) where n number of marked individuals in the population, m = number of marked individuals captured during a given survey, u = number of unmarked individuals captured during a given survey, 5 = summation over all surveys (k). DeLury (1958) gives arguments for preferring this method over the Schnabel method, which is based on maximum likelihood. Momot (1967) compared both techniques in his crayfish study and found the Schumacher method gave more consistent results. Cave populations were permitted free movement, with corrections made for the interference of such movements with population estimates. Sex, size, season, and species-specific movement rates were calculated from the displacements between initial and final captures of crayfish encountered at least twice during a census. The applicable value, expressed as displacement per day, was divided into the average distance to the census area boundaries for each crayfish captured during a survey. In this manner the expected length of time for a particular marked crayfish to remain in the study area was calculated. 40 This "estimated sampling life span" was reassigned after each recapture. If a marked individual was not recaptured within this time interval, it was assumed to have left the study area and was removed from the tally of marked crayfish. If population densities are similar in various longitudinal sections of a stream, then this method should be valid whether movement is random or directed at the population level. Even if population densities do vary in different reaches of a stream, non—directed population movements would not invalidate this technique. The more sophisticated mark-recapture method described by Jolly (1965) was also applied to the data. This stochastic model, which accounts for migration, failed to produce reliable population estimates. This conclusion was based on results indicating that the estimated total population was often less than the number of crayfish encountered during a single survey. The failure apparently resulted from the extremely small crayfish populations being dealt with in this study. Length-Weight Relationships After completion of the field study, specimens of both crayfish species were removed from the cave for measurement. Carapace length, which is referred to by the abbreviation CL, was measured with dividers. Wet weight was determined after lightly blotting formaldehyde-preserved crayfish; dry weight was determined after drying to constant weight at 60° C. Molted skins were either found in the cave stream or obtained from laboratory-held animals. Skins were measured by the same methods employed for crayfish bodies. 41 Population Attributes Tissue Growth Marked crayfish recaptured at annual intervals (i 15 days) were measured for growth of carapace length. Growth measurements were made on crayfish captured during censuses and other sampling occasions from November 1968 to July 1970. Subsequent recaptures monitored growth of individuals for periods of up to 3.5 years. Recapture of marked cray- fish and size frequency analysis were employed to determine growth of epigean C. laevis. Based on the relationship between the annual increase in carapace length and the initial carapace length, a growth curve was constructed. This curve showed the relationship between mean carapace length and age. An individual was assigned to a year group by comparing its length to the growth curve; for example, if the length of a crayfish indicated that it was between 3 and 4 years old, then it was assigned to the fourth (3-4) age group. Growth in length was converted to dry weight increment by means of the length-weight relationship. From this the relative annual rate of growth (h) and the instantaneous growth rate (9) were calculated using the formulae: w - w h= t 0 w o g = loge (h+l) where wb = weight of crayfish at the beginning of the time interval, wt = weight of crayfish at time t, t = time interval of one year. These formulae, derived from Ricker (1958), were used to calculate growth for each year group. 42 An experiment was carried out to compare the growth rates of young-of—the-year C. laevis kept under epigean and hypogean conditions. Crayfish from a single brood were used; they had just detached from the pleopods of the mother and were 5.0 mm in carapace length. Each experimental crayfish was placed singly in a glass quart jar containing some substrate and organic detritus from either the epigean or hypogean stream. The jar was sealed with a plastic screen having 2 mm apertures. Fifteen jars each were placed in representative riffle areas in the cave and surface streams, with contained substrates matching the location. The jars were placed so that the substrates inside and outside the jar were flush and separated by the screen, allowing the exchange of some substrate, detritus, and fauna. These crayfish were sampled periodically for a year, with five jars selected at random from each habitat on each sampling occasion. The dry weight and carapace length of sampled cray— fish were determined. Molting Data on molting was gathered as follows: the annual molting pattern was determined by observing molting and growth of individual crayfish during the three 4-month intervals between the four censuses; molt increment was determined by the above method and also from newly molted individuals found associated with their old skins. Because of its aberrant nature, molting information from laboratory animals and regenerating crayfish in the field was not included. 43 Reproduction Ovarian egg counts were not made because of the disruption this would have caused to the crayfish populations. The populations of these cave-inhabiting crayfish were so small that the removal of even a few individuals would have had a significant effect. The reproductive cycle and production of fertilized eggs were estimated from counts of eggs carried on the pleopods of captured females. Mortality The annual total mortality rate (a) and the instantaneous rate of mortality (1) were calculated using the formulae: N - N a = o t N o i = -lo l—a ge( ) where N0 = the number of crayfish present at the beginning of the time interval, Nt = the number of crayfish surviving to time t, t = time interval of one year, except for 0.5 year interval in calculation of pleopod egg mortality. The above formulae, derived from Ricker (1958), were used to calculate mortality for both sexes, and for each year group. Mortality rates, derived from a modified type of horizontal sampling, were based on the observed annual mortality of year class cohorts. In addition, the mean age composition, as determined from four censuses, was used to establish "mortality" rates based on vertical sampling. 44 Population Size and Structure Crayfish of both species were censused by the mark-recapture technique of Schumacher and Eschmeyer (1943), with recapture rates computed for stratified groupings based on species, sex, age and season. For 0. inermis differences between the sexes and seasons were found to be only random variations, and the recapture data were reblocked into age categories only. The census area consisted of 500 linear meters (750 square meters), and population estimates were ob- tained at 4-month intervals during the year of study. The populations inhabiting the five sequential 100 meter sections of the census area were also calculated. Estimations were based on the proportions of initial captures in the various sections. The population of C. laevis inhabiting two surface streams, one issuing from Shiloh Cave and the other from Sullivan's Cave, were also censused for comparative purposes. Biomass and Productivity Known growth rates were used to follow the various age classes through the four censuses. Annual body tissue production was computed for each class by summing production computed for the three 4-month intervals between censuses. Production was calculated by the formula, + o N loge Nt l ge o logeP = 2 + loge (w - w ). Thus the production for the period was equal to the average number of (loge N + loge NO) specimens, e 2 , multiplied by the weight t increment of a specimen, wt - wo. Computation of the average number is 45 based on the exponential nature of mortality. Production of young-of- the-year was derived for the annual interval with No and E; referring to pleopod eggs. Egg production was based on pleopod egg counts and data on egg weights. Annual production of molted skins for each age class is based on the formula, E. (ile), where fi.is the mean number of the cohort from the four censuses and Ed represents the total weight of skins shed annually by a member of this age group. Information on molted skins was derived from growth rate and molt increment data. Production was computed for the census area of 500 meters and the quantitative study sub-area of 364 meters. The cave stream had a mean width of 1.5 meters; thus the census and quantitative study areas contained 750 and 546 square meters, respectively. Respiration The oxygen consumption of O. inermis and C. laevis from both habitats was studied by measuring oxygen depletion in sealed chambers. Crayfish were collected and individually placed in plastic chambers with sufficient natural food for a 24-hour period of adjustment to experimental conditions. They were then transferred to a clean chamber of similar dimensions that contained cave water with a known oxygen content. A small amount of calcium carbonate was added to buffer against change of pH, and the chamber was sealed. The buffering precaution was taken, despite the findings of Fox and Johnson (1933) that ventilation in Astacus fluviatilis was unaffected by a rise in the carbon dioxide tension and of Helff (1928) that changes in pH and carbon dioxide content had no marked effect on the oxygen consumption of Orconectes immunis. A wide size range of crayfish was studied, and 46 the size of the chamber was chosen so that its volume was approximately 300 times the volume of the crayfish. In order to maintain an accurate temperature, these chambers, together with control chambers, were kept in the cave stream for the duration of the experiment, which was termi- nated after 24 hours when oxygen levels had decreased to approximately half of the saturation value. Studies of the epigean crayfish population were carried out in the surface stream, which maintained the same tem- perature as the cave stream. This oxygen consumption study, which was conducted in the field during May, was followed by a similar laboratory study in November. Because the rate of respiration in crayfish is often correlated with oxygen tension, studies were carried out to ascertain this relationship, and to allow computation of a correction factor forihe change in respiratory rate incurred during the experiment. These rate—tension experiments were carried out in Erlenmeyer flasks that were sealed with a layer of mineral oil; this arrangement permitted periodic sampling by withdrawing a water sample with a syringe. Because of accumulation of nitrite nitrogen as a waste product during the course of the experiment (less than 1 ppm), the azide modification of the Winkler method, as described in Standard Methods for the Examination of Water and Wastewater (twelfth edition, 1965), was used for determination of dissolved oxygen content. A micro-Winkler technique, described by Fox and Wingfield (1938), was employed to analyze the 10 ml samples obtained in the rate-tension study. Lawton and Richards (1970) concluded that the various respiration-measuring methods - cartesian diver, Gilson, Warburg and Winkler-— were equally suitable for ecological work, provided they were used with a sound knowledge of the animal under investigation. 47 Time Frame The study commenced in July, 1968, with construction of the weir and the entrance closures. By October the weir was operational and data collection commenced on stream flow and drift. In November, 1968, a preliminary crayfish census was carried out in which the coding procedure was tested. The study was in full operation during the 1969 calendar year, with data collection on food inputs and crayfish populations. Trophic and respiration studies, which disrupted the crayfish populations, were carried out during the summer of 1970 after other phases of the study had been terminated. Follow-up crayfish surveys were made through 1975 to collect long-term growth data. CHAPTER V THE ENVIRONMENT Results The Stream Hydrology The mean rate of flow at the weir for the study year of 1969 was 2,077 kl/day; the median daily discharge was 776 kl. There was a 186- fold variation in daily flow rate, from a low of 173 kl to a high of 32,300 k1. The two greatest floods occurred in January, but other floods took place in April, July, November and December. The lowest flow was recorded in October (Table l and Figure 14). In general, flow rates throughout the year underwent a dual pattern. Flow rates were high from December to April, and discharge was highly correlated to precipitation; during the remainder of the year, however, flow rates were low, and discharge was not very responsive to precipitation (Figure 15). For example, similar 8 cm rainfalls resulted in a 22—fold increase in discharge rate in January, whereas there was only a lO-fold increase in July; in addition, the discharge rate of the January flood was 6 times as great. There was even a greater dichotomy at lower precipitation levels between the two annual phases. 49 0000000 0000003 0000000 Eoum 000000300 00000 0000000 0005 0000 000.000.0 .. 0.0- 0.000 90020 00909 0000 . . n . n u 100 H 20020 50 00 00 00 + 00 0 00000 + 0000 0000 + 5500 .. .. 90020 00000 0000 00.00 00.0 H 00.0 0000 H 0000 0050 H 0000 0.0: 0.0 .000 00.00 00.00 H 00.0 0505 H 0000 0000 H 0000 0.5- 0.0 .000 0500 00.5 00.0 H 00.0 005 H 5000 000 H 0000 0.0: 0.5 .000 00.00 05.00 H 00.0 0000 H 5000 0000 H 0000 0.0+ 0.00 .>02 00.0 00.0 H 55.0 000 H 000 500 H 000 0.0+ 0.00 .000 05.0 00.0 H 00.0 00 H 000 00 H 000 0.0- 5.0 .000 00.0 50.0 H 05.0 05 H 500 00 H 000 0.0: 0.0 .000 00.5 05.0 H 00.0 000 H 005 005 H 005 0.0+ 0.00 .000 00.0 00.0 H 00.0 500 H 000 000 H 000 0.0: 5.00 .000 0000 50.5 00.0 H 00.0 000 H 0000 000 H 0000 0.0: 0.0 002 00.05 00.00 H 00.00 50000 H 0000 0000 H 0050 0.0+ 0.00 .000 00.0 00.0 H 00.0 000 H 500 000 H 000 0.5- 0.0 .002 00.00 00.00 H 00.00 5500 H 0000 0500 H 5000 0.0- 0.0 .000 00.00 00.00 H 50.00 00000 H 00000 0000 H 0000 0.0+ 0.00 .000 00.00 00.00 H 00.0 0050 H 0000 0000 H 0000 0.0+ 0.00 .000 00.0 00.5 H 00.0 0000 H 0000 0000 H 0000 0.0+ 0.00 .>oz 0000 200000 200000 2000 0000000 9209200 100 9209200 2000 2000 000900 000020 900003 000 000020 200900>00 000000 2090000000: 2090000000: 2500 0:09 0>0909000> «2009090000000 000 H 20020 0003 90 90020 00000 00909 magmd ZOBmMm 92¢ .mwmfimomHQ gfim .ZOHBANBHmHummmII.H mamdm. 2.4: ”and: (v; uvmw-Au //\.I SO .>©suw mo 000% 0:» 00m 0000000 0000:0000 800000 000 :00umu0m000um 00000 aria/v.0: . .. 0 L...) 1" .1. ., 0°: .0... .00 005m00 Sl ( “19) NOIlVlIdIOBHd 9 L __1‘ j y 1. 1 1 9 9 _A L L Y W 0 E I 3 Jun: 0 E o a S’.’ < z _, O a: (D _ U... H _ _¥ > .IIIIIUIII... h< r3 l E‘ r- < 8 SEP L 110-] FEB L - I AUG .000‘ 15,000+ 0 .000 15,000-4 (KOO/DI) BDHVHQSIO WVSELLS Figure 14 52 .500000 0>0o 0:0 00 0000:0000 0:0 :000000m0000m c00300n @030c0000a0m .00 005000 TATION ’ PRECIPI ‘0- FLOW 9" 8000 -20 (”3) 53 NOIlVllleHUd .9 4000 1970 SEE NOV. JAN. JUL. 1969 Figure 15 JAN. MAR. MAY 1968 NOV. 54 Discharge records extended over a l6-month period, and that portion of the annual cycle from November to February, when flooding was maximal, was repeated. It was quite evident that the extent of flooding varied considerably from year to year; the mean daily discharge of the repeated 4-month period underwent a 57% decrease. The observed inter- and intra- annual variations in stream flow are largely the result of precipitation patterns. The average annual precipitation at the Oolitic Weather Station was 116.8 cm. Precipitation for 1968 was 4.0 cm above the norm, but 1969, the study year, was 6.6 cm below. The precipitation of the last two months of 1968 and the first month of 1969 was consistently above the norm, especially in January when precipitation was 83% above the average level. The 44.1 cm of precipitation during this 3-month period, which was 38% of the average annual precipitation, resulted in the extensive flooding of the cave stream observed during this same period. The stream of Shiloh Cave has an estimated watershed of 7.8 kmz, of which 20% is forest and the remainder is primarily farmland (Table 2). This watershed can be subdivided into smaller drainage areas associated with the various tributaries of the cave stream, although the accuracy of these determinations is increasingly uncertain the smaller the area. The three major tributaries ——-the Black Damp, the Graveyard Waterfall, and the Main Stream-— are all in fairly open communication with the surface by either collapse sinkholes or swallow holes. The minor tributaries do not have obvious points of communication with the surface and probably form from water percolating through dolines or intermittent stream beds. The watersheds of these minor tributaries possess the extremes of surface features, with the 285 Meter Tributary draining flat, cleared land with a high sinkhole density, and all others draining forested hills that lack sinkholes. 55 .0000 000 n 00 00 00 00 00 0 mm 00000000000 u 0 .0000 00000: H mm WUZ¢UHmszHm om.A0 om.A0on. om.AOAom. mo.A0Ao0. 0o.AOAmo. 0oo.AOA0o. 0 00.o+ m0.0| om.0+ «0.0+ mv.m+ om.vn 0 m m m m m 0 0 mo.o+ 00.0: om.o+ mm.o+ vm.o+ 00.0: 0 mMZflB 0080mm mMAOmMZHm Emmmom mMZdE UHfimmm mmqomMZHm Emm000 9080 #60020 wBHZHO¢M0¢ 000000000 000 000000000 000000000 BZMBZOU mmeflz 020 mMMDB¢m0 modmmDm Dmmmmm903 zmmzamm 08mmal0 mBHSV ZOHedqmmmoo .. .. .. 00 mm om 05.5 00009 mm.m mvm 00000000000 00 0 05 00.0 00000000 05.0 0m0 00000000000 0 o 000 No.0 0800 0000 00.0 000 00000000000 0 o 000 No.0 0000 0000 00.0 050 00000000000 mm o 00 00.0 0300 00000 00.0 000 0000 3000030 00 on 00 m.m 0800 00000 50.0 mmm 00000000000 00 vm o 00.0 000000 00002 000 mm.0 com 0000 3000030 m 00 Om «.0 000000003 000>0>000 00.0 500 00000000 00000000 00 00 00 0.0 0000008 CV 800000 0002 A0\0mov 00000 2000000202200 00209 00000 000000 Amsxv BZNBZOU wBHZHA z¢mwomwm 0090mm iMZHm w «HMG MfidfiDmHmB wommzm I¢M0¢ Izmmonm N8\>BHmzmo mwdszmD m>¢0 mOQHmm mmqmz¢m mma¢3 m0 $005240 DmmmmmBmZ 0MBumu=nwuuuq ocflonw uwumz o:a mm.0 m.0H FH.0 H.0H mm.0 o.vH mv.0 H.mH mmaHm m» o H MHN >0 0 m 00H m» 0 b How HH H 0 50H 30am gzw>nuo n 0v. mo.0 0.5 m.moH 05.0 ocm Hm.0 HmH 00.0 va Hm.H 00H muouos oom 0N. 00.0 H.m n.mmH «v.0 0H~ 00.H mmH 05.0 va Hm.0 an muwuue 000 m N mH. 00.0 0.0 0.00H no.0 00m 00.0 va Hn.0 oom 05.0 50H muwuws 00v 5 pm. 50.0 ~.HH m.00m mv.0 mHm mo.H mmH 50.0 How no.H mmH muauws oo~ n m~.0 00.0 >.HH m.~0~ 0m.0 0HN ~H.H NmH 00.0 mom mw.0 00H mumums 0 m m * m x xwumcm .xH0 >muwcm .xH< >muwcm .me >muwcm .xH4 HH\HauV»ummzm . >952“. I 54 I H33 2.? .azmzomzoo 0.~ 0.0 0.5H m.mH HoomxH. mHmz a4 scum omszzoo mmqmzH=n HH HHum< MH .cmn mmqmfim mugs; Adzomémw mo BZEZOU >Ummzm 92¢ EHZHg mace II. n mung. 58 .mwflumusnfluu muH 0cm Emmuum :HmE mzu Eoum xaumuumsw UmeEdm umum3 mo muHCHmead “is".a't . . ,. uni: “51“ .h .oH musmflm 59 ma muzmflm ZO_._.m<._.:m_m._. % .r 3 60m . H 7 a . 33:5 22: *wov #33 Cum no 8. nov 9:8 m6 *3». 023 ad 053550 0.0 5 3555 3.233 mace 2259.35 5253 man 44:52; 2:22525 u *0 26.: 23:5 no R oz< 29:80.. £2595 60 because sinkholes occur more commonly on flat land, and it was this land that was cleared and used for human habitation. The causal agent is uncertain, but lower alkalinities would logically arise from waters percolating through acidic woodland soils. Alkalinity values for the Selinite Passage were not included in the correlation analysis, because its high alkalinity level did not correspond to the pattern apparent in the other tributaries. Although this deviation may have been real, there was a strong possibility that the water samples had been contaminated with substrate, since it was physically very difficult to obtain a water sample from this particular site. Some of the minor tributaries, such as the Sand Spit, Half Dome, and Selinite, were only found flowing during the winter months. Compared to the main stream, those tributaries that were active throughout the year showed fairly large fluctuations (a mean range of 33 ppm) in seasonal alkalinity levels. Although the mean alkalinity for the constant flow tributaries steadily increased from January through October, seasonal patterns varied widely among them. This was especially noticeable in the October samples that were taken when the stream was near its base level. The large horizontal tributaries resembled the main stream in that their alkalinities were at a yearly high. On the other hand, the minor tributaries, with water flowing from higher rock strata, had low alkalinity levels compared to samples taken at other times of the year. Water dripping from a stalactite at 250 meters during October was found to have an extremely high alkalinity level of 328 ppm. Although there were many active stalactites over-hanging the stream, their rate of dripping was quite slow, and their combined effect on the cave stream was probably minimal. 61 Temperature The water temperature of the main stream was quite stable through- out the year. The average temperature was 12.5° C, and variation between summer and winter, during periods of normal seasonal discharge, was only 1.5° C. The total annual variation, however, was 11.0° C. This indicates that large temperature fluctuations are of a transitory nature, apparently associated with flooding. Food Resources Food resources in the cave are essentially of an allochthonous nature, since photosynthesis can only occur in the twilight zone and chemosynthesis is thought to be of only minor significance. Although a small fraction of this imported organic matter may be brought into the cave by the movements of air currents, troglophiles and trogloxenes, the great bulk of food enters the cave as vegetative detritus carried in by water. This study of food resources focuses on measuring the quantity and quality of this major source of allochthonous matter, but also analyzes the stream water and its substrate for organic content from all sources. Leaf Input Leaves randomly collected from litter at the cave entrance in January were by weight 51% sugar maple (Acer saccharum), 35% black oak (Quercus velutina), 13% beech (Fagus grandifolia) and 1% other species. In July the composition had changed, with 24% maple, 46% oak, 26% beech and 4% other species. The compositional change reflects the fact that the beech and oak trees do not shed their leaves until spring; these 62 leaves were not available for the January sample and by July had not undergone the decomposition experienced by the maple leaves, which had existed as moist litter throughout the winter. The three most common leaf species were chemically analyzed from samples taken from the living foliage, newly-shed leaves, and January and July entrance litter (Table 4). In the foliage the lignin content was much lower in maple than in the other two species, whereas cellulose and ash were lowest in oak. Crude protein varied within narrow limits, with oak having the highest value and beech the lowest. Maple was highest and beech lowest in "other components"; the values of this category were obtained indirectly as the percentage of the leaf not accounted for by the other categories. Extensive decomposition had occurred in the leaves collected from the litter at the cave entrance. In both January and July leaves the lignin content had increased greatly in all species, but was still lowest in maple. The litter leaves had also increased in ash content, with the highest value found in maple and the lowest value still found in oak. The cellulose content from the litter samples was approximately the same as that from fresh leaves, and was still lowest in oak. Both protein and "other components" underwent a marked decrease during decomposition. The components of newly-shed leaves were generally of intermediate value, with ash content similar to that of fresh leaves and protein levels resembling those of the entrance litter. The energy content of the foliage was 5.06 cal/mg in oak, 4.60 cal/mg in beech and 4.56 cal/mg in maple, but in the entrance litter the values decreased to 4.92, 4.40 and 3.91, respectively. The species ranking probably reflects to a large extent the content of energy-rich lignin. 63 G 3 3 0.00 H00 0.00 0.00 0.00 0.00 H00 000 0.00 .. .. .. umfiom 0.00 0.00 0.H0 0.00 0.00 0.00 H.00 m.H0 0.00 .. .. .. 0.000000% H.00 0.3 0.00 0.00 0.00 0.00 0.00 0.0H 0.00 .. .. .. mmoHsHHmom O 0.00 0.00 0.00 0.00 0.00 H.0m 0.00 0.00 0.00 .. .. .. swan 0.00 0.00 0.00 0.00 0.00 0.00 0.3 0.00 0.00 0.00 0.00 0.00 umfiom 0.0 H0 0.0 H.0 0.0 0.0 0.0 0.0 0.0 0.0H 0.0H 0.0H 036.8% 0.0H 0.0m H.0H 0.0.0 040 0.0H 0.Hm 0.00 0.0H 0.0H 0.0H 0.0H mmoHsHHmom O 0.00 0.Hm 0.00 0.00 0.00 0.00 0.0H 0.00 0.Hm 0.0 0.0H 0.0H fic03N 0.0H 0.0H 0.0 H.0H 0.0 0.0 0.0 H.0 0.0 0.0 H.0 0.0 :00 3002 06000 020 30.0: mommm 020 mamas 00000 020 3.22 mommm 020 m 0.50 00002.00. azmzomzoo m 00.0qu moz ZH mm>¢mq m0 ZOHBHmomEOUNQ 02¢ ZOHBHmomZOU m0¢92mummmll.v mamdfi 64 Percentage composition analysis does not indicate the rate of decomposition of the various components between foliage, shed and litter leaves; this was measured by selecting lignin, an extremely refractory substance, as a marker to measure the change in the other components. Lignin actually decomposes at a low, but unknown rate, and the decreases calculated for the other components must be regarded as underestimates. Absolute and relative changes of component content, measured as the percentage change of the component/lignin ratio, were computed for the three major species of foliage and July entrance litter leaves (Figure 17). Relative to lignin, the ash content underwent the smallest decrease, ranging from 29 to 38%. The decrease in cellulose ranged from 41 to 73%, and in "other components" from 58 to 80%. Protein decomposition was the greatest, with decreases that ranged from 79 to 87%. The relative component decomposition was consistently the lowest in beech and greatest in maple. At least part of the greater decom- position in maple leaves can be accounted for by the fact that they exist as moist litter throughout the winter, whereas beech and oak leaves are not shed until spring. Warm, moist air flowing out of the cave during the colder months maintains a microclimate in the entrance litter that is favorable for decomposition. Some seasonal differences in the decompositional state of entrance litter were noted, although these differences were generally of small magnitude and often not consistent; mixing of the current crop with previous litter, especially in the slowly decomposing oak and beech, probably minimized seasonal differences between samples. In maple seasonal differences in the percentage decrease from the fresh state 65 '_-' ‘I‘tfl‘ ‘ " .L'XI’SA": Figure 17. Decompositional changes between foliage and July entrance litter; based on component/lignin ratios and assuming that lignin is not decomposed. 66 ........ --------- uuuuuuuu ........................................ ....................................... ........................................ ....................................... ........................................ ...... ....... r v ‘5 . 40 so % DECREASE BEECH - - 6 200 ' Wm MG DECREASE/G OF DRIED FOLIAGE Figure 17 67 were minimal for all components, except for cellulose, which was not so extensively decomposed in January (59.8%) as it was in July (73.1%). In beech litter, cellulose, "other components" and especially ash were more highly decomposed in January; beech leaves collected from the litter in January were probably heavily weighted by leaves that had been dead for 15 months, whereas those in the July sample were mostly shed only 9 months previously. Although oak has a pattern of shedding similar to beech, it showed no consistent or extensive seasonal differ- ences, and cellulose, which varied the most, was actually more highly decomposed in July. The overall seasonal decompositional state for the three species could be simply deduced from the percentages of lignin in January and July litter. Lignin was relatively more abundant in the more highly decomposed leaves; therefore, oak and beech were more decomposed in January, and maple was more decomposed in July ——- a pattern which reflected the length of time spent as litter. These seasonal differences, however, were very minor, and the decompositional state of litter entering the cave was approximately the same throughout the year. A more important seasonal difference was the species mix, which was probably highly weighted with maple leaves during the winter, and with oak and beech leaves during the spring and summer. The decompositional process could be followed in leaves, from foliage through shed leaves and litter to cave stream macroseston, by analyzing species mixtures based on the proportions found at the entrance in either January or July (Figure 18). Relative to lignin, the amounts of cellulose and "other components" decreased at a near-constant rate as the leaves proceeded from foliage to macroseston, with "other components" decreasing at a faster rate. The ash content was decreased in newly-shed leaves, but this level was maintained and even slightly 68 .oflumu chmHH\ucwcomEoo ocu mo mmmcmco so @0002 00 .:mu:mcomfioo “capo: paw :fimuoum .mmoHDHHwo .me .mucwcomfioo any mo mmcmcv wmmucmoumm one .>Hm>fluommmmu .mem can mm .omv >Hso :0 Ho Hwom cam ma .mmv >HMSGMh ca uwpufifl wocmuucw on» «o oflumu mwflommm may on mcflcuooum UmucmHQB .mamma 6cm comma .xmo mo wuamomEoo m mum mm>mmq .coummmouomfi cam umuufla mocmuucm mo mmHQEMm >H96 cam >MmSCMb c003 .mo>mma mafimomeoomp CH mmmcmzo ucwcomaou .0H mus0Hm 69 0H 0.33.0 WhCO z. wm<04 wawxm >430: >43“: III. >532: .II 002029.200 3:00 0 23005. D 000.5qu 4 10¢ 0 502.5“. 70 increased in the entrance litter; perhaps this resulted from substrate contamination or a rate of decomposition even less than that of lignin. Ash again decreased as the entrance litter was converted to macroseston, but the net decomposition from foliage to macroseston was much less than that of the other components. Crude protein underwent the most extensive decomposition of any component during the conversion from foliage to newly-shed leaves. The state of protein decomposition in the entrance litter, however, was only slightly greater, and, by the time the leaves had been transformed into macroseston, the protein level, relative to that in the litter, had actually increased by 18%. This increase could have been an artifact resulting from lignin decomposition, although the increased microfloral populations attacking leaf substrates in an aquatic environment were probably responsible. The leaves of the macroseston contained only a fraction of their original decomposable components; only 57% of the ash, 21% of the protein, 19% of the cellulose and 6% of "other components" remained. Lignin increased in importance as the most abundant leaf component. All decomposable macroseston components -—-ash, cellulose, protein and "other components" —— were in a more decomposed state in January than in July (Figure 18). Macroseston Several factors, such as the time interval between floods, the relative magnitude of flooding, and the time of year, seemed to have some influence upon the amount of vegetal macroseston, although the relationships were not well defined. The daily amount of macroseston Passing through the weir was best described by a log-log relationship with daily discharge, with different regression formulae for decreasing 71 and increasing flows (Figure 19). The macroseston load for discharge rates above 650 kl/day was greater during periods of increasing flow. The load increased with discharge during periods of increasing flow, but declined slightly for periods of decreasing flow. In calculating the annual macroseston input, the preliminary daily input for those days not sampled was computed by applying the proper regression formula to the known discharge rate. This computed daily input was corrected by a factor that was derived by calculating the ratio between actual and calculated inputs of the bracketing sample periods. This correction adjusted for a complex pattern of temporal variation, and also for an underestimation of macroseston inherent to the log transformation. The calculated annual input of vegetal macroseston into the quantitative study area was 2405 dry grams (Table l). Macroseston entering the cave during the months of January and April, when the two greatest floods occurred, accounted for 64% of the total annual input. By contrast, the three-month period from August through October, which experienced the lowest mean flow rates, contributed less than 3% of the annual input. The mean composition of the vegetal macroseston was 12.6% leaves, 36.8% wood and 50.6% detritus (Table 5). Chemical analysis of the detritus indicated that it was composed of about 53% plant material, 6% animal material, and 41% inorganic matter. There were seasonal differences in the ratios of the macroseston components. Relative to wood, leaves were less abundant in the summer samples, but the difference was not significant. Similarly, detritus comprised consistently greater percentages in the winter samples, but again the difference was not significant (Table 5). 72 Figure 19. Relationship between vegetal macroseston and discharge for both increasing and decreasing flows. [Y] (dry g/day) VEGETATIVE MACROSESTON 73 INCREASING FLOW LOG v = 1.401 LOG x — 3.912 r2 = 0.80; n = 18 100' ) —"""" DECREASING FLOW LOG Y 8 0.903 LOG X - 2.522 r2 =- 0.72; n = 45 10J . l 11 0.1.1 T T ' I 1 1 Y 100 500 1900 5900 1qpoo DISCHARGE (kl/day) [x] Figure 19 2 0,000 74 TABLE 5.--ANALYSIS OF VEGETATIVE MACROSESTON ENTERING THE QUANTITATIVE STUDY AREA FROM FOUR WINTER AND FOUR SUMMER 24-HOUR DRIFT SAMPLES COMPONENT LEAVES WOOD DETRITUS SAMPLE PERIOD 1/5—9 7/25—29 1/5-9 7/25-29 1/5-9 7/25—29 % OF VEGETATIVE 14.4 10.8 27.5 46.0 58.0 43.2 DRIFT (1'1 daily 5) 14.8 11.4 19.0 19.1 16.6 17.7 ENERGY CONTENT 4.05 4.11 4.13 4.32 2.39 2.52 a (kcal/g) 10.02 10.02 10.01 10.03 10 10.02 S ‘6) 23% CRUDE % t1PROTEIN 13.4 11.4 10.6 10.0 8.0 8.5 O a) + + + + + . i , U +,(N X 6.25) -0.5 -0.2 .0.2 -0.6 -0 3 0 7 1-3 a) .0 u 2 rd g 5; 2 12.5 11.5 16.0 16.8 8.6 7.9 g ELCELLULOSE 10.5 10.7 10.4 10.7 10.1 10 U) :3 m 'U 0 "5 2 53.0 48.7 35.8 40.8 18.0 25.2 g a) LIGNIN 11.3 12.6 11.8 11.4 11.7 10.6 a“ g E: % 15.6 14.4 14.1 11.3 43.7 39.7 s ASH 10.2 10.4 10.3 10.3 10.5 10.4 t-TEST OF SEASONAL DIFFERENCES IN THE RATIOS OF MACROSESTON COMPONENTS COMPONENT WINTER SU R SIGN RATIO E" s Y’ s d f t P ns=not TESTED sign. LEAF 0.60 0.34 0.24 0.07 6 2.06 P>.05 ns WOOD LEAF + WOOD 0.74 0.19 1.38 0.51 6 -2.35 P>.OS ns DETRITUS 75 Based on component/lignin ratios, which showed component changes relative to the extremely refractory substance, lignin, seasonal differences were observed in the state of decomposition of the macro- seston (Figure 20). Ash, cellulose, protein and "other components" occurred at reduced levels in July wood and detritus. Leaves underwent more complicated seasonal changes, with a slightly reduced protein content and an increased content of "other components" in July. The greater proportion of lignin in the January sample indicated that leaves were more highly decomposed in the winter; this agreed with data derived from the study of the decomposition of leaf input, but did not match the seasonal pattern evident in wood and detritus. The seasonal differences in the lignin content of leaf macroseston did not appear to result from seasonal differences in the species mix of litter entering the cave, which would have had an effect just Opposite of that observed. A consistent increase in energy content for the summer samples of all macroseston categories was correlated with a concomitant decrease in ash content. Information from Table 5 was combined with published (Maciolek, 1962) energy values for lignin and cellulose to obtain the annual energy and crude protein budgets for the various categories of the macroseston (Table 6). Leaves, wood and detritus contributed 15%, 47% and 38% of the annual energy input, respectively. The total annual energy input was 7980 kilocalories; however, 46% of this amount was contained in lignin, which is highly resistant to decomposition, and 11% was contained in cellulose, which can only be directly utilized by a few organisms. The annual amount of crude protein entering the quantitative study area was 230 dry grams; this was 9.6% of the vegetal macroseston. Leaves, wood and detritus contributed 16%, 40% and 44% of the crude protein, respectively. Figure 20. 76 Percentage change of component/lignin ratios of vegetal macroseston components between four January and four July 24—hour drift samples; based on the assumption that lignin is not decomposed. s T N E m 0 .pl. P O M R" nu D. C E R D E U H R T by nv 7 7, F. 8 AU .L U .L H I. S E A C DETRITUS OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO WOOD 40H 0 mmma wusHOmnm 02 mm 302 m0 COHHOQ 0 000250 Umcwmuno 0000 502m Um>020© mmsam> 000000022 A0020 “000.0020 209020 0002 0m0.200.2 000 000.000 102022 2000 20000 m 0002 0000 000 0000 022 000 020 0000 w 000 0000 m G 9 m m 303 :0: m M 000 0000 000 0000 00 022 022 000 m I 000 0202 c. m 2020 2200 “M m 020 0000 022 000 00 000 000 0020 m m. 00 000 8 Tu O N 30 82 m 002 0002 0H 00 m0 00H 002 000 .m m0 mom S 02222922 90222 02222922 90022 02222922 90222 02222922 90222 02222922 90222 22909 0922202200 22290 200202220 02220 920 22 . 222290220 022 22909 92 2mmmmm 222020 9229200 902222 mAm>mA AdDZZ< 2222 00090 2>2929H92200 229 22 402222922 02222 20 02222922 92220222 0922202200 209020 20 9229200 902222 022 2229022::.0 22229 79 By employing two drift nets, one to monitor macroseston entering the quantitative study area and the other to measure output, quantita— tive and qualitative changes in the macroseston flowing through the quantitative study area were observed (Table 7). The components of the plant macroseston-—- leaves, wood and detritus (chemical analysis indicated the detritus to be about 53% plant material) —— underwent proportional changes in passing through the quantitative study area. Wood increased from 24.4% to 34.1% of the vegetative drift; on the other hand, leaves decreased from 10.2% to 3.3%, and detritus declined slightly from 65.5% to 62.5%. These changes probably resulted from the higher resistance of woody tissue to biotic processing. The animal drift also showed differences in input and output composition. Animals of terrestrial origin increased from 2.1% to 12.9% of the animal drift, and aerial animals increased from 15.1% to 22.9%. The aquatic components declined, however, with animals decreasing from 74.3% to 60.4%, and exuviae from 8.5% to 3.9%. This shift in component proportions resulted from both an increase in the amount of terrestrial and aerial animals, and a decrease in the aquatic animals and exuviae. The lesser amount of terrestrial and aerial drift intercepted at the 79-meter station probably indicates a restricted non-aquatic habitat associated with the Graveyard Waterfall Tributary and the main cave stream above the entrance breakdown. The greater amount of aquatic animals and exuviae caught at the 79-meter drift station is thought to be a result of the more populous benthos associated with the area immediately downstream of the collapse sink- hole at 0 meters. The percentage of macroseston of animal origin was only 3.0% at the 79-meter station; however, the proportion increased to 18.7% 80 AdBOB 00 O O ' O I. O 0 II I 0 0 00 0 0000 0 000 + 0 000 0 000 + 0 0000 02200 . . . I . . u . . n . . u . 00009 000 0 00 0 0000: 00 0 + 00 00 0 000 + 0 000 00 0 + 00 00 0 000 + 0 0000 00000 000 0.00 0.000- 00.0 2 00.00 0.00 2 0.00 00.0 2 00.00 0.00 2 0.000 0002 000 0.00 0.000: 00.0 0 00.0 0.0 2 0.0 00.0 2 00.0 0.00 2 0.000 00>000 000 0.00 0.000: 00.0 2 00.00 0.00 0 0.000 00.0 0 00.00 0.000 0 0.000 20002000 0 0 0 .0 I 00.00 0.00 I 0.00 00.0 I 00.0 0.0 a 0.00 00009 0 + + + + 028024 000 0.00 0.0- 00.0 0 00.0 0.0 0 0.0 00.0 0 00.0 0.0 0 0.0 000>022 000 0.00 0.0: 00.0 0 00.00 0.00 2 0.00 00.0 2 00.0 0.0 0 0.00 0000002 000 0.000 0.0+ 00.0 2 00.0 0.0 0 0.0 00.0 0 00.0 0.0 0 0.0 00000002209 000 0.00 0.0+ 00.0 0 00.0 0.0 0 0.0 00.0 2 00.0 0.0 2 0.0 000202 22 0.20 20 2.00 20 0 00 22 0.00 03 20 0 02 220 02 03 20 0 02 020 02 00 00 0 02 220 02 222922 000 02 222022 00 92 922200200 2222 00092 2000229 900000 00020 20922202022 OZHmmdm ZH BDmZH 2H mwz¢mu ZOHBdBm ZOHBUMAQOU 209mmmOMU¢Z 0002222 20228 20 2:0 22 Umum002mua0 00022 200 222202m500 00 owumsm©2 2223 20502> 2000202 22225 020 0:2 20m2250202 302 mo ©002mm 2 @20252 029052 20 02220 220Q522 2sonnvm 00222 8020 3000 22023 00022 02202 2:2 200 202250020 223 GoummmOHUME 24HMd NQDBm W>HB¢MA 02¢ UZHMMBZM mfizmzomSOU ZOEmNmOMUdZ m0 ZOmHm¢mSOUII.h mam¢8 81 at the point where it left the quantitative study area. The absolute amount of animal drift in the input and output was remarkably balanced (less than a 0.2% increase), and the change in proportion between the animal and plant components was instead due to an 88.1% decrease in the amount of vegetative drift in passing through the quantitative study area. This represented a daily consumption of 1188 i 223 mg of vegeta- tive macroseston by the biota of the quantitative study area. The equality in the levels of animal drift entering and leaving the study area indicated that animal drift had no net effect on energy flow and should not be included as an energy input component. Microseston Analysis of the microseston load carried by the receding waters of the largest observed flood showed a definite relationship 2 (r = .89) between load and rate of discharge. This relationship is described by the regression formula, microseston load in cal/1 = 0.00922 (discharge in l/sec) i 0.836. This regression indicates the increased load capacity of higher flow rates that apparently results as a function of scouring. The flow of microseston into the quantitative study area was quantified by applying the regression formula to daily discharge rates. The annual input of microseston, which included particles in the 0.5 - 230 micron size range, was calculated to be 1,401,239 kilo- calories (Table 1). Because of the variation in flow rates, the seasonal distribution of microseston underwent extreme fluctuations. For instance, 48% was carried into the study area during January, while the five-month period from June through October only accounted for 4%. 82 The decline in microseston levels between water samples collected quarterly from the 200 and 400 meter stations indicated that only 0.10% of the microseston input was retained in the quantitative study area. Samples taken quarterly from five stations positioned every 200 meters along the main stream indicated that the microseston level re- mained fairly stable throughout the entire length of the cave stream; it averaged 0.79 cal/1 and had a range of 0.42 - 1.51 cal/l for all stations and all seasons. Microseston of the tributaries flowing from higher rock strata was consistently greater (mean = 1.22 cal/l; range = 0.58 - 2.22 cal/l) than that of the main stream, and it seemed evident that the tendency for microseston of the main stream to de- crease as it flowed downstream was counterbalanced by the input from the tributaries (Figure 21). There were seasonal variations in micro- seston levels, especially at the tributary stations (Table 3). Most of the variation, however, was caused by differences in discharge rates, as A.O.V. testing of seasonal microseston levels adjusted for discharge indicated that the variation was not significant (.10 > P > .05). There was not a significant correlation between microseston and any measured surface feature of the various watersheds (Table 2). Another phase of the study, in which microseston and benthos of several nearby caves were compared to various surface features, indicated a strong correlation with septic tanks. That this same correlation did not occur in considering the various watersheds of Shiloh Cave, itself, may have resulted from inaccuracies in delineating the watersheds of the extremely small tributaries, or from rapid processing of the microseston-poor water entering the cave system. 83 .. . ,. ! . I I] . 2. t h < : . . i... 1 u-.|I.H‘.--.- . - 1 ‘J I- A .. I 52.-51%}-.. . >>C _. S.< 2. . .p. .3 x... ,ZZV<|. .3! ‘ A ..DI'... \. IIIHlblil I'I | «I ...‘|I\I. . 1' | l. Il‘.1 .mmflumuonfluu mufi cam Emmuum same may Eoum >Humuumov omHmEMm coumwmouofle mo hmuwcm .HN whoop 84 am musmflm ZO_.—.m.:.\.\:::\ In 1 .COHumum Hoom comm um mmuflm OBu Eouw >HamocchHEom UoHQEmm oumuumfldm 655 m0 Acoflumfl>wo pumpsmum qusntf n4 . r . . .. t + cmoEv xmumcm .NN ousmflm 88 mm musmwm ZO_._.¢0 924 mudmmDm MO ZOmHM£mZOUII.m Manda .4. - . u. ¢. “ v. 92 Close inspection revealed an intermittent stream that issued from a spring and, after flowing only a few meters, entered a swallow hole. This was the probable source of the surface isopods, which apparently entered the cave as drift by way of the Graveyard Waterfall. Troglobitic amphipods were the third major component of the benthos and composed 16.6% of its biomass. There appeared to be two very similar species that differed in adult size and habitat. The smaller-sized amphipod, Crangonyx gracilis packardii, was restricted to riffle areas, whereas the larger species, probably Crangonyx obliquus,was rarely encountered, except for those found in crayfish traps located in pools. Diptera larvae, planaria and stonefly nymphs were numerically insignificant members of the benthos. The planarian was a rare troglobite; it was probably Sphalloplana weingartneri, described by Kenk (1970a) from nearby Bronson's Cave as the first troglobitic planarian known to occur in Indiana caves. The numerous species of diptera larvae and the stonefly nymph appeared to be troglophiles. Several troglophilic species of caddis larvae and mayfly nymphs occurred in drift samples and in other benthic samples, but were not represented in the 26 samples under consideration; they were apparently regular, but uncommon, members of the benthos. Various nematodes, copepods and ostracods also occurred in the samples, but, being approximately the same size as the apertures of the sieve used in the collection of the samples, were quantitatively under-represented and were not in- cluded in the tabulated data. Quantitative benthic samples from the surface stream, which were taken 30-90 meters downstream of the cave exit, produced dramatically different results. The total biomass was 4860 dry mg/m2 —— a level 64 times greater than that in the cave. Epigean Gammarus amphipods v4 .3 1fiarrhw.’ ..:. .. .. ... — ..1IIV....I.)..\. tdfl‘ w -.. fi- ;_ €~ 93 were the dominant component, comprising 90% of the total biomass. A few hypogean amphipods were found which evidently had been washed out of the cave. As in the cave benthos, isopods were the second most abundant component, but they made up only 3.8% of the total biomass of the surface stream, as compared to 22% in the cave. The isopods were predominantly an epigean form, Lirceus sp., but 16% of the isopod biomass was contributed by Asellus stygius, which had apparently left the cave as drift. These hypogean isopods seemed to be established in certain portions of the stream, especially in those samples taken closest to the cave exit, and they had grown to a size not attained in the cave. The planarian, Phagocata gracilis, was the third most abundant member of the stream community. This species is basically an epigean form, but has often been reported to invade caves, where it is often found with reduced pigment and eyes (Kenk, 1970b). In the present study Phagocata was found to only inhabit the surface stream. The fourth most abundant organism (classified as "other" in Table 9) was a fingernail clam, which was also restricted to the epigean habitat. Oligochaetes were the fifth most common component of the biomass of the surface stream benthos. Their weight density of 49 dry mg/m2 was almost equivalent to the level of cave oligochaetes, which were the dominant component of the cave stream benthos. Other minor members of the benthos were stonefly nymphs and caddis larvae, which also occurred at levels similar to those found in the hypogean habitat. I -\ ‘. .«g a 94 Discussion The cave stream ecosystem is almost totally dependent upon the importation of organic matter. Photosynthetic activity is lacking, and this may be responsible, by causing a lack of utilizable carotenoids, for the lack of pigment in many cave organisms (Maguire, 1961). Also, in a study of Mammoth Cave, Barr and Kuehne (1971) suggested that, while chemosynthetic bacteria could be potentially significant, their quanti— tative contribution is actually relatively small. Autotrophic activity, therefore, is thought to play only a minor role, and food resources in the cave stream are based on allochthonous plant debris. However, even surface streams depend heavily on food importation. Minshall (1967) emphasized the importance of allochthonous detritus in the ecology of a springbrook. Nelson and Scott (1962) found that primary consumers of a Piedmont stream derived 66% of their energy from allochthonous organic matter consisting largely of leaf fragments, and Chapman (1964) found that in Coho salmon only 9% of the known energy sources had an aquatic origin. Hynes (1963) stressed the importance of the terrestrial habitat to the production of streams; he stated that, in this respect, running water differs from other lighted environments and resembles dark areas, like the bottoms of lakes and deeper parts of the sea. Perhaps caves would also fit into his concept. Consequently, from the standpoint of food quality there seems little difference in epigean and hypogean streams, although the living vegetation of epigean streams may play a role disproportionate to their abundance. An important difference between surface and subterranean streams is that allochthonous material has unrestricted access along the entire length of surface streams, whereas in cave streams imported organic ..:. u a .3 ‘4'! [labor a .. _ . . ...: .4 ,P>.05 P>.lO P>.lO C. laevis SOURCE SEASON SEX INTERACTION d f 3 l 3 MS 0.060 0.0088 0.048 F 1.12 0.16 0.90 PROBABILITY OF A LARGER F P>.lO P>.lO P>.lO 114 (P > .10) or season (.10 > P > .05). On the other hand, the ratio was obviously dependent on carapace length, with a much higher ratio characteristic of the smaller crayfish; this was related to both their lessened tendency to enter traps and the decreased efficiency of their visual capture. The variation of the ratio with carapace length was not linear on either arithmetic or logarithmic scales, however, and re- quired the use of a French curve to express the relationship (Figure 23). In making population estimates, the length-specific ratio, as determined from this smooth curve, replaced the values derived from the seasonal and sexual grouping of length-blocked data. In C. laevis, A.O.V. testing of the population/capture ratio, which was blocked into nine carapace length categories (Table 10), showed no significant difference related to either season (P > .10) or sex (P > .10). In addition, the magnitude of the movement rates and their erratic variation with carapace length invalidated any linear or curvilinear relationship to carapace length alone, at least for the larger-sized crayfish (Figure 23). In several of the longer carapace length groups, movement was so extensive that, combined with the large proportion of the population captured, the ratio between population and number captured was less than one. In these cases extensive movement during the census created a large turnover in the individuals comprising the population of the census area; this resulted in the capture of more individuals during a census than existed in the population at any one time. The smallest value for the ratio, which occurred at the carapace length category centered at 40 mm, corresponded with maximum movement as measured by displacement rate and wandering (see Figure 30 below). Although neither sexual nor seasonal variations in the ratio were found to be significant, population estimates were made independently for 115 .mwbmm~.0 you poo poo .mwfihmcw .0 How omnmflanmumo ma magmGOHumHmu HmwcHHH>Hso cmauooam m was .mmmSmcoo Moom co comma mum mcofluma>mo oumocmum ocm mama: .numcoa mommmumo ou ofiumu many mo mflnmcoflumamu map can .msmcmo m mcfluso consummo mamoofl>flpcfl mo gonad: amouom map 0» .wHDuQMUmH can xumfi SQ ompmeflumm mm .COHumaomom mnu mo oHumu one 0‘00 .mm magmas 116 mm madman AEEV IbUZm A m0.0 H 0.0H Hmuoe H.mH 0.m H H.0 m.H H 0.0 m.m H m.m mmHmz mHEHmcH .o 0.0H m.m H 0.m m.m H 0.0 0.m H 0.0 m0H020m 0.0 0.m H m.m 0.0 H 0.0 H.N H 0.0 mmHsHmseH Am H.mv Am H.mc Am H.mv mama mm gases mama masmm> msomo mmmommm mumpammo H mm>m0m mmm mammmmao mmmzsz mooflnmm Hmflo m mo sumo Eoum m .QOHHmm amazon comm cwnufl3 .ocm moowuwm Hudson m mo somm Eoum ma £HH3 .m»w>udm mv co owmmn mumo umoomamz mmma ozm AdomH> mm wm>mom mmm ommoemdu mmHmmmmo mo mmmzoz mme mo zomHmmmEOOII.HH mummy 119 TABLE 12.--TEMPORAL DIFFERENCES IN THE NUMBER OF VISUAL CRAYFISH CAPTURES PER SURVEY. Data based on 45 surveys, with 15 from each of 3 annual census periods and, within , 5 from each of 3 diel periods each census TIME PERIOD DIEL ANNUAL SPECIES 9:00 A.M. 3:00 P.M. 9:00 P.M. July '69 Nov. '69 Mar. '70 O. inermis l3.7t3.9 l7.2t5.0 13.3t4.4 l6.7t4.4 12.712.6 l4.8:5.9 C. laevis 4.6:2.0 5.7:2.9 4.5:2.3 6.7:2.4 4.1tl.4 4.0:2.3 ANALYSIS OF VARIANCE TABLE SIGNIFICANCE HS 8 highly Sign. SOURCE SPECIES d f SS MS F P s = sign. ns = not sign. DIEL O. inermis 2 138.53 69.27 3.90 .05 >P S C. laevis 2 12.98 6.49 1.37 P>.10 ins ANNUAL O. inermis 2 116.13 58.07 3.27 P=.05 ns C. laevis 2 69.38 34.69 7.33 .005>P HS INTERACTION O. inermis 4 74.94 18.74 1.06 P>.10 ns C. laevis 4 2.88 0.72 0.15 P>.10 ns ERROR 0. inermis 36 639.20 17.76 C. laevis 36 170.40 4.73 TOTAL 0. inermis 44 .968.80 C. laevis 44 255.64 120 On an annual basis, there were found to be no significant dif- ferences in the capture rates for O. inermis. The capture rates for C. laevis, on the other hand, were significantly different at the 0.5% level; the rate was higher in July than in the other two, nearly equivalent, census periods. Mark—recapture analysis (see page 112) suggested that the capture rate probably reflected, to a large extent, the population level; this indicated seasonal stability in the population of O. inermis and seasonal fluctuation in the population of C. laevis. Annual variation in catchability due to behavioral modification has not been demonstrated. Diel variation of the capture rate was not significant in C. laevis, but was significant at the 5% level in O. inermis. The capture rate was higher for surveys conducted in the afternoon, and the variation probably resulted from a diel activity pattern. This topic will be more fully discussed in the section on activity. Length-weight relationships In all cases linear correlation was found between the log of length and the log of weight, and the regression formulae for these relationships were determined (Table 13). For any given length, the weights of both body and exuvia were less for O. inermis than for C. laevis. The skin/body weight relationship was also less in O. inermis, so that exuvia weight loss per molt was greater in C. laevis. The ratio between wet and dry weights was not constant; increased weight loss on drying occurred in O. inermis and with decreasing length in both species. This variation apparently resulted from species-specific and size-specific differences in the degree of development of the 121 0 mum.0 Hm.m . x 00H mom.m m 00H mHsmmH .0 sumcmH mommmHmu n x o mom.o ma.m I x moH mom.m w moH meHmcw .o cflxm pmuaoa mo usmflm3 ago u w 00 mmm.0 00.m u x 00H 0m0.m m 00H mHsmmH .6 numcwH monHHmo u x mm m0m.0 00.m : x 00H 000.m m 00H mHEH02H .o scon mo HamHmz mum u m Hm nmm.o mm.m I x mod me.m w moH mHSmmN .O cumcma oommmumo n x 00 «00.0 m0.v I x 00H 00m.m m 00H mHEHmcH .0 moon mo uanmz H03 u m ANHV monB onedszmmBmo SE 2H panama wommmumo u x Ax so my nm>mmmmo mo memmm 0H uanmz u m mmmommm mmmmammm> mammmmomm ezmHommmmoo «anamom onmmmmomm mZHMm QMBQOZ QZQ mmHQOm mmHmwfimU mom mZOHmmmmwmm B$UHW3IZBQZMAII.MH Edgy. .laf’v. . 122 exoskeleton. That is, the exoskeleton, which contained relatively little water, constituted a greater proportion of the total body weight in C. laevis and in larger crayfish, in general. Length-weight regressions indicated that, in crayfish of equal carapace length, 0. inermis had a lighter body than C. laevis. This weight reduction, which is typical of troglobites, probably developed in response to energy conservation, although Vandel (1965) regards the slender body of many cavernicoles to be the result of orthogenesis of a phyletic line. The reduced weight of the exoskeleton in 0. inermis also functions in conservation of energy, and it apparently evolved upon the relaxed need of the exoskeleton's protective function in the cave environment. In this regard, exposure to both solar radiation and predation was drastically reduced as a result of the invasion of the cave habitat by the epigean ancestor of O. inermis. Subsequent to the evolu- tionary development of its troglobitic form, 0. inermis was joined by the troglophilic crayfish, C. laevis. Its larger size, robustness, heavily armored exoskeleton with well-developed chelae, and aggressive behavior posed a predatory threat with which 0. inermis was poorly equipped to cope, at least in an offensive sense. The low mortality rates, which were characteristic of the O. inermis population (see Table 27 below), indicated that the troglobite had successfully dealt with the problem by other means. Activity Introduction Most epigean crayfish are regarded as being nocturnal. Penn (1943) found C. clarki to be most active at night, responding positively to 123 twilight conditions. The locomotor activity of O. virilis was inves- tigated by Guyselman (1957), who found maximal activity between 7:00 and 10:00 P.M. Park, Roberts and Harris (1941) studied the troglobitic crayfish, O. pellucidus, under laboratory conditions and concluded that it had an arhythmic diurnal activity pattern. However, Brown (1961), analyzing the same data, found a statistically significant circadian rhythm with minimum activity about 9:00 A.M. and maximum activity about 7:00 P.M. His conclusions, however, were based on a questionable procedure and on data from only one crayfish. Jegla and Poulson (1968) found evidence for a circadian rhythm in O. pellucidus and O. inermis, but emphasized its variable periodicity in different individuals. To shed further light on this situation, crayfish surveys were conducted at different times during the day. Although trap captures were not relevant because of uncertainty of the time of entry, visual capture rates would seem to indicate activity patterns under natural conditions. Surveys were conducted in the morning (9:00 A.M.), after- noon (3:00 P.M.) and evening (9:00 P.M.). The three time periods are indicated by a time that fell within the surveys, although it must be understood that the surveys required a variable length of time of up to four hours. Besides diel activity patterns, also of interest was the longitudinal displacement of crayfish that resulted from locomotive activity. A few studies have undertaken to determine movement of stream- dwelling crayfish in terms of displacement rates and home range. Most of these studies were at the population level, employing mass tagging or weir trapping. A few crayfish studies, notably the cave studies by Cooper (1975) and Hobbs (1973), have had success with a tagging method that permits the recognition of several individuals. In the present study the crayfish were individually tagged by a different method that 124 permitted monitoring the movements of several hundred individual cray- fish. The longitudinal position of a particular crayfish in the census area was noted at each capture and recapture during the four censuses. This procedure was aided by placement of location markings on the cave wall at every five meters. Results Circadian rhythm As mentioned in the section on temporal differences of capture rates, analysis of variance indicated a significantly different (.050 > P > .025) diurnal rate of capture for O. inermis, although the highest rate at 3:00 P.M., instead of 9:00 P.M., does not correspond with the findings of Brown (1961). C. laevis also showed the highest rate of capture at 3:00 P.M., but it was not significantly different from the other two time periods (Table 12). Longitudinal movement Patterns in individuals Troglobite: The pattern of movement in O. inermis was fairly simple (see Figure 24 below, which displays movement during the July census when the crayfish were especially active). Basically, this species ex- hibited the phenomenon of home range. Most recaptures occurred within ten meters of the initial capture site, indicating the equivalency of home range, in its strictest sense, with the local pool. In many cases, however, especially in the males of the larger size categories, I Hum Figure 24. 125 Movement patterns of all 0. inermis captured at least three times during the July census. Data from 56 individuals are shown, and the connecting lines in the capture sequences indicate the existence and extent of wandering. 126 M.AI.EEB FEIMIILEES (mun) MOVEMENT on.” v .030 '°° '0. 50. '(m 0 O Q. .om Ida I00 .o. 2 0000 9 A r: T (7) .000° W S Ibo la) {Hflh>|%9 «nun. v #8“ . O. ‘ Q <" N 0 ‘ .1. a d 58:4 ‘13 .m 9090 130 f N Z'-w no 2 .g . LU > E < a .l 4 m w I c Eggg Ix) Inc 15“” lunflfl’ Iflmflh II) gang Ifidflb I90 v m 2 g S . 3.8 g m 6 Inc In: Enos layman .mflwfi ooooo 'u’lflb uagnfi “IDIKFO I O 6 . INITIAL CAPTURE SITE 0 RECAPTURE CAPTURE Figure 24 SITE SEQUENCE FOR INDIVIDUAL CARAPACE LENGTH CATEGORIES ~ 1 am. k ;‘ ~ 127 crayfish wandered up to 60 meters from the home pool and returned. This extended area should be included as part of the home range, since analysis of movement over a year's time indicated these wanderings to be habitual; some male 0. inermis normally traveled an area of up to 300 meters (Figure 25). There were a few instances in which crayfish underwent a net displacement during a census; at least some of these could be explained by the home range concept as cases in which the crayfish had not yet completed the return trip to its home pool. Possibly two or three examples of annual net displacement were observed (Figure 25), but these exceptions merely accentuated the positional stability and home range behavior of the majority. The complex data on crayfish movement was condensed in order to permit further analysis. This was accomplished by dividing net dis- placement of a crayfish during a census by the elapsed time and ex- pressing movement as a rate in meters per day (Table 14). The rate of movement was related to size with the highest rate occurring in the 17.0 - 20.4 mm size group, although larger adults in the next two size categories also showed a high rate (Figure 30a). There were also sexual and seasonal variations in movement. Males had a greater move- ment rate in all size categories (Figure 26a), and both March censuses had a generally lower rate of movement than the July and November censuses (Figure 26b). Movement rates ranged from a low of 0 meters per day for females over 26.5 mm C L during the March 1970 census, to a high of 6.1 meters per day for 17.0 - 20.4 mm C L males during the November census; the mean rate was 2.3 meters per day. The percentage of crayfish with a net displacement of more than 10 meters during a census was also determined. This, in effect, gave an estimate of that fraction of the population which had entered a 128 Figure 25. Annual movement of five randomly selected individual 0. inermis in different sex and size categories. 31313:. .85. 129 MALES FEMALES I . a9 = i5"o r T .&§; L? 5 loan m... 1 an m u. i 3 T M: 0" ° 0' T a ‘2‘ g '°‘ 2 3 $59 3 I 5% S !-0 3 E 3 5: I a :§1 5% .‘O'ga .°5'l-5 I'é .‘G .5 2% I-"A'fi .Oo..-OQ I'g’ .é '. .A j. :6 g I: : ‘ ' ‘9 ° 3 t “ E I 3 . B 3. "I .- 2°? 0 p. g 4 E :: ". I 55 I 0 gal; "Q'A'o ‘9.8 ....LD ..T.?. ..:: ! ...... D .-0'A'D j'?"o.c ..Q .0 choafl : N I- T ; E Q h. w 4’ 5a " S 3 E ‘ 8 l EL '35. Z l-g‘ 1 2 EV ‘$‘ 3 .29 a. ,0 : 0' a ..uuoO ...A'é .Q~£.6 ..00'. ..Q'é ., ..... a i’O"'-Q {'0"'G ..o-éfl if f g . ID MAR. '69 JUL. '69 NOV. '69 NAR'TO CENSUS ODO- INITIAL CAPTURE SITE AND RANGE OF INDIVIDUAL FOR EACH CENSUS OBSERVED 13;? Figure 25 (mm) CABAPACE LENGTH CATEGORIES A 'U'" Figure 26. 130 Daily displacement in O. inermis. (m/day) DISPLACEMENT NET 131 4. ’0 0——0 MALES // \ emu-o FEMALES 0......0’ \ A COMBINED 10 \ 4 [$3. \ /;’ a \ I; \ I :' \ / :' ‘-. \ 2“ o/ a: ... \ :' ". .0. \ :0. .... ...o. . ... \\ I ’0 O. .0 .0. ... I d ". 0" —- —o’ q .00.. .00..... ...OOIoooooouoo AI 0‘ u 1 1 I I I I I 6* CENSUS cocoon-cu. MAR'BQ -'—' JUL.’69 -4 _ NOV.'69 MAR.70 4‘ O 2‘ O——-h‘ 0 y j v 1 I 011 1-2 2-3 3-4 4-5 5-6 6-7 >7 AGE GROUPS Figure 26 (yrs) w... 5. ..x n . o: .3 in: w; s???‘ .../:2 ...... 132 TABLE 14.--LONGI'1‘UDINAL MOVEMENTS OF 0. inermis INDIVIDUALS Based on not displacement of individual between first and last sighting during_census STAT CARAPACE LENGTH CATEGORIES (mm) census OR on sex GROUP ANAL 5.0- 12.0- 17.0- 20.5 22.6 24.5 25.6 over . 11.9 16.9 20.4 22.5- 24.4 25.5 26.5 26.5 § 0.7 0.3 1.8 1.1 0.6 0.9 1.5 - x s 0.6 0.3 3.5 1.5 0.5 0.9 1.9 -- g g n 4 12 19 17 12 8 4 o §3 31 0 o 36.8 29.4 50.0 50.0 50.0 - 32 -- -- 71.4 40.0 66.7 75.0 100.0 ~- § 1.0 1.8 4.1 5.5 4.6 1.8 0.6 0.2 s 1.4 4.1 10.3 20.0 6.0 1.8 0.6 0.1 :5 g n 2 26 31 19 6 7 3 2 g g; 31 o 15.4 29.0 36.8 50.0 42.9 33.3 0 32 -- 100.0 55.6 28.6 33.3 33.3 100.0 .- m I 0.5 0.4 5.6 1.1 4.1 0.9 1.2 2.6 a s 0.8 0.5 13.6 2.4 4 9 0.7 1.6 2.5 f, >- 3 n 5 14 26 14 15 6 2 3 U gs; 31 20.0 7.1 38.5 28.6 60 0 33.3 0 100.0 32 100.0 100.0 60.0 75 0 55 6 100.0 -- 33.3 E 0.1 2.8 0.9 1.8 1.4 0.6 0.5 0.1 so 5 0.1 7.9 1.7 3.5 3.1 0.6 0.8 0.2 g; n 2 23 27 17 11 6 3 2 ... 31 0 21.7 40.7 58.8 27 3 50.0 33.3 o 22 -- 80 0 27 3 40.0 66 7 66.7 100.0 -- Y - 2.0 3 3 3.4 3 8 1 3 1.2 1.6 5 ~ 6.6 10 3 13.3 5 2 0 8 1.5 2.6 m n - 33 54 44 15 10 6 4 53 31 9.1 31.5 34 1 66.7 60 0 50.0 50.0 g 32 - 100.0 52 9 40 0 60.0 66 7 100.0 0 f: _ "I 2: -~ 1.3 3.1 0.9 1.9 1.0 0.7 0.6 w s - 3.3 7.8 0.9 3.4 1.3 1.0 0.6 n 42 49 23 29 17 6 3 21 - 16.7 40.8 47.8 37.9 35 3 16 7 33.3 h- 32 -- 85.7 50.0 45.5 54.5 66 7 100 0 100.0 U, § 0.6 1.6 3.2 2.5 2.5 1.1 1.0 1.2 g s 0.8 5.0 9.1 10.8 4.1 1.2 1.3 1.9 2 n 13 75 103 67 44 27 12 7 0 21 7.7 13.3 35.9 38.8 47.7 44.4 33.3 42.9 a 32 100.0 90.0 51.4 42.3 57.1 66.7 100.0 33.3 'STATISTIC OR ANALYSIS ; = mean (m/day) s = standard deviation (m/day) n = number of individuals observed 21 a t moving more than 10 meters (& wandering) *2 = x of wandering individuals that were moving downstream 133 wandering phase and left their home pools. The overall wandering frequency was 32.8%. The wandering fraction was size dependent and peaked at 47.7% in 22.6 - 24.4 mm C L O. inermis (Figure 30b). The fraction of the population in a wandering phase did not show much season- al variation. Likewise, there was no overall sexual difference, al- though a higher percentage of females wandered in the smaller size categories, whereas in crayfish larger than 22.5 mm C L, males wandered more often. Of those crayfish moving more than 10 meters, 57% were observed to move downstream. If the home range phenomenon completely controlled all 0. inermis movement, then the number of crayfish moving upstream should have balanced those moving downstream. Chi—square testing indicated that this slight discrepancy was not significantly different from the lack of populational net displacement expected. Troglophile: Positional changes occurring in the C. laevis population during the July census showed that movement was extensive and complex (Figure 27). In some individuals movement was less than 10 meters and was apparently confined to a single pool. In other instances, crayfish wandered up to 100 meters and returned to a home pool. Other crayfish underwent a net displacement during the limited observational period of the census. These movements superficially resembled either migration or kinesis, but observation over the course of a year (Figure 28) provided contrary evidence. These movements, in most cases, were probably fragments of home range behavior, in which the home range was extremely large and the completion of a circuit temporally protracted. The sizes of the home ranges varied greatly among individual C. laevis (Figure 28). «u ' ~54 1 3v." Figure 27. 134 Movement patterns of all C. laevis captured at least three times during the July census. Data from 22 individuals are shown, and the connecting lines in the capture sequences indicate the existence and extent of wandering. captured at least Data £10022 necting lines 11 existence and INITIAL CAPTURE SITE MOVEMENT IN METERS RELATIVE TO 490 “POL o A UPSTREAM DOWNSTREAM 135 8312s ’ FEMALES O 0 g; m 00 00000 c A Q (9 O N O 0 00000 O 9 I000 600..., J 5'25 m .6 A” III l I INITIAL CAPTURE SITE O RECAPTURE SITE V CAPTURE SEQUENCE FOR INDIVIDUAL Figure 27 (mm) CARAPACE LENGTH CATEGORIES fun..- ”1'. ‘ur. —.-.._-.- ‘4‘.- Figure 28. 136 Annual movement of five randomly selected individual C. laevis in different sex and size categories. 137 AEE. IPUZNJ w0.h<.._wc mcmhms Z. h2w3w>05 INITIAL CAPTURE SITE AND RANGE OF We INDIVIDUAL FOR EACH CENSUS OBSERVED } CENSUS I MAR. '69 0 JUL. '69 A NOV. ‘69 I3 MAR. '70 Figure 28 138 They seemed most extensive in the larger sized males, but the fact that in some cases the range approached the length of the 500 meter census area throws some doubt on their validity as examples of home range behavior. It is possible that in some instances the home range was underestimated, because it was of greater size than the census area. This possibility was supported by the fact that many C. laevis, pre- sumably possessing strong locomotive behavior, were encountered in only one census during the year of study. It is also possible that these individuals were undergoing migration or kinesis. Crayfish displaying kinesis behavior in a restricted habitat might have appeared to have a home range, and the extent of available cave stream habitat upstream of the sinkhole entrance is unknown. Home range was certainly operating in the majority of the individuals, but it is uncertain whether it was a universally valid concept for all segments of the population. Analysis of movement in terms of displacement rate (Table 15) indicated that movement was quite extensive, especially in the middle size ranges from 25.5 to 44.4 mm C L (Figure 30a). Males almost con- sistently had a greater rate than females (Figure 29a), and there was also seasonal variation with movement depressed during the March censuses (Figure 29b). Movement rates ranged from a low of 0 meters per day for the 5.0 - 12.9 mm C L size category during the March, 1969, census, to a high of 65.1 meters per day for 25.5 - 30.4 mm C L males during July. The mean rate of displacement was 7.6 meters per day. The percentage of C. laevis with a net displacement of more than 10 meters during a census, which in effect measures the magnitude of crayfish in a wandering phase, increased with carapace length, in general, and reached a high of 80.0% in the 38.5 - 41.5 mm C L class (Figure 30b). There was no overall sexual difference, and seasonal variations seemed to be random. The mean wandering frequency was 49.7%. .....A Figure 29. 139 Daily displacement in C. laevis. (m/day) DISPLACEMENT NET 50< 140 °--0 MALES 9 """ 0 FEMALES /\ A COMBINED 10. / \ I .- \ \ ,. .. ’ ° ’ 3.. / 0_,.......0.. \\ I ‘0... 5‘ I .. .... 0 0.. \ ‘0 0 o 0.. .. .. / .... “O. 0\ ‘00 x ‘ / ..o ‘5‘ 0/ 0 .9 ... 14 :0. .0 A 0.. 0.2 1 T f I T I I 1 1 i m o loos-cs... MAR, .69 I — I — 0 JUL . 969 ml NOV. ’69 " - - MAR. ’70 0 0" I \ d I .\ I '\ I ‘\ 204 I .\ o /\ I - . I \ .I '\ .I \ .0 .... J o o .. a $5 b\. ,l ’3. .... ’0’ \ . ’Qt.‘ A 0.. so 0 -. ""0."”"'O. . ’ .0 o On... ..... O‘I k00:0.§00¢"'~§’. ‘0 ~ - — —G/ gfl 0:1 1-2 2-3 3-4 4.5 5_6 6-7 7-8 ,8 AGE GROUPS ("3) Figure 29 ~-'~ _' U'E' {spvv‘ Figure 30. 141 Inter-specific comparison of net displacements occurring during a census. 142 \A \ \\ \‘I‘ u ‘IIII‘I‘ m 6H\\ 0 III, 0 III m III, u III " \ \ ... x \A 6 \\ S A, mu mm l/l e e / .. mm M6 . . .../ O.C_ ... _ 6/2 u _ ... 00 ...a m _ ... _ 8 .... . . 23>... hzm5m0<4uma hmz .A .8 I > a a 8 M 9 fl \ O \\ x . 7 Al ...o . cw / 0 II III. as .O A / .0 a. [I u III/m MIA 5 ‘ 4 \\ a \\ a \\\\ av 4 A I 6 .... II a l u l a II u If s I n1 OOOOOOA / I v 2 O... I 000 I 000 I to. I :0 [A .. J. a \\ 1 0| \ N \\» . e... A 0 o m . - ol . 0 8 4 mmmhm—z or 205 5.sz mm...— (yrs) AGE GROUPS Figure 30 143 TABLE 15.--LONGITUDINAL MOVEMENTS OF C. laevis INDIVIDUALS Based on net displacement of individual between first and last sighting during census STAT OR CARAPACE LENGTH CATEGORIES (mm) 088508 on 55x GROUP ATAL 5.0— 13.0— 19.5 25.5— 30.5- 34.6- 38.5- 41.6- over ALL 12.9 19.4 25.4 30.4 34.5 38.4 41.5 44.4 44.4 LENGTHS x 0 0.1 0.6 3.3 1.2 3.7 3.1 13.7 1.4 2.8 s -- 0.2 0.5 4.3 2.0 5.2 3.6 18.5 0.3 5.7 5 m n 1 7 8 12 4 9 4 3 3 51 m g :1 o o 37.5 66.7 25.0 77.8 75.0 100.0 100.0 54.9 2 a :2 -- -- 33.3 37.5 o 42.9 6.7 0 33 3 35.7 x 0.7 0 5 8.4 37.8 2 4.6 20.9 0.4 -- 14.2 s 0.7 0 4 19 7 75.6 16 2 7.0 21 0.5 -- 39.0 S 3 n 3 3 9 9 6 3 4 3 0 40 g g 31 33.3 0 44 66.7 50 0 33.3 100.0 0 -- 47.5 22 100.0 -- 75.0 33.3 33 3 0 75.0 -- -- 52.6 g _ g x -~ 1.0 1.2 0.7 18.0 9.8 46.8 2.5 1.7 13.2 8 s -- 1.4 1.9 0 6 28.0 4.6 117.5 1.5 1.5 56.2 5 3 n 0 2 6 9 3 2 8 3 3 36 g g 31 -- 0 16.7 44.4 66.7 100.0 75.0 100.0 66.7 55.6 32 -- -- 100 0 75.0 50.0 50.0 50.0 66.7 100.0 65 0 x 0.3 3.3 0.4 1.2 0.7 0.6 5.8 - 0.6 1 6 s -- 6.1 0.5 1.3 0.6 0.4 9.6 - 0.7 3 8 g g n 1 4 8 10 3 2 4 o 4 36 E g 31 o 0 12.5 70.0 33.3 0 75.0 -- 50.0 38.9 32 -- -- o 85.7 0 -- 66.7 -- 100 0 71.4 x -- 1.6 3.9 17.2 7.4 4.9 33.4 3.75 1.7 11.2 s -- 3.9 13.7 57.7 17 4 5 8 95.9 0.8 -- 46.1 m n -- 10 19 16 8 10 12 3 1 79 g 31 -- o 31.6 68.8 37.5 60.0 83.3 100.0 100.0 50.6 32 -- -- 50.0 45.5 33.3 33.3 70.0 33.3 0 47.5 E m T -- 0.2 1.5 5.2 6.3 3.2 11.5 6.4 1.1 4.5 m s -- 0.3 2.3 12.6 14.1 4.7 18.0 12.9 1-0 10.9 5 n -~ 6 12 24 8 6 8 6 9 79 g :1 -- 0 25.0 58.3 50.0 66.7 75.0 50.0 66.7 50.6 h 22 - - 66.7 64.3 25.0 50.0 50.0 33.3 83.3 57.5 m E 0.5 1.1 3.0 10.0 6.9 4.3 2427 5.5 1.2 7.6 g s 0.6 3.1 10.8 37.5 15.3 5.3 74.6 11.2 1.0 33.1 g n 5 16 31 40 16 16 20 9 10 163 0 31 20.0 0 29.0 62.5 43.8 62.5 80.0 66.7 70.0 49.7 3 32 100.0 -- 55.6 56.0 28.6 40.0 62.5 33.3 71.4 53.1 < ‘STATISTIC 0R ANALYSIS Y = mean (m/day) s - standard dev iation (m/day) n - er of individuals observed 31 - 0 moving more than 10 meters (\ wandering) t2 I t of wandering individuals that were moving downstream 144 During the March, 1969, census, 64% of the wandering crayfish were moving upstream. By the July census there was no obvious trend, with 53% moving downstream. Both the November, 1969, and March, 1970, censuses showed downstream displacement, with rates of 65 and 71%, respectively. Males showed a slight net tendency to move upstream (53%), whereas the females generally moved downstream (57%); the mean displacement was 53% downstream. These directional movements were tested by Chi-square, and none were found to be significantly different from random behavior. There was some evidence for mixing of the epigean and hypogean C. laevis populations. A 39.5 mm C L female that had been last observed in the cave at 34 meters was captured in the surface stream 14 months later. However, eighteen epigean crayfish that were tagged never ap- peared in cave surveys. Effect at pOpulation level Some additional light was shed on movement by analyzing the gradual decrease of marked crayfish in the census area populations. This was accomplished by monitoring for a year the percentages of cray- fish in the various age classes that were first captured and coded during the first census. The initial ratio was determined by applying the Schumacher mark-recapture procedure to the first census data; the resulting estimated population was divided into the number of individuals marked. The final ratio was derived directly from the percentage of first census-marked crayfish in the crayfish captured during the fourth census, which was conducted a year later. These procedures were per- formed independently for different carapace length cohorts, which were followed throughout the year by application of measured growth rates. 145 Although crayfish so marked underwent an absolute decrease during the year, this resulted not only from migration, but also from mortality. On the other hand, the annual decrease of marked crayfish relative to the total number comprising a year class cohort resulted only from mi- gration. There was no evidence that differential mortality of marked individuals or loss of marks were factors. The decrease in the ratio resulted both from emigration of marked individuals and from immigration of unmarked individuals. The percentage decrease in the ratio was a measure Of migration across the census area boundaries. It was, in fact, equal to both emigration and immigration, if these two rates were equal. If there was net immigration, then the percentage decrease would be greater than the actual emigration rate of marked crayfish; if, on the other hand, there was net emigration, then the percentage decrease would be less than the percentage of marked crayfish leaving the census area. In 0. inermis, crayfish of small (5.0 — 16.9 mm) and medium (17.0 - 24.4 mm) carapace lengths had similar annual migration rates of 50.4% and 51.5%, respectively. Larger crayfish had a lower rate of 37.5%. Migration rates were much greater in C. laevis. Crayfish with a carapace length of 25.5 mm and larger migrated at a rate of 68.4%. Smaller crayfish migrated at the even higher rate of 83.1%. Regression analysis (Figure 31) indicated that, for both species, the migration rate, as measured by the percentage decrease in the ratio of marked crayfish in the population, was linearly related to the log of the displacement rate. This relationship, however, resulted in a migration rate that was less than doubled when the displacement rate was increased ten-fold. This fact reflected the home range be- havior Observed in the movement study of individual crayfish. 146 .uwumH “00> m comm msmcwo map mcfluflnmncfl COHDMHDQOQ may ca mamzofl>flocfl meumelww>usm uma mo muuocoo CH mmmmuooo mmmucmoumm on“ so Comma mmoq .COHDMHSQOQ mmum msmcmo wnu Eoum Cmflw>muo omxumg wo mmOH Hmsccm 0CD com 0000“ uswEm>OE HMCACDDHOCOH cwaumn mflnmcoaumamm .Hm mudmflm 147 Hm 0856H6 8 26.6}... #2023535 F P _ IfiHW96~A a 2.56:. d 6.3-0.: 4 6.9-6.6 O l- .0 23m. .0. v.mNA ¢.mN ..O.m AEEVJU HSHAVHO aaxuvw :IO Olivu NI asvauoaa % 1vnNNv [A] 148 The migration rates represent an approximation Of marked crayfish emigration from the census area. The two values would be equivalent only if there was a lack Of net population change across the boundaries. Whether or not emigration was balanced by immigration can best be judged from census data on temporal changes in the size of the crayfish populations (see section on population size and structure below). Popu- lation sizes and structures were, in general, temporally stable. Con- sequently, in most cases migration between the census area and other reaches of the stream probably consisted most often of equivalent two- way flow rates between areas of similar population densities. A notable exception was an apparent net immigration Of smaller- sized C. laevis. This conclusion, based on population changes that are discussed in more detail below, gained a measure of support from in- formation derived from displacement and migration rates. The population group in question consisted of C. laevis with carapace lengths ranging from 5.0 to 25.4 mm. This group had the highest Observed migration rate. Figure 31 illustrates that, unlike the other population groups, its apparent migration rate was greater than the regression line. Thus, migration, which was based on loss of marked crayfish, seemed to be higher than would have been expected from the displacement rate. The actual rate of migration for this group may have been less than that indicated by the measuring procedure employed. This possibility was due to the over-estimation of the migration rate that would have resulted if immigration was greater than emigration. Accordingly, the net immigration for this group, that was suggested by the Schumacher- estimated population data, would have been expected to result in the situation illustrated in Figure 31. Likewise, Schumacher-estimated data indicated that emigration of larger C. laevis may have occurred, 149 and this conclusion was consistent with the position of this group below the regression line of Figure 31. Discussion Diel capture rates indicated that a circadian rhythm of activity may have existed at the population level, but, if so, only to a slight extent. The peak Of activity at 3:00 P.M. may have resulted from an exogenous cue, such as entrainment by light in the twilight zone. On the other hand, differences in the pattern of investigative interference preceding the three time periods may have accounted for the differences in capture rates. Emphasis should rather be focused on the nearly equivalent activity in different segments of the diurnal cycle. Endogenous rhythms of epigean crayfish are cued by Zeitgebers that generally produce a nocturnal activity pattern; this probably functions to reduce predation. In hypogean crayfish an endogenous rhythm probably lacks entraining Zeitgebers, and this would result in a free-running activity pattern not strictly phased to the diurnal cycle. Although the individual would retain a circadian activity pattern, there would not be a synchronized pattern at the population level. This behavior pattern would have the functional advantage of temporal partitioning Of space and food resources. Circadian rhythms at the population level would have no apparent function in the cave stream habitat, unless inter- specific crayfish predation is a significant ecological interaction. The fundamental pattern of longitudinal movement for both species supported the operation of the home range phenomenon, although it was Of a peculiar character. These crayfish resided in a home pool, usually of about 10 meters in extent, in which movement was restricted for extended periods of time; this may be regarded as the primary home range. 150 The crayfish periodically wandered from the home pool and habitually traveled a secondary home range. The extent of this secondary home range showed both inter—specific, and intra-specific variations. Intra— specifically, the home range was larger in males and young adults in both species, and inter-specifically, C. laevis possessed much larger home ranges than 0. inermis. Both species also exhibited seasonal variations, with a decreased rate of travel during the March censuses. Although several dams and waterfalls occur in the cave, they apparently are not significant barriers to crayfish movement. C. laevis individuals were frequently Observed climbing the cave wall to by-pass the weir, and several 0. inermis were found that had fallen into cavi- ties while traversing the top of the weir. Similar cases Of cave cray- fish traveling overland were reported by Cooper (1975) and Hobbs (1973). The high humidity of the cave atmosphere is obviously conducive to this type of behavior. The proportion of time spent wandering the secondary home range varied with size in both species, being much lower in the younger cray- fish; no sexual differences were observed. There was a large difference in wandering frequency between the two species, with a mean value of 33% for 0. inermis and 50% for C. laevis. The greater wandering rate of C. laevis was observationally obvious. Most 0. inermis inhabited pools, and only a very few were found moving through riffle areas. C. laevis crayfish, on the other hand, were frequently encountered moving through riffles, although they also most Often occurred in pools. Trophic studies (see section on diet analysis below) have shown that C. laevis had a higher percentage of animals in its diet than did 0. inermis, and C. laevis, therefore, would be expected to be more frequently encountered in the riffles, which supported a richer 151 benthos than did the pools. In any event, wandering was so prevalent in C. laevis that the concept Of a home pool or primary home range was not as well established as in O. inermis. Nevertheless, the fundamental pattern for both species was a home pool from which the crayfish made periodic forays, the extent of which was partially sex and size dependent. Other studies on the occurrence of home range in crayfish have produced varying results. The pond-inhabiting Orconectes virilis was studied by Camougis and Hichar (1959), who concluded that it did not display a home range. Likewise, Mobberly and Pfrimmer (1967) reported a lack of home range in a population of ditch-dwelling Orconectes clgpeatus. In both studies, however, the investigators were testing for the presence of a highly confined home range, with 40 and 15 meter limits, respectively. In addition, both of these studied habitats lacked a complex structure, such as the riffle-pool sequence character- istic of the cave stream. In contrast, Black (1963) studied two stream—inhabiting species of Procambarus and concluded that they ex- hibited a home range with limits less than 30 meters. A mean home range of 21.5 meters has been reported by Hobbs (1973) for O. inermis Of Pless Cave. Although the home range concept adequately described the behavior of the majority Of the crayfish, there was also good reason to believe that other locomotive phenomena may have been operating, especially in C. laevis. The apparently huge home ranges of certain individuals may actually have been kinesis behavior confined to the limited cave habitat. Uncertainty exists because of a lack of knowledge of crayfish movement beyond the census area. The gradual loss of marked crayfish from the census area indicated that a degree of migration was operating in both species, especially C. laevis. Apparently, home range behavior patterns 152 were applicable to individuals on a short-term basis, but the home ranges were not spatially immutable throughout their life spans. Hobbs (1973) has made similar observations of O. inermis in nearby Pless Cave. Some short-term movements Of cave-inhabiting C. laevis, at least at the level of the individual, appeared to be unidirectional and directed. Several were observed to travel more than 100 meters per day, with the greatest being 337 meters per day. Of those C. laevis moving more than 10 meters during a census, the average displacement was 11.9 meters upstream. The majority of these same crayfish, however, were moving in a downstream direction, so it was apparent that no simple migratory mechanism was operating on the population as a unit. Migration at the population level has been noted in other crayfish. Semiannual migration of Astacus klamathensis has been observed between a river and a creek (Henry, 1951). An upstream migration of Orconectes nais was regarded by Momot (1966) to be a means Of stream repopulation. Migration between epigean and hypogean habitats has also been shown. The only direct evidence was the recovery in the surface stream of a single C. laevis individual that had been previously Observed in the cave. However, indirect evidence, derived from the peculiar population structure of the cave population (see Figure 47 below), indicated that certain members of the surface population were migrating into the cave. The nocturnal nature of most crayfish preadapts them to the cave habitat. Chidester (1908) observed Cambarus bartonii to be negatively phototactic to strong light; this behavior might influence them to enter caves, and negative photokinesis might lead to their incorporation into the cave population. Hobbs and Barr (1960) pointed out that cambarids characteristically explore the sources of springs, 153 and Weingartner (1962) observed Cambarus laevis moving into a spring of a nearby cave. Other factors might also have been operating to initiate movement into the cave. The density of the epigean population was quite high; it was much greater than the cavernicolous population Of C. laevis (see section on population size and structure below). Bovbjerg (1964) stated that, as a result of intra-specific aggression, members of a population of Procambarus alleni disperse at rates directly related to density. It was also reported that the larger and older individuals of this species are more dominant (Bovbjerg, 1956). This behavior, if applicable to C. laevis, might explain the peculiar structure of the cave population, which was inflated by those size categories cor- responding tO young-Of-the-year and year-old epigean C. laevis. Flooding has been suggested as an important agent in controlling the movement of crayfish. Wickliff (1940) reported that crayfish rapidly repopulate riffles denuded by floods. In the present study, although floods may have displaced riffle—inhabiting crayfish, they probably had a negligible long-distance displacement effect; no mass downstream displacement associated with flooding was observed. Black (1963) stated that floods did not appear to disperse crayfish. The functional aspects of crayfish movement within the cave are uncertain. Movement might have been related to sexual activity. Move- ment was quite extensive in both the July and November censuses, and, based on the fact that form I males of both species were proportionally most abundant during the November census (see section on population size and structure below), mating apparently occurred during the fall. In addition, the greatest movement occurred in young adult crayfish, although there was no abrupt change in locomotive behavior with attain- ment of sexual activity. It can probably be more logically assumed 154 that movement was primarily a part of feeding behavior. There was a low rate of movement for both species during the March censuses. At this time of year food was probably quite plentiful, with relatively large quantities of detritus and riffle-inhabiting benthos having been washed into the pools by the frequent winter floods. Evidence for increased drift of benthos and increased allochthonous input during this period of the year was gained from drift net samples (see page 70 above). After flooding subsided in late spring, this reservoir of food in the pools was gradually depleted, and crayfish probably began moving into adjacent riffle areas in search of food. This increased activity was evident in the July and November censuses, which were conducted during, and preceded by, periods of stable, low stream flow. The greater movement and more frequent wandering into riffle areas that characterized C. laevis were also probably trophic responses. As discussed in the following section, this crayfish had a higher ingestion rate and a higher proportion of riffle-inhabiting benthos in its diet than did 0. inermis. Trophic Ecology Results Predator-prey,interactions The various relationships among predator (either 0. inermis or C. laevis), prey (either isopods or amphipods) and habitat (either bare or rocky substrate) were examined. Factorial analysis (Table 16) indicated that only the main effect involving prey was significant (P < .005). Isopods were preyed upon at a higher rate than amphipods by both predators and in both types of substrate. 155 .cmHm mHanmumm .cmflm uocnmc mg mg ms mG ma mm mG WUZAwUHhHZUHm OH. A m CA. A m CH. A m 0H. A m CH. A m m A moo. mo. A m A OH. m mm. 66. mo. mo. 6m.H mm.6N oo.m m 66. NH. am. so. so. so. om.mH om.H m z 66 H H H H H H H m 6 8v 2: :3 mommm 0mm um 04 mm moamammm mmmm memmsmmam momaom monaONmmezH msommmm 2H4: NH NH NH NH NH NH NH NH :.m m ..m m». 0. mm. mm. om. as. am. mm. m m A mN.N oo.m oo.N mm.N mo.N 66.N m6.H 60.N .m moomHmmzm mnemomH maomHmmzm mOONOmH moomHmmzm moom0mH moomHmmzm maom0mH mmmm mwsmmw .0 mwfiummw .0 mw>mmw .0 mHEHOCw .0 moawommm smoom mmmm mammammnm MODBm ZOHBU¢MWBZH wmmmIMOBdammm mo mHm>Q¢Z< Hdeoeodm 024 wBADmmmll.©H mqmda 156 Prey were consistently captured at a higher rate in the rocky substrate, but the difference was not significant (.10 > P > .05). The facts, that the difference was near the level of significance and that all four predator-prey interaction sample means were higher under condi- tions of rocky substrate, indicated the possibility that a more extensive testing program may have shown a significant difference. If there was a substrate effect, it could have been due either to increased artifici- ality of the bare substrate, which might have diverted the predators from food-search to escape behavior,or to non-random distribution of prey by aggregation in the rocks, which might have made prey capture more efficient. This latter factor might also explain the greater susceptibility of isopods, which display a greater degree of thigmotaxic behavior than amphipods. The C. laevis crayfish apparently engaged in a more vigorous search activity, as measured by the movement of rocks; C. laevis noticeably disturbed the rocks in 83% of the trials, as opposed to 3% for O. inermis. Before testing, the troglobite, O. inermis, was assumed to be more efficient in prey capture than the troglophile, C. laevis, under cave conditions of darkness and low prey density. The data, however, did not indicate any significant difference between the two species, and, in fact, the predation rate sample means were slightly higher for C. laevis. It is possible that the test may have actually measured satiation, and O. inermis, with its lower energy requirements, would have been more susceptible to this confounding factor. This seems unlikely, however, because the number Of prey presented to O. inermis and C. laevis was only 27% and 9% of their respective daily ingestion in the field (see following section on diet analysis). 157 Diet analysis Fecal material collected from C. laevis in the field was found to be correlated to body weight by the regression, log mg dry feces/day = 0.3041 log 9 dry crayfish + 1.4677. The slope of this regression was used to adjust the egestion rates of both crayfish species, both in the laboratory and in the field, to the value associated with a crayfish weighing 1 dry gram. It was not possible to compute a similar regression for O. inermis, because there was an insufficient range in their carapace lengths. The individuals employed in this study, especially the 0. inermis crayfish, were selected for body weights of approximately 1 dry gram, and adjustments were, therefore, relatively minor. Troglobite Laboratory-maintained crayfish ate all three kinds of homogeneous diets, with the animal diet ingested at the highest rate and mud at the lowest (Table 17). The assimilated fraction of the animal diet was quite high, whereas the fraction of plant detritus assimilated was low. The energy content of the feces was in all cases less than that of the ingested material; consequently, the efficiency of energy assimilation was greater than assimilation based on weight. The energy assimilation efficiency was 97.4% for a diet of animals and 51.4% for a diet of plant detritus. Crayfish supplied with a complete diet, including mud, plant and animal components, ingested 8.1 mg dry food/day/g dry crayfish, with a ratio of 12.6% mud, 28.6% plant detritus and 58.8% animals. The fecal 158 TABLE 17.-~FEEDING RATES, ASSIMILATION EFFICIENCIES, AND FECAL CONTENT OF HOMOGENEOUS LABORATORY DIETS 35:63 1 24 43 33 46 02 54. 76 37 86 79 1.5 9295200 0 10 60 11 10 50 62 90 31 60 31 71 >Ummzm 04 007 006 10 203 20 203 20 203 20 209 20 S m 2350mm 7 92 73 32 83 94 41 4 6 01 36 13 mQDMUo 04 207 606 70 303 20 303 30 203 30 309 30 I T S N . . 1 MW 2 77 43 35 51 86 73 36 48 12 29 4.8 MmeHDAAMUp L4 L07 206 (m2 803 01 803 00 003 00 909 00 l 1 l 1 1 C E F 1 26 06 56 57 00 46 76 O9 02 69 31 =mm¢emHo meson moomszOZOm mo wozmHonmm oneU3um ocfiusc ceauoE00wmcouu m>flummmao .H o o m o o e oH.o m.o m.o 0.6 n.a o.~ o o o o Hm.a h.m m.m m.v© n.m w.m© N.m m.m h.N H.v m. . . . . . . .. mfl>mm .o v v v v v v A Hula mo Am Navn ma Ao wavv ma AN mavv 6H ~ 0.0 .o .o . . . x m. x . . . . b 16 To ~13 m4 Amexamov szeora mmoasaamo :m< >czmHonlm Am.~ovm.wm AH.H~.o.mm “H.6HV6.- Am.nva.m meMmcw .o ezmezoo woomo w a w meommzm ezcomx ‘ omoeomoemmommuo >00 o\>mc\ge xuc mo pcmmmumxo sun moon» ccm.nocuo ca cmuofigbob mum ZOHmeOzH Dsz<2HEnm mod mDOHZIUmF LO ZOHEmCFmu\me >uo« h h n m m m m on 2 m.m 0.0 ®.NH HH.O mm.H mm.H mm.m NH.mH m mw>mmw .0 m.m m.q v.hm N.mMH m.mm 0.0H m.om mo.H mm.m Hm.v oo.¢o mh.mm .W m0¢mmDm m m m m m m m ON a m.© N.m m.m mm.o v®.H mo.N Oh.HN Hm.ma m ww>mmw .0 m.N m.©H m.om b.0HH 0.0 b.0m v.0m om.a vm.v N©.m wo.Nm mo.Nm .W m>¢0 m m m m m m m cm : m.ma h.mH N.NH mm.o NO.H no.m hN.hH mm.m m m.b m.mN o.h© o.Hv m.mH m.®m m.vm mo.H mm.m mm.h fim.om mm.®H .W meHmCN .O % % % % a. % .Inona d % % % S w a W m m a m Em as a v m m H W 3 m Irma LE 3 S 3 Tu G I x G I [/naao 3 D T H ¥ .1 N W W m N A I G 1 S L Tm 5L N3 0 Tu T r. m m w ZOHBéADmom mmHhW rim)“-- WV 11‘ declhlc .Uosume mammamcm Hmowm on» ma pwumefiumm wumflc cawflw cam xuoumuonoa nuHB uoflp >Houmuonda Hmsuom mo :OmwummEoo if. .r- A“ fll:-:. v .Nm musmflm 164 crayfish; the estimate of the proportion of dietary components was 6.2 percentage points low for plant material, and 0.6 and 5.6 percentage points high for the mud and animal fractions, respectively (Table 18). The field diet of hypogean crayfish, as estimated from fecal analysis, was 110.7 mg dry food/day/g dry crayfish; mud, plant and animal components composed 2.9, 16.3 and 80.8% of the diet, respectively. The ingestion of surface crayfish was greater, with a rate of 138.2 mg dry food/day/g dry crayfish; mud, plant and animal components comprised 8.3, 4.3 and 87.4% of the diet, respectively. The ingestion rates of the field populations were much greater than laboratory maintained individuals -— 6.7 times greater for the cave population and 8.4 times greater for the crayfish inhabiting the surface stream. The proportion of animals in the diet was lowest in the laboratory group and highest in the surface population; the consumption of plant material was con— siderably less in the surface population, and the percentage of mud in the diet varied from a low in the cave population to a high in the laboratory group (Table 19 and Figure 32). Discussion Trophic studies based on fecal analysis proved to be a valid method. When compared with known ingestion rates and components, this technique only slightly underestimated the rate and plant component. One of the most interesting findings of the trophic study was the large difference in ingestion rates between field and laboratory-maintained crayfish. The estimated high ingestion rates of the field populations might be suspect, since they are based on assimilation rates determined in the laboratory and extrapolated to the field. In order to check this possibility, the assimilation fractions of both the field and 165 laboratory crayfish were determined by a method described by Conover (1966), and it was found that the field assimilation efficiency was only 4 percentage points lower than that in the laboratory. It can be con- cluded that the technique developed in this paper for determining dietary components and ingestion rate was valid in the present case, but care should be exercised in extending its use to other species. Hubbell, Sikora and Paris (1965), for example, found that a terrestrial isopod behaved quite differently; isopods in the laboratory ingested food at a much higher rate and had a much lower assimilation fraction than isopods in the field. It was determined that the epigean population of C. laevis, al- though omnivorous, subsisted almost exclusively on an animal diet. This contrasts with the conclusions of other authors that crayfish are primarily herbivores; Tack (1941) found that animals made up only 4.5% of the diet in Orconectes immunis, and Momot (1967) observed Orconectes virilis to be primarily a herbivore and at times a facultative scavenger. However, both authors used the stomach analysis method, which, because of dif- ferential assimilation, underestimates the ingestion rate of animals. In contrast to the epigean population, those C. laevis crayfish in- habiting the cave stream were exposed to a low density of prey, and plant detritus made up a larger portion of the diet. This shift in dependence to a lower trophic level was a more efficient use of available energy. Despite this dietary modification, the ingestion rate was depressed in relation to the value of the surface population. The troglobitic crayfish, O. inermis, extended this trend towards reduced consumption and more efficient use of energy, with a low ingestion rate of a diet high in plant material. 166 Problems of increasing significance with a shift towards a more herbivorous feeding habit are the much lower assimilation efficiency and the decreased proportion of protein in the food. The protein content of leaves at the cave entrance in July was only 6.3%, while leaves in the cave stream had a protein content of 11.4%. This increase in protein content was probably attributable to bacterial and fungal populations that began to grow on the leaves once they entered the water. It was ingestion of bacteria and fungi, along with the plant material, that probably provided the crayfish with significant amounts of assimilable energy and protein. There was evidence from the homogeneous laboratory diet studies that crayfish might have been able to digest maple leaves more effi- ciently than either oak or beech, with beech being the least digestible. Analysis of leaves entering the cave indicated that maple had only a 42% content of the highly refractive substances, lignin and cellulose, whereas both oak and beech had a 52% content. The content of cellulose and lignin in leaves may have affected the assimilation efficiency by: (l) decreasing the content of assimilable substances; (2) physically combining with assimilable substances and preventing their digestion; (3) affecting the biomass of bacteria and fungi growing on the leaf. There was also evidence that crayfish ate leaves selectively, with the preference ranking matching the ranking of assimilation efficiencies. This evidence was based on the fact that uneaten maple, oak and beech leaves, which were left over at the termination of the complete labo- ratory diets fed to O. inermis, occurred in a 1:6:15 ratio; in addition, this ranking was consistent in all four cases. Cummins et al. (1973) cited several papers that indicate that large particle detritivores show preferences for certain leaf types. Kaushik and Hynes (1971) 167 found that all tested species (2 amphipods and l isopod) preferred leaves in the order, maple, oak, beech; this order corresponds with the crayfish results of the present study. In the cave stream both crayfish species were most commonly found inhabiting the pools, where the substrate was composed of mud mixed with a small quantity of plant-derived detritus. This region of the stream primarily provided the mud and plant components of the diet. Although a portion of the mud in the diet may have been ingested accidentally along with other dietary items, crayfish were observed to ingest mud alone in the homogeneous laboratory diets. Gounot (1960) found that clayey silt was necessary for the growth of young hypogean amphipods of the genus Niphargus; it was concluded that the silt, although it has a probable nutritive role, is necessary because of its antibiotic properties. Animal prey species were extremely sparce in the pools and were represented by a few amphipods that, because of a lack of heterogeneity of substrate grain size, were rather homogeneously distributed. From evidence derived from the predator-prey interaction study, it is evident that this combination of a "bare" substrate and lack of isopod prey, together with low prey density, probably resulted in inefficient prey capture in the pools. Prey capture must have been more efficient in the riffle areas, where prey species, especially isopods, were more abundant and had an aggregated distribution based on current and substrate heterogeneity. Accordingly, it is not sur- prising that crayfish, especially C. laevis, which had a higher pro- portion of animals in its diet, were frequently found foraging in the riffles. 0. inermis had a slightly higher energy assimilation rate of both plant and animal material than did C. laevis. This increased 168 efficiency could have been mediated by a genetic change that occurred during the longer length of time that the troglobite had been exposed to the cave environment with its low level of food resources. On the other hand, it may have simply been related to a reduced ingestion rate; this would have decreased the rate at which food passed through the intestine, and this change, in turn, would have permitted more complete digestive processing. The difference in their assimilation efficiencies was partly responsible for a 2.7-fold difference in ingestion rates, despite only a 2.3-fold difference in metabolism. It must be pointed out that assimilation, as measured in this study (ingestion = egestion + assimilation), was probably not a true measure of energy available to the crayfish. Bacteria inhabiting the gut and feces probably siphoned off a portion of the "assimilated" energy. Zhukova (1963) stated that in the chitin-lined crustacean hindgut bacterial activity is high and no digestion occurs. Conover (1966) believed, however, that bacteria did not contribute more than a few per cent to the total organic content of copepod feces. Nevertheless, Newell (1965) has shown that bacteria can rapidly process fresh feces, causing a rise in the nitrogen content and a decrease in the carbon level. For this reason, in order to accurately reconstruct dietary input, the fecal matter had to be collected at frequent and consistent intervals. One incongruous result was the fact that laboratory-maintained crayfish, despite being subjected to abundant food resources, ingested food at a much lower rate than crayfish inhabiting the food-poor cave stream. The explanation probably lies in the fact that the energy expenditure for locomotion was minimal in the laboratory, where the crayfish were confined and where they found abundant food without the necessity of movement. In the cave, on the other hand, food was scarce 169 and much energy must have been expended to search for it. Activity associated with interactions with other crayfish was another factor to which the isolated laboratory crayfish were not subjected. These addi- tional energy expenditures of the field populations required an increased rate of energy procurement and accounted for the higher ingestion rate in the field. It can be seen that a food-poor ecosystem, such as the cave, is potentially subject to a high degree of resource exploitation. This might arise from a negative feedback mechanism in which the increased energy expenditure of food-search activity requires higher ingestion rates. The result would be increased consumption of already deficient food resources. Although the hypogean C. laevis maintained a high ingestion rate, it is apparent that this proposed feedback mechanism was not fully developed. Instead, certain energy-requiring processes, such as growth and reproduction, operated at reduced levels (see section on budget changes below). The troglobite, O. inermis, has evolved an ecological role that even further minimized such a drain of food resources and, in addition, probably increased the stability of the ecosystem; this was accomplished by a decreased metabolic rate (see section on respiration below), which resulted in both a decreased ingestion rate and an increased efficiency of energy assimilation. Population Attributes Results Tissue growth Introduction The recapture of crayfish on a yearly basis provided data on the annual growth increment. This information, collected from November, 1968, 170 to July, 1970, was regressed on carapace length. The regression formulae for the two species provided the basis for growth curve re- lationships between carapace length and age. Based on observed cara- pace lengths, the growth curves indicated the possibility that the crayfish inhabiting the cave may have possessed extraordinary longevity compared to epigean crayfish (see section on longevity below). The following factors, however, might invalidate this method of indirect aging: (1) errors in the regression formulae were compounded in constructing the growth curves, resulting in increasing age un- certainty with increasing length; (2) the decreasing growth increment with greater length also increased uncertainty at the upper end of the growth curves; (3) the growth rates may have undergone temporal change; (4) migrant crayfish that had undergone different growth rates may have entered the population; (5) the growth curves represented the means, and did not account for individual variation. In the case of 0. inermis, the occurrence of a few crayfish that were of greater length (up to 29.5 mm C L) than the maximum length (28.0 mm C L) predicted by the regres- sion formulae showed that at least some of these factors had an effect. The discrepancy between actual and predicted maximum length is very slight, however, and the degree of correspondence basically supports the growth curve. Although the procedures outlined above were satisfactory for establishing the mean values for the annual growth increments and the growth curves, it was found that the variances were quite high. This resulted in confidence intervals that were impracticably broad, at least for individual data points. The primary reasons for the large variances are probably attributable to the following: growth based on molting has a step—like pattern, at least when growth is measured 171 by increase in length, and it would be expected to have inherently greater variance than a system with gradual growth; secondly, trans- formation of the annual growth data into a growth curve is burdened with the assumption that a short-term pattern of extraordinarily high or low growth, which is shown by an individual during a year‘s observation, is maintained throughout its lifetime. Growth data collected on individuals for longer intervals of 2.0 - 3.5 years made a significant contribution to the determination of growth rates. This information, which overcame the above-mentioned short- comings of annual data, was used to check the accuracy of the annual growth rate determinations and the growth curves derived from them. The long-term data were also used to construct confidence intervals for individuals, so that the probable age of a given individual could be determined from its carapace length. The procedure was to compare the actual length of a recaptured crayfish with the length computed from the growth formula. The deviation, expressed as a fraction of the computed growth increment, was regressed against the initial carapace length. The upper 95% confidence limit for individual data points was determined by a one-tailed t-test; this gives, with 95% confidence, the maximum growth deviation to be expected for an individual of any given carapace length. In addition, for any given carapace length, an annual growth increment can be computed. By multiplying this increment by the applicable confidence limit for growth deviation, a factor is derived, which, when added to the increment gives maximum growth, and, when subtracted from the increment gives minimum growth, with 95% confidence. By applying this procedure to two hypo- thetical individuals, one assumed to have maximum growth and the other assumed to have minimum growth throughout their lifetimes, it is possible I v ..I'.‘ m an“ \‘ h; .- 172 to construct a growth curve with a 95% confidence interval for the prediction of individual observations. The confidence intervals so computed were much narrower than those based on annual data, but they would have been even less with additional data gathered over a more extended observation period. Troglobite Annual recapture included 88 crayfish with initial carapace lengths ranging from 9.0 to 28.0 mm. Growth was found to fit the regres- sion, Y = 8.27 - 0.293 X, where Y is the annual growth increment in mm and X is the initial carapace length in mm (Figure 33). There was no significant difference in the regression between the two sexes (Table 20). The regression formula was used to construct a growth curve relationship between carapace length and age (Figure 34); the asymtotic nature of the curve is quite obvious. The predicted mean carapace length and their 95% confidence limits were calculated for each year class (Table 21). Supplementary growth measurements support the validity of the growth curve for the population (Table 22). Measurements of 34 tagged individuals were made over a time interval of 2.0 to 3.5 years, and the results were superimposed on the growth curve (Figure 34). The extended- interval growth measurements were also converted by the growth curve to an annual basis (Table 20). The resulting regression formula was not significantly different from that computed from annual observations. The growth rate was slightly lower, however, resulting in a decrease of approximately one mm in the theoretical maximum carapace length. 1373 w>usu zuzouo . >n Anscct x >H~.o - mm.o n » N.¢~ om mam.o mean: 0» vmuuo>aoo mo.Aona. OH.AmA0m. Ho>uoucn x mmm.o - vo.o u s H.n~ ca ~e~.o moamaom coauu>uomno _ How» 0 x mm~.o - mm.” n » m.n~ cm om6.o cocanaoo m.m-o.~ . m. oa.AaAo~. ow.Aa m T. S x mm~.o - n~.m n s ~.m~ mm 664.0 cmcfinsou x mm~.o - am.a u s o.m~ av vmm.o moan: Hn>uoucn 0m.Aonv. om.nm 20au0>uwmno . x hom.o . em.m n s 6.m~ av Hme.o mmamamm Hmscc¢ D . x ema.0 a mm.m n > 6.~o Hm hav.o cochneoo r p. x NNH.o - 56.5 n » o.vo «H mmn.o moan: Hm>uwucm m Om.Am om.Am cofium>uomno m. x vea.o - vm.m u » ~.Ho as mee.o memEmm Hmscc< mamaHm¢> Amuo x to w Aesvxeozmq moamamao gmmmmo onecznzmmemo xmm momaom s op cosmos leecezmsmmozH mezomo q: 4462 mo 3 ammeuu m.ezmo:em mA¢DOH>HQZH Qm¥mmwmmO A¢>mszH mMUZOA DZ< 443224 20 ammdm IBUZNJ MU‘A‘G‘U 2H ZBSQGU lmHm>mm~ .0 can meHmcw .0 mo gumcma ca £03000 Hooccd .mm musmflm 175 325200 I wm4HD2H 2mm: de2 AflDQH>HQ2H Zémz 2¢m2 2H mBHZHA MUZMDHMZOU wmm mBHZHA MUZWQHMZOU wmm m04 mwbmmw .0 meHmcw .0 ANN manmev vm musmfim mo w>uso msu Bonn £u3oum Awumm> m.mno.Nv EHwUImsoa mo mcoflu Imw>o© map :0 comma mum mamocw>flpcfl MOM muaEHH wocmpflmcoo map coammoummu nuBoum Hanson on» so comma mum muHEHH mocmpamcoo “Hosp pom magma 0:9 mEHZHA MUZMQHMZOU wmm 92¢ m2dmz EBH3 .AEEV mmBGZMA m0¢m¢m¢0 UHhHUmmmlmwdll.HN mqmflfi low magma mo mmHsauom 179 TABLE 22.--COMPARISON OF MEASURED LONG-TERM GROWTH OF CRAYFISH CARAPACE LENGTHS WITH COMPUTATIONS FROM ANNUAL GROWTH DATA TIME CARAPACE LENGTH(mm) GROWTH SPECIES SEX INTERVAL INITIAL FINAL DEVIATION (years) ACTUAL COMPUTED (mm) 0. inermis Female 2.0 10.8 17.5 18.9 1.4 0. inermis Female 2.0 19.0 22.8 23.3 0.5 0. inermis Female 2.0 21.5 23.5 24.3 0.8 O. inermis Female 2.0 12.5 20.2 19.8 0.4 O. inermis Male 2.0 19.0 24.0 23.3 0.7 O. inermis Male 2.0 14.8 20.0 21.3 1.3 O. inermis Male 2.0 13.0 19.7 19.6 0.1 0. inermis Male 2.0 15.5 22.5 21.7 0.8 O. inermis Female 2.2 13.7 21.5 21.3 0.2 0. inermis Female 2.3 24.5 25.0 26.5 1.5 O. inermis Female 2.3 13.3 22.5 21.3 1.2 0. inermis Female 2.3 16.0 24.5 22.6 1.9 O. inermis Male 2.3 25.0 27.2 26.7 0.5 O. inermis Male 2.3 15.5 19.7 22.8 3.1 0. inermis Male 2.5 17.0 24.0 23.4 0.6 O. inermis Male 2.5 22.0 25.5 25.6 0.1 O. inermis Female 2.6 17.5 26.0 23.8 2.2 0. inermis Female 2.6 21.5 25.5 25.4 0.1 O. inermis Male 2.6 24.0 25.5 26.5 1.0 O. inermis Male 2.6 16.5 23.7 23.4 0.3 O. inermis Male 2.6 11.5 18.0 21.3 3.3 O. inermis Female 2.7 15.0 21.5 22.9 1.4 O. inermis Male 2.7 17.5 22.8 23.9 1.1 0. inermis Male 2.7 20.5 23.8 25.1 1.3 O. inermis Female 2.9 24.5 25.7 26.9 1.2 O. inermis Male 2.9 17.7 22.7 24.4 1.7 O. inermis Male 3.0 20.2 23.0 25.4 2.4 0. inermis Male 3.0 14.7 20.8 23.4 2.6 O. inermis Male 3.0 22.0 25.5 26.0 0.5 O. inermis Female 3.3 10.0 22.0 22.3 0.3 0. inermis Female 3.3 11.0 24.0 22.6 1.4 O. inermis Male 3.3 26.0 27.2 27.5 0.3 O. inermis Male 3.3 16.7 22.3 24.5 2.2 0. inermis Male 3.4 16.0 23.5 24.5 1.0 C. laevis Female 2.0 43.0 46.0 47.9 1.9 C. laevis Male 2.3 37.5 45.5 44.5 1.0 C. laevis Male 2.4 28.5 40.0 38.4 1.6 C. laevis Male 3.3 35.0 45.5 45.3 0.2 *based on regression formulae of Table 20 In 180 Figure 34 illustrates the extent of deviation between actual and predicted growth over a period of several years. Deviation, expressed in relative terms as a fraction of the predicted growth increment, was correlated (r = 0.46) with carapace length; the relationship is represent- ed by the regression formula, Y = 0.017 X - 0.084, where Y is the ratio between deviation and the predicted increment, and X is the initial carapace length. For any given carapace length, the upper individual 95% confidence limit for this relationship was applied as a factor to adjust the mean growth increment to maximal and minimal levels. Beginning with the 5.0 mm carapace length of newly-hatched O. inermis, this procedure generated the 95% confidence limits for the prediction of the growth of individuals over their lifetimes (Table 21). For any given individual crayfish, based on its carapace length, age could be assessed with 95% confidence. The ratio of the confidence interval to the mean carapace length ranged from a high of 37% for 2 year-olds to a low of 11% for 12 year-olds (if they occurred in the population); the mean value was 24%. In absolute terms, the confidence interval was broadest at age four. The confidence interval gradually narrowed in older crayfish, as faster growing individuals experienced the highly reduced growth rates characteristic of the longer carapace lengths, while individuals with lower growth rates were undergoing the more rapid growth associated with a lesser carapace length. In terms of weight, the growth rate, which was based on annual data, was quite low for O. inermis (Table 23). The relative annual growth rate varied from a high of 19.04 during the first year of growth to a low of 0.12 during the seventh; the mean was 3.33. The annual absolute growth increment peaked during the fourth year and gradually declined in older crayfish. 181 mama nuzoum msoocmucmumcH u m mama £¢3oum m>HHMHmH Hmscad u s pwumfifixowmmm waao ma npsouu« oumu £u3oum ousaomnw Hmsccd u x no.N 0N.Nh hw.N 5mm.o Nvo.v who.a mNm.o vmm.m moa.o zamz .. .. .. ovN.o HhN.o mmm.a .. .. .. sum .. .. .. mmN.o mvm.o mvm.a maa.o NNH.o mno.o nun .. .. .. mum.o mmv.o Nmm.H moa.o mwa.o Hoa.o cum .. .. .. Hmv.o amm.o 0mm.H mmN.o omN.o mma.o gum m¢.o mm.o mv.v moo.o Ham.o NNo.H nmm.o hmv.o mma.o :uv mm.o mN.H hm.m Nmm.o Nmm.a nmm.o mvo.o mam.o oma.o cum om.a m@.N «N.N mvm.H wmo.m omm.o Nma.a mmN.N voa.o 0cm mo.m o¢.¢mN vm.o mov.m mON.mN mmo.o mam.N ¢¢O.ma mvo.o uma m a x m a x m a x mazomo mo mmm» *mw>mmw .0 zamemm mwsmmw .0 zmmwomwm mHEHm:w Ame emonz smo zH mazomo mmHmmamouu.m~ mamas 182 It was also found that the growth rate varied in different sections of the stream (Table 24). The census area of the cave stream was longi- tudinally divided into five lOO-meter sections, and the growth rates of crayfish living in each section were compared. Analysis by Tukey's honestly significant difference procedure (Steel and Torrie, 1960) provided evidence that growth was not significantly different for four of the sections, but that crayfish inhabiting the upstream section (0-99 meters) grew at a significantly higher rate. Troglophile Annual growth was monitored in 31 individuals of the cave popula- tion, with carapace lengths ranging from 11.0 to 45.5 mm. Growth best fit the regression formula, Y = 8.38 - 0.134 X, where Y is the annual growth increment in mm and X is the initial carapace length in mm (Figure 33). There was no significant difference in sex-specific regressions (Table 20). The regression formula was used to construct a growth curve relationship between carapace length and age (Figure 34); the growth rate gradually decreased with age, but was not asymtotic. The predicted mean carapace length and their 95% confidence limits were calculated for each age group (Table 21). A largely unsuccessful attempt was made to collect long-term growth data. The extreme movement characteristic of C. laevis is probably the reason only four long-term (2.0 - 3.5 years) recaptures were achieved. Premature death, due to insufficient adaptation to the cave environment, is possibly another factor, but trophic studies (see page 164 above) have shown that members of the cave population ingested food at almost the same rate as the surface population, and 183 *2: 2H BZNZMZUZH 2930mm AdDZZ< «2: 2H BzmzmmUZH $83020 A43224 ww>mmN .0 meHmCN .0 Ho.m mN.mo MN macauomm Cwnuwx m: CA. A m mv.o mN.H mN.H H macauomm 0:054 MHSQMN .0 hm.~ mm.VNH on mc0wuowm canvas m: moo. v m no.0 mv.oa N>.H¢ v mcofluowm ocosd meuma .0 .cmwm yo: u m: .cmwm >H£0w£ n m: a m m: mm «0 onede¢> mo momDOm mmHummm m02<0HhH20Hm mqmde m02¢Hm<> mo me>A¢z< camcoa wuommumo cows new £u3ouo ou :oflmmmuowu numcmalucmfimuocfl >9 Umumsmnm« b.NH ma vm.o oh.N m.m~ ma mmvloow m.m 0H m~.o mm.N m.mH 5N mmmaoom NH mm H mm c v.6 ma om.o Hm.N h.HN mN mmNiOON o.ma ma mm.o mN.N m.mH ma mmauooa ma mw.~ vo.m v.5m HN mo.N o~.v N.Nm mm mmno ZOHB Auommumu m.wlo.m0 mwfluokumo LuvcmH m cc wmo.0 0mv.0 NA mommm mc 02. A a m0.a mm0.0 650.0 N ZOHFU.N 00H.0 00H.o H xmm mw>0mN .0 m: OH. A a mm.a vwo.o mNA.0 N msz m: moo. v a 0h.02 00v.0 maw.0 N «ccamszmdu mxooam 0H0.0 m0m.0 VN mommm m: 0A. A m hm.N 0m0.0 050.0 N ZeueodmmPZH Amy m: 0H. A a mN.N hmo.o hmo.o H xmm mehmcw .0 m mNo. v m mm.m Hmo.o HmH.o N NIHF m: moo. v m wh.mH mom.0 «MN.H v acazhozmdv mxuoqm .«cmcmwméuv yo: H m: unauwuwcufim n m m m: mm 00 moxDOm AZOHEMMMWMMW unaoamhamam shaman u m: NUZ‘UHLHZUHM mam‘a m02 ho mHm>A¢2¢ > o < No m:0quwuu~m 0540 can :00 ca cocoau:« 902. m m .3 ma ma N m m w w v = vv mm on am mm 0 Am Nv hm ma 0 m mw>mm~ .0 50 «R NF 05 2. o 00 no 00 mm 02. R 0H 0H 0H mm ma 0 0 0 o w an 2 0m MN MN 0N 0N 0N ha NH HA ON 0 m wwfiumcw .0 mm mm mm mm 05 mm v0 mm v0 N0 00H N m<£ >02 400 quzmm HAoz L56 1&4: m.mN|o.mN oéNquN m.o~uo.: méaooéa mNHIOd m.mlo.m Qonmm ulna .300 holed. mw>00N .0 . mbmzmumMBzH xmm ”Sou 000050 MNEhmcw .0 .wmnmoomhm UZHZOHBHBKa um.ofllo.m "moflu000umo m ou nmmcmucoU m .. mm.~ oa.vv om . momma m: 0H. A m nm.a mo.H mo.H H oneommmezH Aoav m: 3. A m and and and a xmm $32 .0 m: moo. v m H~.oa mm.m~ mm.ma a «.mauzmq .. mm.o cm.Hm . cam momma m: 0H. A m cm.o o~.o o~.o m oneummmsz Ammv m: mo. A m A H. m~.m no.H mo.~ H xmm mHE~mzw .o m: moo. v m qo.mH mm.m «m.HH m «mpozmq .coflm >H£mfl£ n m: AZOHeéoHAmmmv .mmfim uo: u m: m m m: mm m@ mumsow mmHommm mozmonquHm mummy mozmHm<> mo mHmmqmzm mm mm m m m VH va HA 2 mm.H mm.a mm.o ¢~.H m~.H mv.a mm.o mv.o m mw>mm~ .9 mm.m vm.~ m>.H mo.m an.m wm.m Hm.~ oo.a k NNH wmfi Ha mm mm mm mm m : mm.o No.0 mv.o vm.o 00.0 vo.c mv.o o m mwsumcw .o ov.a HN.H nm.o mo.H Hm.a om.H Ho.a om.H M ma<2mm mam: m.nv-o.mv m.¢v-o.wm m.nmuo.am m.om-o.v~ m.m~-o.na m.oano.oa emsm mmHommm m.mmno.m~ o.v~-o.am m.cm-o.ma o.wfl-o.ma m.ma-o.m m.m -o.m A30u um3oH0 mH>mm~ .0 . EE xmm .30u uwmmsv mHEchw .0 .mmHmoumemu A V mewzmq mummmmmu xmm nzm meozmq mm oszoHeHemmm xmm 02¢ abuzmd .mmHUmmm wm QMN»A¢Z¢ 024 QWZOHEHEM¢MUI|.mN mdmdfi 193 Figure 37. Molting in O. inermis 194 2 15 1001 M I ‘ 5 ’o.\ .— ... 21 36 5 O ‘ “... 02 #9 I ék? ' v 5 0°” 0. (a . 5 2°. S In 60 o z I.“ o 9 - E m .... t“- ‘ a oassnvnaous cg —— MAR—JUL. z . -— JUL. — nov. 5° NOV. - MAR. O. I If I l l I 1004 03—-——03 a OBSERVATIONS o w ' ‘-\ —— MALES E E o . , ~02; ...egs FEMALES g z E 17 '00... ~.,‘ -09.. O "- .0... °. 0 E III 80‘ .037 55 \‘o. o ' a K» E V m I 2 mops ANNUALLY o..\g h- 13 % .Ju‘a) (9' O "u z 60‘ 49 o.\ g‘rtu . I .0 o’- 9 1 .°° \ C .0 km E ..0.\ <2: F:- 2 40‘ ’3. \8 g ' 1 MOLT ANNUALLY _ ‘ baa I I T *1 I I 1.8“ - —§— MALES — ‘ )LS; E d ..I ...-I.-.I.... ...§... FEMALES V ‘ wgg.’ ’ ~ ‘ ~ ..COOI.......§ ’5 1.4‘ \\ "... [U q \\ .... :1 ‘ \ “... L2, «I ‘\\ .0 .......... " 1.0-I \‘ ‘ "" x a 1 ‘x ‘E 0.6 5-8 9112 19.-15’ 17-20 21-24 25-28 INITIAL CARAPACE LENGTH (mm) Figure 37 -~ .— Figure 38. Molting in C. laevis. 195 1004 % MOLTING DURING 4-MONTH INTER-CENSUS PERIOD 0! O k I ‘ a OBSERVATIONS \\ .3 a —- MAR.— JUL. \ t. . — JUL.- NOV. \\'-.. ..IIII NOV.—MAR. ‘0.‘ 0+ 10 I 1 I I I I 100+ 2m0IOOIOIIIIIIQ8 fl OBSERVAT|ONS . 7 \\ .00.... — _ MALES (ED I ‘8 ‘I \‘70 0'0..... 10 nun FEMALES g I. Q 1 \\ u... ’0: o 5 m \ ."af’ \ Q7 2 N — 2 MOLTS ANNUALLY A / “0 a"... o o a \O’ .... .. . O. E é a, I 8 0g“ \ 0.. P o 6' Lu 3 501 \ .0. 2 “J 2 \ .0 % LU .4 05 0.. R’- 9 —-1 MOLT ANNUALLY ‘¢ _— m .. o. z u‘ I“! g E5 z * E a 0+ '02 —-—-——. I I I T 1 1 Sq A .45. MM. _ E " X, S; S 0.0;... FEMALES 5 ..:" " I.“ T I - to? \ ...000. 2 l .0’ \ 0.. I“ J ... \ m 3 .....’ \\ .... 2 I — " ...... / .... Z 9”... / .... g E .. a " '{ ’} 11 10-16 17-23 24-30 31-37 38-44 45-51 INITIAL CARAPACE LENGTH (mm) Figure 38 197 Juveniles molted in excess of twice annually, with those in the 13.0 - 16.9 mm C L and 17.0 - 20.9 mm C L size groups molting about every 4.7 months. Crayfish of the 21.0 — 24.9 mm C L size category molted twice annually, while those in the largest size group molted only once. Molt increments were determined from 246 observations (Figure 37c and Table 26). The increment reached a maximum of 1.6 mm in the young juveniles, then decreased to 0.9 mm in the largest size category; these differences were highly significant. Females had a slightly greater molt increment than males, but the difference was not significant. Troglophile No significant differences were observed in either the seasonal or sexual molt frequency in C. laevis (Figures 38a and 38b, and Table 25). Molting frequency generally decreased with increasing size, al- though the older juveniles and young adults (24.0 - 44.9 mm C L) were stabilized at two molts annually (Figure 38b). These differences in molting frequency associated with increasing size were highly sig- nificant (Table 25). From an initial molt of 1.1 mm, the molt increment increased with increasing length, peaking at a 3.8 mm increment in the largest juvenile stage; the growth increment in larger crayfish decreased to a level equivalent to that in the smallest size categories (Figure 38c). Although differences in length-specific growth increments were highly significant,the greater increments observed in females were not sig- nificantly different from those of males (Table 26). 198 Reproduction Troglobite Only a few ovigerous 0. inermis were encountered during the course of the study. During the egg laying season only 19.4% of the mature females (22.6 mm C L or larger) were in berry. There was evidence that only about two-thirds of the females with developing ovarian eggs and cement glands eventually laid eggs, with the others apparently under- going resorption (see section on population size and structure below). Eight pleopod egg counts were made; the range was 23 to 61, with a mean of 37. The number of eggs did not appear to increase linearly with parent size. Egg-carrying females ranged in size from 23.0 to 26.0 mm C L, and the greatest number of eggs were carried by a crayfish 24.0 mm in carapace length. 0. inermis eggs were quite small, having a mean dry weight of 2.56 mg. Of the total of ll egg—bearing females encountered, all but one occurred in a 4.5-month time span, from June 27 to November 11. The one exception was a female with well developed ovarian eggs (the eggs can be seen through the translucent carapace) and cement glands that had been removed from the stream. This crayfish was observed to carry eggs from October 25, 1969, to April 7, 1970. The unnatural experimental conditions may have caused the failure of these eggs to hatch. A total of four females carrying young attached to their pleopods were observed. All were encountered in a 2.5-month interval, from September 1 to November 18. The time required for the eggs to hatch was observed by placing a female with well-develOped ovarian eggs inside a sealed minnow trap. The trap was positioned in the cave stream so that the flow of 199 water and food through the 6 mm-mesh openings simulated natural conditions. This 23.5 mm C L female was placed in the trap June 29, 1969. Eggs were attached to the pleopods August 1, and they began hatching 102 days later, on November 11. By November 18 all young had left the pleopods of the mother. Two of the O. inermis that had produced young were found up to 6 months later; consequently, it is assumed that females were capable of reproducing more than once in their lifetime. The size range of ovigerous females indicated that egg production probably occurred over at least a 3-year period, beginning during their fifth year. The largest (28.0 mm C L) female captured, of advanced but highly uncertain age, had develop- ing ovarian eggs in the spring of 1970 and possibly laid eggs later that year. During the year of study a total of 7 females carrying either eggs or young on their pleopods were captured in the census area; four of these females were found in the quantitative study area. Pleopod egg production was determined from the egg count tally; in the cases of females bearing young, the mean number of pleopod eggs was substituted. It was calculated that the annual pleopod egg production for the census area was 248, and for the quantitative study area 129. Troglophile The capture of ovigerous C. laevis was quite meager. Only three females ranging in size from 43.0 to 45.5 mm C L were found in berry during the year of study. This represented 25.0% of all females 42.0 mm C L and larger encountered during the egg laying season. There was no evidence of egg resorption. The number of pleopod eggs had a wide 200 range, from 7 to 183, and had a mean of 98. The eggs were large with a mean weight of 5.44 mg. Of the total of five egg-bearing females encountered in the cave, all occurred in a 3-month period from July 20 to October 22. A cray- fish, observed to be carrying eggs October 22, was found to be carrying newly-hatched young when recaptured four days later. A 43.0 mm C L female with developed cement glands was placed in a sealed minnow trap July 6. This crayfish was carrying eggs 14 days later, and by October 25 the young crayfish were beginning to temporarily detach from the pleopods. Young remained associated with the parent for another 24 days. The total developmental time was 121 days. C. laevis females did not die after reproduction, as do the females of many crayfish species. Of the three females encountered carrying eggs, two were recaptured 6 to 19 months later. The annual pleopod egg production for the census area was observed to be 260, and for the quantitative study area 70. Mortality Troglobite Mortality rates in O. inermis were very consistent, with no excessive mortality associated with any particular age group (Table 27). Mortality was generally greater at the age extremes and lowest for the 2-3 age class, which was transitional between juvenile and adult. There was, however, a definite pattern of differential mortality between the sexes (Figure 39). Mortality, based on horizontal sampling, was higher for males in the 2-3 and 3—4 age groups; female mortality was greater in both younger and older age groups (Table 27). The general pattern 201 moxwm noon new unvao>wsvm ma >ufiamuuoe wow oomoon uonu :Oaumsdmmu any :0 Ovmonn _ momsmcoo v no come :0 vwmancc uuozou umw> some no mcofluo>ummno amsccm co ommmn .EuOu coaufiooe a. Amhm.0. .00m.0~ “Ohm.ov AvNo.Ov Aoav.o. Avom.ov z «UZHAmlim AdBZONHmO: ucUZHAmZ «OZHdm8H>qu AWDOmz8HAH>mDm 024 m9<¢ >9HA¢B¢O£II.FN Hdmdb 202 Figure 39. Comparison of sex—specific mortality rates in O. inermis, based on both horizontal and vertical sampling. 203 535 O a I .. / o I / l/ I IIIIIIII o \‘0 IIIIIIIIIII LI \\\ IIIIIIII \ IIIIIIIIIII \\\ IIIIIIIII \\ oooooooooooo \\ IIIIIIIIIII \\ ttttttttt \\\ III. 0 IIIIII \0‘ IO. IIIIIIIII \\ nnnnnnnnnn \ IIIIIII \ I... \ \‘l IIIIIII IIIII \\ ........ \ nnnnnnnn \ ........ \ IIIIIIII G 0“,, IIIII 00 me III, C... LN III of: PI I, 00 ML ’1' to A I 0 SM '1'" 000 I O A I! to A II.’ TL 00"" om IIIII \\ mm ...... \\ WW. 000000 0‘“ I o‘“?\\ -l \ — \‘Joav 0 0. . .... .l . .... . ... . .. . . m .. m 0 cl 0. O 0.. 232.3 39 0.2: cosivUO\m a Q a A3330 3.: 22:0. :05: “\b mmhh3 come we mcoHuu>uumnO Hansen :0 woman a . K :OHUMRONEO + >uHamquE A :OHuMNUHEEa H mwxom :uon uOu unmac>w=qw we >umuwunofi moo comowdm umzu :oHumfiamnu :o cummn_ H mwmsmcou e no cave co o0mmn.« HEHOu UOaufivoE ac 83.8 33.8 230.8 23.: SS8 83.8 25.2 23.0 mama H38 :08 280 03.0 NH To ....H o o o .. .. .. .. .. .. m; 8H H.~ TH To To H.o Amwmnwc AWMWHMC H SH SH 2: To . . Ham. mm. mm. Hm... HE. . me m6 0.... mg «g 82 Him H ”Mummy H H AMMMHWC H Tm ~.v Tm NH mm 93 m... H H H xwmmumv H Hmmwnmc TH H... 3.: 3.: HH 2.: 3.: Ammumc H ”mums H H Ammmumv To GOA OCH OCH OCH OCH OCH AWWMHMV .. .. AMWMHMV .. .. DOMWWMQ IBOQ MMJ IIOZHJQZ‘M 440H91N> IUZHJQZ¢m AKFZONHKOI IIUZHQQI iUZHQQSH>mDm AMDOHZ¢BZ8qufim02 443224 mwsoam m~=wm0>H>MDw 92¢ ab‘z rkHA¢Emo:ll.wN nant? 206 Figure 40. Inter-specific comparison of survivorship curves, based on both horizontal and vertical sampling. NUMBER SURVIVING 1.. 05% 5:4 100* 501 101 0.3 PLEOPOD 6.1 EGGS 207 1L2 213 ’3Th Ids —— HORIZONTAL SAMPLING "H" VERTICAL SAMPLING 0 9199.525 A gems. 'tz ‘\ ‘hs (I ‘\‘. \.'L \.‘E . \\Ra \ °.A..quu"'%.. . \ \ " O O O I O O I AGE GROUPS Figure 40 (yrs) 5T6 6:7 718 208 Survivorship curves based on horizontal and vertical sampling were slightly dissimilar, especially in the older age groups where the horizontal-based survivorship was less. It was judged that study- related mortalities, which occurred primarily in the younger age groups, did not contribute to this difference in survival. Longevity The longevity of the crayfish living in Shiloh Cave could only be approximated. Based on growth rates measured during the year of study, age was equated to length. Because of variation in the growth rate of individuals, this aging method was accurate only at the popula- tion level, as a mean value,and then only under certain assumed condi- tions. It was especially fallible for accurately aging very large individuals. The maximum carapace lengths attained by O. inermis were 29.5 mm for a male and 28.0 mm for a female. Because these individuals probably had higher than average growth rates, they could not be used in estimating the maximum age. A more reliable estimation was reached by taking the mean length of the 12 largest crayfish of both sexes; this number represented 5% of the sampled population. The resulting carapace lengths of 27.2 mm for males and 27.0 mm for females, indicated a probable age of 9 years. In the case of C. laevis, carapace lengths of 47.8 mm for the males and 46.7 mm for the females were the means of the largest 5% of the sampled population. According to the growth curve, this indicated an age of approximately 9 years. The asymptotic nature of the growth curve, which was especially well-developed in O. inermis, caused severe problems in accurately determining longevity from maximum carapace length data, since crayfish could have lived for several years with no measurable increase in length. 209 The life span was best judged from long—term recaptures of tagged individuals, although this also relied in part on the age-length rela- tionship. Using this method, the oldest O. inermis was a male that was originally tagged when its carapace length was 26.0 mm, and its age, by length-age conversion, was assumed to be 6.8 years. This individual was recaptured 3.3 years later, and was assumed to be 10 years old (with 95% confidence that the crayfish was at least 7.6 years). The oldest female was monitored for 2.9 years, and it was judged to be over 8 years of age (with 95% confidence of at least 6.4). There were very few long—term recapture of C. laevis; this probably resulted from the higher mortality rates and greater movement that were characteristic of this species. The oldest, followed for only two years, was a female judged to be over 9 years of age at the end of the period (95% confidence of at least 6.8). The oldest male, which was monitored for 3.3 years, attained an estimated age of almost 8.5 years (95% confidence of at least 6.8). These age estimates for C. laevis are especially questionable, on account of the possible movement between the epigean and hypogean streams. Such movement would invalidate initial aging by the length conversion method, because the length-age relation- ship differed greatly for crayfish inhabiting these two habitats. Population size and structure Troglobite According to growth data gathered during the year of study, the population, based on carapace length, was partitioned into seven year groups (Table 29). Another group, represented by crayfish greater than 7 years of age, was not an annual age grouping. Information from 11.. - 210 >uuwn :H a n _m. OHME w u o mvcmHm ucwEwo awaoHo>mo zqu a u Amv EHOm umun a n H40 coHumHamom mo w n U mmHmeu mo umnfisc u m memE mo nonfisc n < ADNZHQEOU mN w 0mm mNm Nmm mNm mmm mmxmmv ZOHBNADmOm adeoa N.mNH v.0“ Ho.0Nv Av.HNv Ho.om0 H0.00H0 A0.00Hv w.vw 0.0 H N N.H N N H.H H m 0.0 N H m.o 0 H m.oNA NA N.mn m.0w AQ.ONV AH.va “0.0mv Ho.omv A0.00H0 A0.mN0 A0.00Hv m.wN m.mv N.H m N N.H N N m.o H H m.H m N m.H v N Iw.mN le mm.N_ _o.mN_ n.0H v.0“ A0.MNV AN.m®V Ho.mNV AN.wo. Am.NHV A0.00H0 A0.mN0 Hm.mmv Aw.mNV A0.mN0 m.mN w.Nm m.m w m o.v N o H.m m m N.m m o m.m N ¢ Im.VN mum 3.2 SH: 3...”: m.NHH v.HH AH.vH0 AN.mmV AN.NNO Hw.mNV Am.mm0 Am.mHv Hm.mHv Hm.Nm0 v.vN m.Nm m.o mH m N.m 0H N m.N NH MH m.m NH m m.N NH 0 I0.NN mlv H.0R N.HH Am.wv Av.mN0 AN.w0 AH.HNV AH.NV Ho.oN0 AH.mV HH.mV Am.0HV Ho.oNv m.NN m.Nm m.0H mH HN N.HH 0H HN 0.0 mH mH m.m NH vN 0.NH 0N NN Im.0N vlm m.mw v.Hu AN.mH0 Am.vHv AN.®NV A0.0HV v.0N m.vm o.mH VN mN o.VH HN mN 0.NH mN 0m m.VH mN NN H.VH mH 0N I0.NH MIN m.mfl N.vw AH.HV Am.mv 0.0H N.vv m.mN ow mm w.mN Nm Nv m.mH Nv mN N.wN mv Mm w.HN Nv on IO.NH NIH N.OH m.m« m.HH H.vv m.mm NN Ho w.vm vm 0m N.HV Na MN m.Nm ow no w.mm NN Nm no.m HIo .8 E. E. :2 E. Hmv A40 Amv Adv Am. A40 Am. A40 An. Am. a o m < O m m o m m u m m u m < .35 Sub .H 0 mm... 0N. mom<2 mw. mmmlm>oz mo. wflbh mm. =Um<2 D m H zdm: mamzmo >ODBm m0 m¢mfi HIP UZHMDQ (mm< >QDFm NE omN Hie UZHBHQ4IZH mwfihmfiw .0 mo mmDFUDKPm Dz¢ WNHm ZOHB‘ADQOQ omfidtHBmMII.mN mqmda 211 long-term growth studies (see page 208 above) indicated that longevity may have been greater (longevity Of approximately 10 years) than that indicated by the above method, that is, age categorization of the population based on the relationship between carapace length and growth rate. The proportions of the various age classes in the O. inermis population showed a consistent decrease with age (Figures 41 and 47). The decrease was very gradual, and the young-of—the-year crayfish made up only 38.6% of the population. Sexual maturity in males, indicated by presence of the form I gonopod, was generally reached during the third year, when crayfish in- creased to carapace lengths greater than 18.0 mm (Figure 41). One exception was a form I male with a carapace length of 16.0 mm, which indicated an age of less than two years. Upon reaching maturity during the third year, form I males commonly occurred in all succeeding age groups. The percentage of mature males in the form I condition generally increased from the third (15.2%) to the sixth year (63.2%). Based on the capture of only a few individuals, the two oldest age categories seemed to show a leveling off of this trend. Mature females often possessed developed cement glands and maturing ovarian eggs visible through the translucent carapace. They first ap- peared during the fourth year, with the smallest mature females having a carapace length of 21.0 mm. This indicated a pattern in which female maturity lagged a year behind the attainment of maturity in males. Cooper (1975) found a similar, but even more pronounced, lag in the cave crayfish, O. australis, in which male maturity was reached at a carapace length only 74% of that of females reaching maturity. The percentage of females with developed cement glands increased from a 212 .uuonoo 0 mo cuzoum muMOHOCH mBOHHm ocm Umxwm uoc HIo mum “Nosum mo How» on» mcHuso ANS ova mmum msmcwo map :H :OHumHsmom mHEHOCH .0 MO mwumEHumm musummowulxumz .Hv musmHm H3 39V (SM) 00. >433 00. 20m<§ 3.5.. 25:00 nV EEO“. ...mz_m m..<=o.>_oz_ Ow n O m or >¢¢wm 2. moz<40 hzwiwo mw4<§ wm4<2wu SOVdVUVO H19N31 WW 90) (SIUGWOJOUI 214 mean of 8.5% of fourth year females to 23.0% for those in their sixth year, while the two older age categories were stabilized at the 20.0% level. The pattern of change of crayfish in the active sexual state was similar in both sexes, but the level was always lower in the females. The production of pleopod eggs was observed in only the fifth and sixth age groups, but it is assumed that the seventh and over-seven age groups would have shown reproductive capacity if a larger population had been observed. The sex ratio varied greatly among the age classes (Table 29). The proportion of males in young-of-the-year crayfish was 44.1 i 0.2%. This value only represented the ratio of crayfish 10.0 mm C L or greater, since smaller individuals could not be reliably sexed. It is uncertain whether this value represented the ratio at hatching or resulted from a differential rate of mortality in the unsexed sizes. The proportion of males steadily increased with age, reaching a high of 57.9 i 6.1% during the fourth year. The ratio underwent an abrupt change during the next year, and the prOportion of males declined to its lowest level of 32.9 i 12.9%. The ratio of males again increased during the remaining age groups and reached a high of 64.6 i 29.2%. This complex pattern was the result of sex-specific mortality rates that, except for young adults, favored males. There were seasonal changes in the sexual condition of both sexes. Form II males predominated in July, with only 6.6% of males Older than two years with form I gonopods. By November the percentage of form I males increased to 49.4%. Intermediate levels occurred in March, with 25.8% in 1969 and 27.0% in 1970. March appeared to be a molting period with males changing from first to second form. The reverse molt occurred between July and November. It should be emphasized that the annual 215 alternation of form did not involve the entire population of mature males (Figure 42). In females the cement glands and oocytes underwent seasonal changes in development. During March,l969, cement glands were developed in 19.5% of females older than three years. By July cement glands were developed in only 15.0% of the mature females, but the change was par- tially accounted for by the presence of females in berry (5.1%). By November developed cement glands had decreased to only 5.1%,and 9.9% of mature females were in berry. Of those individuals with developed glands in July, by November approximately a third were carrying eggs, another third were unchanged, and the final third had apparently under- gone resorption. In March of 1970 the proportion of females with developed glands had increased to 19.7%, which was comparable to the level of a year previously. These changes in the sexual conditions of both sexes were phased to the reproductive pattern of this species. Mating apparently occurred in late fall when the breeding form male (form I) was most abundant. Ovarian eggs and cement glands developed during the spring, and pleopod eggs were carried in summer and early fall. Seasonal fluctuations in the carapace length structure of the 0. inermis population were minimal. Recruitment was spread over a period of several months, and this, combined with differential growth, resulted in a near constant flow through the age groups. Only one definite cohort pulse was observed passing through the population; this consisted of an increase in the mean carapace length of the young—of- the-year crayfish (Figure 41). The population size for the census area was 356 i 28; this was equivalent to a density of 4747 crayfish per hectare. The estimated 216 Figure 42. Seasonal percentages of form I male 0. inermis. 217 E34666 M2345> 100+ 801 1 0 6 _2 11 I“ 4 mo”. 5m... HR 20‘ 0- JUL'69 NOV'SS MAR37O MARHSQ CENSUS Figure 42 218 population showed only a slight numerical variation seasonally. The lowest population levels occurred in the March censuses -—-333 in 1969 and 329 in 1970. The largest population, 382, occurred during the November census. These changes in population level paralleled similar slight changes in young-of-the-year crayfish, which probably generated the fluctuations. From knowledge of the time of egg-laying, it would be expected that the young-of-the-year would be most numerous during the March census, instead of least numerous. However, at this time they were mostly quite small, and this probably permitted them to escape representative detection. The population size also varied in different longitudinal sections of the stream (Figure 46b). The highest density was found in the up- stream section, where 32% of the census area population resided. The middle section, 200—299 meters, was inhabited by 22% of the population, and the other three sections contained equivalent levels of 15%. The population levels of the various sections were compared to the following variables: (1) the annual growth rate; (2) the number of C. laevis; (3) the extent of optimal habitat (subjectively judged to be the non- riffle areas); and (4) the food level, as determined by measurement of the benthos, microseston and mud energy content (Table 31). Positive simple correlations, with correlation coefficients above 0.83, existed between the number of O. inermis and three of the variables. Only the food level, with a value of -0.43, did not fit the general pattern. Partial correlations were computed for the three variables showing positive correlation. Both optimal habitat and the number of C. laevis still showed high positive correlations (r = 0.80) with the number of O. inermis; the growth rate, however, had a coefficient of -O.54. 219 Because of the small number of stream sections observed, none of these partial correlations, although some were high, were found to be signifi- cant by the t-test. Troglophile All sizes, from 7.0 to 47.0 mm C L, were well represented in the C. laevis population. There was an isolated occurrence of one exception- ally large individual of 58.5 mm C L. With the aid of growth data this population was categorized into 9 age groups, with the last group in- cluding all crayfish over 8 years of age (Table 30). Long—term growth studies (see page 208 above) reinforced longevity estimates of approxi- mately 9 years. Aging was based on the relationship between the mean annual growth increment and the initial carapace length of members of the cave population; this relationship provided an estimation of the age of individuals. In contrast, based on population structure and limited growth data, the epigean population was roughly estimated to be composed of four year groups (Figure 45). The population structure of the cave—inhabiting C. laevis was quite complex (Figures 43 and 47). Several age groups, such as the third, fourth and seventh, were numerically inflated above the levels of the preceding age groups. This would superficially appear to result from variations in annual recruitment. The survivorship curve based on horizontal sampling showed a similar pattern, however, with several cohorts increasing numerically with time (see Figure 40 above). In all such cases the data had overlapping 95% confidence limits, and the increases in numbers could simply be ascribed to sampling error. The following observations, however, provide arguments against drawing this conclusion: (1) the patterns of the horizontal and vertical 220 iuon :H o I 2: vol: an... OH. 0 I n nccoHu acolou £3950 cuH) o I Am. . EON awn: o I 2. cons—Eon :IIOOANL no 9.5.1... can HIo: :oHuIHsnon mo 0 I U IIHIEON no heal-E I m MNHII uo moan I 4 :oHuIHaaon :IIuHmn- . I . . 39.—HAUL woman. m 3 + N : 3 00 on 9N zonagaom .262. 413. 3.0: H.mHn oNu 36H. 8.9.. 863 8.03 3.02 303 . 8.03 80: ...va mA N.0N N.m m.N 02H Nd n N m.N v 0 HA N H o.« N H 3.0: H62 N8“ 3.3 80.3 30.2 .062 _. Finnv 8°: vévIoNv mIN H.Hn via. o.N o.H o o 0 H.@ N m H6 9 0 m6 N H Tm“ H.NR :4: 3.68 863 :68 8a: 8.93 3.2; Ném m6 m.N MN 03 N N H.m N v Hé N N ¢.N n m m.Hva.mm NIw «.63 may . wéN v.0 m4 m.m m6 H N mé. o m N.m N n 0..: v m v.mmIo.mm mum m.m« véw wiHm v.3 oé né 0.9 N N N.mH N m m6 n n N.MH v o ménumdm ml... Lmfi. .06 «:43 «8.0: N.mHH H.NH ..on cme 3..va adv N.mH o.N 0.0 N.mH w v N.HN m 0 N4: 0 w H.NH v m v.0nlm.mN cum . co.m 10.02 .388 v.oH« N.HN umNn uNNN av.¢0Im.Nm H.wm N.mH 0.m 02w v.0H v m 06H 0 m N.wH m HH N.HH m m v.mNIm.mH HIN .w.m .365 7va v.2 cva «02. «v.NnIm.NN w.VN N.vH m.m m.N 0.0H N m Hiw o v NA: v NH m.mH N m «.mHIodH NIH 0.?» 3:8 uNNmm .ommw .¢.NNIo.m de nH N.mH NH H6 v H.NN oN N.MH 0H m.NHIo.m :HIo _S _E 3.: ME 2: H5 2: 30 2: :3 H5 3. 2: 3: 2: 1E: Amp: 0 U m < U m 1 o m < U m c U m < .H U m0< 2.. :92: mo. mango: mo. .53». mm. :05:- o m + ands. . I mamzmo g Egan—Ham z< OZHHHEZH gnfijag gag E ho ZOHhflr-Hamfl £0301 ‘ 92‘ NEH—Fm m0 Ex NE. 02350 at imam E OmN E UZHFHEH hw>0§~ ..u gr: m0 ”macaw 024 ”Nam zOnB‘Hn—g aimkmflIIAun MARSH N 221 .Umxmm uoa filo mom “macaw mo Mama mnu mcflusw ANE omnv mwum msmcwo may :H cofiumasmom wwbmmw .9 mo mmumefiumm musummomulxumz .mv wusmflm 222 39V (SJA) Oh. IU¢zxmn z. , mcz<40 hzuiwo mmu<2 mwa<§wu ”on Wm .oo BOVdVHVO H19N31 (“‘“‘) 223 survivorship curves were quite similar; (2) the standard deviations in age class percentages among the four censuses were small, especially when it is considered that a portion of this variation was due to real seasonal changes in the population structure (Figure 47); and (3) even if the increases were spurious, the numerical stability of an age class during a year's time seems contrary to an expected mortality-based decrease in numbers. Accepting these arguments, and with fluctuations in annual recruitment being ruled out as the prime causal agent, the increases in certain age classes can only be explained by immigration. The survivorship curve based on horizontal sampling indicated that the cave C. laevis population was probably supplemented by immigration in the second, third and fourth age groups. The significance of immigration, either from other reaches of the cave stream or, more probably, from the surface stream, was demonstrated by the fact that young-of-the-year crayfish made up only 17.3% of the cave population. The influence of immigration was also observed in the sex ratio. The sex ratio of the cave pOpulation varied in a very complex manner (Figure 44) that seemed to be primarily influenced by immigration in the younger age classes and other factors in the older groups. Immigra- tion was assumed to be Operating when the age class in question was both numerically inflated and locomotively active. Although the measurement of age-specific movement rates of the cave population did not constitute a direct measure of immigration, extensive movement, expressed as a displacement rate, appeared to be associated with an influx of crayfish into the census area. There is a logical connection. Assuming move- Inent patterns to be specific for various carapace length categories thether the crayfish were residing in the epigean or hypogean stream, Figure 44. 224 Relationship between movement, sex ratio, and relative change in the size of succeeding age groups in hypogean C. laevis. 225 22.)... pzmsmojama hmz Some 93 ozamomma 20:“. 325.0 43.5552 : 4 2 m 42.. 6 0 LP F P r bi h P a». .0 \ 8 o 0000000 ‘.“‘o fr 00 ‘- ...... ““ .. “ ...... ‘III‘\II- on” o" ..o f 'l' o "'0' 6 11"", 0 I'll, 0000 6 fl. 00000 5 ooooo O. T 000000 N 0000 5 E P r. “M meU .4 E T O Cnmn A nu L S. anE 3 I- F. G D S A u _ .... — 2 2 fir J o m 4 o 1% q 7 5 3 m 352m“. x + m d: 1 ‘1 1 4 2 O w W + . . (ms) AGE GROUPS Figure 44 226 extensive non-directed displacement would result in a net movement from the high density surface population to the low density cave population. The young-of-the-year C. laevis were not sexed because of the un- certainty of proper sex identification. Crayfish of the second year class were 69.8% male, and this ratio gradually decreased to a value of 49.1% in the fourth year class; moreover, the sizes of the age classes and the length-specific movement rates were concurrently in- creasing. All of these changes were probably associated with immigration of females into the population. After the fourth age class the propor- tion of males increased until it reached 65.0% in the sixth. This was associated with both decreasing movement rates and decreasing age class sizes. For these age classes immigration seemed to be curtailed, and differential mortality favoring males apparently accounted for the shift in the sex ratio. A decline in the proportion of males to 55.0% of the seventh age class was correlated with increases in both the size of the age class and movement. Once again immigration may have been involved, but it is uncertain whether the increased proportion of females was due to immigration of females or to other factors that were apparently operating in older age groups. The eighth and over eight age groups had sex ratios of 33.3 and 26.7% males, respectively. This decreasing proportion of males was associated with both decrease in movement and decrease in size of the eighth age class; this relationship runs counter to trends in younger classes. Immigration seemed to be insignificant in these groups, and the changes in the sex ratio may instead have been (:aused by higher mortality rates in males. Another possibility is 1:hat slight differences in the growth rate favoring females may have Ilad a cumulative effect, and this may have been responsible for the ILower proportion of males reaching the two largest size groups. 227 The length-specific sex ratios of trap-captured epigean C. laevis were similar to those of the cave population. Females were in the ma- jority, except for the second year class. The ratio of the first year class, which corresponded in its length range to the age classes of the cave population experiencing immigration, was only 43.9% male. The preponderance of first year class females in the surface population probably resulted in their entering the cave to the greatest extent; this migration would explain the inflated numbers and shifted sex ratios of corresponding size groups in the cave population. Both males and females in the cave population reached maturity during their seventh year. The percentage of males with form I gonopods varied in the mature age classes from 66.6 to 100.0%, but no trend with length was obvious. The proportion of females with developed cement glands increased with increasing length, from 11.1% of seventh year females to 22.9% of those over eight years old. Pleopod eggs were observed being carried by females in the eighth and over eight age groups. In epigean populations maturity was probably attained when the crayfish were two years of age (Table 30). Not only did epigean popu- lations reach maturity at a much younger age, but also at a smaller size. In the cave the smallest form I male was 38.5 mm C L and the smallest female with developed cement glands was 41.5 mm C L. In the surface stream flowing from Shiloh Cave the lengths were 34.0 and 34.5 mm C L respectively, and in the stream that issues from Sullivan's Cave the lengths were 32.0 and 38.5 mm C L. The sexual condition of hypogean males varied to some extent with 1:he seasons. During November form I gonopods were present in 100% of nnales older than six years. Form I males were present throughout the fifear, however, and never were less than 58.4%. The 66.7% presence 228 of first form males in July was comparable to the level of first form males in the surface population. In hypogean females the presence of either cement glands or pleopod eggs was found throughout the year. Cement glands occurred only during the March censuses, varying from 23.8 to 35.0% of the females among the age groups older than six years. Pleopod eggs were present in both July and November, varying from 22.5 to 13.3%, respectively. In July cement glands and pleopod eggs combined were present in a larger proportion of mature epigean females than in their hypogean counterparts (48.5 versus 22.5%). This pattern of change in the sexual state of the cave population indicated that mating probably occurred in late fall, when the first form males were most abundant. In females, mating was followed by the developing of cement glands in the spring and the carrying of pleopod eggs during late summer and early fall. There were only small seasonal fluctuations in population structure. The low growth rate and extreme longevity characteristic of the cave population tended to minimize seasonal changes in the structure of the population. Although young-of—the-year crayfish were present through- out the year, the largest pulse occurred in July. These crayfish were approximately eight months old, and had entered a size range that was more representatively sampled than the smaller sizes. The mean size of the estimated population in the census area was 74 i 16, which was equivalent to a density of 987 crayfish per hectare. 'The density of the epigean population was roughly estimated to be $174,000 crayfish per hectare, or 176 times the density of the cave I?opulation (Figure 45). The size of the cave population underwent Substantial seasonal changes that probably resulted from fluctuation Figure 45. 229 Size and structure of epigean (approximated) and hypogean populations of C. laevis. 230 ANNUAL GROWTH OBSERVED IN 2 EPIGEAN INDIVIDUALS " EPIGEAN — — HYPOGEAN O AGE Q .......... w ll'\'._0 IIIIII fl‘ ......... .0 My OI»... ll lllllllllll m mu. 0 ..: O 0 0 1 O 1 1 :5 Sh \ 5:38. :5qu CARAPACE LENGTH ( mm ) Figure 45 231 of the first year class and migration. In March of 1969 the population size was 76. Measurements of movement showed that 64% of wandering C. laevis were moving upstream at this time. It is reasonable that, if the denser epigean population was undergoing similar movements, this would result in increasing the size of the upstream cave population; the July census did show an increased population of 96. By July, how- ever, 53% of wandering crayfish were moving downstream, and, apparently as a consequence, the population had decreased to 64 in the November census. Movement of the majority of the crayfish at this time was still downstream (at a rate of 65%), and by the following March the population had decreased to 61. Population densities varied in different longitudinal zones of the cave stream. Density was highest in the upstream section, from 0 to 99 meters. This section, which was a fifth of the census area, contained an annual mean of 57.4% of the population. Density was lowest in the middle section, with densities progressively increasing in both upstream and downstream sections (Figure 46b). The population levels of the various sections were compared to associated variables (Table 31). The number of C. laevis showed positive simple correlations with the number of O. inermis (r = 0.90), the growth rate of O. inermis (r = 0.94), and the extent of non-riffle habitat (r = 0.71). Only the food level, with a value of -0.30, did not fit the pattern. The growth rate of C. laevis was not tested, because insufficient numbers remained within one stream section throughout the year. Partial correlations, which Inere calculated for the three variables showing positive correlation, Vvere positive for a number of O. inermis (r = 0.80) and the growth Itate of O. inermis (r = 0.87); the extent of non-riffle habitat, Ilowever, was transformed to a negative correlation (r — -0.66). 232 Figure 46. Inter-specific comparison of longitudinal population levels. (#Imz) CRAYFISH PER TOTAL HABITAT 0.8- .o 9’ 0.4+ 0.2‘ 04 233 H mm TOTAL HABITAT —— OPTIMAL HABITAT Q. inermis a Q. laevis I DE) ... o..\ I " . I .29 O .....IIOIIIIUIIICI G o. \ / A\\ /’ \\ I/ / "'n\.AfiufiflI.W-m-?A"'" nuance-00"'.A O 0-99 100:199 200-299 300-399 400-499 LONGITUDINAL STREAM SECTION (m) Figure 46 (ll/III2 ) CRAYFISH PER OPTIMAL HABITAT 234 TABLE 31.--LONGITUDINAL VARIATIONS IN CRAYFISH NUMBERS VARIABLE ::::Igu 1 2 3 4 5 (meter MEAN NO. MEAN NO. 0. inermis NON-RIFFLE FOOD . C. laevis O. inermis GROWTH* HABITAT(m) LEVEL** location) 0-99 42 115 4.2 39 2.82 100-199 9 55 2.3 15 3.14 200-299 6 77 2.5 28 2.78 300-399 7 54 2.9 27 3.55 400-499 9 54 2 8 13 2.71 MEAN 15 71 2.9 24 3.00 STD. DEV. 15 26 0.7 11 0.35 *mean of annual length increments (mm) that were adjusted for variation in length-specific growth rates ** + r1 r2 r l 2 3 + r3, where ratio of the measured food level in a stream section versus the mean for the study area = microseston (cal/l) mud (cal/g) = benthos (mg/m2) SIMPLE AND (PARTIAL)* CORRELATION MATRICES VARIABLE 1 2 3 4 5 1 0.90 0.94 0.71 —0.30 " (0.80) (0.87) (-0.66) 2 0.90 ._ 0.84 0.86 -0.43 (0.80) (-0.54) (0.80) 3 0.94 0.84 .. 0.77 —0.19 (0.87) (—0.54) (0.64) 4 0.71 0.86 0.77 _, 0.00 (-0.66) (0.80) (0.64) 5 -0.30 -0.43 -0.19 0.00 -- *variable 5 not included 235 .mwumaflumo cofiumHsmom accommom v Eouw ©0>flumo mcoflumfl>oo oucpcmum 0cm mcmoE mCNBOBm .moflamu>m 000 m0 COmemmEoo DamaowmmluoucH INDIA”! 7’s... Ian! I! .64 6.66.8 236 :V wusmfim ZO_.—.H26A64. unmoflmflcmam. A¢DQH>HozH ammomaq no oneaooq mqHozH emmumgq oa mo mmeozmq moHGCN 0mm uo OHQEmm co ~00me Guano “m949Hm<: Bzamhmha 2H DZH>H.H mZOHHSDmOQ mw>¢mN .U Mrm QMZH7 -0.18 0.85 0 " °° " 7—8 '° -' ° 5.24 8.17 1.03 >8 -- '° -° 2.77 9.54 0.38 TOTAL 32.15 55.44 0.63 71.11 84.50 1.41 -lav- F _I‘...“ : Figure 48. 272 Interspecific comparison of biomass and numerical densities. (dry g/ha) BIOMASS 400' 200' 273 600‘ — BIOMASS . ......- DENSITY o 9. ingrmig * A Q. laevus A . .1500 L. 4000 . ~500 1 )- 0. '0 0'1 1'2 2'3 3'4 4'5 5'6 6'7 7'8 >8 AGE GROUPS (yrs) Figure 48 DENSITY ( crayIish/ ha ) 274 Figure 49. Interspecific comparison of the production rates of body tissues, molted skins, and eggs. 275 _ BODY TISSUES IIIIOII EXUV'AE 9.132115. 0 QLine A . 2:3 3:4 INITIAL STAGE OH AGE ‘FZ o'-1 P150900 sees :00 018' v a; .J O 0.06J 2.32:9. :3 552.0362... PRODUCTION (yrs) Figure 49 276 The annual production rate was 2.09 dry kg/ha for the census area (Table 33). The production consisted of 45.3% body tissues, 53.8% molted skins and 0.9% pleopod eggs. In contrast to the older stages, body tissue production was slightly greater than production of molted skins in younger crayfish (Figure 49). Age-specific body tissue produc- tion rates were generally high throughout the intermediate age groups. Discussion Despite a population density almost 5 times greater, the standing crop of O. inermis was less than 36% that of the cave C. laevis. This difference resulted from the disparity in body size of the two crayfish species. Productivity was likewise less in O. inermis, but not to the same extent as the difference in standing crop; this was reflected in a larger P/B ratio for O. inermis. This greater ratio resulted despite a generally greater length-specific relative growth rate for C. laevis. The C. laevis pOpulation, however, possessed a size structure skewed toward the larger individuals with reduced growth rates, resulting in lower growth relative to biomass. This peculiar size structure apparently resulted from augmentation of the cave population by migrating epigean C. laevis. The P/B ratios of both species are within the range typical of animals requiring two or more years to complete their life histories (Mann, 1969). Turnover ratios expressed for the entire life cycles of the two cave crayfish are quite high compared to data presented by Waters (1969). The populations of both species showed a similar pattern of production partitioning; body tissue production was slightly greater than production of molted skins in smaller crayfish, but there was a 277 reversal of this relationship in the larger crayfish. The decrease in the proportion of production funneled into body tissue with increasing body size was especially pronounced in O. inermis. Molted skins comprised the major portion of production in both cases, although C. laevis was more efficient overall in channeling production into body tissue. This higher efficiency of body tissue production resulted even though the exoskeleton made up a greater percentage of body weight in C. laevis. For instance, in a 0.50 dry gram crayfish, 0. inermis has a carapace length of 23.5 mm and the exoskeleton is 39.0% of the body weight, whereas C. laevis is 20.4 mm in carapace length with the exo- skeleton making up 43.0% of body weight. Visual observation of the appearance and relative thickness of the exoskeleton leads one to expect an even greater difference, but the thinner, more elongate shape of O. inermis apparently increases the surface-volume ratio. The heavier exoskeleton of C. laevis was more than compensated by a greater mean relative growth increment per molt ——-0.60 for C. laevis versus 0.56 for O. inermis. Molted skins represent a sizeable loss of energy and minerals from the population. This loss is minimized by mobilization of a portion of the carbohydrates, proteins, and minerals from the exoskeleton prior to molting. Schurr and Stamper (1962), for example, reported a 20-30% withdrawal of calcium into body reservoirs. The extremely low standing crops of the cave populations are evident when they are compared to the standing crOp of the epigean C. laevis population. The biomass per unit area for the cave population of C. laevis was only 0.7% of the level for the surface population. The estimated standing crop of the surface population was extremely large, and it is possible that crayfish movement between various sections of the surface stream, despite erected barriers, may have 278 interfered with the census and resulted in an overestimation. It is my belief, however, that the population estimate is reasonably accurate. Food items, especially detritus (70.9 ash-free grams per m2) and amphipods (4.4 dry grams per m2), were quite abundant and could reasonably support a large stable population. Poulson (1965) reported that the population density of a troglophilic amblyopsid cavefish was 240 times greater in a spring than in an associated cave. The population might also have been inflated beyond its stable level by seasonal movement of C. laevis into these headwaters below the springs of Shiloh Cave. Very little literature exists on crayfish production. Momot (1967, errata) studied 0. virilis inhabiting a marl lake and determined their annual production to be 100.4 wet kg/ha. The production of both cave species together was only 6.9% of this value. The productivity of the river-dwelling O. propinguus was reported by Vannote (1963) to be 415 wet kg/ha/yr ——-a rate 60 times greater than the combined production rates of the two species of Shiloh Cave crayfish. More information is available on the standing crops of crayfish occupying various habitats (Table 34). Of all these studies, none reported standing crops as low as the 15.2 wet kg/ha that was found for the combined cave populations of the present study. Only one station on a biologically unproductive epigean stream with shale and sandstone bedrock approached the low standing crop of Shiloh Cave (Slack, 1955b). On the other hand, the 1538 wet kg/ha biomass density of the epigean C. laevis population was extremely high when compared to other studies. Only a fish-farm pond studied by Langlois (1935) showed a higher standing crOp of crayfish. Unfortunately, no information on standing crop or productivity of Cambarus species was found in the literature. 279 mpsum ucmmmum Emouum w>wu SOHHSm mwbmmN .0 Ha mcsum ucmmwum Emmuum o>mu noaflnm mewmsw .0 m mmma .mfioamamq mason summIrmwm oflso 63080638 .0 mwna I amp mmma .umnoflm 0cm mewsosmo 0:00 manomDSUMmmmz ENNNHHS .0 Apmumfiflpmmv 0NN Emma .0080: mxma Hume cmmwnoflz mflmwuws .0 oma I ma mmma .Hamusq 0000 06005042 6423880 .0 mom I mks mpcom mea .xome ammunsmflm xuow 302 MHCBEEN .0 0mm I mm mvma .Amcaamow mccom cmmflnoflz qusesw .0 mema I 0m ovma .mmeaxoflz Emowum oano pwflwflusmpflcs 0mm I om Hw>wg ammH£ONz moma .wuocsm> umuszEumz UmSOHucm mzswcwmoum .0 omv Qmmma .xomHm mamwuum MEMHUGH 0:302HQOHQ .0 ovm I ma 26g\mx 0033 mUmDOm mmDBdmmeHA BdEHmflm mmHUmmm momU UZHQZdBm wQDBm Bzmmmmm HEB 20mm 00Q u x u:\m 003\oo ca coflumesmcoo ammmxo » wuwzz .Aomeamseamv H\0o m.s mo onmzme zmowxo 24 ea .emonz woom zo ememE 9H2: mmm onemzsmzoo zmwwxo mo onmmmmomm H\oo CH coflmcwu cmmwxo n x u£\m 003\oo coflumESmcoo commxo n w mmHUmmm OHOSB mmHmwde .ZOHmsz ZMUNXO ZO ZOHBmZDmZOO mewxo m0 ZOHmmmmOmm O om.NH Ed EmHmwummno may um>o coflpmuflmmmn mo COmflHmmEoo oawaowmmumucH .om musmflm 284 Om demHh CG :3 50.25 53 CAT: 0% _. v.0 w0.0 $00.0 .\o 06/. 0‘00 M%WN%%39 .0\\.w..... Xx“ @(V 0000 \ I II V .090 \\z700 O 00.00000 \ m o fiPO-O o \ JV... \\0) 0...(..... \ o 70 0 000 \ 1 00 \ \\ / ooooo \ KO 3 .... \000 H $ .6000. \) o .0 000 \ o n r (a 0090 \ o $6 0000 \\ + H @6 00900 \A( f 00000 \ (0 IF. 0000 \ O 0009 \ \v J J \ xx .. n. \ .O \o ‘0 \\Ao . \ ....» f0. l \ 285 and oxygen tension (Table 36). Neither sexual nor seasonal variations were significantly different. Troglophile The rate of oxygen consumption varied with oxygen tension, although the log-log linear relationship indicated the probability that respiratory regulation occurred at higher oxygen tensions (Table 35). These crayfish survived oxygen tensions as low as 0.20 cc/l, although temporary paralysis occurred in one case. Respiration also varied with body weight, with an inverse linear regression between the log of body weight and oxygen consumed per unit weight (Table 35); this resulted in oxygen consumption being proportional to the 0.76 power of wet body weight (Figure 50). In 12.5° C, oxygen- saturated water, respiration varied from 0.143 to 0.020 cc/wet g/hr for extrapolated weights ranging from hatching to the largest (58.0 mm C L) individual encountered. Statistical testing of respiratory rates, adjusted for oxygen tension and body weight, indicated no significant differences between the two sexes, or between the epigean and hypogean populations (Table 36). Seasonal differences were significant for both sexes, with the respiratory rates greater in May than in November. However, the seasonal factor was probably confounded by the change of the test site from the field to the laboratory; for this reason, the source of this variation can not be identified. The relatively high respiratory rate of C. laevis permitted sampling intervals which allowed the determination of diurnal variation in oxygen consumption. The three 8-hour periods were significantly dif- ferent, with the 1600-2400 period having a higher respiratory rate than the other two periods of the day. 2286 m oHo.A0AmNo. mm\N m~.v .>.o.< o pofluwa Hmcusflo m: ow.Am om o~.o . u m youanm: m No.AmANc. Nm MN.N u N Humane .MN>00N .00 cemowm m mAHoo. ov m~.m u 0 A00H080w .6w>mmN .00 cemmmm m: om.A0 mm am.o u m Ammams .mNEATCN .00 Cowmmm m: ov.A0Aom. om MN.o u e Ammawsmu .mwsuch .00 cemmmm m: om.Am mm mv.o u m «mw>0mw .00 xwm mc om.Am mm vH.o u N «mehch .00 xom m .780. 8 $12 8 A 63060.6. ucmoHuACONM uoc n we ucmoduwcvfim u m >PHAHm4mOmm 0 mDA<> Emma mDOmU 009mm8 mqdem¢> mz00 zuom mwbmmN .0 x N vHHo. momo. oowanomo damn o>mu nuom mw>0m~ .0 x mm ono. VOmQ. oomOIOOVN Hack 0>00 zuom mwsmmN .0 0H Nmao. N000. mason «N ocfiumm cmwmflmm mean: wwsmmw .0 m oamo. oHNo. muaon vN mcwumm :aomwmm moanewm mw>mmw .0 x Na HNHo. vao. undo: VN Hawk 0>00 moan: mwsmmN .0 x ma Nmao. ovvo. mucon vN Adam 0>00 mmamswm ww>mmN .0 x NH ANNo. ammo. mayo: VN mcwumm o>mo moan: ww>0tw .0 x MN Gwao. memo. muse: oN mcwumm 0>00 mmdmawm ww>muw .0 x AN @000. NmNo. mason vm Hank o>00 moan: mwshwcw .0 x cm 0000. quo. mason vN Adam «>00 moamemm meumcw .0 x x S N N00. SS. 3:0: 3. 9.30m 260 6262 6.3.205 .0 x x ma mNHo.o NoNo.o musoz vN onwuam o>00 madmaom mNEuocw .0 o N o m c N c m M oonmm «zom¢Mm ax0 :mHL>Zm 02¢ 4(100209 0 om.NH um couosucoo mumou “armflc3 >000 o 003 o.n 0:0 Aomumuaumm. ~\Nc oo m.N Ou noumsnvo .u£\o u03\uo .4sxm amfimsmuo N.v m.NH mm. mm.H 68800:“ .0 mmw>sxm rmfimsmuu H.m 0.44 N0. N0.m 666664 .0 Nuom :mflmsmno 0.4 ~.Nv mm.o om.m mwsumcw .0 swam smhmsmuo m .m m .w 20x29 mammHa msomo szeomm moomo w .N08\Hmo 00 swmmzm onezmm BZWBZOO MDmmHB momezmm mmmBO 02d EmHmwmmu m>¢0 m0 mBZWBZOU ZHHEOMQ MQDmU 02¢ %GMMZMII.hm mqm¢8 294 .mwumu Hmomm cawflw Ucm >Houmuoan CH mocmumMMHw may >9 @wumsnwm mmsam> coflumuflmmmu “NE mvm mo mmmm >Usum w>fiumuflucmsv wnu cwsounu :Hmuoum wwsuu 6cm >muwcw mo 30am Hmscc¢ .Hm wusmflm (88) 3963 295 NOILVZIWIlfl am 000000 1 (19) 9989 3...: «.N 0000 m .0 l 3.0. 4.0 2.0. 0.: I ‘llllllulldll . , .000: n .3: ¢.> x0 I. ‘Jdlll: m 300. z w m .2300 0.32.0903 m 0:: w u. .403 :00 :03 304. , ,I 03 .3300: lizolimudfl :::... : A. i u 1.3 ... 20:05.3. 9 u l .03 :3. u l 024.... .0. 2.05.... 00:00 :3 .. ... :8... 50020 t “8.30.... a 30.: no 20.0.20 Iawl “.504 05>:me .0 2.30:... 30.: . , .1. lij: = 0 9.04000A u ... .0 xf s , 50.500 8.80... 0000 02 2.30 _H— 1.000 20.20.0000 .. 20.5002. 1 00: u , ; 12.... 20.5083. 2 .0: m 183. 0004 .500000000 04.. 100 (Embed .2000 H F ..<:.z< < 0000200 VII _. .00. 504. 000.5...le040 f 2.203 I 0000 I} H A. .r .02. 002. 00:00.: 30.: ( 2050000012 4 300.44: 08.00.... i kr E 33.44: 02.54.. 4‘ Im=dmzh 30.: c < ZOhmme¢0§a 296 limivorous habits and also to a slight extent by the crayfish. The benthos, in turn, was ingested by the crayfish and, together with possible cannibalism and interspecific crayfish predation, composed the animal component of the crayfish diet. Quantitatively, mud was consumed at an insignificant rate, and animal material was the major dietary component of both crayfish populations. The calculated ingestion rate was reduced to correct for a 4% error in weight assimilation, as discussed in a previous chapter. Measurement of energy pathways within the crayfish accounted for a large portion of ingested energy --96.3% in O. inermis and 92.1% in C. laevis. These results required, however, that the laboratory measured respiratory rate be multiplied by the difference between the respiratory rate in the laboratory and field for both crayfish; this factor was estimated from the difference between the laboratory and field fecal rates (see section on diet analysis above). Justification for this procedure is founded in a similar situation reported by Odum (1962); based on the fact that the excretion rate paralleled oxygen consumption, he monitored excretion rates of a snail in the two situations and concluded that the snail was more active in the field than in the laboratory. The excretion of nitrogenous waste products was not measured, but it is assumed that this metabolic pathway accounts for only a very limited fraction of the total energy flow. Crustacea are primarily ammonotelic (Meglitsch, 1972), and ammonia is the end product of the wet oxidation technique used for energy measurements; thus, unmeasured excretion energy is derived only from small amounts of low-energy urea and uric acid that are produced. Its calculation, however, would have improved the balance between energy influx and efflux. Unlike energy, 297 measured protein usage accounted for only a small part of the crude protein ingested by the crayfish; this is attributable to the apparently large quantities of excreted crude protein. Because of the method used to measure egestion, bacteria mixed with the feces partially processed the fecal substrate, and this rate, as measured under experimental conditions, is expressed separately in Figure 51; for the calculation of assimilation, however, it is combined with egestion to express the total non-assimilated rate. Bacterial decomposition of fecal matter is a natural process, which after the two day duration of the experiment, had already occurred to an appreciable extent. The fecal decomposition results, however, are in disagreement with those of Davies (1964), who in a similar experiment, found no significant bacterial processing of goldfish feces after 23 days. The remaining energy pathways within the crayfish are concerned with production; they involve energy storage, either in the form of body tissues, exoskeleton material lost during ecdysis (exuviae), or structures, such as eggs, that are associated with reproduction. Their quantitative relationships were, in respective order, 70.6%, 26.9% and 2.5% for C. laevis, and 59.9%, 38.5% and 1.9% for 0. inermis. Energy efficiencies for the crayfish species are shown in Table 38. Those efficiency ratios involving respiration should be regarded as estimates, because of the indirect method used to arrive at the respiration rates in the field. The net growth efficiency [production/ (assimilation)(lOO)] values of 4.5% for 0. inermis and 3.6% for C. laevis are quite low, whereas the assimilation efficiency [assimilation/ (ingestion)(lOO)] values of 83.2% for O. inermis and 79.3% for C. laevis are fairly high. 298 .moumu Hmomm pamflm pom muoumuoan coo3umn oocmummmap m>fiumamu on“ 00 pcmam>flsvo HOUUMM cowmum>coo m an pmumSmwm .oumu coaumuflmmou Um>flumplwuoumnoan m :0 comma moam>« coaumwwcfi n H coflumaasfimmm n d «cowumufimmwu u m mmmw paw mafi>sxm .mnmmwu wpon mo :ofluosvoum u m mm.0 Hm.o m.m0 m.wm v.0m m.m mw>mm~ .0 om.v Hm.o m.mm m.m0 H.Hm m.4 mwfinmcw .0 .one0mHmmmm on0000omm H\« «H\m ««\m ¢\0 .us\~e\amox. .0. medm wwmmzm MUZMHUHmmm mammZM mmHUmmm mZOHE¢ADmom mmHmwdmu Zflmwomwm mom mflBdm Q24 mmHUZmHUHmmm wmeZMII.wm mqmdfi 299 Discussion of the Pattern Mud was ingested in only small amounts by both crayfish species. Although a large absolute quantity of energy is stored in the mud substrate, organic matter is present in only very low concentrations and cannot be efficiently assimilated. Mud, however, may furnish minerals or vitamins that are an essential part of the crayfish diet, but that require only a low rate of ingestion. It is also a possibility that mud is accidentally ingested, in whole or in part, in conjunction with the eating of other food items that are associated with the substrate. Nevertheless, crayfish maintained in the laboratory on a diet of mud alone ingested this dietary component at a rate similar to that observed in the field. Plant material made up a significant part of the diet and, as discussed in a previous chapter, the cave-dwelling C. laevis has shifted its diet from animals to plants in relation to the surface-dwelling population. This shift towards a dietary dependence on a lower trophic level apparently is advantageous in the food-poor cave environment, and the troglobite, O. inermis, has an even greater dietary dependence on plant detritus than does the troglophile, C. laevis. This shift to a larger food base is of limited extent, however; this is probably because it is accompanied by a decreased assimilation efficiency. Unlike other members of the benthos which are of small enough size to graze the microfloral film, crayfish, in addition, ingest the plant particles which are composed mostly of indigestible cellulose and lignin. Of the total caloric content of the macroseston plant material, 57% is derived from cellulose and lignin, and an unknown fraction of the remainder may be rendered unavailable by the structural organization 300 of the cellulose and lignin. This unavailability may apply to the crude protein contained in the plant material, which, at least in its total amount, appeared sufficient to satisfy the needs of the crayfish. It is assumed that, because of the small particle sizes of the microseston, plant detritus ingested by the crayfish is derived princi- pally from the macroseston. The crayfish, 0. inermis and C. laevis, ingested 13% and 15% of the total available macroseston, respectively, and this is much greater foraging impact than indicated by other studies, such as the 5% level of kelp-feeding isopods (Hayes, 1974), and the low utilizations indicated for a salt marsh herbivore and detritivore (Odum and Smalley, 1959). Lee and Inman (1975) suggested that the fact that herbivores consume only a low percentage of available net primary production reflects their role as an ecosystem regulator. The detritus- based trophic ecology of the cave, however, lacks the structure neces- sary for such a control function, since the foraging activity of the crayfish has no effect on future food influx. The rate of plant consumption for the combined crayfish populations was almost twice as great as the calculated macroseston retention rate, and this does not even take into account macroseston channeled through the benthos or mud substrate. It seems obvious that the macroseston retention rate, which was calculated during base stream flow, was a gross under-estimation, since it was inadequate to supply the needs of the crayfish populations. This implies that macroseston washed into the cave during periods of flood was retained in large quantities and that this stored food base was utilized throughout the year in con- junction with current seston inputs. The calculated retention rate measured only that portion of utilized energy based on macroseston re- tained during periods of low discharge and did not measure the stored 301 macroseston utilized by crayfish. That fraction of the macroseston that did not go into storage or that was not consumed by the biota flowed through the quantitative study area to downstream areas of the cave and surface stream. The retention rate calculated for the microseston may also have been in error, but flooding probably did not lead to the degree of storage that is suspected in the macroseston. Animals were the most important food item in the diets of both crayfish, especially C. laevis. This food is easily digested and contains high levels of crude protein that, in comparison to plant material, probably has an amino acid content more in balance with the needs of the crayfish. Although part of this animal dietary component was possibly obtained by inter— and intra-specific predation on crayfish, the bulk was derived from the benthos. The animal component of the diet remains important, not only because of its assimilable nature, but also because it allows the crayfish to utilize, at least indirectly, the energy of the microseston, which is fed on by the benthos. Although both the benthos and crayfish depended upon the macro- seston as a dietary component, competition seemed to be tempered by interactions that suggest some degree of protocooperation. To a large extent the diet of the benthos is obtained by grazing the microflora, which is renewable. On the other hand, most of the crayfish-processed macroseston, because of its high content of cellulose and lignin, is recycled to the substrate where it is available to the benthos of coprophagous habit. Nevertheless, predation on the benthos not only provides a ready source of energy and crude protein, but also increases the quantity of macroseston available to the crayfish. Although both the benthos and the crayfish utilize the macroseston, the microseston is probably only significantly utilized by the benthos. 302 For the benthos the calculated yield/biomass ratio, where yield is based on the animal component of the crayfish diet, was 55. This value is actually too high, because the yield does not take into account either cannibalism or predation on crayfish. In addition, the benthic biomass was under-estimated. This was due to certain components of the benthos that were not quantitatively sampled because they were either too small in body size (e.g., copepods, ostracods and nematodes) or too deep in the substrate. Waters (1969) stated that turnover ratios calculated by the predation method, as in the present case, are invariably high. Hynes (1970) suggested that a P/B value of 10 is probably charac- teristic for populations dominated by arthropods. Although O. inermis comprised only 32% of the total crayfish biomass, this troglobite had a slightly disproportionate impact on the ecosystem by ingesting 35% of the energy flowing through the crayfish populations. This resulted despite the fact that, in considering crayfish of equiva- lent body size, C. laevis experiences the greatest energy flow; at the population level, however, the mean smaller body size of O. inermis favors energy flow through this species due to the inverse correlation between ingestion and body size. The growth efficiencies for both crayfish were extremely low. This resulted from the fact that a very high percentage of assimilated energy is used up in respiration, which is apparently associated with foraging activity. The relationship between net production and main- tenance metabolism indicated an efficiency much lower than the regres— sion for poikilotherms formulated by Engelmann (1966). In a study of three arthropods, Van Hook (1971) found in all cases that growth ef- ficiencies were higher and respiration efficiencies lower than in the present study. Thus, in this ecosystem where energy is extremely 303 limited, growth is not only low in an absolute sense, but also receives a lesser share of energy flowing through the population. 0. inermis, the troglobite, attained a higher growth rate than the troglophile, C. laevis. The troglobite has achieved this increased efficiency by decreasing locomotive activity. 0. inermis also had a higher assimilation efficiency than C. laevis, and this resulted in spite of the fact that O. inermis consumed a higher percentage of highly indigestible plant material. Both cray- fish had high assimilation efficiencies. Odum and Smalley (1959) believe a high assimilation efficiency would benefit a population whose food is produced at a low, but continuous rate; this they concluded from a study of a slow-growing, long-lived, detritus-feeding snail population, in which they found that the snails assimilated a consid- erable portion of ingested food, but used a relatively small part of the assimilated matter for growth. Consequently, there is not only ecological and population dynamic similarities between the snail and crayfish, but also similarities in their patterns of energy flow. Welch (1968) believes that net growth efficiency is negatively cor- related with assimilation efficiency; he suggests that a decrease in net growth efficiency and an increase in assimilation efficiency is in response to a depressed ingestion rate. For the cave-inhabiting crayfish, net growth efficiency, when plotted against assimilation efficiency, falls far below the regression relationship presented by Welch (1968); this again demonstrates the dominance of respiratory energy flow over growth pathways for these cave populations. As discussed in a previous chapter, crayfish maintained in the laboratory with abundant food had a much lower ingestion rate than crayfish inhabiting the food-poor cave. This foraging response to 304 a change in food abundance characterizes the strategy of a time minimizer, as opposed to an energy maximizer (Rapport and Turner, 1975). This strategy leads to a more efficient utilization of re- sources, because a minimum quantity of resources is used for maintenance and self-replacement. This is contrary to the strategy of an energy maximizer, which favors a higher feeding rate in order to maximize the energy surplus available for reproduction. This implies that the cave-inhabiting crayfish, as time minimizers, target only a certain limited amount of energy for reproduction. The ability to govern reproduction would certainly be an asset to a population living in an ecosystem where energy is extremely limited. In fact, only a small percentage of adult female crayfish produced eggs, and these were apparently the individuals that were able to secure energy sufficient to carry out the high energetic cost of egg production. In comparison to epigean crayfish, reproduction was drastically reduced, and the low percentage of storage energy diverted to egg production (1.9% for O. inermis and 2.5% for C. laevis) suggests that production energy in cave-inhabiting crayfish is preferen- tially funneled into tissue growth and exuviae production. Unfortunately, comparable energetic data is not available for epigean crayfish to substantiate this opinion. The strategy of a time minimizer has the effect of minimizing depletion of the food base by decreasing both the ingestion rate and the population size. This trophic strategy is well-suited to the cave ecosystem, and its operation seems to be substantiated by the study results. Although it was initially assumed that protein might be a critical factor, the study revealed that protein was available in surplus quantities. Microflora processing the protein-poor seston 305 input apparently augment the supply by incorporating and converting inorganic nitrogen dissolved in the water. The low energy influx seems instead to be the most critical factor in the trophic ecology of the hypogean crayfish. Budget Changes Associated with Adaptation Eg_the Cave Habitat The decreased energy flow is important, not only in itself, but also in the changes in energy partitioning that have resulted. These changes are shown in Figure 52, which indicates sequential partitioning modifications associated with epigeal—hypogeal and troglophilic- troglobitic transformations. Because of a lack of literature on the energetics of epigean crayfish, the proposed energy scheme is based in part on data from other animal groups and on certain supposi- tions, although the limited study of the epigean C. laevis population has provided some data. Modifications occurring in the energy flow of the hypogean population of the troglophile are assumed to be of a physiological nature, whereas in the troglobite genetic factors are assumed to be involved. In this scheme a significantly decreased food base is thought to trigger a sequence of energy flow modifications. Initially, the scarcity of food depresses the ingestion rate. In C. laevis, the troglophile, the reduction is modest, but in O. inermis, the troglobite, there is a low rate that represents a major decline from epigean levels. The depression of the ingestion rate is much greater in the troglobite, even though evolutionary modifications have increased its food-detecting capabilities (Packard, 1888). The decreased rates result from a lower level of available food, and, at least in the troglobite, the decrease probably also involves a shift in trophic strategy from energy 306 .UmumoflUCH ma mcoflumahowmnmuu AmNEHmCN .0. UHuHQOHmonuIUHHHSQOHmOHu no Amw>mmm .0. Hammom>nlammmwmw Hmsuflm nua3 Uwumfioommm mwmcmzo ucflcofiufipumm >9 vmuo>mw mH umzu mao>ma mafimmmooum mcflpmoUOSm cwo3uon >mznumm 30am wmnmcm one .moflommm amomomwn Hmnuo wnu op w>flumHmH .mmcmno mo wmummp mmHMOflccfl Auwmmwa u HHmEm paw .kumon n mound. muflm 3ouum new .mmsmno mo cofluomuflp mouMOflpsfl Ammmmuomp u £306 paw .mmmmuocfi u may COflwomuflp 3ouu¢ .Hm>wa mafimmmooum comm um mpcmEuHmmEoo ucmHGMMAp maaumucm wmumcm mo mucsoem m>flumamh 0:» CH mwbmmH .U Gmwmwmm CH :uwuumm msu scum mmmcmno mumoapcfl mzouum ucweuummsoo .mbmo msu mo mmmn poom 000000000 050 Scum uddwwu umnu mcowUMOHMAUOE mGHCOHuHuHmQ Hafiucwswom mcflzonm .gmflmmmuo summon»: nmdousu 30Hm mmumcm .mm 000000 mm 000000 cfl 0.8.0:. owm0><3Ih0 30...... >Ommzw I‘ll ¢¢ Gt P2w5w>0$ 05:05 .0 O A 0300. .w . ‘ m20_._.3xw hzmfihmaaioo >Ommzm 307 .Av LIV 5m30m000 w0<¢0hw Adv Air 20_._.<.=!_mm< 9D 20.».mm02. AV. .0'4 Auv A.V Av. .‘Iv Shggwc m02