— , 7 ii ,_, — 7-, ,, é , _ , I, _ , f ,i f .7, g , WM 9| i; ‘HH; v! Q); '--1 IE? '0) THfi DUMMY WATERFOWL NEST AS AN {ND'éX TO PREDATEON Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY Don Ray Perkuchin 1964 IIIIIIII W W \‘MMW IIIIIIIIIIIIIIIIIII \\\\\\\\\\1 '\\‘i\\\\\\\\\\\\\l\g\\\\fi * f 3 1293 10575 L I B R A R Y Michigan State Univegsity ‘ ABSTRACT THE DUMMY WATERFOWL NEST AS AN INDEX TO PREDATION by Don Ray Perkuchin Rates of destruction of artificial waterfowl nests by predators were studied at the Agassiz National Wildlife Refuge in northwestern Minnesota between April 18 and August 11, 1960. The predator removal area was based on a dike- transect route on the west half of the refuge and the check area on a similar route in the east half of the refuge. Predator numbers were reduced on the removal- zone dikes by setting out strychnine—poisoned eggs. Fol- lowing predator control, waterfowl breeding-pair and brood counts were determined along the dike-transect routes. Nearly llOO dummy nests were placed in nearly equal pro- portions on the two areas as a test of predator control effects. Increased survival of dummy nests occured where predator control was undertaken. Both before and during the control period, nest survival was significantly higher in the predator-removal zone than in the check area. Fol- lowing the centrol program, there was no apparent difference for the two zones. The low survival in the check zone Don Ray Perkuchin before control work is attributed to the preponderance of crows there then. Nearly equal artificial—nest survival in both areas following predator reduction efforts is be— lieved to have been due to the greater abundance of food and cover then and to the more nearly equal predator popu- lations occurring after the control campaign stopped. Pseudo-nests located on dikes suffered higher losses than those situated elsewhere, except during the predator reduction period. Presumably the higher destruction of dike nests before and after the control program was due to the predators' ability to locate these nests more easi- ly on the restricted areas of the dike-tops. The reversal in this dummy-nest destruction pattern during the reduc- tion campaign was almost certainly due to the predator- removal efforts then underway on the removal—zone dikes. The degree to which artificial nests were exposed to View had no clear-cut effect on their survival. A heavy covering of dried vegetation placed over half the poisoned nests seemed to enhance their survival slightly. In contrast, however, lightly covering the non-poisoned test-nests with a mixture of dried grass and duck feathers did not enable their greater success. Reproductive success of wild blue—winged teal was higher in the area of predator control. Comparable data for mallards and gadwalls, however, revealed higher suc- cess in the check area for the former species and no dif- ference between the two areas for the latter. No decisive Don Ray Perkuchin effects of predator control on waterfowl production during 1960 were evident in this study. The dummy-nest data of this investigation provided indexes to the relative abundance of egg—predators and to the degree of predator control achieved. THE DUMMY WATERFOWL NEST AS AN INDEX TO PREDATION By Don Ray Perkuchin A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife fly; $47—94: a w» ,/<)¢fj7( 1:; €>¢1A4’é%7 f {f /\ 4% v-'.’1«.’Ja{'4 / 196A ACKNOWLEDGEMENTS I would like to express my gratitude to Messrs. Forrest Carpenter, Harvey Nelson and Herbert Dill of the Bureau of Sport Fisheries and Wildlife for their useful suggestions and assistance in providing equipment. Sin- cere thanks are due, too, to Merrill Hammond and Donald Balser of the Bureau for helpful advise. Other Bureau employees who I wish to thank are E. Marvin Mansfield and Ralph Town for compiling and summarizing the waterfowl breeding pair and production data given beyond and David Olsen for describing the habitat types along the study transects. The entire predator-poisoning campaign was directed by Berkeley Peterson of the Branch of Predator and Rodent Control of the Bureau. He was assisted by Messrs. Carl Burrell and Glen Bernstein, temporary employees of the Minnesota Conservation Department. This study was made in connection with thesis re- quirements for the Master of Science degree at Michigan State University. Special appreciation is due Dr. George A. Petrides of the Department of Fisheries and Wildlife of that institution, for his counsel and considerable interpretive and editorial assistance. Dr. Phillip J. Clark of the Department of Zoology assisted greatly in ii the design of the study and in statistical analysis. I am grateful also to Drs. Miles D. Pirnie and Peter I. Tack of the Department of Fisheries and Wildlife and to. Dr. George J. Wallace of the Department of Zoology for their review of the manuscript and for subsequent sug- gestions and comments. The aid and encouragement of my wife, Carol, was instrumental in the completion of the study. 111 TABLE OF CONTENTS ACKNOWLEDGEMENTS LIST OF TABLES LIST OF FIGURES Chapter I. INTRODUCTION The Study Area and Transects II. METHODS Artificial-Nests Poison-Nests Waterfowl Reproduction Indexes III. RESULTS IV. SUMMARY BIBLIOGRAPHY . . . . . iv Page ii vi 12 22 23 Table 1. LIST OF TABLES Page A comparison of habitat percentages within one- eighth mile of dike- transect routes-- Agassiz National Wildlife Refuge, Minnesota, Summer, 1958 . . . . . . . . . . . . . . 5 Summary of predators poisoned and recovered—— Agassiz National Wildlife Refuge, Minnesota, May 3 to June 10,1960 . . . . . 13 Seasonal extent of predation on dummy-nests within predator-removal and undisturbed check zones-—Agassiz National Wildlife Refuge Minnesota, 1960 . . . . . . . . . . . . . 1“ Percentage survival of dummy-nests on the dike and non-dike locations as related to predator removal on the dikes of one area (between May 3 and June 10 only)--Agassiz National Wildlife Refuge, Minnesota, 1960. . 1? Proportion of newly-hatched broods and numbers of breeding pairs observed for principal nesting ducks on the waterfowl census routes of the removal and check zones--Agassiz National Wildlife Refuge, 1960 . . . . . . 2O LIST OF FIGURES Figure Page 1. Agassiz National Wildlife Refuge study area transect- routes . . . . . . . A vi CHAPTER I INTRODUCTION Between April 18 and August 11, 1960 on the Agas- siz National Wildlife Refuge, Minnesota, I studied dummy, or artificial, waterfowl nests as indicators of duck nest predation. The effort was made to test the values of such nests in measuring (1) the level of abudance of egg preda- tors both with respect to the effectiveness of predator control programs and at other times, and (2) the effect of predator control on waterfowl nest survival and popu- lation reproduction. The effects of covering dummy-nests to discourage predation also were studied. My investi- gation was but a part of a more comprehensive predation study at the Agassiz Refuge which was started in 1958 and is scheduled to be completed during the late summer of 196A. While conducting this study, I was employed as a Wild- life Aid by the U.S. Bureau of Sport Fisheries and Wild- life. Predation studies using dummy nests comprised of chicken eggs were conducted by Merrill Hammond, Arthur Hughlett, David McGlaughlin and Howard Woon in 1951 on several national wildlife refuges in North Dakota and Nebraska. This was done to test the effect of land use and nest concealment on predation. 1 These studies indicated that, although predators do not ordinarily depend upon visual clues to find the nests, concealment did reduce predation on the dummy nests (Hammond, unpublished manuscript). The Study Area and Transects The Agassiz Refuge comprises about 61,000 acres of the lakebed of ancient Lake Agassiz in Marshall County, northwestern Minnesota. Low earthen dikes have been built to create 1“ artificial pools containing about 25,000 acres of shallow marsh and open water. Bulrushes (Scirpus sppo) and cattails (Typha spp.) are the princi- pal emergent aquatic plants. Crowfoot (Ranunculus spp.) and pondweeds (Potamogeton spp.) are the most common of the submerged flora. The dikes, which crisscross the area, are travel- ways for predators and refuge personnel. Personal obser- vations indicate that uplands on and adjacent to the le- vees afford important nesting places for blue-winged teal (Anas discors), gadwall (A. strepera), mallard (A. piggy: rhynchos) and pintail (A. aguta) ducks. Vegetation occurring on these upland sites includes quackgrass (AgrOpyron repens), bromegrass (Bromus sp.), bluegrasses (Egg spp.), Canada thistle (Cirsium arvense), sow thistle (Sonchus sp.), nettle (Urtica sp.), sedges (Carex spp.), red-osier dogwood (Cornus stolonifera), willows (Salix spp.), trembling aSpen (Populus tremuloides) and balsam poplar (P. balsamifera). Red raspberry (Rubus idaeus), blueberries (Vaccinium spp.), chokeberry (Prunus Virginia) and serviceberry (Amelanchier sp.) also are present and may seasonally bear abundant fruits. A 17.7 mile transect route was established on dikes in the west half of the refuge (Fig. l). A similar route plotted on the east half of the area extended for 15.6 miles (Fig. 1). The west transect route and adjacent lands comprised the predator removal area. No predator control work was done on the east side dike-transect route during 1960; it served as the check zone. From 1958 aerial photos of the refuge, Olsen (Agas- siz Refuge files, 1961) computed the habitat-type pro- portions of the area within one-eighth mile of both dike- transect routes (Table 1). He defined the several habi- tat types as follows: 1. Open Water — emergent vegetation lacking. 2. Open Marsh - large open water areas interspersed with small islands or patches of emergent aqua- tic vegetation. 3. Closed Marsh - an area heavily interspersed with extensive zones of emergent aquatic vege- tation such as cattail and phragmites. A. Ditch - borrow canal adjacent to road or dike. 5. Upland Brush - wooded areas including extensive willow patches. .mousomluoomcmne mop< mcsum mmsmmm ouwaeafiz Hmcofipmz Nammmwoamm ol¢lolo. 33m Table 1.--A comparison of waterfowl habitat percentages within one-eighth mile of dike-transect routes--Agassiz National Wildlife Refuge, Minnesota, Summer, 1958. Habitat Type Removal Zone Check Zone Open Water 4.1 7.9 Open Marsh 5.3 5-3 Closed Marsh 7.5 8.2 Ditch A1.9 “1.9 Upland Brush 11.2 Open Upland A1.2 25.“ 1 6 From data of David Olsen, Agassiz Refuge files, 19 1. 6. Open Upland - large grass or sedge covered areas which may be lightly interspersed with small willow patches. The two study areas showed a general similarity except that the uplands of the check zone were less open (Table 1). CHAPTER II METHODS Artificial Nests.--Dummy-nests consisted of three fresh chicken eggs each. A total of 1,08A such nests were set in locations which appeared to be potential nesting sites for dabbling ducks. The vegetation was spread apart and a shallow depression scooped in the soil. The eggs were placed in the depression and the growing plants restored, as closely as possible, to their original condition. These artificial nests were partly concealed by standing vegetation. They were similar in general appearance to a natural duck nest. Between April 18 and 23, 190 of these dummy-nests were placed in the removal area and 201 in the check zone. Study of these sets terminated on May 10. Be- tween May 10 and 18, 171 new test-nests were distributed in the removal zone and 192 in the check area. Those were observed until June 16. A third group of 31A arti- ficial-sets was put out between June 22 and July 5. Of these, 130 were in the removal area and 18A in the check zone. This group was studied through August 11. Nests of each group for each period were set out during as short a time as possible, the time varying between 6 and 1A days. Sixty feet was the minimum distance between nest- sets. The average spacing was 125 feet. Three to four- foot long willow wands were placed about thirty feet from each set to assist in relocation. An orange tag with a number and detailed location for each test-nest was at- tached to each wand. The pseudo-nests were distributed in lines within a half mile of the dikes in both zones. Lines on the dikes ran parallel to the tops while those off the le- vees had no particular placement pattern. From four to eight nests were set out in each strip. Most non-dike lines of nests nevertheless were within 200 yards of a dike. Half of the test-sets distributed along each dike- route were put on the dikes, the other half on upland sites adjacent to the levees. Under natural conditions, mallard eggs are usually exposed to adverse environmental factors, such as weather and predators, for 32 to 38 days prior to hatching. This is based on an incubation period of between 23 and 29 days, usually 26 (Kortright, 1957:154), an average clutch size of 9 eggs, and an average of one egg laid per day. It was intended therefore that data on artificial nests be gathered for about 35 days after all the nests in a group had been distributed. Study of the first group of nests was terminated after only 15 days due to the nearly total destruction recorded on May 10. Observation of the second group of sets stopped after 37 days on June 16. For the third group of nests study was completed after 50 days on August 11. Personal involvement in other duties during the last two periods prevented the completion of observations on these two groups of nests within the 35-day goal. All sets of each group for each period were checked twice. Nest-surveys were at least eight days apart. Any effects of the observer in possibly leaving trails for predators to follow were assumed to be proportional in both the removal and check zones. In an attempt to determine the effects of hiding the dummy-nests, alternate sets within the two zones, regardless of site, were lightly covered with a mixture of dried grass and duck feathers. A dummy-nest was recorded as destroyed when any of its eggs was damaged by a predator or was missing. Sets which were trampled by cattle, or which could not be relocated, were eliminated from the study. Poison-Nests.-—Poison-nests consisted of either two chicken eggs or one turkey egg which had been in- jected with strychnine. In appearance, the poisoned sets were similar to the dummy—nests. A total of 258 such sets were distributed along the dikes of the re- moval zone. 10 On May 3 and A, 212 poison-nests were put out. The remaining 71 were distributed on May 9. All poison- ing operations were terminated on June 10. The strych- nine-injected nests were set out along the dike trails where it was believed predators would be most likely to find them. Each poison set was marked by a two to three foot metal rod placed within a few feet of its location. All of the stations were checked daily and any destroyed, missing or spoiled eggs were replaced. Each damaged egg was examined to determine whether a bird or mammal was the plunderer. As a further check to determine the effects of covering artificial nests, half of all treated sets were covered with a 3 to 6 inch layer of driedxege- tation. The remaining nests were not covered (Agassiz Refuge files). Waterfowl Reproduction Indexes.--Duck pairs and lone males were tallied in estimating the waterfowl breeding pOpulations of the two study zones, as sug- gested by Williams, gt_al. (l9u8). A marked movement of broods from small to larger water bodies in prairie areas has been reported by Evans g£_al. (1952) and Berg (1956). Evans g£_§l, (1952) also found that as broods became older their mobility increased. Keith (1961) indicated that tendencies for broods to move may be prevented or at least delayed by 11 the relative permanency of water in small potholes in his southeastern Alberta study area. Based on these findings, it was assumed that any newly-hatched ducklings observed in a particular zone on the Agassiz Refuge were born there. On May 27, refuge waterfowl breeding-pair counts were made from roads by counting ducks seen within one-eighth mile on either side of the removal and check zone dikes (Fig. l). Broods of ducks were counted each week between June 6 and 19 on these same transect—routes. These counts were started when the first group of newly-hatched young was observed in June and were continued until the last newly-hatched ducks were seen in August. Breeding pair and brood data were obtained by refuge personnel for the principal breeding ducks. Since the blue-winged teal, mallard and gadwall were most abundant on the refuge these were selected for use in this analysis. The water- fowl production data are in the process of being written up for publication (Balser, in letter). Statistical Analysis.-—Chi-square in two-by-two contingency tables and the t-test were used in analysis of all data (Snedecor, 1956). CHAPTER III RESULTS The predator-control campaign resulted in the killing of at least A0 mammals and 9 avian predators on the dikes of the removal zone (Table 2). About 650 strychnine—injected eggs were at least partially con- sumed by predators (Agassiz Refuge files) and the total kill of carnivores was higher than the number found dead. Eighty-two poison-sets were considered to have been bro- ken-up by mammals and eighty-one by birds. Evidence from the remaining 90 plundered poison-nests was incon- clusive. In addition to the animals recovered, coyotes (Canis latrans), Franklin ground-squirrels (Citellus franklini) and woodchucks (Marmota monax) also were potential egg eaters. They occurred on the refuge and could have participated in egg predation. Bull snakes (Pitquhis catenifer) have not been seen there. The dummy-nest investigation yielded four princi- pal conclusions: 1. The survival rate for dummy nests was greater in the control zone.—-Both before and during the predator control period (see Table 3), dummy nest survival was significantly higher on the predator-removal area transect l2 13 Table 2.--Summary of predators poisoned and recovered dead as a result of eating poisoned eggs--Agassiz National Wildlife Refuge, Minnesota, May 3 to June 10, 1960 _ - Species Number” Striped skunk (Mephitis mephitis) 27 Raccoon (Procyon lotor) 5 Mink (Mustela vison) 5 Badger (Taxidea taxus) 2 Red fox (Vulpes fulva) l Marsh hawk (Circus cyaneus) 8 Crow (Corvus brachyrhynchos) 1 “It is known that not all the animals which consumed some of the poison were recovered. Raccoons and crows are often able to travel long distances before death; on the other hand, they may get less than a lethal dose of the strychnine solution. .szo OH oCSh cam m an: cmozumn mmxfio ocom Hw>oEop may no cmposccoo was Hoppcoo powwompmm 1A mouse Hm m sHsm 0» :wH guano scum: mm HH pmsmsa mm mesh OMH *Hm>oemm mman mm mmH guano ms-mo HA SH mess mH on OH Hm: HAH .Hmsoemm ©-H H mm 0p wH Hom sumac HmumH :m m Hm: HHsa< omH .Hm>OEmm mSHEHH emsosummo poz mummz UOHsmm Assam emomHm mocmofimcoo mmm mo homopcmohmm mo cam mumoz mmpmm Lonadz mma< ommH .mpomoccHz .owsmom oMHHcaflz HmCOHpmz NHmmmwothnnoumompa cfisufiz mumo:|%553v no coaumnmmg mo ucmuxm Hmcommmmnl.m oHnme 15 than on the check-area. Following control efforts, there was no apparent difference in nest survival for the two zones. These results were contradictory to those ex— pected, but the much higher crow population in and near the check zone before the predator control campaign is believed to be the probable cause of the significantly higher extent of nest destruction there then. Two to three hundred crows stopped at the refuge in April and practically all were concentrated in the east half of the refuge. Hammond (19A0) found that crows may destroy 30 per cent of the waterfowl nests on some federal re— fuges in the prairie states. Preston (1957) observed that a pair of crows discovered all early and most late mallard nests scattered in a 100 acre pen. Hence at the Agassiz Refuge, the crows were at least a likely prospect as a nest survival influence. All but a few pairs of these birds migrated further north by mid-May. By the beginning of the control campaign, there must have been more nearly equal predator populations in the east and west portions of the refuge. A higher nest survival rate prevailed in the removal zone during the May 10 to June 16 period (Table 3) and cannot be accounted for except as a result of the May 3 to June 10 predator reduction program. 16 The apparently equal dummy nest survival in both the removal and check zones during the test period following predator reduction (Table 3) is considered possibly to have been due to a greater abundance of food for preda- tors which occurred then everywhere on the refuge, to vegetative cover becoming more dense, and/or to pre— dator populations reoccupying the narrow study transects from which they were removed. Whichever of these fac- tors may have been the cause, survival of dummy—nests improved even in the absence of predator control in the check zone as the nesting season progressed. The evi- dence here supports the conclusions of Sowls (1955) that renesting can be of major importance in maintaining an area's waterfowl production. 2. Dummy—nests located on dikes suffered higher losses than those situated elsewhere.--Both before and after predator reduction work, dummy nest survival was higher off the dikes than on them (Table A). During the control interval, however, test-nests on the dikes had a higher survival rate. It seemed likely that predators searching for food along the 50 to 150 foot-wide dike-tops were more success- ful in finding nests than those foraging for food off those travel lanes. This is presumably the reason that there was higher survival for dummy-nests located off 17 Hm>oEop Loamcoua mo GOHLom * mm mm m: mm HH pmswsanmm mesh om mm om so *mH mc:HIOH as: m NOH H mm OH HaguwH HHEQH moh¢ AHm>othlcozv xomco ow Am a: mu HH pmswzaumm mesh am pm . mu :m *mH mezH-OH Hm: mH Hm H mm OH HmZImH HHsQH mop< Hw>oEom Umzoppmoo poz oomOmeoQ poz mpmoz mo mpmoz mpwoz mo mpmoz UOHpom mowmucooaom mo .02 mowmucoopmm mo .02 mxHaueoz mme .ommH .mpommeeHz .owsmom oMHHcHfiz HmCOHumz NHmmmw¢11Hw>oEoMucoc U:m_Hm>oEoALhopmomna on oopmHop mm wCOHpmooH mxfiouzo: ocm,oxflc.o:u Go.mumm:-meesc ho Hw>H>n2m owmucmohomln.: mHnt 18 the dikes of both areas both before and after the control period. The reversal of this nest survival pattern during the control interval seems almost certainly to have been due to the predator—reduction campaign then underway on the removal-area dikes. 3. The degree to which artificial-nests were ex- posed to view had no clear-cut effect on their survival.—— Of the covered dummy nests studied in the removal zone from May 10 to June 16, eighty-three per cent survived, compared to 78 per cent of the uncovered nests. This difference was not significant. Of the covered poisoned- sets put out during the predator control program, seventy- eight per cent escaped predator-destruction, while only 67 per cent of the uncovered nests were unplundered. This difference was statistically significant, but only at the 90 per cent level. In view of the results of the non-poisoned dummy nest data and the low level of significance for the poisoned-egg study, it does not seem likely that the visibility of nests had much effect on their survival under the study conditions. A. Predator control seemed either to have no ef— fects or to have opposite effects on the numbers and sizes of broods of blue-winged teal and mallards during l960.-- Comparable data for gadwalls indicated that predator re- duction had no effect on their reproductive success. The l9 breeding pair populations of these three species was several times greater in the removal zone (Table 5) pro- bably indicating that the water-fowl breeding habitat was better there, particularly for mallards. For blue-winged teal, the number of newly-hatched broods per breeding pair was higher in the predator- removal zone than in the check area (Table 5). This difference was significant at between the 90 and 95 per cent level. For the mallard, however, the number of newly-hatched broods per breeding pair in the check area was very significantly higher than that in the re— moval zone. There was no statistically significant dif- ference between the two areas, though, for comparable data of gadwalls. The relation of these varied results to the predator control program is not clear. The mean sizes of 106 broods of downy-plumaged young blue-winged teal on the removal-area census route (8.“ young per brood) was significantly larger at the 97 per cent level than that for 93 similar groups on the check- area route (7.7 young per brood). These broods were of size classes Ia, Ib and lo as described by Gollop and Marshall (195“). Mallard downy broods had a mean size of 7.6 young for 56 broods in the check zone and 6.8 young for 38 similar groups in the removal zone. This difference in brood size was highly significant. For gadwalls, the mean downy brood sizes were 7.6 young for 2O .AzmmHv HHmnmgmz cam QOHHoo an omnfipowoc mm mooopo mH mmmHo** .COHmH> Imp mHnHmmOQ o» poonQSm new mopooop mwsmmm Nfimwmw< on» Eopw :mxmp sumo msmcHEHHmpms oH. om m ow HH HHmzcmu om. omH am 25m mH ppmHsz Hm. wOH mm HMH mm Hume emwcqumsHm LHmm wcflommpm mLHmm zmooopm pfimm mcHUmopm mpfimm ammooopm poo mnoopm wchoopm consummlesoz conoummlzazoz mo mLoDESZ mcomlxooso pom mcoopm go mamoesz conopmmleSoz mo msonssz wchmmpm cocopmmansoz mo whooesz ocomuam>osmm wmfiomam [I‘ll HI‘ |I \I‘ II I .ome .oCShnmmz .muowoccHz .owzmom mmHHoHaz Hmcoflpmz NHmmmw< Insmocou xomno cam Hm>oson on» no mopson mamcmo Hzompmpmz onp co mxoso wcfipmmc HMQHocfipo pom Um>hmwno mhfima wcaooopn mo mpmnESC Ugo chOAQ omnoumnnszmc mo COthoaonmll.m magma 21 25 broods in the removal area and 7.“ young for 20 broods in the check zone. This difference was not significant. Each of the three species discussed appeared to have been affected differently by the predator control efforts, if predator control indeed had any affect on the survival of wild nests and young during that year. No conclusive effects of predator removal on waterfowl pro— duction on the Agassiz Refuge for 1960 seem evident from this study. CHAPTER V SUMMARY Study of 1,08“ artificial waterfowl nests placed in waterfowl habitats on the Agassiz National Wildlife Refuge indicated that dummy nest destruction rates are inversely correlated with predator control. They served as indexes to the relative abundance of egg predators on one area as compared to another and to the degree to which predator control is effective in reducing predators. Predator removal efforts increased the survival rate for dummy-nests. Test-nests located on dikes suffered higher losses than those situated elsewhere, except during the predator reduction campaign. Artificial-nests covered with grasses and duck feathers showed no clear-cut tendency toward higher survival. The reproductive success of wild blue-winged teal was higher in the area of predator reduction. Comparable data for mallards and gadwalls indicated higher success in the check zone for the former where predators were not limited and showed no difference between thetwo zones for the latter. No conclusive effects of predator control on waterfowl production during 1960 seem evident from this study. The dummy-nest data of this investigation did not provide an index to wild nest survival. 22 BIBLIOGRAPHY Berg, P. F. 1956. A study of waterfowl broods in eastern Montana with Special reference to movements and the relationship of reservoir fencing to production. J. Wildl. Mgmt., 20(3):253-262. Evans, C. D., Hawkins, A. S. and Marshall, W. H. 1952. Movements of waterfowl broods in Manitoba. U.S. Fish and Wildl. Serv., Spec. Sci. Report--Wildl., No. 16. “7 pp. Gollop, J. B., and Marshall, W. H. 195“. A guide for aging duck broods in the field. Mississippi Flyway Council Tech. Comm. 1“ pp. Hammond, M. C. l9“0. Crow-waterfowl relations on federal refuges. Trans. N. Amer. Wildl. Conf., 5:398—“0“. Keith, L. B. 1961. A study of waterfowl ecology on small impoundments in southeastern Alberta. Wildl. Mono-. graphs, Wild. 800., No. 6. 88 pp. Kortright, F. H. 1957. The ducks, geese and swans of North America. Amer. Wildl. Institute., Washington, D. C. “76 pp. Preston, F. W. 1957. The look-out perch as a factor in predation by crows. Wilson Bull., 69(“):368~370. Snedecor, G. W. 1956. Statistical methods. 5th ed. Iowa State College Press, Ames. 53“ pp. Sowls, L. K. 1955. Prairie ducks. Harrisburg: Stackpole Co. 193 pp, Williams, C. S. l9“8. Waterfowl breeding conditions-- summer l9“7. Special Scientific Report No. “5. U.S.D.I., Fish and Wildl. Serv. 23 ROOM USE 0m! ROOM USE ONLO 30 Ill111111111111111 1111111 “1 1111111111 293105750289