■(/ SURFACE-CURRENT STUDIES OF SAGINAW BAY AND LAKE HURON, 1956 SPECIAL SCIENTIFIC REPORT-FISHERIES Na 267 ';!^'> UNITED STATES DEPARTMENT OF THE INTERIOR FISH AND WILDLIFE SERVICE EXPLANATORY NOTE The series embodies results of investigations, usually of restricted scope, intended to aid or direct management or utilization practices and as guides for administrative or legislative action. It is Issued in limited quantities for Official use of Federal, State or cooperating agencies and in processed form for economy and to avoid delay in publication . United States Department of the Interior, Fred A, Seaton, Secretary Fish and Wildlife Service, Arnie J. Suomela, Commissioner SURFACE-CURRENT STUDIES OF SAGINAW BAY AND lAKE HURON, 1956 by James H. Johnson Fishery Research Biologist Great Lakes Fishery Investigations Bureau of Commercial Fisheries Special Scientific Report — Fisheries No, 267 Washington, D, C. December 1958 The Library of Congress catalogue card for this publication is as follows: Johnson, James H Surfacc-cunciit studies of Saginaw Bay and Lake Huron, 1956. AVashington, U. S. Dept. of the Interior, Fish and AVildlife Service, 1958. 84 p. maps, diagrs., tables. 27 cm. (Special scientific report — fislieries, no. 267) Bibliography : p. 27-30. 1. Saginaw Bay. 2. Huron, Lake. i. Title. (Series: U. S. Fish and Wildlife Service. Special scientific report : fisheries, no. 267) [SH11.A335 no. 267] Int 59-12 U. S. Dept. of the Interior. Library for Library of Congress The Fish and Wildlife Service series, Special Scientific Report — Fisheries, is catalogued as follows: U. S. Fish and Wildlife Service. Special scientific report : fisheries, no. 1- [Washington, 1949- no. illus., maps, diagrs. 27 cm. Supersedes in part the Service's Special scientific report. 1. Fisheries — Research. SH11.A335 639.2072 59-60217 Library of Congress Table of Contents Page Drift-bottle designs 2 History 2 Drift bottles released in Saginaw Bay and Lake Huron . . 4 Releases and recoveries 7 Wind data 10 Water movements of Saginaw Bay 11 General features of the bay 11 Previous studies 12 Drift-bottle movements in 1956 12 Water movements of Lake Huron 16 Previous studies 16 Drift-bottle movements in 1956 17 Dynamic heights 18 Appraisal of the general pattern 19 Rate of drift 21 Quantitative relationship between surface drift and wind ... 24 Factors influencing drift-bottle movements 25 Saginaw Bay 25 Lake Huron 2 6 Recommendations for study of currents on the Great Lakes ... 26 Acknowledgments 27 Summary 27 Literature cited 27 Appendix 30 ABSTRACT Surface currents of the waters of Saginaw Bay and lower Lake Huron area were studied in the summer and fall of 1956. Drift bottles were used in Saginaw Bay and drift bottles together with the dynamic-height method were used in Lake Huron. A total of 2,650 drift bottles were released; 1,843 (69.5 percent) reply cards from the recovered bottles were returned. Correlation appeared to be high in Saginaw Bay between direction of surface currents that moved these bottles and direction of winds. In Lake Huron this correlation was less apparent, although the drift of bottles was generally from west to east, seemingly under the influence of the prevailing westerly winds of this area. figure 1.— Lake Huron and Saginaw Bay. SURFACE- CURRENT STUDIES OF SAGINAW BAY AND LAKE HURON, 1956 It is increasingly apparent to students of aquatic biology that water currents and movements of large water masses play a para- mount role in the life cycle of many aquatic organisms. Water movements influence the distribution and ultimate survival of eggs, larvae, and adult aquatic organisms. The relation is direct when organisms are car- ried along actively by currents, and indirect when currents cause changes in environmental factors such as temperature, salinity, and other physical and chemical conditions. Although the direct effects are more obvious, the indirect influences may play an impor- tant role in survival and distribution of aquatic life. At times movements of water masses can alter so adversely the habitat of fishes, especially those that live in a narrow environmental range, that kills of catastro- phic proportions occur. The disaster which overtook the tilefish off the northeastern coast of the United States in 1882 was due to a sudden but temporary flooding of cold polar water into the warmer waters normally inhabited by this fish (Bigelow and Welsh 1925). An estimated 1 1/2 billion dead fish were sighted on the surface waters shortly after this calamitous event. Not only was fish life affected but certain invertebrates were exterminated by the cold mass of water. Outbreaks of red tide off the western coast of Florida are probably initiated by water masses which differ in salinity and chemical characteristics from the normal water off the Florida coast (Slobodkin 1953). Slobodkin believed that prediction of red tides would depend on more detailed knowledge of coastal drainage and hydrography, and that prevention of red tides may be possible, to some extent, by altering certain coastal drainage patterns. The adverse effects of movements of water masses upon fish populations are probably matched by an equal number of favorable incidents. The 1904 year class of herring in the North Sea dominated the commercial herring fishery in that area from 1908 to 1919. This year class was prominent for other species also. Evidence indicated that the success of this year's hatch was due to an abnormally intense inflow of Atlantic water into the North Sea that carried with it either an abundance of the actual food required by newly-hatched fishes or provided certain nutrient salts resulting in a high abundance of basic food organisms (Tait 1952). According to Tait , there seems little room for doubt that the essential causes of fishery fluctuations lie in hydrographic conditions and that adequate observations of these conditions affords the surest means of anticipating these fluctuations. He perceived that the rela- tionship of hydrography to fisheries is analogous to that of meteorology to agricul- ture. In recent years, certain commercial fisheries of Saginaw Bay, Michigan, have deteriorated at an alarming rate. The annual commercial catch of the walleye (Stizostedion v. vitreum) has decreased in the last decade to such an extent that the economy of this fishery has been greatly weakened (Hile 1954). Many fishermen attri- bute the scarcity of the walleye in the bay to pollution. Production of lake herring (Leucichthys artedi) and whitef ish (Corego- nus clupeziformis) is also low. The yellow perch (Perca f lavescens) , on the other hand, are at such a high level in numbers that their growth is stunted (El-Zarka 1958). A fishery survey conducted with the study upon which this report is based revealed an abundance of alewives (Pomolobus pseudo- harengus) and smelt (Osmerus mordax) but there is little commercial production of these species. The U. S. Fish md Wildlife Service and the Michigan Department of Conservation conducted a cooperative limnological survey in Saginaw Bay and adjacent Lake Huron waters in the summer and fall of 1956 (fig. 1). The objectives of the study were to gain basic information on species composi- tion and species inter- and intra- relation- ships, and to develop the possible causes of the fluctuations in the Saginaw Bay fisheries. As part of this project drift bottles were used to obtain information on the current systems and to determine the amount of water interchange between the bay jind lake. This report analyzes the drift bottle movements in Saginaw Bay and Lake Huron and summarizes briefly the use of drift bottles by other workers. DRIFT-BOTTLE DESIGNS History Among the first recorded accounts of the use of drift bottles is that of Bernar- din De Saint-Pierre who in 1784 recommended releasing floating bottles from time to time with each bottle carrying a note telling the day, latitude, and longitude of release (Rouch 1954). A French naturalist, Aime', shortly before the middle of the nineteenth century released 50 bottles off the Algerian coast and subsequently obtained returns of 3 of them (Schmidt 1913). An interesting early record of drift bottles is that reported by Prince Albert I of Monaco; a bottle released toward the end of the nine- teenth century in the Atlantic Ocean south- west of Ireland was recovered 646 days later on the coast of Tunis in the Mediterranean Sea (Schmidt 1913). Evolution of current indicators since these experiments have been along two main lines: indicators of a stationary mechani- cal type that measure the current at a certain point; and passive objects carried along by currents. Since this project utilized the latter type, the following report is limited to passive drift units. In 1892, 1893, and 1894 masters of some merchant vessels released nearly 5,000 drift bottles at various points throughout the Great Lakes (Harrington 1895). These bottles had no ballast or drags and each contained a reply card. Although the bot- tles floated low in the water, enough Wcis exposed above the surface to cause Harring- ton to remark that the wind may have influ- enced their movements. He reasoned that bias to the results wjis not impxartant since the wind that drifted the bottles would move the surface water in the same direction. He did feel, however, that wind caused the bottles to drift faster than the water but that the effect was slight. From the use of plain stoppered bottles, it was but a short step to reduce the wind effect by the inclusion of a ballast to make the bottles float with only a small portion exposed above the surface. Garstang (1898) used ballasted "egg-shaped" soda-water bottles on the English Channel. They were 9 inches long and the upper half was painted red to make them conspicuous. Ballast con- sisted of lead shot held stationairy in the bottle by paraffin to minimize displacement of the center of gravity. He concluded that movements of ballasted bottles were princi- pally due to the force that local winds exerted upon the surface of the water, sub- ject, in certain areas, to modification by tidal currents. Drift bottles used in the Danish Ocea- nographical Expeditions to the Mediterranean Sea in 1908-1910 were ordinary champagne bottles, well corked, with the mouth dipped in pitch (Schmidt 1913). Some bottles were ballasted with sand; others had no ballast. Any difference in travel between bottles with and without ballast was not given in the results of the experiment. Schmidt believed that the wind had considerable direct effect upon the bottles. Platania (1923) in a further report on the Danish Expeditions concluded that drift bottle movements in the western Mediterranean did not reflect true currents, but were influ- enced primarily by the prevailing winds. The travel of well-designed surface floaters, when interpreted properly, gives reasonably reliable information on surface currents at a particular time and place. Investigators, however, are frequently interested also in subsurface currents. To obtain information on subsurface currents. Bidder (Carruthers 1927) developed a "bottom trailer" bottle for use in the North Sea. The bottom trailer used by Nelson (1922) was a stoppered glass bottle, the neck of which carried a straight wire tail pointing in the direction of the long axis of the bottle. The bottle was weighted to have a small negative buoyancy in sea water. When released it sank to the bottom. Its descent stopped as the tip of the tail touched the bottom, and it drifted with the current in that position. The weight of the bottom trailer was adjusted so that the bottle weighed 1.7 grams more than the volume of sea water (at 8" C. , specific gravity 1.0275) displaced by it. These bottles were expected to become entangled within the nets and trawls of fishermen and the reply cards subsequently to be returned by them. Carruthers (1947) reported that suc- cess in the use of "bottom trailers" had not been great because they "take sanctuary" between the sand ridges. Various other methods have been used to reduce exposure of drift bottles to winds. Gilson (Carruthers 1930) experi- mented with coupled systems consisting of pairs of bottles, one bottle of the pair with positive buoyancy and the other of negative buoyancy, linked together by cords 3 meters long. In his experiments in the North Sea he noted a marked difference in rate of travel between simple surface float- ing bottles and his coupled systems. On occasion simple floating bottles and coupled bottles released at the same time traveled significantly different routes and direc- tions. A similar coupled system was described by Sverdrup, Johnson, and Fleming (1942) in which the lower of the two coupled bottles contained a weak acid which in time corrodes a metal stopper, thus permitting sea water to fill the bottle and sink it. Fishermen are depended upon to return reply cards from bottles that become entangled in their nets. (ballasted and unballasted) , drag-fitted bottles, and coupled bottles systems. Drag-fitted bottles were of two kinds: those with a metal drag suspended from a surface-floating bottle by a wire 3 feet long; those with a 9- inch-high, 6- inch- diameter toffee tin suspended by a 3-foot wire from a surface-floating bottle. The toffee tin contained a drift bottle and both the surface-floating bottle and the one in the tin contained reply cards. The coupled system consisted of two bottles, a bottle of negative buoyancy suspended from a surface-floating bottle by a 3-foot piece of stout sash cord. Carruthers found that in some instances unballasted surface floaters, ballasted surface floaters, and drag-fitted bottles put out at the same time and place showed significantly dif- ferent movements. Only two replies were received from the coupled system — not enough to allow a valid compjirison with the other returns . One of the most popular methods of reducing direct influence of the wind is by using a metal drag that is suspended from the drift bottle by a length of wire. This arrangement not only reduces the surface area exposed to wind but also causes travels of the bottle to be affected by currents between the surface and the depth of the drag. Length of suspending wire ceui vary but most workers have used a wire in the neighborhood of 3 feet long. Mavor (1922), however, (Bay of Fundy) used drags suspended by wire 5.5 meters long. Webster and Buller (1950) in studying ocean currents off the New Jersey coast used both free bottles and bottles with drags suspended by a 4-foot wire. Their bottles released with drags attained a greater speed of transport than those without; furthermore, prevailing winds had little effect upon the direction of drift. Deason (1932) who released bottles with a drag suspended 3 feet below the bot- tle (Lake Michigan) concluded that the action of the prevailing westerly winds had much to do with the rate of and direction of surface currents. In Hudson Bay experi- ments, Hachey (1935) used a 3-foot galva- nized wire to suspend a metal drag. He made no remarks concerning the circulation of waters other than the general circulation seemed to be counterclockwise. Carruthers (1930) experimented with drift bottles on the North Sea to ascertain difference in the travels of surface floaters In recent years drift cards in pis tic envelopes, as developed by Olson (1951), have been looked upon with favor by some as a substitute for drift bottles. Olson used a polyethylene envelope 0.004 inch thick with the return card hermetically sealed within. His Lake Erie experiments indicated that the travels of these envelopes were not at the complete mercy of the wind. Some cards that were returned 18 months after release were still in good condition. In remarks on Olson's work, Verber (1953) wrote that drift cards were better than drift bottles since the cards are inexpensive and give greater accuracy in interpreting the surface flow because they are not exposed to the wind. He concluded that Olson's work proved a direct correlation between wind and surface flow in western Lake Erie and that the movements of surface water were wind controlled. In Georgian Bay, Lake Huron, 3,000 drift cards similar to those developed by Olson (1951) were dropped from an airplcuie (Fry 1956). Polyethylene material, however, was only 0.002 inch thick and proved to be only moderately satisfactory because pin- holes developed in the plastic from sand abrasion. Drift cards were used by the Fish and Wildlife Service on Lake Superior in 1953 Eind on Lake Michigan in 1954. Few returns were obtained from these releases, and of those cards returned many were found water-soaked inside the polyethylene enve- lope. The polyethylene envelope was in some cases not sealed properly; and since a number of cards were found on the bottom off shore, it is possible that many of them sank before they reached land. Bougis and Ruivo (1953) added ballast to the polyethylene-envelope type of float. Their "siphonophone" consisted of three parts: a poleth- ylene envelope of 0.004 inch thick- ness that floats on the surface; the reply card within the envelope; and a drift with ballast. The drift was a ribbon of polyethylene 1.2 meters long and 8 centimeters wide attached to the polyethylene float. About 20 grams of lead ballast were placed at the lower extremity of the ribbon to make the drift sink into the water. Experiments on the Bay Banyuls, with drift cards only and "siphonophore" drifts, proved that the former fol- lowed the course of the wind closely, whereas the latter traveled at vari- ous angles with the wind and also moved much more slowly. Drift bottles released in Saginaw Bay and Lake Huron The drift bottle with metal drag was the design chosen for the Saginaw Bay-Lake Huron study in 1956 (fig. 2). It WcLS evident from work by the Fish and Wildlife Service on Lake Michigan in 1955 that bottles with drags resisted direct effects of wind and presumably gave a better indication of water currents near the surface than did the ballasted bottles. Plastic envelopes were rejected for reasons already noted. A disadvantage of a bottle with the drag suspended several feet below the bottle is that the drag hits bottom in the surf zone and resists being washed ashore by the small waves characteristics of the Great Lakes. Observations have revealed that these bottles can be carried many miles in the surf, sometimes against the prevailing offshore current, before they are washed ashore. This disadvantage was elim- inated, for the most part, by suspending the drag only 1 foot beneath the bottle instead of the usual 3 to 4 feet. The change was made after repeated tests with dye markers showed no discernable dif- ference, under ordinary conditions, in water movement between 1 and 4 feet below the surface. Water level Reply card Brass ring m. - P^^^ ron suspension wire Metal drag Figure 2. — Drift bottle as it appeared at time of release. Great Lakes Fishery Investigations Ann Arbor, Michigan, U. S. A. NOTICE TO FINDER Drift Card No This card is being used to study currents of the Great Lakes. Please fill in blank spaces. Mail every card you find. Canadian postage will be replaced. You will be told the time and place this card was released. Thank you. a. ni. -. p. in. Time of recovery: Date Hour . Was a metal fin attached to the bottle? Cj Yes; LJ No Exact location card was found -_ CO "1 ft s o n n se D 01 >1 e n ^* o D Near Remarks: (City) (County) (State) Please print : (Name) (Address) U. S GOVERNMENT PRINTING OFFICE 16 — 71459-1 Jigure 3. — Reply card similar to that placed in each bottle. Reply cards (fig. 3) were placed in 4-ounce Boston-round bottles after which the bottles were stoppered with corks and the stoppered ends dipped into beeswax. Several drops of beeswax were then placed in the bottle caps and the caps screwed on the bottle. Drags were squares of 28-gage galvanized metal (4" X 4") so cut arid bent that water movement from any direction struck areas of the three planes of the drag. During periods of high winds cind heavy seas the horizontal fin of the dreig would in- hibit vertical movement so the bottle would be under the surface much of the time. Studies on Ljike Michigan in 1955 disclosed that many bottles lost their drags before they were washed eishore. The comments of finders indicated that the drags were lost because the soft iron suspension wire was broken at the neck of the bottle. The break was caused, most likely, by the bending of the wire as the bottle was moved by the waves. In the Saginaw Bay-Lake Huron project, loss of drags was reduced by placing a brass ring in the suspension wire at the neck of the bottle (fig. 2). The bottle could then move freely without bending the wire. As satis- factory as this arrangement proved to be, it did not end loss of drags. Of 1,076 bottles recovered within 29 days after release, only 18 (1.7 percent) had lost their drags. Of 523 bottles recovered after more than 29 days, 168 (33.3 percent) had lost their drags (table 1). Undoubtedly some of the 523 bottles recovered after Table 1. — Loss of drags from drift bottles released in 1956 in relation to number of days between release ajid recovery Number of days ' Total number Bottles for which Bottles that had lost drags and recovery recovered on drags Number Percentage 0- 9 510 8 4 0.8 10-19 349 5 3 0.9 20-29 217 6 11 5.2 30-39 147 2 26 17.9 40-49 114 3 43 38.7 50-59 88 3 33 38.8 60-69 64 3 26 42. C 70-79 34 2 15 46.9 80-89 28 0 16 57.1 90-99 8 0 2 25.0 100-109 9 1 3 37.5 110-119 11 1 2 20.0 > 119 20 3 2 11.8 Not determined 4 0 1 25.0 29 days had actually landed long before their recovery. The small loss of drags from bottles out more than 119 days (11.8 percent — table 1) can be explained in part by the fact that many of these bottles were recovered at unfrequented places aind may have landed many days before they were found. Duck hunters returned a number of these bottles in the fall from marsh areas. Even though the time out for the bottles was around 4 months, total travel from release point was less than 10 miles. Remarks made by persons returning cards from bottles with lost drags indicate that the second weakness of the unit is in the attachment of the suspension wire to the drag. The weight of the drag and the stress imposed during the bottle's journey even- tually cut through the wire. Because bottles in this study moved relatively short distances, the percentage of drags lost was small. In an experiment where bottles might be expected to be out for an average of over 30 days, the weakness in the suspen- sion of the drag should be remedied. The question often arises whether or not a reward should be paid for return of Ccirds. A reward might increase the prob- ability of the return of a reply card that has been found, ajid might encourage active searches for bottles. On the other hand, in projects where a large number of bottles are released the cost of rewards becomes prohibitive. Although the data (table 2) are far from conclusive, it appears that where rewards have been offered, the re- turns have not been consistently (12 to 57 percent recovery, average 29.0) greater than in experiments where rewards were not offered (3 to 67 percent, average 28.6). This may not, however, be an entirely fair comparison. Poor returns are to be expected in some experiments and those on which rewards were offered may have been this type. Table 2. --Comparison of numbers of "bottles' released and percentage return for each of several areas Investigator Area Number released Percentage i-eturn Type of drift unit Reward Alme 1845 1/ Mediterranean 50 6 . . . .2/. . . None reported Ayers et al. 1956 Lake Huron 1,641 10 Plastic toothbrush containers None Carruthers 1925 North Sea 1,275 67 Bottles plain None reported Carruthers 1927 English Channel 500 33 Bottles bottom-trailing and bottles plain None reported Daniel and Lewis 1930 Irish Sea 1,180 51 . . . .2/. . . None reported Deason 1932 Fry 1956 Garstang 1898 Lake Michigan Georgian Bay English Channel 283 3,000 430 64 10 27 Bottles with drags Drift cards Bottles ballasted None None None reported reported Hachey 1935 Harrington 1895 Mavor 1922 Hudson Bay Great Lakes Bay of Fundy 500 5,000 396 5 14 18 Bottles with drags Bottles plain Bottles with drags and bottles plain None reported None reported 25 cents Platania 1923 Mediterranean 515 26 Bottles ballasted and bottles plain None reported Ruschmeyer, Olson, Lake Superior 1,000 33 Bottles ballasted None reported and Bosch 1957 Schmidt W13 Mediterranean 200 29 Bottles ballasted and bottles plain None reported Tait 1930 Tibby 1939 Uda 1932 Waldichuck and Tabata 1955 North Sea Pacific Ocean Wakasa Bay Strait of Georgia 4,825 5,943 740 237 23 3 31 57 Bottles ballasted Bottles ballasted . . . .2/. . . Bottles ballasted None None None reported Small monetary award 3/ Webster and Buller 1950 Atlantic Ocean 489 12 Bottles with drags and bottles plain 50 cents Wright 1955 Lake Erie 98 55 Bottles with drags None reported 1/ Reported by Sclunidt 1913; not seen by me 2/ Type of drift unit not specified 3/ Exact amount of reward not stated No monetary rewards were offered in this study since remarks on returns of cards released in Lake Michigan in 1954-55 clearly indicated that finders were more interested in learning when and where bottles were released than in remuneration. Inquiry in shore areas showed that once a bottle was found word spread fast in the vicinity and searching for them quickly became a loccil pastime, merely for the satisfaction of finding a bottle that had drifted from an unknown point and for the feeling of parti- cipating in a scientific study. Business reply cards were used in this study so no postage was required when they were mailed in the United States. Where return of a STATUTE MILES Figure 4. — Drift bottles released at 42 stations in Saginaw Bay and Lake Huron during 1956 and rela- tive abundance of recoveries along the shoreline. Triangles indicate stations where 30 bottles were released, squares 40 bottles, X*s 80 bottles, zuid circles 160 bottles. card was at the expense of the finder, as viftien these cards were mailed in Canada, the sender was returned the postage due him. A letter was sent to the finder of every bot- tle telling of the time, place, and purpose of release. Any postcige due the finder, was included with this letter. RELEASES AND RECOVERIES From June 5, 1956, until November 14, 1956, during nine cruises of the Fish and Wildlife Service research vessel Cisco, 2,200 drift bottles were released at 27 dif- ferent stations on Saginaw Bay and adjacent areas of Lake Huron. In addi- tion, on three synoptic sur- veys of Saginaw Bay on June 7, August 10, and October 30, 1956, the Fish and Wildlife Service research vessel Musky and a Michigan Department of Conser- vation patrol boat dropped an additional 450 bottles at 15 different stations. Altogether, 2,650 bottles were released at 42 stations (fig. 4). In Sagi- naw Bay the distance from any one point to a release point did not exceed 6 miles. Re- leases were more widely spaced in the adjacent waters of Lake Huron. It was not the intent of the investigation to make an intensive study of Lake Huron proper; rather we wished to study Saginaw Bay and its relation to the lake. It is common in drift- bottle work to receive reply cards from bottles recovered months and even years Eifter the date of release. The value of a recovery in the determina- tion of currents decreases the longer the bottle is out in excess of actual drifting time. In areas where ice forms yearly, the possible effects of the spring ice breakup upon bottle movements preclude sensible analysis. To eliminate the latter problem and to eliminate other questionable records, bottles recovered after Febru- ary 28, 1957, were not used in analysis. As of that date. StClair River Figure 5. — Location of recoveries of drift bottles found after February 28, 1957. returns had been received on 1,603 bottles; 60.5 percent of the total released. By Decem- ber 1, 1957, an additional 240 reply cards had been returned (fig. 5) giving a total return of 69.5 percent--a very high percentage in comparison with returns in other studies (table 2). The percentage re- turn of bottles released during any one cruise in this investi- gation decreased as the season progressed (table 3). Recoveries of bottles were for the most part highest over weekends (table 4). This trend was especially noticeable in late summer and early fall. During June, July, and August, the beaches and shores of the lake were apparently well covered throughout the week. After the vacation season ended, however, the shores were visited more frequently on weekends. Table 3, — Returns of bottles released in 1956 from each of nine different cruises of the Cisco and from three cruises each by the Musky and Michigan Department of Conservation Patrol Boat [Returns from bottles recovered after February 28, 1957, are not included] Cruise Date Number of bottles released Number returned Percentage return li/ June 3-11, 1956 350 262 74.9 11 June 19-July 2, 1956 310 232 74.8 III July 11-23, 1956 240 171 71.3 ivi/ July 31-August 13, 1956 460 299 65.0 V August 21-September 2, 1956 240 155 64.6 VI September 11-24, 1956 310 145 46.8 VII October 2-15, 1956 240 125 52.1 villi'' October 23-November 5, 1S5G 460 211 45.9 IX November 13-21, 1955 40 3 7.5 Total 2,650 1,603 ... 1 1/ Cruises during which the Musky and Michigan DeparLiiient of Conservation patrol boat participated Recoveries before February 28, 1957, were made from a point 5 miles north of Sturgeon Point on the western shore of Lake Huron down through Saginaw Bay, the lower lake area, and up the eastern shore of Lake Huron to Cape Kurd (fig. 4). Several areas, however, yielded surprisingly few recovei)- ies. Returns were especially light from the southeasterly corner of Lake Huron (fig. 4). Table 4.^ — Percentage distribution of recoveries of bottles within the week, in the various months of recovery in 1956 Day June July August September October November December Monday 9.0 6.3 10.8 4.7 11.7 2.1 12.9 Tuesday 15.7 8.8 7.7 6.3 3.9 5 . 3 12.9 Wednesday 12.4 9.2 7.0 9.0 7.8 3.7 9.7 Thursday 15.7 10.1 9.2 8.6 9.7 17.0 16.1 Friday 11.4 33.7 22.1 11.4 7.0 7.6 9.7 Saturday 19.4 16.2 20.6 27.9 35.9 34.0 2.6 Sunday 16.3 15.8 22.7 32.0 24.0 30.3 36.1 Total recoveries 196 246 314 293 305 200 35 The number of returns from any pstrti- cular area depends on two major factors. First, of course, bottles must actually wash ashore. The numbers that do are determined by water moyements and shore configuration. Second, the bottle must be seen and picked up. Apparently all shore areas were covered to some extent, especially during the vaca- tion period, but the distribution of resorts and beaches is by no means uniform. Conse- quently, the recovery records do not give a precisely accu- rate measure of actual landings. I believe, however, that within reasonable limits they are quantitatively dependable, at least for central and southern Lake Huron. Five returns made after the "cut-off date" of Febru- ary 28, 1957, are of particular interest because the recovery points were so far removed from the rest. All five were re- leased during the fall of 1956. Apparently, travel of these bottles was influenced by water circulation resulting from the strong southeaist, east, and southwest winds that blew during the fall which would account for their travel to Drununond and Manitoulin Islands (fig. 5). It is conceivable that a number of bottles landed on these and other islands that rim the northern boundaries of Lake Huron, but because the shores are so little frequented only a few were found. WIND DATA After many years of work on the North Sea, Carruthers (1947) emphasized the impor- tance of keeping suitable records of wind for use with studies of water movements. Wind data for this investigation were taken from Coast Guard Stations on Lake Huron and Saginaw Bay. (Other stations were held to be too far distant for the records to be useful in this study.) At no time were bottles dropped more than 50 miles from a source of wind information. The Tawas Point, Bay City, and Harbor Beach Coast Guard Stations submitted wind data consist- ing of six observations daily, that is, an observation every 4 hours. The wind direc- Table 5. — July 31 j 1956, wind data from Coast Guard Stations on Saginaw Bay and Lake Huron Hour Tawas Point Bay City Harbor Beach 0400 S 5 NNW 18 S 16 0800 S 10 NW 16 SE 10 1200 SW 5 NW 10 Calm 1600 SE 2 N 6 SE 16 2000 N 2 N 12 S 5 2400 NW 5 N 6 NW 5 HARBOR BEACH w- TAWAS POINT lOmph tion~and velocity were recorded for each observation at each station and a prevail- ing wind vector for the day at each station determined. A wind track based on these vectors was used in interpreting drift- bottle movements. Even though the area studied lies within the belt of prevailing westerlies, frequent wind changes occur with the pas- sage of the racuiy pressure systems through the lakes cireas. These changes are of paramount importance in affecting surface-water flow. In addition, there may be some tendency for off- shore breezes at night and onshore breezes during the daytime. Because of these frequent wind changes, local winds at stations no more than 50 miles apart may be blowing from nearly opposite directions at the same time (table 5). These differ- : ences in wind direction among the ■ stations are more common during periods of light, variable breezes. During strong blows and gales the wind direction is less likely to differ between stations. BAY CITY Figure 6. — July 31, 1956, wind vector at the Tawas Point, Bay City, and Harbor Beach, Michigan Coast Guard Stations. Although the prevailing winds can differ between stations (fig. 6), the wind tracks in 1956 taken over periods of a month, were simi- lax (fig. 7). Similarity among stations was greatest for August, September, October, and November and least in June and July when winds were usually light. This seasonal trend has prime significance in 10 31 OCTOBER TAWAS POINT 10 OCTOBER BAY CITY 10 OCTOBER 31 OCTOBER HARBOR BEACH 10 OCTOBER Figure 7. — Prevailing wind vectors for October 1956, at Coast Guard Stations bordering Lake Huron. determining the procedure to be followed in the interpretation of effects of wind on water move- ment. The direction of the wind and direction of the movement of a bottle were correlated over periods of several weeks, regard- less of the station for which the track was used. On the other hand, for study of short-term movements, it is desirable to use wind data from the station closest to the path of drift. It may be of some significance that movements of bottles that agree least with wind movements were those dropped farthest from any source of wind information. WATER MOVEMENTS OF SAGINAW BAY General features of the Bay According to the Great Lakes Pilot, 1956, "Saginaw Bay, the largest indentation along the west shore of Lake Huron, has a width at its entrance between Pointe aux Barques and Au Sable Point of 26 miles, and from this line southwesterly to its head at the mouth of Saginaw River the distance is 51 miles. Its mini- mum width is 13 miles, between Sand Point on the east and Point Lookout on the west in the outer portion of the bay; but, owing to the very shallow bank extend- ing from the easterly shore to beyond the Charity Islands, and to the shoal projecting from Point Lookout, the deep channel at this point of least width is contracted to a width of about 1 3/4 miles. The water level in Saginaw Bay is subject to sudden changes due to the wind, a north- east gale driving the water into the bay so as to raise the level at the mouth of Saginaw River 3 to 4 feet sometimes in less than as many hours , while a southwest wind lowers the level at times sufficiently to cause large ves- sels to ground in the channel." Of the several tributaries 11 to Saginaw Bay, the Saginaw River at the southwestern end is the largest. It follows that the net flow of water must be from the southwestern end of the bay northeastward into Lake Huron. north side and a similarly distributed outflow on the south side. They thought that the winter circulation might be the same as that in the spring or fall. The greater portion of the bay is less than 20 feet deep. It is only near Lake Huron at the mouth of the bay that a well- formed, persisting thermocline is present in the summer and early fall. Temporary stratification does, however, occur within the bay during the Rummer. In 1956 a ther- mocline developed in certain inner areas but it was ill-defined and temporary. Approximately 25 percent of the total area of the bay became stratified and almost all of this area was at or near the mouth. Previous studies Both Harrington (1895) and Ayers et al . (1956) were concerned primarily with circu- lation in Lake Huron proper, although the latter released "bottles" in the bay, and both had recoveries here. According to Harrington (1895) the most marked feature of the drift in Lake Huron is the stream passing southward along the west shore and crossing the mouth of Saginaw Bay. The few bottles that entered the bay during his investigation landed in the northwestern and southeastern sections. He did not com- ment on circulation within the bay. The Saginaw Valley Project is of inter- est here even though no current studies were made (Adams 1937). For the summer and fall of 1935 and the summer of 1936, it is apparent from the study that there was little or no correlation between chloride concentrations in different areas of the bay and wind direction. Ayers et al. (1956) noted changes in circulation at the mouth of the bay in different months. They believed that in June 1954 there was inflow on the north side and outflow along the south side. Some inflow in July was thought to be sub- surface with outflow spread over much of the surface. Conditions in August were held to be similar to those of July. They conjectured from their limited data that Saginaw Bay may behave like a simple estu- ary of the same geographical orientation; hence, in the autumnal circulation there would be inflow at all levels along the Drift-bottle movements in 1956 In our investigations, analysis of the drift-bottle returns from releases in Sagi- naw Bay confirms the belief of Ayers et al. (1956) that no one stable surface-current pattern exists within the bay. In fact, results disclose more variability of the surface currents than their studies were able to show. It appears that the dynamics of the bay are closely related to the highly variable meteorological conditions of this area and that the surface currents are in a continuous state of change. For this reason, we must state specifically under what conditions any particular surface- current pattern was found. It is possible in Saginaw Bay for com- pletely different current patterns to exist Figure 8. — Typical surface-current flow for Saginaw Bay in the summer of 1956. 12 on succeeding days. In the summer of 1956 surface currents did at times approach a state that might be called "typical" for the bay (fig. 8). Yet this system was in a continuous state of readjustment to chang- ing winds and no single surface-current pattern persisted over an extended period. Movements of bottles, released on August 10, give a good indication of surface currents under fairly stable westerly winds (fig. 9). Of 230 bottles released on this day only one was recovered on the western shore. Its travel cannot be determined since it was out 99 days before recovery. Very likely it first traveled easterly under the west winds, became entangled in marsh weeds along the eastern shores and then was re- floated by strojqg easterly winds and carried across the bay to the western shore. Under the influence of prevailing easterly winds, surface currents travel westerly in the bay as is demonstrated by the recoveries from bottle releases on October 12 and 13 (fig. 10). Although releases were made at fewer stations on these dates than on August 10, it is clear- ly evident from the recoveries that south- east, east, and northeast winds caused a general westerly surface drift. To indicate more clearly the relation- ship between local winds and drift-bottle travel, the release and recovery points of the drift bottles have been plotted, along with the winds that blew a short time prior and subsequent to release (Appendix). As exajnples, three typical stations in the bay have been chosen [an inner-bay station, fig. 11; mid-bay station, fig. 12; and outer-bay station, fig. 13 (see pages 14 and 15)], and drift-bottle travel will be discussed on the basis of information from releases at these stations. Nineteen recoveries were made from the STATUTE MILES "Figure 9. — Surface currents in Saginaw Bay determined from travel of drift bottles released on August 10, 1956, during a period of moderate westerly winds. Figure 10. — Surface currents in Saginaw Bay determined from travel of drift bottles released on October 12-13, 1956, during a period of strong southeast-northeast winds. 13 30 relejises at the inner-bay station (fig. 11). It is apparent that no single surface circulation could have existed throughout the summer and fall seasons to give such a pattern of recoveries. On the June 7 (fig. 21 A) and August 10 (fig. 43 A) releases, the direction of drift was to the northeast, presumably resulting from prevailing south- west winds at these times. Two days after 10 bottles were released on October 30 (fig. 68 A), moderate northeast winds caused currents that carried all but 3 of the bot- tles ashore to the southwest. The 27 recoveries from the 30 releases at a mid-bay station show a greater disper- sion than those recovered from the releases at the inner-bay station (fig. 12). Subse- quent to June 7, at which time 10 bottles were released, the net wind vector was to the north. The resulting water movement accounts for the recovery of the 2 bottles found to the north of the release point (fig. 23 B). The 7 bottles recovered at least 49 days after release to the east and southeast appear to have drifted in this direction because of the net winds to the southeast for the period of time they were adrift (fig. 23 B). The 8 recoveries from the 10 releases on August 10 were to the east and southeast (fig. 45 B). The net wind vector for the period of time these bottles were drifting was to the east. The 10 bottles released on October 30 landed to the northwest of their release point (fig. 69 C). Apparently these bottles were car- ried with water moved by the winds that blew the day of and the day after release. The 60 recoveries of 80 bottles re- leased from ein outer-bay station substan- tiate the theory that surface-current flow was unstable during this study (fig. 13). Recoveries from releases at this station 41 MILES ABOVE PORT HURON STATUTE MILES Figure 11. — Location of 19 recoveries from 30 drift bottles released at an inner Saginaw Bay station. Ten releases were made on June 7, Aigust 10, and October 30, 1956. A triangle shows release point; X's mark recovery points. Figure 12. — Location of 27 recoveries from 30 drift bottles released at a middle Saginaw Bay station. Ten releases were made on June 7, August 10, and October 30, 1956. A triangle shows release point; X's mark recovery points. 14 were made not only from widely scattered areas in Saginaw Bay but also from many points along the shore of Lake Huron. All 10 bottles released from this station on June 7 were recovered from the Tawas Bay area to the northwest within a short dis- tance of one another (fig. 24 B). After release of these bottles, winds were vari- able for 3 days; next the wind was to the northeast for 3 days; the wind to the southwest a week after release ^parently caused the bottles to land in and around Tawas Point. Bottles released on June 29 were recovered for the most part on the eastern side of the Michigan Thumb (fig. 30 B). The net wind to the southeast most likely caused currents that Ccirried the bottles to this cirea. The widely scattered points of the 7 recoveries from the 10 releases on July 18 follow well, with one exception, the wind track (fig. 36 B). The only logical explanation for the recovery STATUTE MILE5 Figure 13. — Location of 60 recoveries from 80 drift bottles released at an outer Saginaw Bay station. Ten releases were made at 8 different times from June through October 1956. A triangle shows release point; X's mark recovery points. on the Michigan Thumb after 27 days is the influence of a current entering Saginaw Bay from Lake Huron. The recoveries from releases on August 10 (fig. 46 B) and Au- gust 30 (fig. 52 B) indicate that surface currents flowed toward the east in August and September. Westerly winds prevailed these months. Bottles recovered from releases on September 21 (fig. 59 B) and October 12 (fig. 63 A) apparently were car- ried by currents caused by east and south winds, respectively. The recoveries from October 30 releases (fig. 70 A), although scattered widely, correspond well with wind direction. Apparently the bottles drifted into the bay where one was deposited on the west shore. A reversal in wind direction then drove the bottles over to the Michigan Thum where they landed near the mouth of the bay. It is evident from the foregoing dis- cussion that the surface currents of Sagi- naw Bay are closely related to the winds cind consequently are highly variable. This dependence of currents on wind wzis closest for the inner reaches of the bay. Near the mouth of the bay a few bottles were found, the movement of which could not be explained by local wind action (fig. 33 B). Wind direction here at the time of release was to the northeast. The 3 bottles that landed within a day of release, however, traveled to the southeast at least 90° to the right of the wind vector. This variation most probably was brought about by a strong current entering the bay from Lake Huron, causing the bottles to drift at right angles to the wind. Additional variations were noted around Tawas Point and the area north- east of Sand Point. Surface currents in the bay apparaitly orient to changing winds in a short period of time. For 2 days prior to release and on the day of release of 10 bottles north- east of Point Lookout, winds were from the south (fig. 32 B). The day after release the south winds diminished and chcinged to north. Surface currents had to reorient to the north winds in a very short time in order to cause the bottles to land as indi- cated. For 4 days prior to release of 10 bottles south of Point Au Ores on October 30, strong southerly winds blew (fig. 69 A). The day after release the south winds moder- ated and the second day after release changed to north. The surface currents had to reorient rapidly to the north wind to 15 cause southwest drift of bottles, one of which was recovered 4 days after release. The time required for the current to orient to the wind is dependent, of course, upon the strength of the wind and the existing flow pattern. In an area such as Saginaw Bay where a consistent current pattern does not exist, it seems possible that a surface current might change in response to a rapidly changed strong wind in a matter of hours. No attempt was made to study subsur- face currents. The Ekman spiral (Sverdrup, Johnson, and Fleming 1942) is frequently mentioned in explanations of surface and subsurface currents. According to this theory, surface currents on the northern hemisphere are directed 45° to the right of the wind, while at greater depths the current turns more to the right and the velocity decreases. Near the bottom of the friction layer the currents are low in velo- city and move opposite to the wind direc- tion. This theory, however, presupposes conditions of equilibrium, a state that is not reached in Saginaw Bay because of the influence of variable winds. Thus it would seem that the theory of the Ekman sprial does not apply. The relative shallowness of the basin also would seem to be an ad- verse factor. Because of the prevalence of shallow water and the consequent transi- tory thermocline development, I suspect that subsurface currents are highly influ- enced by surface currents and may be similar in direction to them. Modifying factors exist in the bay to complicate the simple wind-dependent sur- face flow. Outflow of the streams and rivers in the area must have some effect upon surface currents. The largest of these, the Saginaw River, enters the south- ern end of the bay. Some of this river water must diffuse into the bay water but a discrete mass of water has been found to follow the eastern shore of the bay out into Lake Huron proper (Adams 1937). This mass well might be the flow of Saginaw River water. A theory proposed by Steele (1957) interprets the hydrography of the northern North Sea in terms of the possible effects of lateral eddy diffusion. This diffusion depends upon the principle "that when a jet issues into a motionless fluid there is turbulent mixing along its edges malting the jet gradually spread out. An important feature is that as a result of this mixing, the jet draws in fluid from its surround- ings." If this theory holds true for the North Sea and other bodies of water, it may apply to Saginaw Bay also, especially if all the streams and rivers entering the bay and currents entering from Lake Huron be- have as jets drawing in water laterally. Circulation at the mouth of the bay must be affected to a large extent by move- ments of Lake Huron water. Harrington's (1895) work indicated that a strong current flows down the western shore of Lake Huron across the mouth of Saginaw Bay. The pene- tration of Lake Huron water into the bay is still a matter to be resolved through chemi- cal and physical data collected during this study. WATER MOVEMENTS OF LAKE HURON Previous studies The first account of drift-bottle work upon Lake Huron was that by Harrington (1895). He recognized a variability in surface currents of the lake when he stated, "While the winds from the Great Lakes are westerly in their prevailing direction, this is the region of variable weather, and the actual directions of the wind change from day to day. There will, consequently, be considerable variation in the currents from time to time, and this undoubtedly causes a wayward motion of the current bottles." He found that the courses taken by the bottles in Lake Huron exhibited a somewhat more complicated drift than did bottles released in Lake Superior and Lake Michigan. Ayers (1956) adapted the oceanogra- phers' dynamic-height method of determining currents to freshwater conditions. Find- ings on Lake Huron in 1954, based on this method, seemed to be in good agreement with results obtained by other methods (Ayers et al . 1956). Analysis of the data from three synoptic runs in 1954 revealed dis- tinct differences in surface circulation at the times (spring, summer, and fall) of the runs. They concluded, further, "The fundamental surface circulation pattern in the upper and central portions of the lake appeared to be counterclockwise. In the 16 lower end of the lake outflow to the St. Clair River appeared to consist of a mean- dering surface current, near or east of the midline of the lake, which approached the entrance of the river from the northeast." Drift-bottle movements in 1956 In the present investigations, only one bottle was found below the head of the St. Clair River that flows out of Lake Huron. Of the many bottles that rounded the Michigan Thumb, most landed on the east- ern side of the Thumb before reaching the river mouth. These recoveries lend some support to Ayers' contention that outflow to the St. Clair River was from the north- east. Recoveries of 416 bottles from 760 releases along 3 transects in the southern part of the lake (fig. 4) are considered in the analysis of Lake Huron water movements. A striking feature of the returns from these releases was the scarcity of recover- ies from the Saginaw Bay area (fig. 14). Only 16 bottles were found within the bay and none of these had penetrated more than 10 miles. Penetration into the bay was correlated with easterly winds during the time the bottles were adrift. As Harrington (1895) and Ayers e_t al. (1956) have indi- cated, however, a strong current may at times pass down the west shore of Lake Huron and set up counterclockwise rotation at the mouth of Saginaw Bay. Had bottles been released near the west shore above Au Sable Point, it is possible that many would have been carried into the bay by such a current. The remainder of the recoveries from the Lake Huron shores were scattered widely. The tendency was marked, however, for the bottles to drift to the east (fig. 15). All recoveries from stations 2-8 were from the Michigan Thumb area, mostly on the ..northern and Ccistern sides and from the STATUTE MILES Figure 14. — Location of 16 drift-bottle re- coveries in Saginaw Bay from 760 releases on Lake Huron. Eighty bottles were re- leased each at stations 1-6 and 40 each at stations 7-13. X's mark recovery points. Figure 15. — Location of 45 recoveries from 80 drift-bottle releases at a Lake Huron station. Triangle marks release point; X's mark recovery points. 17 eastern shore of LaJce Huron. All recoveries from 200 releases at stations 9-13 were from the eastern shore of Lake Huron. Apparently the surface current on Lake Huron during the summer and fall of 1956 had a net circu- lation from west to east. Although the direction and intensity of local winds were important in explaining surface drift in Saginaw Bay, they appear less significant in Lake Huron proper (fig. 42 D) . On August 3, 20 releases were made, 10 at each of the indicated stations, within 2 1/2 miles of one another. The large dif- ference in direction of drift from the two stations of bottles that were out approxi- mately the same length of time and released at nearly the same time indicates forces other than wind at work in the formation of currents. Certainly, wind conditions play a prominent role in formations of surface currents in Lake Huron. However, the rela- tionship between wind and currents is not STATUTE MILES Figure 16. — Location of 22 recoveries from 40 drift-bottle releases at a Lake Huron station. Triangle marks release point; X's mark recovery points. nearly so obvious as in Saginaw Bay. According to Millar (1952) the energy input into a lake from a day's wind may not be completely dissipated until 12 days later. If this relation holds in Lake Huron, the prevailing winds assume a prominent role in formation of the general surface current pattern in the lake. As was true in Saginaw. Bay, location of Lake Huron returns from a particular station can vary widely throughout the season. The wide scatter of the 45 returns from 80 bottles released off Harbor Beach (10 bottles each at 8 different times from June through October 1956) is strong evi- dence of the instability of the lake cur- rents (fig. 15). The drift throughout the investigation from some stations, however, could be much more stable (fig. 16). Dynamic heights The use of the dynamic-height method of determining current flow depends upon the availability of a subsurface reference plane at which currents are absent. Ayers et al . (1956) has indicated that Lake Huron has certain characteristics concerning circulation that are psuedo-oceanic. In calculating the relationship between wind and depth of mixing for oceans, Sverdrup, Johnson, and Fleming (1942) derived the formula D = 7.6 where D is ^~Sl the depth in meters, W the wind velocity in meters per second, and v the latitude for which the calculation is made. If 44° (the approximate average latitude for southern Lake Huron) is substituted for