18 Copy I DEPARTMENT OF COMMERCE BUREAU OF FISHERIES HUGH M. SMITH, Commlwioner EXPERIMENTAL STUDY OF THE GROWTH AND MIGRATION OF FRESH-WATER MUSSELS By Frederick B, Isely Professor of Biology^ Central College^ Fayette^ Missouri APPENDIX m TO THE REPORT OF THE U. S. COMMISSIONER OF FISHERIES FOR I9I3 Bureau of Fisheries Document No. 792 WASHINGTON GOVERNMENT PRINTING OFFICE; 1914 ^ \- DEPARTMENT OF COMMERCE BUREAU OF FISHERIES HUGH M. SMITH, Commissionet EXPERIMENTAL STUDY OF THE GROWTH AND MIGRATION OF FRESH-WATER MUSSELS By Frederick B, Isely Professor of Biology^ Central College, Fayette, Missouri APPENDIX III TO THE REPORT OF THE U. S. COMMISSIONER OF FISHERIES FOR 1913 Bureau of Fisheries Document No. 792 WASHINGTON GOVERNMENT PRINTING OFFICE I9I4 .Is D. OF W, JUL 28 1914 \^ EXPERIMENTAL STUDY \m GROWTH AND MIGRATION OF FRESHlMR MUSSELS By pREDERICi'^ Professor of Biology,m -^^1^9^' Fayette, M Appendix III to the Report of Fisheries I Commissioner bbll m ,'^'- m. — 4 CONTENTS. Introduction 5 Plan of investigation 6 Marking of specimens 6 Field records 7 Planting sites 7 Planting of tagged specimens 7 Shoofly Creek 7 Chikaskia River 8 Growth results 9 Table 1. A year's growth in young Quadrulae 10 Table 2. Two periods of growth, of one year and three months, respectively 11 Table 3. Seasonal growth 12 Tables 4-7. Rate of growth in different sizes of Quadrula undulata l3 Table 8. Rate of growth by species 15 Discussion of data 15 Arrested -growth rings 17 Migration 19 Shoofly 19 Chikaskia 20 Discussion of migration data 21 Economic bearing of experiments — Related problems 21 Summary 22 Literature cited 22 Explanation of plates 23 EXPERIMENTAL STUDY OF THE GROWTH AND MIGRATION OF FRESH-WATER MUSSELS. By Frederick B. Isely, Professor of Biology , Central College, Fayette, Missouri. INTRODUCTION. The growth lines of the Unionidse have long been considered by many observers as "annual rmgs," marking the yearly increase in the shell diameter. On the other hand, proof of the con'ectness of this assumption has been lacking, and not a few mvestigators have ques- tioned its validity. If the so-called annual rings do mark yearly additions, the rate of growth may readily be ascertained in many species by inspection; if, however, two or three of these hnes appear in one season, or prominent Imes appear only at in-egular yearly mtervals, the impor- tance of ''growth Imes" as definite mdicators of rate of growth loses much of its significance. The economic importance of fresh-water mussels has added a new stimulus to the study of the growth problem. Investigations ^ under direction of the Bureau of Fisheries durmg the past five years concerning various phases of and questions related to the problem of artificial propagation of the species valuable for use m the manufac- ture of pearl buttons, knife handles, etc., has promoted mquiry con- cerning the time required for an economic species to reach marketable size. Israel and Haas, among German investigators, have recently given the growth question some attention in connection with their extensive study of the fresh-water mussels of streams of Germany. In this country Lefevre, Curtis, and Coker^ have gathered experimental data concerning growth. In my study of the ecology of the Unionidae during the past five years, one of the perplexmg problems has been the rate of growth and a Various papers by Lefevre, Curtis, Coker, and other workers, in Bureau of Fisheries publications, the Journal of Experimental Zoology, and the Biological Bulletin for 1909, 1910, 1911, and 1912. b Through the kindness of Dr. Coker, I have been permitted to read in manuscript the results of his experiments and observations. . 5 6 GROWTH AND MIGRATION OF FRESH-WATER MUSSELS. the age question. After carrying on experimental work for some time on my own initiative, I was given opportunity in 1910 to undertake the work on a larger scale under the direction of the Commissioner of Fisheries. PLAN OF INVESTIGATION. In the experiments to be undertaken I proposed to ascertain the rate of growth, to mquire into the meanmg of growth Imes, to inves- tigate the relation of age, maximum size, etc., and, as a secondary problem, to gather data concerning migration. The plan of work was, in brief, to collect a thousand specimens of as many species and sizes as could be secured in. the region where the work was carried on ; in some way mark them individually, weigh, measure, and make any other necessary records; and then return them to their usual habitat and at suitable intervals reclaim, weigh, and measure again. To carry out this work, two questions of method had to be solved: (1) The marking of specimens and (2) the reclaiming of them. MARKING OF SPECIMENS. In 1909 I experimented with two ways of marking: One was to scratch a symbol (pi. i, H), as a Roman numeral, on the sheU, and in this way identify it for future records; the second method was to fasten a serial-numbered tag to the shell and keep records by these numbers. In the final work both methods were used, the former for light shells, the latter for heavy ones. The method of marking a shell by scratching a number on it is simple enough, as a mark cut through the epidermis of a mussel valve will be carried indefinitely. To get a satisfactory series for a large number of specimens, however, is difficult. The tagging with serial numbers seemed to me to be more exact. For this method I used brass tags about the size of a dime and fastened them to the specimens with a light copper wire passed through a small hole made in the posterior edge of the valve. The hole in the valve was made with a very fine button-eye drill about 2 millimeters from the posterior edge of the shell (pi. i, ii, and iii). To hold the button-eye drill, a geared hand drill was used, and only a few seconds were needed to drill a hole even in a thick shell. By making a little hook on the wire it could be passed through the hole and out between the valves, usually without much difficulty. In large specimens it was found that the work of tagging could be more readily accomplished by wedging the valves open slightly before inserting the wire. A slight injury to the animal often resulted from the fact that the mantle was not sufficiently drawn back to avoid the drill point. While the irritation at the moment was GROWTH AND MIGRATION OF FRESH-WATER MUSSELS. 7 doubtless severe, the ultimate effect upon the animal's future growth and activity was insignificant (p. 19). FIELD RECORDS. A tabular record was kept of all specimens tagged. The weights were taken in grams, a Harvard trip balance being used; the measure- ments, length, height, and breadth in millimeters with a steel caliper. The field record form, kept in duplicate, was ruled in columns with headings for number, weight, length, height, breadth, species, and remarks. PLANTING SITES. In most cases, after tagging and taking records, the mussels were placed directly back in their original habitats, but in some instances transplanting from river to creek or from pond to river was practiced. Four different sites were selected for planting: One on Shoofly Creek, on the Corn farm in the north end of Kay County, Okla.; two on the Chikaskia River, on the Brewer and Esch farms near Tonkawa, Okla.; and a pond site on the Browne farm near Autwine. As the pond dried up, due to the unusual drought of 1910, this series does not figure in the results. The Chikaskia River is a small, clear-water, sandy stream. The sand is coarse, and frequently there are stretches of gravel and occasionally mud banks and small, mud-bottomed side channels. In certain portions of the Chikaskia, Unionidse are abundant. Shoofly Creek is a tributary of the Chikaskia. In very dry weather the water stops flowing over the shallow, gravelly stretches; but the ponded sections, often a mile in length and -with water 2 to 6 feet deep, have a constant water supply. In certain of these ponded portions mussels are fairly abundant. PLANTING OF TAGGED SPECIMENS. In all, about 900 specimens were tagged, weighed, measured, and listed for future observation. For convenience in further dis- cussion these specimens may be grouped into seven lots. SHOOFLY CREEK. Shoofly, lot A, 140 specimens, and lot B, 80 specimens. — Nearly all of these were Qitadrula undulata (three-ridge) and were taken from the direct site where they were planted after tagging. All of the Q. undulata (three-ridge) secured in the Shoofly were large; out of some 500 specimens handled in two days, only 4 weighed under 200 grams. 20225°— 14— 2 8 GROWTH AND MIGRATION OF FRESH-WATER MUSSELS. Shoofly, lot C. — Twenty-two specimens of Anodonta grandis (floater) were planted in a mud bank similar to the environment selected by this species in this creek. Shoofly, lot D. — As already indicated, the Shoofly species were mostly undulata, with a few grandis, and all were large. In order to get a larger number of species and smaller specimens, I collected 168 mussels in the Chikaskia and planted them in the Shoofly. In this lot, Q. undulata (three-ridge), Q. lachrymosa (maple-leaf), Q. pustulosa (warty-back, pimple-back), and Q. ruhiginosa were represented by fairly good numbers, and nearly all of the specimens were under 200 grams in weight. It should be stated here that a few specimens of all of the QuadrulaB were found native of the Shoofly, and in addition to these, Lampsilis gracilis (paper-shell), Lampsilis anodontoides (yellow sand-shell), Symphynota complanata (heel- splitter), and Anodonta imhecilis were found. Not counting grandis (floater), the ratio would be 25 undulata (three-ridge) to one of another species. This, however, is not unusual in dominance of species in certain streams. In planting specimens, the bottom was cleared of the original occupants and the tagged specimens were put in their places. For example, for lot D, an area of bottom about 12 feet in diameter was cleared and the Chikaskia specimens spread out on the cleared bottom. The Shoofly bottom, where the sites were located, is made up of a mixture of broken blue shale, coarse sand, and mud. The water is still and cloudy and from 2 to 5 feet deep. The specimens were planted in water about 3^ feet deep. Results that follow indicate that the above is a favorable type of habitat for the species used. As already indicated, the specimens were free in the stream; no obstruction of any kind was placed in their way, nor any effort made to confine them. The planting operation consisted in turning them out of a sack and spreading them around on the stream floor. In 24 hours after planting it was noticed that most of the specimens had righted themselves and were stuck in the bottom, foot end down, but seldom was a Quadrula found with the foot extended. The Shoofly specimens, lots A, B, C, and D, were tagged, weighed, etc., June 1.3 to 16, 1910 ; reclaimed and first checked over in part June 14 and 15, 1911 ; and a second time some wei'e reclaimed and checked over September 19, 1911. «■ CHIKASKIA RIVER. In the Chikaskia two lots were planted. These were inclosed with a wire netting, as I was a little doubtful in regard to reclaiming free Unionidse in the Chikaskia. GROWTH AND MIGRATION OF FRESH-WATER MUSSELS. 9 Chikaslcia, lot E. — On the Brewer farm were planted 120 speci- mens of various sizes of Quadrul^, and a few representatives of the other species mentioned for the Shoofly, with the addition of a few specimens of Lampsilis purpuratus and Tritogonia tuherculata (buck- horn, pistol-grip). The inclosure consisted of a triangular pen made of 3-foot, 1-inch mesh mre netting, run out from the bank and back again, 40 feet of netting being used in its construction. The bottom was a mud bank along the side and medium coarse sand farther out. The water was from 2 inches to 3 feet in depth, a portion of the main channel running across the lower end. The specimens were collected in part from the immediate vicinity of the inclosure, but, as they were by no means mmierous in this portion of the river, about 90 of the 120 mussels were secured a mile farther down the stream. Lot E specimens were planted, weighed, etc., June 23 and 24, 1910, and checked over in part September 26, 1910, and June 22, 1911. Chikaslcia, lot F. — The second planting in the Chikaskia was on the Esch farm, and consisted of 330 specimens, collected mostly from the immediate vicinity, as the mussels in this portion of the river were abundant. The range of size was good, although really small speci- mens were rare in 1910. The species were about the same as noted under lot D, mostly Quadrulse, as these are the only common species in this stream. Fifty specimens of TJnio tetralasmus , a pond form never found in the Chikaskia, were placed in the Esch inclosure. Tliis pen was made by fencing across a side channel formed by a long bar. The channel was 40 yards long and from 4 to 6 yards wide. The bottom was mostly coarse sand. At the time of its construction there was a regular flow of water through the channel, the depth of water varying from a few inches to 3 feet. Lot F mussels were planted June 23 to 28, 1910; examined and rechecked in part September 26, 1910, April 11, 1911, and June 20 and 21, 1911. GROWTH RESULTS. In discussing experiments and results, we will first consider growth, and second, migration (p. 19). The results have shown that much could have been learned from a smaller number of specimens. It was not known, however, that we should be able to reclaim so large a per cent of the specimens first planted. Then, too, it was necessary to guard against loss by acci- dents, such as changes in course of streams, drifting sand, drying up of water, and other possible environmental changes. The Shoofly specimens were absolutely unmolested; only three dead specimens were found in the whole lot, and these had all started to grow, showing that the tagging certainly had no bad effects. The Chikaskia speci- mens suffered somewhat from all of the hostile environmental factors 10 GROWTH AND MIGRATION OF FRESH-WATER MUSSELS. mentioned above; and, further, some were lost to the small boy inter- ested in collecting brass tags. On the whole, however, the specimens were reclaimed in such large numbers that all of the material could not be worked over for records in the time availal:)le when the rechecking and reclaiming was done. In rechecking material as many specimens were handled as time would permit. As small specimens were few in number, these were always rechecked and care was taken to include representatives of all species. Aside from these influences in selection, specimens were rechecked as found. The left-over material was returned to the stream when it could not be handled. Where a lot was checked over several times, as lot F, naturally the ones worked over the first time they were reclaimed, September 26, 1910, were again followed up in subsequent work. Far more data were gathered than can be included in this paper. In the tables given below I have stated the reasons for selecting the data presented. Nearly all the material used is selected from lots D and F, as the latter was more available for frequent recheckings, and lot D of the Shoofly material represented a larger number of species and more range as to size. Table 1. — -A Year's Growth in Young Quadrul^e from Lot D. [Note. — All starred numbers in tables represent specimens shown in plate figures.] Speci- Species. Weight. Length. Height. Breadth. men No.o June, 1910. June, 1911. June, 1910. June, 1911. June, 1910. June, 1911. June, 1910. June, 1911. 323* 319 Q. undulata (three-ridge). . . . . . .do Grams. 20 49.5 66 22.5 26 80 52 60 104 10 54 67 Grams. 43 77 102 47 53 112 65 76 115 20 63 77 Mm. 45 64 70 45 45 67 64 64 76 30 56 64 Mm. 62 75 80.5 56 59 76 68.5 69 79 37 58 65 Mm. 31 42 46 35 36 55 48 50 60 25 45 52 Mm. 42 50 51 45 46 60 50 52 65 30 46 53 Mm. 17 24 27 19 20 29 26 26 32 15 28 29 Mm. 23 27 341 do 31 158* 149* Q. lachrymosa (maple-leaf), do 25 25.5 159 do.. 33.5 368 Q rubiginosa 27 351 do . 28 353 .do 33.5 349* Q. pustulosa (warty-back or 20 375 do . 29 383 (Jo 30 AVERAGE GAIN AND PER CENT OF AGGREGATE GAIN FOR EACH SPECIES. Species. Weight. Length. Height. Breadth. Grams. 28.8 27.8 13.3 9.6 Perct. 63 64 18 22 Mm. 12.8 11 4.1 3.3 Per ct. 21 21 6.1 6 Mm. 8 8.3 3 2.3 Per ct. 20 19 5.6 5.7 Mm. 4.3 5.3 1.5 2 Per ct. 19 23 5.3 9.7 a Specimens listed in all tables according to size for each species. GROWTH AND MIGRATION OF FRESH-WATER MUSSELS. 11 The foregoing tables relate to Unionidse from lot D (p. 8) that were planted June 15 and 16, 1910, and reclaimed June 15 and 16, 1911. In the figures given, the first series under each number is the initial record, the second series the record a year later. These specimens show manner and extent of growth in one year's time for four species. They were killed at the time of reclaiming, and are preserved for future study should anyone wish to examme them. It can be seen at a glance that undulata (three-ridge) and lachrymosa (maple-leaf) are rapid growers m comparison with pustulosa (warty- back, pimple-back) and ruhiginosa. Further averages and per- centages are worked out for each species listed in table 1 and later tables on page 8. Table 2. — Two Periods of Growth, op One Year and of Three Months, Respectively, in Specimens Taken prom Lot D. [Note. — Starred numbers represent specimens shown in plate figures.] men No. 135 165 146 168 283 317 329 306 171 166 307 322 163 374 363 384 377 358 355* 371 Species. Weight. June 1910. Q. laehrymosa (ma- ple-leaf) do --..do --..do Q. undulata (three- ridge) ....do ....do ....do ....do ....do ....do ....do ....do Q. pustulosa (war- ty-back, pimple- back).... ....do ....do ....do do Q. rubiginosao do Gms. 92.5 152 164 166 95 120 125 140 145 166 183 185.5 200 June, 1911. Gms. 124 182 216 185 132 147 149 164 175 192 217 207 226 45 73 102 107 145 75 97 Sept. 1911. Gms. 142 194 237 197 150 162 159 176 190 214 230 220 236 50 80 112 114 152 84 104 Length. June, 1910, Mm. 70 93 82 93 72 83 85 90 88 92 98 93 June, 1911 Mm. 76 96 83 96 94 97 102 97 102 Sept. 1911. Mm. 81 97 94 92 90 100 97 102 105 99 104.5 57 59.5 66 85 90 70 80 Height. June 1910. Mm. 57 57 68 57 46 51 55 56 52 56 64 00 58 June, 1911. Mm. 63 61 70 61 51 56 60 60 55 Sept. 1911. Mm. 66.5 63 72 63 56 61 62 62 58 Breadth. June, 1910. Mm. 32 37 41 37 34 34 34 35 37 38 35 41 42 June, 1911. Mm. 36 39 4.3.5 39 38 37 37 37 39 40 37 42 43 Sept., 1911. Mm. 37.5 40 45 40 39 38 38 38 39.5 41 38 42.5 44 27 32 36.5 33 35.5 31 33 a It should be noted that five of the specimens for the two last species were more nearly mature for these species than the specimens of undulata (three-ridge) and laehrymosa (maple-leaf). The specimens shown in the preceding table were planted June 15 and 16, 1910, and were reclaimed June 15 and 16, 1911, put back in the stream, and reclaimed September 9, 1911. All of the specimens of lot D that were reclaimed twice, measured, and weighed, are shown in table 2. In the figures given above, the first is the initial record, the second the record one year later, and the third a three months' summer record (June 15 and 16 to Sept. 9, 1911). Table 2 simply adds to the data of table 1 in regard to annual growth, growth of species, etc. It adds the fact of seasonal growth, 12 GEOWTH AND MIGRATION OF FEESH-WATER MUSSELS. and gives data for comparison of yearly growth and summer months. The average yearly growth for 20 specimens shown in table 2, as well as the growth for three summer months (85 days) is shown below : Weight. Length. Height. Breadth. Average gain for one year, 1910-11 Average gain for three months, 1911 6 Grams. 23.2 12.1 Per cent.a 18 8.2 Mm. 4.2 2.9 Per cent.a 5.2 3.4 Mm. 3 2.4 Mm. 2 1 a Per cent of aggregate gain. b It should be noted in this comparison that the 1910 and 1911 summer months were different in weaf^^r conditions. During 1910 there were no rains heavy enough to raise the creek and wash out the food sup- ply of micro-organisms, while in 191 1 there were two periods of high water, one in July and one in August. Per cent of gain gives a truer basis for this kind of comparison than the average net gain. The lack of conformity in the height averages, when compared with other measurements, is doubtless due to error on account of the great difficulty in getting this dimension in rapid field measurements because of the circular ventral margins of these species. Table 3. — Seasonal Growth. Speci- Species. Weight. Length. men No. June, 1910. Sept., 1910. Apr., 1911. June, 1911. June, 1910. Sept., 1910. Apr., 1911. June, 1911. 594 734 Q. undulata (three-ridge). . . do Grams. 210 255 261 290 104 105 154 160 163.5 164 187 91 141 148 Grams. 222 269 273 305 116 115 167 173 174 180 195 98 146 154 Grams. 222 271 272 305 114 117 . 167 171 176 179 200 98 146. 5 156 Grams. 225 270 277 305 116 123 172 180 182 180 200 99 147 156 Mm. 102 116 110 118 76 66 84 86 84 86 90 64 84 84 Mm. 104 122 112 125 76 67 86 88 86 89 91.5 64.5 86 86 Mm. 104 122 112 125 76.5 68 86 88 86 89 91.5 64.5 86 80 Mm. 104 122-f- 678 do.. 112 573 644 753 do Q. lachrymosa (maple-leaf), do 125 77 70 665 do 86-1- 89 660 do;:..:::.: 687 .. do 86.5 694 .... do 90 622 do 91.5 617 Q. pustulosa (warty-back, pimple-back) 64.5 689 86 721 do 86 While table 2 has given some good data concerning seasonal growth, table 3 gives more detail and permits a more exact location of the growth periods. Table 3 specimens are from lot F, Chikaskia River. These specimens were checked up, approximately, at the third, ninth, and twelfth months that they were under observation, and the results are shown by successive records. The initial records were taken June 23-28, 1910; second, September 26, 1910; third, April 11, 1911; and fourth, June 20, 1911. Only the weight « and length records appear in this table. The average gain for the dif- ferent periods we find to be as follows: a Some observers report weight measurements subject to a great deal of variation. In this investigation I have always kept tlie specimens out of water for short intervals and always under cover. Under these conditions I have found weight measures very satisfactory and stable under repeated reweighings. GEOWTH AND MIGKATION OF FRESH-WATER MUSSELS. 13 Weight. Length. June 28 to Sept. 26, 1910 Sept. 26 to Apr. 11, 1911 Apr. 11 to June 20, 1911. Orams. Mm. 10.9 2.4 .5 .1 2.5 .3 For comparison with these averages we have the results from lot E, which I have not tabulated, for 14 specimens: 7 lachrymosa (maple-leaf); 4 undulata (three-ridge); and 3 pustulosa (warty-back, pimple-back). In this case the April reclaiming was not done. Time. June 23 to Sept. 26, 1910. Sept. 26 to June 23, 1911. Weight. Grams. 9.1 5.4 Length. Mm. LO .55 The second period shows a gain in the growth for lot E over lot F. The explanation is one of food and possibly oxygen, and appears later (p. 24) under the discussion of migration. The following four tables show proportional rate of growth, at different ages, of a single species, Q. undulata (three-ridge). The fact that lot D mussels were transplanted (p. 5) brings in an additional factor (footnote, p. 23), but I am not sure that this is material. In making these comparisons weight and length are used. Table 4. — Increase in One Year op Specimens prom Lot D, Weighing Less THAN 100 Grams. Weight. Length. Specimen No. Weight. Length. Specimen No. June, 1910. June, 1911. June, 1910. June, 1911. June, 1910. June, 1911. June, 1910. June, 1911. 323 Grams. 20 49.5 50 66 Grams. 43 77 70 102 Mm. 45 64 67 70 Mvi. 62 75 76 80.5 313 Grams. 70 85.4 95 100 Grams. 101 107 132 124 Mm. 74 79 72 82 Mm. 83 319 287 85 299 283 . . 83 341 282.. Table 5. — Increase in One Year op Specimens prom Lot D, Weighing under 200 Grams. Wei ght. Length. Specimen No. Weight. Length. Specimen No. June, 1910. June, 1911. June, 1910. June, 1911. June, 1910. June, 1911. June, 1910. June, 1911. 336 Grams. 120 120 125 131 134.5 Grams. 135 147 149 160 160 M7n. 79 83 83 87 89 Mm. C) 88 86 94 («) 304 Grams. 145 149 163 187 195 Grams. 163 172 182 222 225 Mm. 89 94 91 95 99 Mm. 92 317 292... 329 331 290 310 . 333 339 . 102 a No record. 14 GEOWTH AND MIGRATION OF FEESH-WATER MUSSELS. Table 6. — Increase in One Year of Specimens from Lot B, Weighing OVER 200 Grams, Selected at Random from 20 Specimens One Year from Date of Planting. Weight. Length. Specimen No. Wei ght. Length. Specimen No. June, 1910. June, 1911. June, 1910. June, 1911. June, 1910. June, 1911. June, 1910. June, 1911. 216 Grams. 193 226.5 240 250 25S.5 Grams. 215 246 263 265 270 Mm. 110 101 115 110 110 Mm. Ill 101 116.5 111 111 244 Grams. 282 291 308 323.5 378 Grams. 300 302 322 335 394 Mm. 114 119 114 115 119 Mm. 115 239 201 120 211 251 242 1-14 205 116 249 120 Table 7. — Increase in One Year of Specimens from Lot A, Weighing over 200 Grams, Selected at Random from 68 Specimens Reclaimed One Year prom Date of Planting. Weight. Lei] gth. Specimen No. Weight. Length. Specimen No. June, 1910. June, 1911. June, 1910. June, 1911. June, 1910. June, 1911. June, 1910. June, 1911. 70 Gram.t. 199.5 223.5 240 252 256.5 Grams. 216 235 260 270 274 Mm. 96 107 106 125 106 Mm. 97 108 107 125.5 107 9 Grams. 266.5 270 273.5 316 328 Grams. 277 285 304 342 340 Mm. 125 114 120 120 125 Mm. 125.5 80 5 115 10 15 120.5 12 2 120+ I 23 125.5 Averages for One Year. Table 4 Table 5 Table 6 Table 7 Average increase in weight. Grams. 27.5 24.55 16.15 17.75 Average length. Mm. 66.1 112.7 114.4 in length, i^^'ength. Mm.. 11.6 3.5 Per cent.a 17.5 3.9 .7 .6 a Per cent of aggregate increase. The specimens under 100 grams make the largest increase in weight and length; those under 200 grams, the second; and those over 200, or fully mature specimens, fall short. A pomt of special mterest in connection with tables 6 and 7 hes in the fact that while per cent of gain in length is almost negligible, weight goes steadUy forward m fairly good proportion. From averages in tables 4 and 5, an average yearly growth for a young undulata (tlu-ee-ridge) might be put at 25 grams in weight and from 5 to 25 millimeters in length; at this rate it would take about eight years to reach the weight of 200 grams and a length of 90 to 100 millimeters; as later growth is slower, about 12 to 15 years would be a fair estimate for the age of a Shoofly Creek undulata (three-ridge) of 300 grams in weight and 110 to 120 millimeters in length. GROWTH AND MIGRATION OF FRESH-WATER MUSSELS. Table 8.— Rate of Growth by Species. 15 [Note.— This table shows species not mentioned in other tabulations, because the numberof individuals was too small for comparative study. The records are useful as indicative of what may beexpected from these species. The Quadrulse here listed are given because of their use in plate figures.] Speci- men No. 109 113 111 114 281* 564* 413* 430 431 XI X H* 566* 650 495 V LI Species. A. grandis (floater). ...do .do. .do. Q. undulata (three-ridge). Q. lachrymosa (maple-leaf). .do. rV^Tien planted.. \1 year later /When planted.. U year later (Whenplanted.. \1 year later I TVlien planted.. U year later When plan ted.. 1 year and 3 months later. {'When plan ted.. 1 year and 3 months later. fMTienplanted.. \1 year later {When planted.. 1 year and 3 months later, r When planted . \1 year later When planted.. ■\3 months later.. (When planted.. )3 months later.. (When planted-. 1 year and 3 months later. ■UTien planted. . 3 months later. . /When planted.. \ 1 year later i When planted.. \1 year later tetralasmus ^{Y^:^fl^^''^- T. tuberculata (buckhorn, pistol-grip).. .do. L. gracilis (paper-shell). ....do .do. L. anodontoides (yellow sand-shell) ....do .do. Time. j„ I When planted.. ■ ™ \1 year later Weight. Length. Height. Grams. Mm. Mm. 143 120 63 193 127 66 144 120 62 192 124 65 211 132 72 269 139 74 281.5 140 82 320 142 82 16.5 40 27 51 65 41 5 22 17 20.5 41 32 8 27 22 18 42 35 128 87 56 138 90 68 470 136 76 489 136.5 76.5 11 51 26 29 69 35 13 55 27 55 82 44 22.5 63 32 84 97 50 97 93 46 116 97.5 49 125 103 48 138 106 49 180 110 51 186 111 51.5 25 65 33 30 71 36 31 70 35 43 72 37 Breadth. Mm. 41 45 41 45 48 52 53 54 16 25 10 20 13 19 31 32 50 50 14 20.5 18 28 19 33 34 35.5 37 37.5 43 43 20 22 22 23 DISCUSSION OF DATA. The results set forth iii the tables given above speak for them- selves and need no extensive explanation. One striking fact is the cessation of growth during the winter months of the Chikaskia specimens, especially in table 3. There are a number of cases in my field records where a loss is shown in we^ht for the six months from late September to early April, although in general they hold their own. It has been my observation for several years that the Chikaslda mussels, in the shallow water at least, burrow down mto the sand m late October and become abundant again in late AprU. For undulata (three-ridge), I have given rather complete averages of weight, and comparisons of rate of growth accordmg to size in tables 5-8. From tables 1 and 2 (all lot D), I find the averages given below for the four species named: Undulata (three-ridge), 12 speci- mens; lachrymosa (maple-leaf), 7 specimens; pustulosa (warty- back, pimple-back), 8 specimens; ruhiginosa, 5 specimens. From 16 GROWTH AND MIGRATION OF FRESH-WATER MUSSELS. table 8, grandis (floater), 4 specimens; tetralasmus, 2 specimens; gracilis (paper-shell), 2 specimens; anodontoides (yellow sand-shell), 1 specimen. Average Increases in Weight and Length, by Species. AVERAGES FOR ONE YEAR. Species. Average Average increase Average weight. in length. weight. Grams. Grams. Mm. 100 30.8 70 '124 28 82 74 11.4 68.8 67 11.5 60 194 48.6 128 28 8.5 67 Average Increase in length. Lachrymosa (maple-leaf) Unduiata (three-ridge) Rubiginosa Pustulosa (warty-back or pimple-back) . Grandis o (floater) Tetralasmus b Mm, l.h 7.2 3.4 2.7 5 4 AVERAGES FOR THREE SUMMER MONTHS. Gracilis (paper-shell) AnodontoidTes (yellow sand-shell) . 22.5 4.5 o Large mature specimens, still the lncrea.se is good. b In an unusual environment lor this species (p. 9). Juvenile Quadrulae " of the above species double in size in a year, as shown by a number of examples (149, 158, 349, 323, in table 1; 281, 564, 413, in table 8). Averages here given and those taken from other sources indicate that a 100-gram, 75-millimeter unduiata (three-ridge) (p. 14) or lachrymosa (maple-leaf) can develop in the Shoofly in about four years (averages for tables 1, 2, 4, and 5). A 300-gram specimen of unduiata (three-ridge) or lachrymosa (maple- leaf) would doubtless be close to 15 years old. The largest unduiata (three-ridge) rechecked weighed 407 grams in June, 1910, and 421 in June, 1911. This specimen was 123 millimeters long, and increased 1 millimeter in length. In unduiata (three-ridge) and lachrymosa (maple-leaf), after the 100-millimeter length is reached, the increase in length is slow, and growth Imes follow one another so closely that the differentiation of lines is difficult. Pustulosa (warty-back, pimple- back) and rubiginosa grow more slowly than the two preceding species. A 50-millimeter 'pustulosa (warty-back, pimple-back) has passed the age of rapid growth, and from this size on additions come slowly. A light-shelled form, as L. gracilis (paper-shell) grows very rapidly; this would seem to indicate that the shell is built up at greater expense of food and energy than the soft parts of the mussel. In table 8, specimen X, during the three summer months, shows an increase of over four times in weight and 27 milUmeters in length; 566, L. anodontoides (yellow sand-shell) , in the same time, makes an increase of one-fifth its original weight, and 4.5 millimeters increase in length. a Early juvenile forms grow even more rapidly as experimentally found (foot-note, p. 5) by Coker. GROWTH AND MIGEATION OF FRESH-WATER MUSSELS. 17 ARRESTED GROWTH RINGS. "Growth lines," the conspicuous dark concentric rings of the shell, may be due to (1) thick epidermis, (2) double epidermis, and espe- cially (3) to double epidermal and prismatic layers. It is well known that the epidermal and prismatic layers are formed by the .-"•'^ Fig. 1-4.— Cross-sect ions of shells of Q?;a(?rM?a showing structure inregion of rest rings, all figuresenlarged 20 times; e, epidermis; p, prismaticlayer; n, nacreous layers; m', successive layers of nacreous structure; g, position and width of rest rings (growth lines). Fig. 1. — Section from near edge of shell, showing double layers and long underlying tongue of epidermal and prismatic structure, which formed the dark wide rest ring of a young rapid-growing Qua'^rula lachry- mosa. Rings 1 1 or i -, specimens 149 or 15S, plate n, would give similar sections. Fig. 2.— Section of shell of mature mussel, taken about the middle of the valve. Rest ring not so wide as in 1, tongue shorter, prismatic layer thicker, and thick successive layers of nacreous struciure. Fig. 3.— a rest ring due chiefly to unusually thick, double layers of epidermis. Fig. 4. — Two successive rest rings near together, undoubtedly formed in the same season, and probably only a few weeks apart, as specimen was young and at the rapid-growing age. (See description of specimen 413, pi. n.) edge of the mantle only. Thus increase in shell diameter begins with the formation of the epidermal, followed by the prismatic layer. However, if the mantle is withdrawn from the edge of the shell, it often puts down new layers of epidermal and prismatic material 18 GKOWTH AND MIGRATION OF FEESH-WATER MUSSELS. underneath older layers "^ of the same structure, and sometimes underneath older layers of nacre as well. This is well shown in figures 1-4, showing sections of shells made through the growth Imes. In watching the growth of tagged specimens it was noted that a distinct growth line was formed at the time of tagging; the work of tagging, i. e., the drilling of the hole through the shell and the placing of the wire, caused considerable irritation, probably a strong con- traction and partial breaking loose of the mantle from the edge of the shell, and, as a result, new epidermal and prismatic layers were put down underneath those already formed, as the mantle worked to its old position, and formed new layers over the wire holding the tag (pi. Ill, 763) . Young specimens, especially, show a conspicuous and well-defined ring passing through the outer side of the wire holding the tag. The so-caUed annual rings had better be called "arrested growth or rest rings," as they represent retarded growth, which may be very temporary, as in the case of the tagged specimens, and still leave a very marked ring. Ordinarily the prominent rest rings are presumably winter rings, ^ representing delayed growth, due to inactivity, a withdrawal of the mantle from the extreme edge of the valves, and the forming of double epidermal and prismatic layers as a result of renewed active growth in the following spring. Other rings may follow arrested growth, due to various unfavorable con- ditions that may arise in the life of the mussel, such as water shrinkage, temporary stranding due to migration, especially at flood periods; in the fighter species perhaps washing at flood times. It is possible that in certain pond forms, as V. tetralasmus and others that five in ponds that go dry for short periods during the summer season, the more prominent rings are summer rings.^ That the concentric rings are by no means dependable as abso- lute annual rings is well shown in many specimens under observation in this investigation. A few specimens shown photographically (pi. i, ii, iii) will clear up some points in regard to growth. Rest rings are not always brought out clearly by photographic methods on account of the unequal or convex surface of the sheU. A sfight ridge that may have no connection with a rest ring wiU show as a shadow line beyond the ridge. This is shown in photograph H, plate i, in the line marked "o." Additional explanations are given with the plate figures. a This explanation was first suggested to me by Dr. Coker, and later verified in connection with my study of shell sections through arrested growth lines. 6 While in this paper we have emphasized the point that " annual rings ' ' are not annual rings absolutely, the statement that the prominent rest rings are usually, under stable enviromnental conditions, winter rings is clearly within the evidence of this investigation. e Live specimens of this species were plowed up in the Browne pond (p. 7) three months after the pond had gone dry. GROWTH AND MIGBATION OF FRESH-WATER MUSSELS. 19 MIGRATION. No end of speculation has been carried on as to the traveling ability of mussels. The long undulation tracks, often found upon the pond or stream floor, together with other field observations, and the active movements of specimens kept in aquaria have afforded data for discussion. While gathering information concerning growth, I have constantly kept in mind the migration question, as it was easy to carry on the two together. As already indicated, the main reason for tagging a large number of mussels was the feeling that many would be lost through migration; and further to guard agamst this migration in the Chikaskia I inclosed the specimens in good-sized pens. SHOOFLY. The extent of actual ttiigration is best shown in considering definite plantings. Lot D (see p. 8) of the Shoofly is good for tliis purpose. The 168 Quadrulas planted here were from the Chikaskia, where they were collected from the sand bars in shallow water. From track-mark evidence these specimens had been actively moving about on the sandy bottom, stimulated to activity by unfavorable environmental conditions. The 164 specimens were spread out on a small portion of the bottom of the Shoofly (p. 8), June 16, 1910, and left free to move. June 14 and 15, 1911, I reclaimed 139 of these specimens, or 84.8 per cent, in about three hours' work. Twelve of these specimens were reserved for records. On September 11, 1911, a cold rainy day, I again checked over lot D, and this time I secured 93 specimens in about one hour's time; the water was so cold that collecting was exceedingly difficult. Three specimens not found in June were found on this date, bringing the total number reclaimed from the original planting up to 142. When we consider that the water was cloudy and from 3 to 4 feet deep the experienced field collector wiU know that specimens could not be recovered in these numbers unless they were on the very spot ^ where they were planted. Lots A and B in the Shoofly gave similar results, although I did not attempt to recover these as thoroughly as in the case of lot D. Lot C in the Shoofly is of especial interest, as these specimens were all A. grandis (floater). Of this lot 12 specimens were found directly on the site where they were planted. Wliile this is rather a small per cent in comparison with lot D, I was surprised to find 46 per cent of A. grandis (floater), as it is well known to be an active a It has been suggested that smce lot D specimens were transplanted the inactivity may have been due to the changed environment. The relatively rapid growth and Uke inactivity of lots A and B, Shoofly specimens, which might very well be considered as control lots, should quiet any apprehensions on this point. 20 GROWTH AND MIGRATION OF FRESH-WATER MUSSELS. form. In all my reclaiming work in the Shoofly, which amounted to about 10 hours' actual hunting and collecting on three different days, I secured only one specimen that may be said to be off the planting plot; this was no. 141, found 15 feet from plot D. CHIKASKIA. In the Chikaskia lot E gave results similar to D. The Chikaskia Kiver specimens were disturbed somewhat by curious people, and for this reason proportional figures can not be emphasized. The inclosure for lot E (p. 9) was rather small, but there was ample chance for movement. The results in this instance substantiate what I have often noticed m field work, namely, that the mussels, especially Quadrulae and related species, are unable to help themselves if con- ditions become unfavorable, but, on the other hand, the power to endure these unfavorable conditions is remarkable. Since the Chikaskia is a fairly swift stream, the lower end of the inclosure was undermined, making escape easy, while a sand bar was formed across the upper portion. Two-thirds of the mussels were caught in the drift bar and when I examined them on September 26, 1910, were helplessly stranded. Conditions of oxygen and food sup- ply must have been unfavorable, and as a direct result the summer's growth was below the average for lot F. These stranded specimens were now taken and put in the outer corner of the pen, where they could escape under the whe net through an opening 5 feet long. Six months later, June 22, 1911, I again examined lot E. A large number of the specimens were recovered, about half being along the wire net inside, and half along the net outside; not a specimen was found over 5 yards from the pen. Clear water and sandy bottom made the fhiding of specimens easy. Lot F spechnens of the Esch inclosure (p. 9) had a good opportu- nity for migration; in fact, the inclosure was not needed, as only two specimens reached either lower or upper cross fences. One of these was a gracilis (paper-shell), and the other a grandis (floater). The Quadrulae did not come within 25 yards of the lower fence. If they were placed m water over 3 feet deep, the migration was slight in any case, as far as the Quadrulae were concerned. Those placed in water as shallow as 1 foot moved to deeper water, which was easily reached in this case. The Lampsiles were more active, and the per- centage recovered was small by comparison. Of the 50 U. tetralas- mus, not a single specimen or shell was found at the first examination, September, 1910; but in the June, 1911, examination, three speci- mens were found. I am not able to explam the disappearance of the tetralasmus; however, they are great burro wers, and may have escaped my extensive digging for them. GBOWTH AND MIGRATION" OF FEESH-WATEB MUSSELS. 21 DISCUSSION OF MIGRATION DATA. The migration results came as a surprise to me. The very fact that I was willing to risk specimens free in the Shoofly would indicate that I hoped to make some kind of recovery; but to go back and find specimens by the score — apparently in the exact spot where they were planted — was not to be expected. The Quadrulse m these plantings show little migration; the Shoofly specimens may be said to be nearly stationary in water over 3 feet deep. Those placed in shallow water in the Chikaskia always moved until water 2 to 3 feet deep was found. Specimens found on shoals and bars in nature are there by chance distribution, not choice, although breeding reactions may cause migra- tion m some species. ECONOMIC BEARING OF EXPERIMENTS— RELATED PROBLEMS. While the scientific interest in the growth and migration problems was the real motive that prompted this investigation, it was the relation of the problem to the practical question of artificial propa- gation of mussels for commercial purposes that made funds available to carry on the work. That the results will be of service as preluni- nary to further investigation is a matter of satisfaction. The man interested in commercial propagation will continue to ask the ques- tion, How long will it take to grow a mussel to marketable size? Much more work will be needed to get at all the facts in the case of the various economic species; and where the work of propagation is to be conducted on a large scale, preliminary experiments will be of value in testing the fitness of a particular region for commercial opera- tions. Some of the methods of procedure have been indicated by the series of experiments here ou timed. I wish to call the attention of the field students of the Unionidse to the transplanting of adult mussels from the Chikaskia to the Shoo- fly, where young specimens were not to be found, yet these trans- planted mussels averaged higher in rate of growth than the regular Chikaskia specimens. If the Shoofly is so favorable a habitat for mussels, why are young specimens absent from the beds ? Again, notice the transplanting of over a hundred specimens from the vicmity of lot F, where mussels are so abundant, to lot E : The transplanted specimens did as well in growth as the specimens that were near the region of abundant mussels in the Chikaskia. This difference in abundance was by no means slight. It is a fact that in the vicinity of lot F one could, in a few hours, collect a wagonload of mussels, while for a quarter of a mile above and below the Brewer inclosure two of us were able to get only 30 specimens in several hours' careful collectmg. 22 GEOWTH AND MIGRATION OP FRESH-WATEE MUSSELS. The distribution of mussels within individual streams, and in the Chikaskia in particular, is not easy to explain. Little or negligible migration among the Quadrulse, at least, has not cleared the situation. SUMMARY. 1. Rate of growth is exceedingly variable for individuals of a single species in the same stream and in different streams, depending, as in other invertebrates, upon season, food, oxygen supply, and other con- ditions. Juvenile mussels grow much more rapidly than adult or near-adult individuals. Lampsilis species grow very much faster than Quadrulse. Specimens in stable conditions seem to have a fairly definite rate of growth from year to year. The rate, after sexual maturity, is slowed down, but growth goes on steadily, though the proportional increase in length is so slow as to make appreciable additions very slight, so that growth lines in Quadrulse, after a size of 100 millimeters (4 inches) has been reached, can not be ascer- tained by inspection. 2. From April to September may be designated as growth months, most specimens showing very slight increase during winter. 3. Lmes of arrested growth may be called rest rings, the conspicu- ous ones being usually winter rest rings; very often, however, the rings may be two or more years apart, or several equally prominent rings may be formed in one year. Prominent lines are generally due to double prismatic and double epidermal layers. Winter rings, es- pecially where environmental conditions are stable, are usually suffi- ciently regular for use as indicators of age in estimating roughly the time required for a commercial species to reach marketable size. 4. Under favorable conditions there is little migration among the Quadrulae. Some of the Lampsiles and other light-shelled species move about quite actively, but probably seldom migrate far from the point where they were dropped from the fish, although their total wan- derings may be considerable. Water of sufficient depth is essential to optimum conditions. The minimum depth seems to be at least 2 feet; the range up to the maximum has not been studied. Quadrulse prefer water over 2 feet deep. The reason they are found on the shoals in many of our streams may be explained as chance distribu- tion, due probably to the fact that the particular individuals have never found optimum conditions after their parasitic development, and the reason they move about is that they do not find the right environment. LITERATURE CITED. « CoKER, R. E., and Surber, T. 1911. A note on the metamorphosis of the mussel Lampsilis Isevissimus. Bio- logical Bulletin of the Marine Biological Laboratory, Woods Hole, Mass., vol. 20, p. 179-182, 1 pi. a Dr. A. E. Ortmann and Dr. F. Haas have called my attention to Dr. Haas's monograph, "Die Union- idae in Martini-Chemnitches Conchyliencabinet," 1910. I have not had access to this paper. GROWTH AND MIGRATION OF 'ATER MUSSELS. 23 ISELY, F. B. 1911. Preliminary note on the ecology of t^ Ibid., vol. XX, p. 77-80. Israel, W. 1909. tJber die Najadeen des Mittelelbeg^ Gesellschaft von Freunden der p. 1-38. 1910. Die Najadeen des Weidagebeites. poisclien Najaden. Beilage zumj zoologischen Gesellschaft, p. 49-56 1911. Najadologische Miscellen. NachricJ gischen Gesellschaft, p. 10-17. Lefevre, G., and Curtis, W. C. 1908. Experiments in the artificial prop| ceedings of the Fourth Internat Bulletin of the Bureau of Fisherij 1910a. Reproduction and parasitism in the Zoology, vol. 9, p. 79-115, 5 pi. 1910b. Studies on the reproduction and art sels. Bulletin of the Bureau of VI-XVII. EXPLANATION Ot All photographs are about natural size. Fc already explained (pp. 10, 11, and 15), tags wire passed through a small hole made about ally along the edge of the posterior umboidal the ring formed when the specimen was formed without exception in all specimens h^ the mussel at the time of tagging and making natural processes between tagging time and due to a ridge on the valve and in photograpl Plate I 3 Specimen H, Lampsilis gracilis (table 8), sho'v (June, 1910, to Sept., 1911); x marks the rest' marked and measured; a;' is probably the 1910-11^ Specimen 323, Quadrula undiilata (table 1), sli growth (June, 1910, to June, 1911). A new ring.; the two rings, y^ and y'^, near together, formed Specimen 281, Quadrula undulata (table 8)| (June, 1910, to Sept., 1911); x^ is probably the Plate I] Specimen 355, Quadrula rubiginosa (table (June, 1910, to Sept., 1911) in a mature slow- rings beyond x. Specimen 349, Quadrula pu^tulosa (table 1 (June, 1910, to June, 1911). , A rest ring, not shows very near the margin; o, on the anteriori Specimen 564, Quadrula lachrymosa (table 8 15 months (June, 1910, to Sept., 1911). Tagpul! Specimen 413 (table 8), shows 16 mm. incr| 1911). Two rest rings, x^, x^, in addition to the If we count the one at the tagging line we have t3ii livenile life of the Unionidae. jl u. 52. Jahresberichte der bnschaften in Gera, Reuss., ir Kenntniss der Mitteleuro- sblatt der deutschen malako- ler deutschen malakozoolo- fresh-water mussels. Pro- lery Congress (Washington), cvui, p. 617-626. Journal of Experimental Jpagation of fresh-water mus- vol. XXX, p. 105-201, pi. is. leasurements see tables. As led on with a small copper the edge of the shell, usu- larks the tagging ring, i. e., measured. This ring was thus marks the diameter of 3rds; x', rest rings formed by [ot specimen; o, shadow line • be mistaken for a rest ring. rowth in length in 15 months led when the specimen was Br rest ring; o, shadow line. "17 mm. increase in one year's jt z' near the margin. Notice specimen was tagged. mm. growth in 15 months fcst ring. mm. increase in 15 months jecies. No evidence of rest lO mm. increase in one year ^ght out by the plate figure, shadow line. [9 mm. increase in length in ifter specimen was reclaimed, year (June, 1910, to June, BJIformed at the time of tagging, rest rings for one year. 24 GEOWTH AND )N OF FEESH-WATER MUSSELS. Specimens 149 and 158, Quaa thirds natural size). These apf from the Chikaskia. The rest ri 1910, to June, 1911) is the same. Specimen 158 is the left Aalvji Specimen 566, Lampsilis cnoc nearly three months (Jun men growing in rather uiifci Specimen 763, Quadrula lad (t) nearly three months after [ Specimen 200, Quadrula 193 grams was found in Jun( be called young. No. 200 if larity of rest rings, four in ni according to estimates workec was 4 or 5 years old when tal ichrymosa (table 1 and as shown here about two- • are from the Shoofly, the two lachrymosa above ■ are about the same as 413 and the time (June, , 80 does not show the tag or mark. f ^Plate III. ii ft'.s (table 8) , shows 5 mm. increase in length in i lo Sept. 19, 1910). A mature Chikaskia speci- ■ feViAdronment. hows how the tag wire is overlaid with nacre ; [^une 27, 1910, to Sept. 19, 1910). In the Shoofly (p. 7) no Q. undulata under i in 1911 two specimens were secured that may ese and is interesting on account of the regu- ' he specimen measures 80 mm. in length and, , n kno'H'n growth of this fspecies in the Shoofly, , o ,31 ^-,.- - U. S. B. F.— Doc. 792 Plate I. SPECIMEN Hi' SPECIMEN 281. U. S. B. F Doc. 792. Plate II. SPECIMEN 355. SPECIMEN 349. SPECIMEN 149. SPECIMEN 564. SPECIMEN 158. SPECIMEN 413. U. S. B. F.— Doc. 792. Plate III. SPECIMEN 566. SPECIMEN 763. SPECIMEN 200. LIBRARY OF CONGRESS 002 903 626 3 ■■M