Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Winter habitat for dabbling ducks on southeastern Vancouver Island, British Columbia Eamer, Joan 1985

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1985_A6_7 E24.pdf [ 4.84MB ]
Metadata
JSON: 831-1.0096084.json
JSON-LD: 831-1.0096084-ld.json
RDF/XML (Pretty): 831-1.0096084-rdf.xml
RDF/JSON: 831-1.0096084-rdf.json
Turtle: 831-1.0096084-turtle.txt
N-Triples: 831-1.0096084-rdf-ntriples.txt
Original Record: 831-1.0096084-source.json
Full Text
831-1.0096084-fulltext.txt
Citation
831-1.0096084.ris

Full Text

WINTER HABITAT FOR DABBLING DUCKS ON SOUTHEASTERN VANCOUVER ISLAND, BRITISH COLUMBIA by Joan Earner B.Sc, University of V i c t o r i a , 1977 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE i n THE FACULTY OF GRADUATE STUDIES (Department of Zoology) We accept t h i s t h e s i s as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA JUNE 1985 © Joan E l i z a b e t h Earner, 1985 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available for reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the head of my department or by h i s or her representatives. I t i s understood that copying or publication of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of ^-o otp<\ KJ\ The University of B r i t i s h Columbia 1 9 5 6 Main Mall Vancouver, Canada V 6 T 1 Y 3 i i ABSTRACT This study i s an examination of the use of coastal estuaries and nearby farmland as habitat by dabbling ducks (mallard and American wigeon) during migrating and wintering periods. I t s aim was to i d e n t i f y aspects of B r i t i s h Columbia coastal habitat of importance to dabblers through an anal-y s i s of the ducks' movements among habitat types and through a d e s c r i p t i o n of where and on what ducks feed. Data were c o l l e c t e d i n 1979 and 1980 along a 30 km stretch of c o a s t l i n e on southeastern Vancouver Island. Results are presented i n 3 parts. Part 1 examines the r e l a t i v e use of farm and coastal habitat through a se r i e s of censuses conducted weekly at 8 farm and 8 coastal s i t e s . The strong negative c o r r e l a t i o n between counts at farm and coastal s i t e s indicates that dabblers t r e a t them as a l t e r n a t i v e h a b i t a t s . The numbers of ducks on farms was p o s i t i v e l y correlated with the area of stand-ing water on the f i e l d s . Farm habitat, apparently preferred during warm, wet weather, was not used when f i e l d s were dry or frozen. Part 2 i s a d e s c r i p t i o n of feeding l o c a t i o n on f i e l d s , at estu-a r i e s and at a shallow, nonestuarine bay. I t i s based on observations at selected s i t e s at high and low t i d e l e v e l s . Each duck i n each observation period was c l a s s i f i e d as to l o c a t i o n and a c t i v i t y . Both species fed p r i m a r i l y i n shallow water, t h e i r feeding l o c a t i o n s h i f t i n g with the t i d e s . Both marsh and marine sections of estuaries were used extensively for feeding. The shallow bay was used e s p e c i a l l y by American wigeon at low t i d e i n f a l l and early winter. The high marsh areas at estuaries were p a r t i c u l a r l y a t t r a c t i v e to mallards when flooded by exceptionally high i i i t i d e s . Feeding i n t e n s i t i e s were consistently high at farm s i t e s for both species. In Part 3, 23 mallards and 40 American wigeon were shot while feeding i n estuarine locations commonly used f o r feeding. Analysis of g u l l e t contents revealed that both species ate a wide v a r i e t y of items. Main foods were, for mallards, seeds, invertebrates and green algae and, f o r American wigeon, green algae, roots, seeds and green vegetation. Algae and marine Invertebrates are not usually considered to be important dabbler foods i n estuaries. Major conclusions and recommendations are: 1) Both farm and coastal s i t e s are important to dabblers, with f i e l d s being favoured as feeding locations under good floo d conditions and coastal habitat being v i t a l during dry or freezing periods. As dabblers move among s i t e s , assessment and management of wintering dabbler habitat should be by wetland complexes rather than by i n d i v i d u a l estuaries. 2) Dabblers feed i n or near shallow water. F i e l d s that do not flood are not worth maintaining as dabbler habitat. Assessment of estu-arine marshes should consider the a v a i l a b i l i t y of food at a l l points of the t i d a l c y c l e . 3) As both species feed on a wide v a r i e t y of items, factors a f f e c t i n g shallow water flooding and thus food a v a i l a b i l i t y are more impor-tant than plant species composition. 4) Marine deltas and beaches are important as refuges from disturbance and as feeding grounds. They form an i n t e g r a l part of dabbler coastal habitat. i v TABLE OF CONTENTS Page Abstract i i L i s t of Tables v i i L i s t of Figures v i i i Acknowledgements i x Introduction • 1 Study Area • 7 Part 1: Selection of Site Introduction 8 Study Area 13 F i e l d Methods 15 Sources of Error 16 S t a t i s t i c a l Methods and Results 18 Discussion 25 Farm-Coast Movement and Degree of Flooding 25 Selection Among Farm Sites 28 Hunting Pressure 29 Freezing Weather 30 V Page Part 2: Selection of Feeding Location Introduction 32 Study Area 34 F i e l d Methods 37 Coastal Sites 37 Farm Sites 40 Disturbance 40 "Very High" Tide Observations 40 Sources of Error 41 S t a t i s t i c a l Methods and Results 43 Duck Counts i n Relation to Tide 43 Duck Counts i n Relation to Season 45 Ac t i v i t y - L o c a t i o n Data 45 Farm Sites 47 "High" and "Low" Tides 55 Ebb and Flow Tides 55 "Very High" Tide Data ( L i t t l e Qualicum River Estuary) 57 Disturbance Data 57 Discussion 62 Farm and Coastal Habitat 62 Feeding Locations at Coastal Sites 63 Disturbance 69 Movement among Feeding Locations i n Relation to Tide Level 70 v i Page Part 3: Selection of Food Items Introduction 72 Study Area and Methods 73 Results 78 Discussion 84 Conclusions and Recommendations 89 References 91 Appendix 1: Preliminary Study, February to A p r i l , 1980 96 Appendix 2: Census Data f o r Each S i t e 99 Appendix 3: Flooded Area Estimates at Farm Sites 103 v i i LIST OF TABLES Table I Counts of American wigeon at L i t t l e Qualicum River estuary, 1975-76 (N.K. Dawe, unpublished data). Table I I P a r t i a l c o r r e l a t i o n analyses between farm and coas t a l numbers, p a r t i a l l e d on date. Table I I I Multiple regression analyses: coastal numbers as a function of date and flooded area at farm s i t e s . Table IV Lengths of major features surveyed f o r a c t i v i t y - l o c a t i o n observations at co a s t a l s i t e s . Table V T i d a l range d e f i n i t i o n s . Table VI A c t i v i t y - l o c a t i o n c l a s s i f i c a t i o n s . Table VII Duck, counts at coas t a l s i t e s , "high" and "low" t i d e s . Table VIII Ducks on land: distance from water. Table IX Use of de l t a areas f o r feeding at "low" t i d e . Table X Feeding i n t e n s i t y at coas t a l s i t e s , "high" and "low" t i d e s . Table XI Feeding i n t e n s i t y at ebb and flow t i d e s . Table XII Summary of f i e l d notes f o r duck samples. Table XIII Gullet contents of 7 mallards shot while feeding i n marine zones. Table XIV Gullet contents of 16 mallards shot while feeding i n marsh zones. Table XV Gullet contents of 17 American wigeon shot while feeding i n marine zones. Table XVI Gullet contents of ?3 American wigeon shot while feeding i n marsh zones. v i i i LIST OF FIGURES Figure 1 Study area. Figure 2 Census t o t a l s f o r coastal and farm s i t e s and flooded areas at farm s i t e s . Figure 3 Mean duck numbers and mean flooded areas at farm s i t e s . Figure 4 Seasonal trend i n duck numbers at Rathtrevor Beach. Figure 5 A c t i v i t y - l o c a t i o n data for mallard and American wigeon at farm s i t e s . Figure 6 A c t i v i t y - l o c a t i o n data for mallard at coastal s i t e s during "high" t i d e s . Figure 7 A c t i v i t y - l o c a t i o n data f o r mallard at c o a s t a l s i t e s during "low" t i d e s . Figure 8 A c t i v i t y - l o c a t i o n data f o r American wigeon at c o a s t a l s i t e s during "high" t i d e s . Figure 9 A c t i v i t y - l o c a t i o n data f o r American wigeon at coastal s i t e s during "low" t i d e s . Figure 10 "Very high" tides ( L i t t l e Qualicum River estuary): number of ducks and feeding i n t e n s i t y as a function of gauge height. Figure 11 Where ducks went when disturbed (coastal s i t e s ) . Figure 12 Percent of daylight hours at d i f f e r e n t t i d a l ranges. i x ACKNOWLEDGEMENTS Funding f o r t h i s study was provided by the B r i t i s h Columbia Minis t r y of Environment, Ducks Unlimited (Canada) and the Natural Sciences and Engineering Research Council of Canada. Invaluable advice and a s s i s -tance were provided by E.J. Osmond-Jones, N e i l and Karen Dawe of the Canadian W i l d l i f e Service, Rodger Hunter of the B.C. Minist r y of Environ-ment and Ken Summers of Ducks Unlimited (Canada). - 1 -INTRODUCTION Rel a t i v e l y l i t t l e research has been c a r r i e d out on North American duck wintering habitat. Studies related to breeding, dominating the l i t e r -ature u n t i l the 1970s, were motivated by concern over a g r i c u l t u r a l drainage i n p r a i r i e nesting grounds and by the well-established l i n k between duck pro d u c t i v i t y and population l e v e l s . The r e a l i z a t i o n that wintering habitat i s being eroded at an alarming rate (Ladd et a l . 1974) and the recognition that factors a f f e c t i n g adult s u r v i v a l are poorly understood (Anderson and Burnham 1976) have broadened the scope of waterfowl research. Wintering habitat i s now being studied within 2 frameworks: (1) waterfowl population biology and (2) land management. Fluctuations i n duck abundance have been related primarily to fluctuations i n conditions on the breeding grounds (Anderson 1975) and longterm declines i n populations have been at t r i b u t e d mainly to loss of breeding habitat (Bellrose and Low 1978). However, changes i n wintering habitat may also a f f e c t population dynamics. Heitmeyer and Frederickson's (1981) study of mallards i n the M i s s i s s i p p i flyway established a d i r e c t l i n k between winter habitat and population f l u c t u a t i o n s . They demonstrated that recruitment rates were affected more by wetland conditions (related to r a i n f a l l ) on wintering grounds than by those on breeding grounds. Other evidence i s i n d i r e c t . The ducks' condition i n winter and early spring i s known to a l t e r s u s c e p t i b i l i t y to disease (Bellrose and Low 1978) and repro-ductive c a p a b i l i t y (Krapu 1981). On a longterm basis, declines of North American populations of several species, notably canvasback (Aythya  v a l i s i n e r i a ) and black duck (Anas rubripos), have been attributed to increased natural mortality and decreased breeding success r e s u l t i n g from - 2 -degradation of wintering habitat (Canadian W i l d l i f e Service 1980). The concerns of land managers r e l a t e more to the d i s t r i b u t i o n and l o c a l abundance of waterfowl than to population dynamics. Whether degrada-t i o n or loss of habitat i n a region k i l l s ducks or merely displaces them, the ducks are l o s t to that region. There are many documented cases of changes i n waterfowl d i s t r i b u t i o n patterns related to a l t e r a t i o n s i n wintering grounds (Greenwalt 1976). Coastal habitat i s r a p i d l y being l o s t . Ladd et a l . (1974) e s t i -mated that, by 1970, 73% of U.S. estuaries had been moderately to severely degraded through f i l l i n g , dredging, draining or use as garbage dumps. They predicted annual losses of 0.5 to 1.0% of remaining wetlands. In Canada, rates of loss of coastal habitat are comparable (Canadian W i l d l i f e Service 1980). Since waterfowl are a resource highly valued by North Americans (Hammack and Brown 1974, Canadian W i l d l i f e Service 1980), research which a s s i s t s i n making habitat management decisions i s badly needed. Waterfowl habitat management consists of (1) protection of habitat, (2) r e s o l u t i o n of land-use c o n f l i c t s r e s u l t i n g from i n d u s t r i a l , r e s i d e n t i a l or a g r i c u l t u r a l development and (3) habitat enhancement. Decision-making i n a l l 3 areas requires an understanding of the ducks' d i s t r i b u t i o n and h a b i t s . Midwinter censuses, conducted annually over U.S. wintering grounds, give only a general picture of the d i s t r i b u t i o n of species (Bellrose 1976). Studies of s p e c i f i c areas provide information on l o c a l d i s t r i b u t i o n and habitat requirements (Hartman 1963, Owen and Williams 1976, White and James 1978), important waterfowl foods (Yocom 1951, Landers et a l . 1976, Hughes and Young 1982), the role of a g r i c u l t u r a l lands ( H i r s t and Easthope 1981, Baldassarre and Bolen 1984) and the e f f e c t - 3 -of enhancement techniques such as water control and burning (Chabrek et a l . 1975). A few studies have examined biomass relationships between waterfowl and t h e i r food supplies (Sincock 1965, Burton and Hudson 1978). An estimated 80,000 dabbling ducks overwinter along the B r i t i s h -Columbia coast, f a r more than anywhere else i n Canada (Canadian W i l d l i f e Service 1980). Many more pass through the region during the protracted, variable migration periods (Summers and Campbell 1978). The Fraser Delta supports large numbers of mallard (Anas platyrhyncos platyrhyncos), American wigeon (A. americana), p i n t a i l (A. acuta acuta) and American green-winged t e a l (A. crecca C a r o l i n e n s i s ) . In the smaller estuaries that dot the B.C. coast, mallard and American wigeon are most abundant. B.C. estuaries, and i n p a r t i c u l a r those of Vancouver Island, are under considerable pressure for development from industry, housing and logging i n t e r e s t s (Hunter et a l . 1980). Habitat losses are already su b s t a n t i a l , with only 30% of the Fraser Delta remaining suitable for waterfowl (Canadian W i l d l i f e Service 1980). Despite the importance of the area to waterfowl and the recog-nized threats to habitat, our knowledge of dabbler d i s t r i b u t i o n and habits on the B.C. coast i s sketchy. D i s t r i b u t i o n data are l i m i t e d to sporadic a e r i a l censuses of sections of the coast (Summers and Campbell 1978 and Hunter et a l . 1980). Most of the habitat use research i s s p e c i f i c to the Fraser River Delta (Burgess 1970, Vermeer and Levings 1977, H i r s t and Easthope 1981), which, i n i t s si z e and degree of freshwater input, d i f f e r s greatly from other estuaries i n B.C. (Hunter et a l . 1980). Several conclu-sions from research i n the Fraser Delta are consistent with conclusions from studies of other large estuaries with extensive brackish to freshwater - 4 -marshes. Main conclusions are: (1) Dabblers feed i n marshes and on nearby a g r i c u l t u r a l land (Burgess 1970, Thomas 1976, H i r s t and Easthope 1981). (2) Bays and mudflats are used mainly for r e s t i n g (Benson 1961, • Burgess 1970, Tamisier 1976, Vermeer and Levings 1977). (3) Water l e v e l s play an important r o l e i n habitat choice (Burgess 1970, Chabrek et a l . 1975, Thomas 1976, H i r s t and Easthope, 1981). (4) Mallards eat mainly marsh plant seeds and waste grain (Yocom 1951, Burgess 1970, Hughes and Young 1982). (5) American wigeon eat mainly green leaves and stems of aqua-t i c , marsh and c u l t i v a t e d plants (Yocom 1951, Burgess 1970, Bellrose 1976 p. 206). Burgess (1970) also concluded that fluctuations i n numbers were rela t e d to migration patterns and the f i e l d s and marshes were favoured as feeding grounds at d i f f e r e n t times of the year. H i r s t and Easthope (1981) disagreed with the l a t t e r point, concluding that f i e l d s were used as an extension of marsh habitat. The few studies of other B.C. coastal s i t e s i n d i c a t e that marine and s a l t marsh habitat may also be important to both species for feeding (Munro 1943 and 1949, Hatler 1973, Dawe 1980 and Dawe and Lang 1980). These findings are consistent with those of Lynch (1939) and Yocom and K e l l e r (1961) for American wigeon and of Olney (1964) and Cronan and Halla (1968) for mallard. This study i s concerned with dabbler wintering habitat within the framework of coastal land management. The objective was to a s s i s t managers - 5 -i n i d e n t i f y i n g valuable coastal habitat by studying d i s t r i b u t i o n and habi-tat requirements of dabblers along a section of the coast of Vancouver Island, B r i t i s h Columbia. The study was i n i t i a t e d i n conjunction with a program of a e r i a l . censuses of the Vancouver Island c o a s t l i n e and vegetation mapping of i t s estuaries, undertaken by the B.C. M i n i s t r y of Environment (Hunter et a l . 1980). In order to i n t e r p r e t census data and marsh vegetation descriptions i n terms of dabbler habitat values, one must know something about the move-ments and feeding habits of the ducks. An important objective of t h i s study was to determine i f the conclusions drawn from Fraser Delta studies may be applied to other B.C. e s t u a r i e s . The thesis i s presented i n 3 parts, outlined below. Coastal B r i t i s h Columbia can be regarded, from a dabbling duck's point of view, as a large area of mainly unsuitable habitat comprised of dense f o r e s t s , mountains, steep, rocky shoreline and towns, interspersed with islands of suitable habitat (hereafter referred to as " s i t e s " ) . Three types of s i t e s are considered i n t h i s study: estuaries, non-estuarine shallow bays and farmland. Part 1 of the thesis i s an i n v e s t i g a t i o n of the f l u c t u a t i o n s i n dabbler numbers at coastal s i t e s . I proposed movement between coastal and farm s i t e s as an explanation for these f l u c t u a t i o n s and I examined factors governing that movement. This section i s based on a series of censuses at 16 s i t e s . Research and conservation e f f o r t s have focused on the marsh sections of estuaries, where plants, such as sedges, rushes and grasses grow. Dabblers also frequent the marine foreshore areas of estuaries, - 6 -however, as well as marine f l a t s i n shallow, non-estuarine bays. Nearby a g r i c u l t u r a l lands are also p o t e n t i a l feeding grounds f o r dabbling ducks. Part 2 of the study i s a desc r i p t i o n of where and when feeding occurs. I examined the r e l a t i o n s h i p between water l e v e l s and choice of feeding l o c a t i o n and assessed the value of f i e l d s , estuarine marshes, marine deltas and non-estuarine shallow bays as dabbler feeding habitat. This section i s based on observations of the ducks' behaviour and l o c a t i o n at several coastal and farm s i t e s . I stress the r e l a t i o n s h i p of time of feeding to the t i d a l c y c l e . P o t e n t i a l food items are very p a t c h i l y d i s t r i b u t e d i n estuaries i n the winter. Seeds f a l l o f f plants and are washed about with the r i s i n g and f a l l i n g t i d e s . Channels and f l a t s of various elevations and s a l i n i t i e s ensure a d i v e r s i t y of vegetation and invertebrate l i f e . The ducks are presented with a d i f f e r e n t s e l e c t i o n of p o t e n t i a l food items i n each part of the estuary. The aim of Part 3 was to i d e n t i f y major food items selected by ducks In the important feeding l o c a t i o n s . Analyses of g u l l e t contents of ducks c o l l e c t e d at estuarine s i t e s form t h i s s ection of the study. To summarize, t h i s i s a study of habitat s e l e c t i o n which examines 3 l e v e l s at which the ducks must make choices: (1) s i t e , (2) feeding loca-r t i o n and (3) food item. I t i s e s s e n t i a l to keep t h i s hierarchy i n mind. Each l e v e l of s e l e c t i o n can only be interpreted i n r e l a t i o n to the l e v e l s above i t . I t i s , f o r example, of l i t t l e use to c o l l e c t a sample of wigeon feeding i n one part of an estuarine marsh i f one has no idea how many wigeon feed i n that part of the marsh and how frequently wigeon are at the estuary at a l l . - 7 -STUDY AREA The study area i s situated i n the Nanaimo Lowland of southeastern Vancouver Island (Holland 1976), a h i l l y region l a r g e l y covered i n c o n i f e r -ous f o r e s t s , l y i n g between the mountains and the sea. The c o a s t l i n e i s mainly steep and rocky, interspersed with shallow bays and estuarine wet-lands. A g r i c u l t u r a l land use consists of small-scale c a t t l e production with some sheep farming and market gardening. Pastures are often poorly drained and subject to fl o o d i n g i n the winter. Winters are wet and mild with periods of f r o s t most often occur-r i n g i n December and January. Longterm c l i m a t i c data, taken from Environ-ment Canada's climate s t a t i o n on the Big Qualicum River (10 km northwest of L i t t l e Qualicum River estuary, at 8 m elevation) provide the following average values ( A i r Management Branch, B.C. Min i s t r y of Environment): 1275 mm r a i n f a l l annually and 47 cm snowfall; year-round average temperature of 9.2°C with maximum and minimum d a i l y temperatures i n January, the coldest month, of 4.9 and -0.4°C r e s p e c t i v e l y . The winter of 1980-81 was excep-t i o n a l l y mild and about average i n p r e c i p i t a t i o n , with a t o t a l of 1027 mm from October to March compared to the longterm mean of 1045 mm for these months. The t i d a l cycle i s mixed, semidiurnal: there are 2 complete t i d a l o s c i l l a t i o n s i n a t i d a l day, with i n e q u a l i t i e s both i n heights of highs and lows and i n time periods between them. T i d a l range i s about 2 to 4 m (Fisheries and Oceans S c i e n t i f i c Information and Publications Branch 1981). - 8 -PART 1: SELECTION OF SITE INTRODUCTION Ducks move about a great deal In winter. Any understanding of . their ecology must begin with an understanding of their movements — where do they go and under what influences? Direct study of the ducks' home ranges would be ideal but expensive and d i f f i c u l t , due to the problems involved with capturing and tracking migratory ducks in mountainous terrain. Information about movements, however, can be inferred from duck counts taken at different sites under different conditions. This is the approach taken here. Previous studies have shown that the numbers of dabbling ducks in migrating and wintering periods at any 1 site along the Br i t i sh Columbia coast fluctuate considerably. Fluctuations have been described on 2 scales: seasonal and day-to-day. Burgess (1970) detected a seasonal pattern in 2 years of aer ia l counts of dabbling ducks in the Fraser Delta and adjacent uplands. Peak numbers of the 4 most abundant species (mallard, American wigeon, p inta i l and American green-winged teal) occurred in November and December, followed by a sharp drop in January. Numbers increased again in late winter, reach-ing a lesser peak in A p r i l . He attributed these fluctuations to migration. He postulated 3 seasonal components of the winter duck population: a f a l l migrating component (late August through December), a wintering component (January) and a spring migrating component (end of January to early May). He assumed that low counts in the short wintering period occurred because - 9 -most of the ducks had migrated south. Similar patterns of seasonal fluctuations i n dabbler numbers were reported from a 10-year study of the Fraser Delta (Vermeer and Levings 1977) and from a 5-year study on the L i t t l e Qualicum River estuary (Dawe 1980) and the Nanoose Creek estuary (Dawe and Lang 1980). In a l l of these, peak dabbler numbers occurred i n November and December, followed by a sharp decline. Numbers at Nanoose and L i t t l e Qualicum appeared to remain low u n t i l the end of the spring migration period. Data were i n s u f f i c i e n t i n the Fraser Delta study to detect any spring Increase. Agencies of the p r o v i n c i a l and federal governments have conducted a e r i a l surveys of the Vancouver Island c o a s t l i n e at infrequent i n t e r v a l s since the mid 1960s. One of the highest counts of dabbling ducks occurred during January of 1971 ( T e r r e s t r i a l Studies Branch, B.C. Ministry of Environment, unpublished data). Dawe (1976) and Dawe and Lang (1980) also reported abnormally high counts of mallard and American wigeon at L i t t l e Qualicum River estuary and Nanoose Creek estuary i n January of 1979. Both of these months had long periods of freezing weather. Dawe suggested that during the freezing periods a g r i c u l t u r a l lands are unavailable and inland ducks are driven to the coast. This contradicts studies i n the Fraser Delta, where low coastal numbers i n January have been a t t r i b u t e d to the high incidence of freez i n g temperatures during that month (Burgess 1970 and Leach 1972). There are few studies with s u f f i c i e n t data to d i s t i n g u i s h day-to-day f l u c t u a t i o n s from seasonal ones. Dawe's (1980) report on the L i t t l e Qualicum River estuary provides the most extensive set of dabbling - 10 -duck counts — once a week for 5 years. However, his reduction of the data to monthly averages to display a seasonal trend masks wide day-to-day f l u c t u a t i o n s , as may be seen i n Table I. Fluctuations i n census data would r e s u l t i f the ducks displayed a . d a i l y rhythmicity of movement and i f the time of day of the counts varied. Daily patterns i n movements of waterfowl have been described by many authors (eg., Swanson and Sargeant 1972 and Campbell 1978). Local studies are contradictory. Hatler (1973) i n 8 days of morning and afternoon a e r i a l dabbler censuses on t i d a l f l a t s found consistently higher numbers i n the morning. Munro (1943) and Burgess (1971) described regular d a i l y farm-coast f l i g h t s f o r dabblers. H i r s t and Easthope (1981), however, found no' s i g n i f i c a n t d i f f e r e n c e between morning and afternoon censuses f o r mallards and American wigeon i n f i e l d s on the Fraser Delta. A preliminary census study was conducted from February to A p r i l , 1980 at L i t t l e Qualicum River and Nanoose Creek estuaries (Appendix 1). Wide day-to-day f l u c t u a t i o n s i n dabbler numbers occur i n the study area. Time of day does not appear to be a major f a c t o r . Numbers of mallards and American wigeon at one estuary and at a nearby farm were negatively correlated — when fewer ducks were i n the estuary, more ducks were inland. High counts on the farm occurred when large areas of the f i e l d s were flooded. This preliminary work suggested that coastal and inland a g r i c u l -t u r a l s i t e s may be a l t e r n a t i v e habitat types, and that duck movements among s i t e s may be regulated p a r t l y by the degree to which inland s i t e s are flooded. This i s contradictory to the conclusions of H i r s t and Easthope - 11 -Table I. Counts of American wigeon at L i t t l e Qualicum R. estuary 1975-76 (N.K. Dawe, unpublished data).  weekly counts average October 638 22 86 249 November 68 27 7 0 97 December 117 0 124 75 79 January 2 29 0 1 8 February 19 0 1 0 5 March 9 24 1 4 10 - 12 -(1981) i n a study of the Fraser River estuary and nearby farmlands. They found a p o s i t i v e c o r r e l a t i o n between dabbling duck numbers on coastal and farm s i t e s and concluded that ducks treat a g r i c u l t u r a l lands as an extension of coastal habitat, with f l u c t u a t i o n s i n numbers being related p r i m a r i l y to migration patterns. The purpose of t h i s part of the study, c a r r i e d out from October, 1980 through A p r i l , 1981, was to c l a r i f y the r e l a t i o n s h i p between coastal and farm s i t e s f o r southeastern Vancouver Island. I asked three questions: (1) Can movement between the 2 types of habitat account for some of the day-to-day and seasonal f l u c t u a t i o n s In dabbling duck numbers at coastal s i t e s ? (2) If so, i s the amount of standing water on the f i e l d s an important factor? (3) What happens i n freezing weather: do ducks leave the area or move from the f i e l d s to the coast? - 13 -STUDY AREA The study area consisted of 8 s i t e s located along a 30 km section of c o a s t l i n e , from Nanoose Harbour to the mouth of the L i t t l e Qualicum River, and 8 nearby a g r i c u l t u r a l s i t e s ( F i g . 1). The coastal s i t e s were of 2 types: estuaries, and non-estuarine shallow bays. Each estuarine s i t e included the i n t e r t i d a l marsh at the mouth of the r i v e r or creek, surround-ing beaches and deep water areas seaward to the l i m i t of v i s i b i l i t y with a 20-power spotting scope. Bay s i t e s , which contained l i t t l e or no vascular plant growth, consisted of beach and deep water areas. The farm s i t e s were blocks of pastureland ranging i n s i z e from 19 to 138 hectares. Vegetation varied within and among s i t e s from heavily-grazed pastures of mixed grasses and forbs to f i e l d s of t a l l , coarse canary grass (Phalaris arundinacea) interspersed with clumps of sedge (Juncus effusus). Two of the estuaries, L i t t l e Qualicum River and Nanoose Creek, form the Qualicum National W i l d l i f e Area, managed by the Canadian W i l d l i f e Service (see Dawe 1976 and 1980, Dawe and Lang 1980 and Dawe and White 1982 for descriptions of vegetation). Small mowed f i e l d s are maintained adja-cent to each estuary, primarily as habitat f o r dabbling ducks. These f i e l d s were treated as 2 ad d i t i o n a l farm s i t e s . - 14 -F i g . 1. Study area. Scale: 1:150,000. Numbers indicate coastal s i t e s : 1, 2, 4, 6 and 8 are estuaries; 3, 5 and 7 are bays. Letters indicate farm s i t e s ( l o t numbers from map sheets 92F/7, 1976 and 92F/8, 1977, Surveys and Mapping Branch, Department of Energy, Mines and Resources, Ottawa): A=lot 8; B=lots 18 and 37; C=lot 46; D=lot 24; E=lots 100 and 25; F=lot 62; O l o t 169; H=lot 44; I=meadows at Nanoose Unit, L i t t l e Qualicum National W i l d l i f e Area; J=meadows at Marshall-Stevenson Unit, L i t t l e Qualicum National W i l d l i f e Area. - 14a -- 15 -FIELD METHODS Censuses were conducted once a week from October 6, 1980 to A p r i l 28, 1981. Add i t i o n a l censuses were c a r r i e d out on freez i n g days, defined as days that followed a night of f r o s t and for which freezing temperatures were forecast f o r at l e a s t most of the day. Five freezing days occurred during the study period. Each census consisted of counts of mallards and American wigeon at a l l farm and coas t a l s i t e s . P r i o r to the study period, a census route was established at each s i t e . Observations were made on foot and from ve h i c l e s , using binoculars and spotting scopes, at predetermined points along the routes. Three observation towers were constructed at s i t e s where there were no natural vantage points. Ducks were counted i n d i v i d u -a l l y whenever possible. When many ducks were disturbed and took f l i g h t , numbers were estimated. Data were recorded on cassette tapes. Several routes were a l t e r e d i n the early weeks of the study i n order to minimize disturbance to the ducks. Coastal and farm s i t e s were censused concur-r e n t l y by 2 observers, from approximately 0900 to 1600 hours. Due to a tape recorder malfunction, most of the coastal census f o r A p r i l 21 was l o s t . Thirty-one complete censuses were obtained. Appendix 2 contains census r e s u l t s f o r each s i t e . At each farm s i t e , the area of flooded land was estimated f o r each census day. Detailed maps of the s i t e s were drawn from a e r i a l photo-graphs. Each census day, boundaries of a l l flooded areas were sketched on tracin g paper placed over the maps. When standing water was very broken up, the proportion (out of 10) of each f i e l d that was covered i n water was - 16 -estimated. Gauges (2 m s t r i p s of wood painted i n a l t e r n a t i n g 10cm black. and white s t r i p e s ) were i n s t a l l e d at low elevation points at the farm s i t e s to a s s i s t i n comparing water l e v e l s . On freezing days, the areas of unfrozen standing water were estimated. When the flooded area was too small and scattered to measure accurately, i t was recorded as 0 or 0.1 ha, whichever seemed c l o s e s t . This method was not su i t a b l e f o r s i t e s I and J (Figure 1), the 2 f i e l d s included i n the National W i l d l i f e Area, due to t h e i r smaller s i z e . Dimensions of puddles were estimated i n metres and the flooded areas 2 recorded to the nearest 10 m , r e s u l t i n g i n a greater degree of p r e c i s i o n f o r these s i t e s . Data from these s i t e s are not included i n the s t a t i s t i c a l analyses, but are plotted on F i g . 3 for comparison. Sources of Error This study i s not concerned with the estimation of numbers of ducks i n the region but rather with the re l a t i o n s h i p between numbers on farms and on the coast. As such, errors made i n duck counts and i n flooded-area estimates should not a f f e c t the conclusions i f the errors are eit h e r randomly d i s t r i b u t e d or consistent. Observer error was made as consistent as possible by following standardized procedures and by assigning 1 observer to coastal s i t e s and 1 to farm s i t e s . A l l flooded-area estimates were made by the same observer. V i s i b i l i t y v aried greatly with weather conditions at a l l s i t e s ; In fog and heavy r a i n , duck numbers were probably underestimated. However, as both farm and coastal s i t e s were affected i n the same manner, errors related to - 17 -weather c o n d i t i o n s s h o u l d not b i a s the r e s u l t s . Two s o u r c e s o f b i a s e r r o r were c o n s i d e r e d . V i s i b i l i t y , p a r t i c u -l a r l y a t the farm s i t e s , was a f f e c t e d by the h e i g h t o f v e g e t a t i o n . T h i s c o u l d have c o n t r i b u t e d t o lower c o u n t s a t the farm s i t e s i n the l a t t e r p a r t of the census p e r i o d , when g r a s s was t a l l e r . Ducks were a l s o h a r d e r t o s p o t on the f i e l d s when t h e r e was l i t t l e o r no s t a n d i n g w a ter. T h i s c o u l d have c o n t r i b u t e d t o lower counts a t farm s i t e s a t times o f low f l o o d i n g . An attempt was made t o a s s e s s the e x t e n t o f t h i s problem by w a l k i n g through the farm s i t e s a f t e r c e n s u s i n g i n the s p r i n g and a t o t h e r times of low f l o o d i n g . As few uncounted ducks were f l u s h e d o u t , I f e e l t h a t b i a s e r r o r was not s i g n i f i c a n t . - 18 -STATISTICAL METHODS AND RESULTS For each census, counts of mallards and American wigeon were t o t a l l e d f o r farm and for coastal s i t e s . Results f o r a l l 32 censuses are shown i n F i g . 2. In order to test the hypothesis that f l u c t u a t i o n s i n duck numbers on the coast are due, i n part, to movement between the coast and farmland, l i n e a r c o r r e l a t i o n analyses were performed (Table I I ) . P a r t i a l c o r r e l a t i o n c o e f f i c i e n t s between coastal and farm counts, p a r t i a l l e d on date, were calculated. These analyses assess the l i n e a r r e l a t i o n s h i p between numbers of ducks on the coast and at farms, taking into account v a r i a t i o n due to an o v e r a l l l i n e a r decline i n numbers during the study period. As may be seen i n F i g . 2, there were serious departures from l i n e a r i t y at both ends of the study period. Coastal numbers increased f o r both species through October, while farm numbers l e v e l l e d o f f to near-zero i n October and A p r i l . Accordingly, the analyses were performed on the 24 censuses from November to March. I t i s cl e a r from F i g . 2 that the extremely high coastal and low farm counts from the 5 censuses on freez i n g days contribute greatly to these s i g n i f i c a n t negative c o r r e l a t i o n s . To test i f the r e l a t i o n s h i p held f o r nonfreezing days, the c o r r e l a t i o n analyses were performed omitting these 5 censuses. For a l l data sets, duck numbers at coastal s i t e s are s i g n i f i c a n t l y negatively correlated with numbers at farm s i t e s . The area of each flooded patch sketched on the maps of the farm s i t e s were calculated on a microcomputer programmed to operate as a d i g i t i z e r . When an estimate was i n the form of the proportion of a f i e l d - 19 -F i g . 2. Census t o t a l s f o r c o a s t a l and farm s i t e s and f l o o d e d a r e a s a t farm s i t e s . Shaded s t r i p s d e s i g n a t e f r e e z i n g p e r i o d s . - 19a -- 20 -Table I I . P a r t i a l c o r r e l a t i o n analyses between farm and coastal numbers, p a r t i a l l e d on date. Data set N P a r t i a l c o r r e l a t i o n c o e f f i c i e n t between farm and co a s t a l numbers 3 MALLARD a l l census days 24 -.68** nonfreezing days 19 -.64** freezing days 5 -.99** AMERICAN WIGEON a l l census days 24 -.83** nonfreezing days 19 -.84** freezing days 5 -.91** a Significance i s assessed with a one-tailed t - t e s t with N-3 df. **P<0.01 - 21 -that was flooded, the f i e l d area was calculated and then m u l t i p l i e d by t h i s proportion. For each census, flood areas were t o t a l l e d f or each farm s i t e and then for a l l farm s i t e s . In F i g . 2 the t o t a l flooded areas are displayed for a l l censuses. Flooded areas f o r each s i t e are l i s t e d i n Appendix 3. To test the hypothesis that movement by dabbling ducks between the coast and farms i s related to the amount of standing water on f i e l d s , l i n e a r regression analyses were performed (Table I I I ) . In each t e s t , the indepen-dent variables are date and t o t a l flooded area at the farm s i t e s ; the depen-dent v a r i a b l e i s duck numbers on the coast. As i n the preceding c o r r e l a t i o n analyses, only censuses from November through March are included since a l i n e a r model i s used to adjust for the seasonal decline i n o v e r a l l numbers. Nonfreezing and freezing days were also examined separately. The tests of s i g n i f i c a n c e f o r the p a r t i a l regression c o e f f i c i e n t s are one-tailed because I hypothesized that (a) large flooded areas should r e s u l t i n a movement of ducks to the farms, and (b) when the f i e l d s dry up, ducks should return to the coast. This would r e s u l t In negative c o e f f i c i e n t s . The o v e r a l l regressions and the p a r t i a l regression c o e f f i c i e n t s are s i g n i f i c a n t f o r both species f o r a l l census days and f o r nonfreezing days. As the regressions f o r the freezing days are n o n s i g n i f i c a n t , further tests on these data sets are not warranted (Zar 1974). The c o e f f i c i e n t s of determin-ati o n express the percentage of the v a r i a b i l i t y i n c o a s t a l numbers that may be explained by the date and by the flooded areas at the farm s i t e s . If the ducks choose farm habitat l a r g e l y on the basis of amount of standing water, the mean number of ducks at each farm s i t e should depend upon - 22 -Table I I I . Multiple regression analyses: coastal numbers as a function of date and flooded area at farm s i t e s . Data set N F(df=2 and N-3) Co e f f i c i e n t of determination (R 2 x 100) P a r t i a l regression c o e f f i c i e n t f o r flooded a r e a 3 MALLARD a l l census days 24 6.38** 37.8 -8.25** nonfreezing days 19 7.70** 49.1 -2.05** freezing days 5 2.20 n s - -AMERICAN WIGEON a l l census days 24 22.21** 67.9 -13.30** nonfreezing days 19 11.35** 58.7 -11.16** freezing days 5 10.31 n s '- -a Significance i s assessed with a one-tailed t - t e s t with N-3 df. **P<0.01 n s n o t s i g n i f i c a n t (P>0.05) - 23 -the mean flooded area at that s i t e , with s i t e s with higher flooded areas supporting more ducks. Duck numbers and flooded areas at each farra s i t e were averaged over a l l 32 censuses. The r e s u l t s are displayed i n Fig. 3. Linear regression analyses were performed on these data. For American wigeon, the r e s u l t i n g regression c o e f f i c i e n t of 18.7 was s i g n i f i c a n t l y greater than zero i n a one-tailed t - t e s t with 8 df(P<0.01). The c o e f f i -c i e n t of determination was 90.0%. For mallards, the regression c o e f f i c i e n t of 6.6 was not s i g n i f i c a n t l y greater than zero (P>0.05). I t i s apparent from F i g . 3 that one data point, s i t e B, greatly reduced the slope of the regression l i n e . Exclusion of s i t e B re s u l t s i n a s i g n i f i c a n t (P<0.01) regression c o e f f i c i e n t of 21.6 and a c o e f f i c i e n t of determination of 92.3% fo r mallards. - 24 -F i g . 3. Mean duck numbers and mean flooded areas at farm s i t e s . Letters r e f e r to farm s i t e locations ( F i g . 1). A l l means are calculated from the same 32 censuses. - 24a -1 0 0 h « |80h 3 C • E • C » A •D je60|- MALLARD u 3 •o c 4 0 h a l 2 0 U ^ G * « B 1 1 1 • • J 1 2 3 4 5 100 |80| 3 *60| u > a> S 2 0 Mean flooded area (ha) • E • A • B • c AMERICAN WIGEON • D G1 2 3 4 5 Mean flooded area (ha) - 25 -DISCUSSION Farm-Coast Movement and Degree of Flooding The very strong negative c o r r e l a t i o n s between t o t a l counts at farm and c o a s t a l s i t e s (Table II) suggest that the 2 types of habitat are used a l t e r n a t i v e l y . As there were frequent s h i f t s i n the r e l a t i v e use of coastal and farm s i t e s throughout the study period, I concluded that the wide day-to-day f l u c t u a t i o n s at any 1 s i t e are due i n part to t h i s r e l a t i o n s h i p . The extent of these f l u c t u a t i o n s (Appendix 2) suggests that movement among farm s i t e s and among coastal s i t e s a lso must be common. This i s discussed further i n Part 2. It i s c l e a r from F i g . 3 that the number of ducks inland i s strongly t i e d to the area of flooding on the f i e l d s . This means that under conditions of high flooding, there i s a movement of dabblers away from the coast (Table I I ) . Flooding, r e s u l t i n g from the accumulation of rainwater i n poorly drained f i e l d s and from overflow of low-gradient sections of creeks, follows a seasonal pattern that depends on the seasonal pattern of p r e c i p i t a t i o n . It appears, then, that seasonal f l u c t u a t i o n s i n duck numbers at i n d i v i d u a l coastal s i t e s are due i n some measure to f l u c t u a t i o n s i n flooding of f i e l d s . The extent of t h i s dependency may be underestimated because of the r e l a t i v e amounts of farm and c o a s t a l habitat surveyed. The coastal census route probably included most dabbler habitat from Nanoose Bay to L i t t l e Qualicum River estuary. However, several farms i n the v i c i n i t y were not censused but were frequented by dabbling ducks. This assessment i s J j - 26 -based on conversations with l o c a l residents, observations by the author and an a e r i a l survey of the study area conducted on November 26, 1980. When f i e l d s were a t t r a c t i v e to ducks, then, a smaller proportion of the true number of ducks i n the v i c i n i t y would have been counted than at times when coa s t a l habitat was favoured. This i s supported by the high t o t a l counts of both species on freezing days, when farm habitat was the l e a s t a t t r a c t i v e ( F i g . 2 ) . This means that seasonal f l u c t u a t i o n s i n t o t a l counts may s t i l l r e f l e c t the coastal-farm r e l a t i o n s h i p as w e l l as migration and m o r t a l i t y . As the ducks' home ranges are completely unknown, i t i s not possible to separate these e f f e c t s . I f the same inverse r e l a t i o n s h i p between farm and c o a s t a l habitat occurs i n the Fraser Delta area, i t could a f f e c t the seasonal fluctuations that have been a t t r i b u t e d to migration patterns (Burgess 1970, Leach 1972, Vermeer and Levings 1977). Studies i n the Fraser Delta have included most, of the coastal habitat and some adjacent a g r i c u l t u r a l lands. However, many farms i n the Lower Fraser v a l l e y f l o o d and are frequented by ducks but were not censused. The midwinter dropoff i n numbers observed i n these studies i s during the season of extensive f i e l d flooding. Burgess (1970) reported that the highest number and proportion of ducks on farms coincided with the mid-winter decline i n t o t a l duck counts. H i r s t and Easthope (1981) concluded that degree of flooding was the most important f a c t o r i n choice of farm s i t e f o r dabblers i n the Fraser Delta area. However, they also concluded from a p o s i t i v e c o r r e l a t i o n between farm census t o t a l s and counts on adjacent c o a s t a l habitat that dabblers treat flooded f i e l d s as an extension of coastal habitat, with - 27 -migration patterns accounting for f l u c t u a t i o n s i n numbers. I believe that t h i s conclusion i s not warranted, as flooding i n the f i e l d s remained stable through most of t h e i r short study period (October to December). A s i m i l a r p o s i t i v e r e l a t i o n s h i p presumably would have resulted i n the present study had flooded area on the f i e l d s remained st a b l e . H i r s t and Easthope's f i n a l conclusion, that eliminating or reducing f i e l d f looding would have l i t t l e e f f e c t on migrating and wintering dabbling ducks, i s not substantiated. Studies i n B r i t a i n and Louisiana indicate that changes i n d r a i n -age patterns on coas t a l farm lands can a f f e c t wintering dabbling ducks. Owen and Williams (1976) found inland pastures to be the most important feeding habitat for European wigeon (Anas penelope) i n B r i t a i n . Owen and Thomas (1979) conducted a study i n England on the Ouse Washes (a large pasture area subject to extensive winter f l o o d i n g ) . Before reserves were created, f l u c t u a t i o n s i n European wigeon numbers were t i e d c l o s e l y to l e v e l s of flooding; the creation of permanent water bodies increased dabbler usage and reduced f l u c t u a t i o n s i n numbers. In coastal Louisiana, Tamisier (1976) concluded that wintering p i n t a i l s and American green-winged t e a l fed mainly In flooded f i e l d s , while Chabrek et a l . (1975) found that improving drainage of a g r i c u l t u r a l lands greatly reduced t h e i r use by a l l dabblers. Selection among Farm Sites The r e s u l t s of the s i t e - b y - s i t e analysis depicted i n F i g . 3 are somewhat s u r p r i s i n g . The farm s i t e s d i f f e r e d i n many respects: s i z e , - 28 -proximity to roads and houses, hunting pressure, vegetation type, flooding regime (see Appendix 3) and the depth, number and shoreline length of the puddles. Presumably these factors greatly influence the types of food a v a i l a b l e and the degree of disturbance. Yet a l i n e a r r e l a t i o n s h i p with, mean flooded area accounted f o r 90.0% of the v a r i a b i l i t y among mean American wigeon numbers at the farms and 92.3% of the v a r i a b i l i t y among mean mallard numbers (excluding s i t e B). The presence of water appears to be both necessary and s u f f i c i e n t to make a f i e l d a t t r a c t i v e to dabblers. Proximity to coastal habitat i s not enough. S i t e I, adjacent to Nanoose Creek estuary, c o l l e c t e d standing water i n one small area only and received very l i t t l e use. Si t e J , adja-cent to L i t t l e Qualicum River estuary did not flood and was not used. American wigeon apparently require a larger minimum flooded area. Sites F and G, with low mean flooded areas, received f a i r l y regular use by mallards but only occasional use by American wigeon at times of highest flooding. Sites H and I, which r a r e l y were flooded to any measurable degree, were used occasionally by mallards only. Why s i t e B should be so unpopular with mallards i s d i f f i c u l t to say. I t was close to a main road and was hunted, but so were several other s i t e s . Its greater water depth i s the most l i k e l y explanation. Studies have indicated that water depth i s important i n habitat s e l e c t i o n f o r breeding mallards (Joyner 1980, Mack and Flake 1980) and for wintering mallards (Chabrek et a l . 1975). Thomas (1976) found tha^ mallards were unable to feed i n water deeper than 40 cm. Part 2 of t h i s study indicates - 29 -that mallards r a r e l y fed on land, while American wigeon frequently grazed on land near standing water. Hunting Pressure Movement between coa s t a l and inland s i t e s i n the Fraser Delta area (Benson 1961, Burgess 1970, Leach 1972) and on southeastern Vancouver Island (Munro 1943) has been rel a t e d to hunting pressure. Evening f l i g h t s from the coast to f i e l d s f o r night feeding and morning f l i g h t s to return to the coast to loaf offshore have been observed during the hunting season. This did not appear to occur i n the study area, possibly because only a few of the s i t e s (farm and coastal) were hunted. Regular morning and evening f l i g h t s were not observed and residents of the farms that were censused did not believe that such f l i g h t s occurred. Ducks of both species were observed on several occasions feeding and r e s t i n g at night both at farm s i t e s and at the L i t t l e Qualicum River estuary. Tamisier (1976) concluded that hunting pressure may influence the d i s t r i b u t i o n of ducks among s i t e s i n Louisiana. H i r s t and Easthope (1981), however, reported that dabblers did not avoid f i e l d s subject to hunting disturbance i n the Fraser Delta area. This would appear to be the case i n t h i s study a l s o . Farm s i t e s B and C were hunted occasionally and s i t e D was hunted f a i r l y frequently; hunting was not allowed at the other farms. Duck numbers at a l l s i t e s were t i e d c l o s e l y to flooded area, with the si n g l e exception of mallard numbers at s i t e B ( F i g . 3). - 30 -Freezing Weather The most dramatic s h i f t s between farm and coast use occurred during freezing weather. In both the December and February cold s p e l l s , large numbers of ducks occupied the coastal s i t e s while few remained inland ( F i g . 2). There was no i n d i c a t i o n that large numbers of ducks l e f t the study area during or following freezing periods. During fr e e z i n g periods, when fewer ducks were on farms, more were found at the coast (Table I I ) . The regressions between numbers at the coast and unfrozen flooded areas were not s i g n i f i c a n t (Table I I I ) . Any true r e l a t i o n s h i p with flooding, however, i s l i k e l y to be non-linear; as more water freezes, le s s of the remaining water i s shallow. As discussed above, water depth Is probably important, p a r t i c u l a r l y f o r mallards. The ducks' behaviour when f i e l d s are frozen might be expected to depend upon the extent of the food resources that remain. I t i s possible, then, that dabblers would migrate south i f coastal habitat were i n s u f f i -c i e n t to meet t h e i r needs or i f the cold s p e l l continued f o r a longer period. In England, at the Ouse Washes, i t has been reported that large numbers of wintering mallards and European wigeon remain i n the area f o r freez i n g periods of a few days (Thomas 1976) but move to the coast during longer freezing s p e l l s (Owen and Thomas 1979). During the short periods of f r o s t they feed on waste potatoes and leaves of winter wheat. This type of food resource i s scarce i n my study area but may be abundant i n other parts of coastal B r i t i s h Columbia, notably the Fraser Delta. The winter of 1980-81 contained fewer than average freezing days. To estimate the number of days with freezing i n the f i e l d s , days with - 31 -maximum temperatures of <3°C at the climate s t a t i o n on the Big Qualicum River were counted ( A i r Management Branch, B.C. M i n i s t r y of Environment). There were 8 such days during the study period compared to an annual average of 19 (SD=10.1) from 1964-1980. Results from t h i s study ind i c a t e that c o a s t a l habitat may be c r i t i c a l l y important when f i e l d s are frozen. However, t h i s conclusion cannot be assumed to hold f o r other areas, since the use made of coastal s i t e s during times of f r e e z i n g i n nearby f i e l d s may vary depending on a g r i -c u l t u r a l practices and the r e l a t i v e amounts of farm and coastal habitat. No l o c a l dabbling duck studies containing farm and coastal census data have recorded the extent of f r e e z i n g i n the f i e l d s . I recommend that researchers assessing dabbler habitat i n c o a s t a l areas subject to f r o s t apportion t h e i r sampling e f f o r t s to include freezing periods. A set of dabbler counts at a coastal s i t e taken on freezing and nonfreezing days would y i e l d two important types of information: (1) how many ducks use the s i t e as part of t h e i r winter habitat (from high freezing counts) and (2) the importance of nearby farms as a l t e r n a t i v e s i t e s (from the magnitude of the d i f f e r e n c e between freezing and nonfreezing counts). - 32 -PART 2: SELECTION OF FEEDING LOCATION INTRODUCTION Part 2 deals with s e l e c t i o n of feeding l o c a t i o n . I demonstrated i n Part 1 that mallards and American wigeon on southeastern Vancouver Island move between coastal and farm habitat types, influenced by water conditions on the farms and by the presence or absence of freezing temper-atures. But are both types of habitat used for feeding? Coastal habitat i s very patchy, ranging from dense, high-marsh plant communities to unvege-tated marine mudflats. Which of these patches are important for feeding? T i d a l l y influenced habitat i s continuously changing. How do the s h i f t i n g water depths at c o a s t a l s i t e s a f f e c t the ducks' feeding patterns? To answer these questions, I attempted to systematically describe dabblers' a c t i v i t i e s and locations at coastal and farm s i t e s . Most pre-vious studies of ducks wintering i n coastal areas of North America and B r i t a i n have based t h e i r conclusions regarding feeding l o c a t i o n on counts of ducks i n d i f f e r e n t habitat u n i t s , on crop or gizzard contents of c o l l e c t e d ducks and on casual behavioural observations. The general con-c l u s i o n has been that dabbling ducks feed primarily on vascular plants i n t i d a l marshes and a g r i c u l t u r a l lands while most diving ducks feed on invertebrates i n marine estuaries and bays (Yocom and K e l l e r 1961, McMahan 1970, Kerwin and Webb 1972, Vermeer and Levings 1977, Hughes and Young 1982).. Some authors (Lynch 1939, Munro 1943 and 1949, Olney 1964, Dawe 1980), however, have observed dabblers, including mallards and American wigeon, feeding on marine f l a t s . - 33 -T i d a l f l u c t u a t i o n s are considered to a f f e c t feeding dabblers. Burgess (1970) concluded that dabblers feed more i n t i d a l marshes at low receding t i d e s . Munro (1949) also considered receding tides to be preferred by American wigeon feeding on marine f l a t s . In t h i s study I examine feeding habitat preferences through a d e s c r i p t i o n of feeding l o c a t i o n s . However, e x t r i n s i c factors such as predation and competition may also influence the animals' choice of l o c a -t i o n (Wiens 1976). As disturbance (related to predation) has been reported to a f f e c t d i s t r i b u t i o n i n dabblers (Hatler 1973, Tamisier 1976, Owen and Williams 1976) an assessment of the e f f e c t of disturbance on feeding l o c a -t i o n i s included. - 34 -STUDY AREA The study area consisted of 3 estuarine s i t e s ( L i t t l e Qualicum River, Englishman River and Nanoose Creek), 1 bay s i t e (Rathtrevor Beach) and a l l farm s i t e s censused i n Part 1 ( F i g . 1). Species composition of marsh vegetation and algae varies among the 3 estuaries (see Dawe 1976 and Dawe and Lang 1980 for species inventories of Nanoose and L i t t l e Qualicum, and Dawe and White 1982 for an analysis of the vegetation ecology of L i t t l e Qualicum). Certain features are common to the 3 estuaries, and to others i n the region. They are outlined below. In the marsh zone of the estuary, stream and t i d a l channels mean-der through vegetated f l a t s and sometimes form small ponds. Channel and pond bottoms are sparsely vegetated, t y p i c a l plants being Ruppia maritima and filamentous algae. The lower-elevation f l a t s and lower channel edges are often sparsely vegetated i n winter. Vegetation might include S a l i c o r n i a spp., Glaux maritima, D i s t i c h l i s spicata, Plantago spp. as well as Fucus spp. and filamentous algae. Upper f l a t s often support dense stands of vegetation i n winter. T y p i c a l l y present would be plants of the grass (Graminae), sedge (Cyperaceae) and rush (Juncaceae) families and P o t e n t i l l a p a c i f i c a . Dense, uniform stands of sedge (Carex spp.) occur along upper channel edges or as patches. In winter, the leaves and stems of most plants are dead; seeds and plant debris d r i f t about the marsh with t i d a l f l u c t u a t i o n s . The d e l t a fans out into s h i f t i n g , sometimes i n d i s t i n c t channels seaward of the marsh zone. These channels often support heavy growths of - 35 -filamentous algae i n the spring. Substrate type and amount of a l g a l growth on the beach vary widely from s i t e to s i t e . D r i f t seaweed and eelgrass (Zostera marina) wash up and down the beach with the t i d e , accumulating i n banks at high t i d e mark. The amount of d r i f t seaweed varies from day to day. The bay s i t e , Rathtrevor Beach, i s a shallow, sandy beach i n Craig Bay. The farm s i t e s are described i n Part 1 and mean flooded areas are l i s t e d i n Appendix 3. Area measurements at coastal s i t e s would be mis-leading, as surveys were l i n e a r , following c o a s t l i n e s , streams and chan-nels . Table IV presents approximate lengths of the main features surveyed. Measurements are of s t r a i g h t l i n e s drawn along sections of the shorelines on a 1:50,000 map. The numerous smaller channels surveyed are not included. - 36 -Table IV. Lengths of major features surveyed f or a c t i v i t y - l o c a t i o n observa-tions at coastal s i t e s . o Site Length of coast-l i n e surveyed (km) Distance surveyed along main stream(s) or t i d a l channel(s)(km) L i t t l e Qualicum R. 1.5 Lower r i v e r and channels 1.2 Englishman R. 1.3 East channels 0.8 Lower r i v e r and west channels 1.1 Nanoose Cr. 0.9 Nanoose Cr. 0.4 Bonell Cr. 0.4 Rathtrevor Beach 2.1 0 - 37 -FIELD METHODS Coastal Sites Observations at coastal s i t e s were made from October, 1980 to A p r i l , 1981 at times chosen to correspond to "low" and "high" tides (defined i n Table V). A schedule of observation periods was drawn up using t i d e tables and the periods were randomly d i s t r i b u t e d among the 4 s i t e s . As r i v e r flow and weather factors such as wind v e l o c i t y affected water l e v e l s , t i d e tables were used only as rough guides. Gauges were i n s t a l l e d at the 3 estuarine s i t e s to a i d i n standardizing water l e v e l s at which observations were made. The gauges consisted of s t r i p s of wood 3 m long, painted i n a l t e r n a t i n g 20 cm black and white s t r i p e s . During prelim-inary studies, s p e c i f i c gauge readings were established as corresponding to about the middle of "high" and "low" t i d a l ranges. Observations were started when I estimated that the water would reach t h i s l e v e l on the gauge during the observation period. Methods were established during preliminary studies. The route followed at each s i t e was chosen as a compromise between maximum coverage and minimum disturbance. Parts of Englishman River estuary were surveyed by v e h i c l e , as t h i s appeared to dist u r b the ducks less than did people on foot . From set vantage points along the census route ducks were scanned (Altmann 1974) for a c t i v i t y and l o c a t i o n (Table V I ) . Each duck was observed f o r about 2 seconds through 8 X 10 binoculars or a 20X scope. Data were recorded on cassette tapes. Adjustments frequently had to be made f o r movement. I f , for example, a c t i v i t i e s were recorded f o r 10 - 38 -Table V. T i d a l range d e f i n i t i o n s . T i d a l Approximate Marsh water l e v e l s Marine water l e v e l s range t i d a l height (m) a Very Above 4.4 Upper f l a t s shallow. L i t t l e shallow water high Other zones deep. at t i d e l i n e . Beach and delta zones deep. No tidepools. High 3.9-4.4 Upper f l a t s dry. Lower f l a t s and sedge stands shallow. Similar to "very high". Mid 3.0-3.8 Tr a n s i t i o n between "high" and "low". Low 2.0-2.9 Very low elevation zones only shallow. Other marsh zones dry. Wide band of shallow water at t i d e l i n e . Gravel bars exposed. Delta zone shallow. Some tidepools Very Below 2.0 A l l vegetated zones Similar to "low", low dry. more tidepools. a From.tide tables ( F i s h e r i e s and Oceans S c i e n t i f i c Information and Publications Branch 1981). - 39 -Table VI. A c t i v i t y - l o c a t i o n c l a s s i f i c a t i o n s C l a s s i f i c a t i o n Description ACTIVITY Feed Any behaviour connected with feeding, i n c l u d i n g up-ending, swimming with head under water and pecking. Move Any movement not obviously associated with feeding, including s o c i a l i n t e r a c t i o n s . Rest Any p o s i t i o n not moving. LOCATION Land No v i s i b l e water; marsh, d e l t a or marine zones. Distance from water estimated (within 1 m, 1 m to 5 m, or more than 5m). Shallow water Water les s than about 30 cm deep. - Marsh Stream and t i d a l channels and surrounding f l a t s supporting well-established, t i d a l l y - i n f l u e n c e d marsh vegetation. - Delta The seaward extension of stream channels, beyond the marsh zone to as f a r as the channels remain d i s t i n c t . - Marine Beach areas seaward of the marsh zone, unvegetated or with growths of algae or eelgrass. Deep water Water deeper than about 30 cm; marsh, delta or marine zones - 40 -mallards i n 1 l o c a t i o n then the ducks moved to another l o c a t i o n yet to be censused, an adjustment was required. I surveyed the second l o c a t i o n and l a t e r subtracted 10 mallards i n proportion to the a c t i v i t i e s of a l l mallards i n that l o c a t i o n . If movement was too extensive to keep track of, the observation was abandoned. Farm Sites During the census study (Part 1), a c t i v i t y - l o c a t i o n observations were recorded at farm s i t e s whenever at l e a s t 5 ducks of 1 species were present. The same c l a s s i f i c a t i o n system was used (Table VI), except l o c a t i o n was recorded only as "land", or "water". Disturbance During farm and c o a s t a l observations, any disturbance causing at l e a s t 5 ducks to f l y up was recorded. Apparent cause, number of each species of duck and where they went ("marsh", "bay" or " l e f t s i t e " ) were noted. "Very High Tide" Observations During the early winter I observed that, at tides i n the upper range of "high" and i n the "very high" range (Table V), generally occurring i n e a rly mornings, there was an i n f l u x of ducks at the 3 estuaries and feeding was more intense. Accordingly, an a d d i t i o n a l set of observations was added, .from January to A p r i l , 1981, at L i t t l e Qualicum River estuary. Part of the marsh, including upper f l a t s , sedge stands and channels, was surveyed from an observation tower. A c t i v i t y - l o c a t i o n scans - 41 -were made at 0.1 m Intervals i n water l e v e l , as the t i d e receded. Obser-vations started at the highest water l e v e l f o r which there was s u f f i c i e n t l i g h t and continued u n t i l the reference gauge l e v e l f o r "high t i d e " ( c a l l e d gauge l e v e l 1) was reached. At gauge l e v e l 8, most of the high marsh area was flooded, while by gauge l e v e l 1, most of i t was dry. A c t i v i t y and l o c a t i o n were recorded (Table V) f o r each duck at each gauge l e v e l . In addition, a l l movements out of the marsh were noted. The number of ducks i n each group that flew up, whether they went to the bay or l e f t the s i t e , d i r e c t i o n of f l i g h t and any apparent disturbance were recorded. S t a r t i n g times of observations ranged from 0620 to 1055 hours, with observation periods l a s t i n g from 70 to 140 minutes. As a c o n t r o l , on 7 occasions when tides were below the "high" range, half-hour observations of the upper marsh were c a r r i e d out s t a r t i n g at 0900 hours. Sources of Error Inevitably, parts of most s i t e s were surveyed inadequately — ducks could have been missed or were disturbed before the observer could record a c t i v i t y and l o c a t i o n . I f 20% of the t o t a l f o r a species i n an observation period was u n c l a s s i f i e d , the data were discarded for that species. Deep channels and upper f l a t s with dense, high vegetation were more d i f f i c u l t to survey than were open areas. This introduced some degree of bias error to a l l estuarine observations. To assess t h i s error, a f t e r several observation periods I walked through upper f l a t s and along deep - 42 -c h a n n e l s t o f l u s h o u t any u n c o u n t e d d u c k s . When few had o r i g i n a l l y b e e n o b s e r v e d , few i f any e x t r a s were s p o t t e d . However , when d u c k s were a b u n d a n t i n t h e a r e a , p a r t i c u l a r l y a t E n g l i s h m a n R i v e r , many u n c o u n t e d d u c k s w e r e o f t e n f l u s h e d o u t . Thus t h e r e s u l t s p r o b a b l y u n d e r e s t i m a t e t h e u s e made o f t h e s e z o n e s b u t c o n c l u s i o n s r e g a r d i n g t h e t i m i n g o f t h i s u s e s h o u l d n o t be a f f e c t e d . - 43 -STATISTICAL METHODS AND RESULTS Data are in the form of proportions: number of ducks in a certain class/number of ducks at the s i te . This enables one to compare across s i tes , tides and species when absolute numbers are very different. Propor-tions were calculated only when there were at least 5 ducks of the species at a site and at least 80% of the ducks were class i f ied as to act iv i ty and location. When not a l l ducks were c lass i f ied , the total of c lass i f ied ducks (not of ducks present) was used to calculate proportions. Whenever a mean number of ducks or mean proportion of ducks is presented, i t s standard error (se) and sample size (N) are given as a measure of the accuracy of the sample mean. Sample sizes vary considerably, as many observations were discarded because of excessive disturbance. In a l l s ta t i s t i ca l tests, data were checked for homoscedasticity using F tests and transformations were applied where appropriate. Unless noted otherwise, variances were equal (P>0.01). Normality of distributions was assessed graphically. There was a tendency for distributions involving proportions to be skewed to the r ight , particularly when means were close to 0. The skewness was not excessive, except where noted. Duck Counts in Relation to Tide If the ducks' movements among sites are correlated with tide leve l , mean numbers of ducks at "high" and "low" tides at each site should differ.- To test this , two-tailed t-tests were performed for each coastal s i te and each species (Table VII) . In 2 cases, as indicated, logarithmic transformations were used to correct for unequal variances. At Rathtrevor - 44 -Table VII. Duck counts at co a s t a l s i t e s , "high" and "low" t i d e s . Species/Site "High" t i d e "Low" tide t - t e s t s f o r difference Mean + se(N) Mean + se(N) between "high" and "low" tide numbers ( t - s t a t l s t i c ) MALLARD L i t t l e Qualicum R. 140+43(19) 129+22(23) 0.46 ns ,a Englishman R. 106+14(21) 80+11(17) 1.29 ns Nanoose Cr. 7+3(17) 17+4(22) 1.29 ns ,a Rathtrevor Beach 4+2(18) 42+12(21) 2.86 ** AMERICAN WIGEON L i t t l e Qualicum R. 150+33(19) 155+24(23) 0.11 ns Englishman R. 244+28(21) 197+38(17) 1.02 ns Nanoose Cr. 136+21(7) 162+25(22) 0.76 ns Rathtrevor Beach 26+20(18) 255+53(21) 3.82 ** a Logarithmic transformation used. ** P<0.01 n s not s i g n i f i c a n t (P>0.05) - 45 -Beach, f o r both species, variances were unequal and "high" t i d e sample d i s -t r i b u t i o n s were highly skewed, with and without transformation. However, the conclusions are probably v a l i d , as the sample means d i f f e r by a factor of 10 for both species. Dabblers were r a r e l y present at "high" t i d e : i n 18 observation periods, there were 15 zero-counts f o r mallard and 14 for American wigeon. Duck Counts i n Relation to Season Numbers of ducks at each coastal s i t e and t i d e l e v e l were plotted against date to check f o r seasonal trends. A l l exhibited wide day-to-day f l u c t u a t i o n s . At the 3 estuarine s i t e s , seasonal changes corresponded to the pattern of seasonal v a r i a t i o n f o r coastal s i t e s i n the census data ( F i g . 2). However, at the 1 bay s i t e , the seasonal trend was markedly d i f -ferent ( F i g . 4). Low tide observations only are pl o t t e d . Peak numbers, occurring i n November, were 143 for mallard and 676 for American wigeon. The l a t t e r was the highest number of American wigeon recorded at a coastal s i t e . A c t i v i t y - L o c a t i o n Data In t h i s section, the a c t i v i t i e s and locations of mallards and American wigeon at 4 coastal s i t e s and 7 farm s i t e s are described. A t o t a l of 158 observation periods at coastal s i t e s provided 9411 c l a s s i f i e d counts of mallards and 26,032 counts of American wigeon. Eighty-one observation periods on farm s i t e s provided 4833 c l a s s i f i e d counts of mallards and 5157 counts of American wigeon. - 46 -F i g . 4. Seasonal trend i n duck numbers at Rathtrevor Beach. - 46a-ioc4 0) u J o 015 •^o 60f l e 40| 20F O MALLARD • AMERICAN WIGEON T • • • © N 1980 F M A 1981 - 47 -A c t i v i t y - l o c a t i o n data for a l l farm s i t e s (combined) and each coastal s i t e are presented i n F i g s . 5 to 9. The bars forming each row of each f i g u r e represent a f u l l breakdown of a c t i v i t y and l o c a t i o n , i n terms of mean percentages, for the stated s i t e and tide l e v e l . Means of le s s than 1% were not graphed. Tables VIII and IX present a d d i t i o n a l information obtained from the a c t i v i t y - l o c a t i o n observations. In Table IX, only "low" t i d e observa-tions are presented, as the deltas were mainly under deep water at "high" t i d e . Farm Sites The n u l l hypothesis: feeding i n t e n s i t y (percentage of ducks present that are feeding) i s the same at a l l farm s i t e s , was tested using data from farm s i t e s with at l e a s t 5 observation periods. The data sets were: 73 observation periods at 7 s i t e s f o r mallards, and 50 observation periods at 4 s i t e s f o r American wigeon. Variances were unequal and sample sizes were very d i f f e r e n t , ranging from 6 to 16 for mallards and from 5 to 15 for American wigeon. Consequently, a nonparametric test (Kruskal-Wallis single f a c t o r analysis of variance by ranks) was applied to each data set. The r e s u l t i n g t e s t s t a t i s t i c s : H = 9.94 for mallards and H = 4.30 for American wigeon, are nonsignificant (P>0.05) when compared with the chi-square d i s t r i b u t i o n . For both species, then, feeding i n t e n s i t y did not d i f f e r among farm s i t e s . Mean feeding i n t e n s i t i e s thus calculated f or a l l farm s i t e s combined: 66% f o r mallards (se=3.2, N=73) and 79% for American wigeon (se=3.1, N=60). - 48 -F i g . 5. A c t i v i t y - l o c a t i o n data f o r mallard and American wigeon at farm s i t e s . Bars are means; v e r t i c a l l i n e s are standard e r r o r s . - 48a -MALLARD pH N=73 AMERICAN WIGEON N =50 ACTIVITY land water LOCATION - 49 -F i g . 6. A c t i v i t y - l o c a t i o n data f o r mallard at coastal s i t e s during "high" t i d e s . Bars are means; v e r t i c a l l i n e s are standard e r r o r s . S i t e s : Qual=Little Qualicum R. estuary; Eng^Englishman R. estuary; Nan=Nanoose Cr. estuary. - 4 9 a -SITE QUAL N=14 60 40 20 «0 c o 60 ENG N =20 « u NAN N=8 FEED • MOVE REST = 40 £20 o o w a a 60 40 20 -ACTIVITY land marsh marine shallow water deep water LOCATION - 50 -F i g . 7 . A c t i v i t y - l o c a t i o n data for mallard at coastal s i t e s during "low" t i d e s . Bars are means; v e r t i c a l l i n e s are standard e r r o r s . S i t e s : Qual=Little Qualicum R. estuary; Eng=Englishman R. estuary; Nan=Nanoose Cr. estuary; Rath=Rathtrevor Beach. SITE QUAL N = 22 60 40 20 ENG N=17 60 40 NAN CO - 2 0 (0 JL O 3 •o «^  O - 60 N = 13 c « u w a 40 20 RATH N =14 60 40 20 ACTIVITY - 50a -rh r r +T7+ EEL rrh la nd marsh marine shallow water deep water LOCATION - 51 -F i g . 8. A c t i v i t y - l o c a t i o n data f o r American wigeon at coastal s i t e s during "high" t i d e s . Bars are means; v e r t i c a l l i n e s are standard e r r o r s . S i t e s : Qual=Little Qualicum R. estuary; Eng=Englishman R. estuary; Nan=Nanoose Cr. estuary. - 51a -SITE QU AL N = 14 60 4 0 -20 ENG N=20 co _60 CO CO o40 20 c Cj> o NAN N=17 60 40 20 f , H+ ACTIVITY — • * I «• land marsh marine shallow water deep water LOCATION - 52 -F i g . 9 . A c t i v i t y - l o c a t i o n data f o r American wigeon at coa s t a l s i t e s during "low" t i d e s . Bars are means; v e r t i c a l l i n e s are standard e r r o r s . S i t e s : Qual=Little Qualicum R. estuary; Eng=Englishman R. estuary; Nan=Nanoose Cr. estuary; Rath=Rathtrevor Beach. - 52a -SITE QUAL N = 21 6 0 -40-20 ENG N=17 NAN N=20 60 40 - 2 0 CO co o 3 °60 c 0) o « 4 0 a 20 RATH N=16 80 60 40-20 FTV-1-ACTIVITY land Jifetfc : r f h • I • marsh marine shallow water LOCATION J ± L deep water - 53 -TABLE VIII. Ducks on land: distance from water. Species/site T o t a l number of Percent Percent from ducks recorded within 1 m 1 m to 5 m on land of water from water Percent greater than 5 m from water MALLARD Coastal s i t e s 2568 Farm s i t e s 1119 81 74 17 26 AMERICAN WIGEON Coastal s i t e s 1737 Farm s i t e s 1195 79 87 14 8 - 54 -TABLE IX. Use of de l t a areas f o r feeding at "low" t i d e . Species/site Number of ducks feeding i n d e l t a zone x 100 Number of ducks feeding i n d e l t a and marine zones Mean + se(N) MALLARD • L i t t l e Qualicum R. 63 + 7(18) Englishman R. 37 + 12(8) Nanoose Cr. 39 + 12(10) AMERICAN WIGEON L i t t l e Qualicum R. 51 + 8(21) Englishman R. 50 + 11(16) Nanoose Cr. 40 + 9(20) - 55 -Farm observations were examined f o r dependence of feeding inten-s i t y on time of day. The 70 observation periods for mallards and 69 obser-vati o n periods for American wigeon ranged from 0915 to 1700 hours. As any r e l a t i o n s h i p might be expected to be nonlinear, the data were examined as scatt e r plots of feeding i n t e n s i t y as a function of time. No nonlinear re l a t i o n s h i p s were apparent for e i t h e r species. Linear regression analyses resulted i n the following regression c o e f f i c i e n t s : b = 0.003 for mallards and b • 0.047 for American wigeon, both nonsignificant i n two-tailed t - t e s t s . "High" and "Low" Tides Table X indicates that the r e l a t i o n s h i p between feeding i n t e n s i t y and t i d e l e v e l varied among the c o a s t a l s i t e s . For the 1 instance i n which the variances were unequal, the t - s t a t i s t i c was well above the P=0.01 l e v e l and I f e e l that the conclusion i s v a l i d . Ebb and Flow Tides "Low" tide data sets for each coastal s i t e were examined for differences between feeding i n t e n s i t y along the tide l i n e at ebb and flow t i d e s . The s t a t i s t i c used was: number of ducks feeding i n the shallow water marine zone/the number of ducks at the s i t e . The data sets were divided into ebb and flow t i d e observations. When the midpoint between s t a r t and f i n i s h times f e l l within a h a l f hour of low t i d e , the observation was disregarded. The n u l l hypothesis of equal feeding i n t e n s i t i e s at ebb - 56 -TABLE X. Feeding i n t e n s i t y at c o a s t a l s i t e s , "high" and "low" t i d e s . Species/Site Percentage of ducks present t- t e s t s f o r d i f f e r -that were feeding. Mean ences i n feeding + se(N) i n t e n s i t i e s at "High t i d e " "Low" ti d e "High" and "low" tid e s MALLARD L i t t l e Qualicum R. 12+4.2(14) 47+4.6(22) 5.25 ** Englishman R. 12+3.1(20) 38+5.2(17) 4.47 ** Nanoose Cr. 28+11(8) 46+8.4(13) 1.37 n s Rathtrevor Beach ducks rare 60+9.5(14) -AMERICAN WIGEON L i t t l e Qualicum R. 29+6.7(14) 63+3.0(21) 5.25 ** Englishman R. 61+4.8(20) 42+4.3(17) 2.82 ** Nanoose Cr. 49+6.5(17) 56+3.9(20) 1.00 n s Rathtrevor Beach ducks rare 81+3.6(16) -a Variances unequal ** P<0.01 n s not s i g n i f i c a n t (P>0.05) - 57 -and flow tides was accepted i n two-tailed t - t e s t s f o r a l l data sets. Test s t a t i s t i c s are l i s t e d i n Table XI. "Very High" Tide Data ( L i t t l e Qualicum River Estuary) Number of ducks present, expressed as a proportion of the peak number, and feeding i n t e n s i t y , are plotted against gauge height i n F i g . 10. Feeding i n t e n s i t y was calculated when at l e a s t 5 ducks were present. The average number of mallards present ranged from 95 at gauge height 5 to 20 at gauge height 1. There were fewer American wigeon present, with average numbers ranging from 21 at gauge height 5 to 9 at gauge height 1. The greatest numbers recorded at 1 scan were 192 mallards and 43 American wigeon. During the 7 c o n t r o l observations, when water l e v e l s were below "high", 4 or fewer mallards and no American wigeon were recorded. A t o t a l of 1498 mallards i n 102 groups l e f t the marsh during the 13 observation periods. Of these, 21% ( i n 15 groups) s e t t l e d i n the bay and the remainder l e f t the s i t e . Ninety-eight percent of mallards leaving the s i t e flew west. Of the 228 American wigeon that l e f t the marsh ( i n 38 groups), 47% ( i n 14 groups) s e t t l e d i n the bay and the rest l e f t the s i t e , a l l f l y i n g east. Disturbance Data As the observers were often the cause of the disturbance and as observation times were chosen to avoid periods of heavy disturbance, the data do not r e f l e c t the true frequency of man-related disturbances at the s i t e s . However, they are of i n t e r e s t i n terms of the behaviour of ducks - 58 -Table XI. Feeding i n t e n s i t y at ebb and flow t i d e s . S i t e E L : No differ e n c e between o feeding i n t e n s i t i e s ebb and flow t i d e Mallard American wigeon df t - s t a t i s t i c df t - s t a t i s t i c L i t t l e Qualicum R. 18 1.28 a' n s 18 0.02 n s Englishman R. 15 1.28 n s 15 0.08 a' n s Nanoose Cr. sample too small 15 0.32 n s Rathtrevor Beach 9 0 . 3 0 a » n s 11 0.44 n s t - s t a t i s t i c c a l c u l ated using arcsine transformation to correct f o r unequal variances. not s i g n i f i c a n t (P>0.05) - 59 -F i g . 1 0 . "Very high t i d e s " ( L i t t l e Qualicum R. estuary): number of ducks and feeding i n t e n s i t y as a function of gauge height. Intervals between gauge heights are 0 . 1 m. Gauge height 1 corresponds to the reference "high" t i d e l e v e l . Number of ducks i s expressed as a percentage of the highest number recorded f o r the observation period; feeding i n t e n s i t y i s the percentage of ducks present that were feeding. Bars are means; v e r t i -c a l l i n e s are standard errors; numbers are sample s i z e s . - 59a -MALLARD •o ;co: >Z>J. --• ---C 1 1 1. , A I * 1 C ~ " ' * ' ' 1 ' ' a • u 3 "° AMERICAN WIGEON GAUGE HEIGHT d e e p e s t — • receding tide — • s h a l l o w e s t - 60 -when disturbed. Causes of disturbances were, i n order of decreasing frequency: people, v e h i c l e s , raptors, dogs, a i r c r a f t , gunshots and boats. F i g . 11 presents a breakdown of where ducks disturbed at coastal s i t e s went. At farm s i t e s , 46 mallard and 31 American wigeon disturbances were recorded. For 70% of mallard disturbances and 48% of the American wigeon disturbances, most or a l l of the ducks l e f t the s i t e . - 61 -F i g . 11. Where ducks went when disturbed (coastal s i t e s ) . Each sector represents the proportion of a l l recorded disturbances i n the stated l o c a t i o n , a l l s i t e s combined. Numbers of ducks i n f l o c k (mean + se) are indicated with arrows. - 61a -LOCATION MALLARD AMERICAN WIGEON WHEN DISTURBED MARSH N=102 N =50 (flocks disturbed) MARINE N =26 N =56 left site flew to marine zone flew to marsh zone - 62 -DISCUSSION Farm and Coastal Habitat Mallard and American wigeon fed r e g u l a r l y and extensively i n both farm and coastal h a b i t a t . Burgess (1970) concluded that American wigeon used the t i d a l marshes and marine foreshore of the Fraser R. delta mainly f o r r e s t i n g , while mallard fed extensively i n the marshes only at c e r t a i n times of the year. Other authors have reported wintering dabblers using coastal s i t e s l a r g e l y as r e s t i n g grounds (Owen 1973, Tamisier 1976). There were no such patterns of d i f f e r e n t i a l use i n t h i s study. When dabblers were at the c o a s t a l s i t e s , at "high" and "low" tides and throughout the study period, feeding occurred ( F i g s . 6 to 9). The choice of feeding l o c a t i o n within s i t e s was c l e a r l y r e l a t e d to the d i s t r i b u t i o n of shallow water. Both species displayed a stronger tendency to feed on land at farm s i t e s than at coastal s i t e s ( F i g s . 5 to 9), but they r a r e l y moved more than 5 m from water (Table VIII) and did not frequent s i t e s without water ( F i g . 3). The few ducks recorded feeding i n deep water appeared to be pecking at f l o a t i n g plant fragments. The conclusion that shallow water i s e s s e n t i a l for most feeding i s consistent with the observation that shallow water d i s t r i b u t i o n i s a major determinant of dabbler d i s t r i b u t i o n (Thomas 1976, Wheeler and March 1979, Heitmeyer and Vohs 1984, Part 1 of t h i s study). Dabbling ducks have 2 basic methods of feeding: f i l t e r - f e e d i n g and plucking (Goodman and Fisher 1962 pp. 19-37). The former method requires water, the l a t t e r does not. Even when water i s not needed for feeding i t may be needed for predator - 63 -avoidance, as dabblers often respond to disturbances by f l y i n g to nearby open water (Tamisier 1976, F i g . 22 of t h i s study). On the basis of an apparent preference f o r farmland, given the r i g h t water conditions (Part 1) and the c o n s i s t e n t l y high feeding i n t e n s i -t i e s at a l l farms, i t seems reasonable to conclude that for mallards and American wigeon flooded f i e l d s are prime feeding s i t e s . Feeding i n t e n s i t y did not vary among farm s i t e s and the number of ducks at each farm s i t e was l a r g e l y a function of the flooded area ( F i g . 3). This indicates that, although the s i t e s were very d i f f e r e n t , a l l f i e l d s covered i n shallow water were equally desirable as feeding l o c a t i o n s . Wheeler and March (1979) found that f i e l d s flooded by shallow water were selected at random by breeding mallards. In other studies, however, food type (Johnson and Montalbano 1984) or food density (Bossenmaier and Marshall 1958) were major factors i n dabbler habitat s e l e c t i o n . Although they may provide excellent feeding grounds, farmlands can become unavailable for long periods when f i e l d s freeze over or dry up. Coastal s i t e s provide more r e l i a b l e h a b i t a t . They freeze l e s s r e a d i l y and, as water conditions are c o n t r o l l e d mainly by t i d e s , most parts of each s i t e are covered i n shallow water for 1 or 2 periods each day throughout the migrating and wintering period. Feeding Locations at Coastal Sites Both estuarine marshes and marine deltas and beaches were important feeding habitat for mallards and American wigeon. Tide l e v e l was all-important i n s e l e c t i o n of feeding l o c a t i o n . At "high" and "very high" - 64 -t i d e s , ducks fed mainly i n the marsh f l a t s and channel edges that were covered i n shallow water (Figs. 6 and 8). At "low" t i d e , ducks fed mainly along the beach t i d e l i n e , i n tidepools and i n marine d e l t a channels. "Mid" ti d e range was a t r a n s i t i o n period during which shallow water i n estuaries covered some marsh and some marine h a b i t a t , making parts of each a v a i l a b l e for feeding. Marine d e l t a channels occupied only small portions of the 0.9-1.5 km of co a s t l i n e surveyed at each of the 3 estu a r i e s . The high proportion of marine-feeding ducks concentrated i n the deltas (Table IX) indicates that these were preferred feeding l o c a t i o n s . This may be related to the growth of filamentous algae, as discussed i n Part 3. Ducks feeding i n the marine zones did not d i s t i n g u i s h between ebb and flow tides (Table XI), i n d i c a t i n g that the common hunters' b e l i e f that waterfowl feed only on receding tides i s unfounded. Burton and Hudson (1978) also reached t h i s conclusion f o r snow geese (Anser caerulescens  caerulescens) i n the Fraser Delta. Although almost a l l feeding at coastal s i t e s occurred i n shallow water, land and deep water were used f o r other purposes. The dry banks of estuarine channels and ponds were used, e s p e c i a l l y by mallards, as r e s t i n g locations ( F i g s . 6 to 9). Unflooded parts of the marine zone were l i t t l e used. Deep water sections of coa s t a l s i t e s were important to both species as refuges following disturbances ( F i g . 11) and at some s i t e s as re s t i n g habitat at "high" t i d e ( F i g s . 6 and 8). Ducks also rested i n shallow water alongside feeding and moving ducks. - 65 -Each coastal habitat type, then, i s a v a i l a b l e primarily at s p e c i f i c t i d e l e v e l s . An evaluation of the r e l a t i v e importance of these habitat types to dabblers should therefore take into account the amount of time each i s a v a i l a b l e . As F i g . 12 i l l u s t r a t e s , there are strong seasonal fl u c t u a t i o n s i n t i d a l ranges, with low tides occurring more frequently i n the f a l l and spring and high tides occurring more frequently i n mid-winter. These f l u c t u a t i o n s are l e s s over 24-hour days than over d a y l i g h t -hour days (Dawe and White 1982); s p e c i f i c a l l y , frequent low tides occur at night during winter. Thus the r e l a t i v e use of marsh and marine habitats also depends upon the extent to which the ducks feed at night. I t was not possible to carry out systematic observations at night, but on several nights during f u l l moon periods, I observed (with the aid of a night scope) ducks r e s t i n g and feeding i n marine zones at "low" t i d e and i n marsh zones at "high" t i d e at L i t t l e Qualicum River estuary. Night feeding among dabblers has been reported both on the breeding grounds (Swanson and Sargeant 1972) and on c o a s t a l winter habitat (Tamisier 1974, Baldassare and Bolen 1984) . The following discussion i s based on feeding during daylight hours. If night feeding i s extensive, seasonal v a r i a t i o n s i n coastal habitat a v a i l a b i l i t y should be l e s s marked. Marine deltas and beaches provided important feeding habitat, p a r t i c u l a r l y i n early f a l l and i n spring. During these periods, low tides were common i n the day ( F i g . 12) and dabbler use of coastal s i t e s was r e l a t i v e l y high ( F i g . 2). This pattern of use can be expected to occur most years as i t was r e l a t e d to the seasonal pattern of water accumulation - 66 -F i g . 12. Percent of daylight hours at d i f f e r e n t t i d a l ranges. T i d a l ranges defined i n Table V. Adapted from submergence/emergence r a t i o s calculated f o r plant communities at L i t t l e Qualicum R. estuary i n 1978 by Dawe and White, 1982. Data points are monthly means. - 66a -- 67 -i n f i e l d s (Part 1). From l a t e November to February, marine habitat was less- a v a i l a b l e and many dabblers were often at farm s i t e s f o r long periods. Dabbler use of Rathtrevor Beach, which contains marine beach, habitat but no d e l t a channels, dropped o f f sharply i n midwinter ( F i g . 4). As discussed i n Part 3, t h i s may r e s u l t from a preference f o r marine deltas i n the spring. Marine zones are considered to be unimportant as dabbler feeding grounds i n the Fraser Delta (Burgess 1970, Vermeer and Levings 1977, H i r s t and Easthope 1981). However, t h i s conclusion has not been substantiated by any systematic observations of dabbler a c t i v i t y . I t i s based p a r t l y on assumptions about dabbler food habits that may be in c o r r e c t (Part 3). Estuarine marshes provided important feeding habitat throughout the study period, but the r e l a t i v e importance of the d i f f e r e n t marsh zones var i e d . Lower marsh f l a t s and lower parts of channel edges, flooded by shallow water through much of the "mid" and lower "high" t i d a l ranges, were av a i l a b l e as feeding habitat f o r about half of most days from October to A p r i l . Sedge stands on upper channel edges and middle-elevation f l a t s were a v a i l a b l e f o r feeding at "high" t i d e . This t i d a l range occurs l e a s t i n the spring and most i n midwinter. The high marsh f l a t s were flooded only at "very high" tides which occur only s p o r a d i c a l l y , mainly i n midwinter. O v e r a l l , a broader range of marsh habitat i s a v a i l a b l e i n f a l l and midwin-ter than i n spring. I f vegetation mapping i s used as a technique f o r dabbler habitat assessment i n estua r i e s , consideration must be given to the timing of - 68 -a v a i l a b i l i t y of food plants, not just to t h e i r abundance. The high marshes (at L i t t l e Qualicum, these are the Juncus high marsh and Deschampsia f l a t s of Dawe and White 1982) are densely vegetated areas containing species that are commonly l i s t e d as dabbler food plants (Martin and Uhler 1939, Yocom 1951). However, they are a v a i l a b l e as feeding grounds r e l a t i v e l y r a r e l y . The lower marsh zones, on the other hand, while sparsely vegetated and with lower species d i v e r s i t y (Dawe and White 1982), are a v a i l a b l e f o r about h a l f of most days. Goss-Custard and Charman (1976) cautioned against the use of food abundance as a measure of habitat q u a l i t y when foods are a v a i l a b l e only under c e r t a i n conditions. Chabrek et a l . (1975) found that the abundance of marsh plants bore no r e l a t i o n s h i p to duck usage of Louisiana coastal marshes because water depth made much of the vegetation unavailable for much of the time. The best approach may be to catalogue a v a i l a b l e feeding habitat at various t i d a l ranges. Variations among estuaries i n patterns of use at "high" and "low" tides ( F i g s . 6 to 9) probably are linked to v a r i a t i o n s i n the r e l a t i v e amounts of low, middle and high e l e v a t i o n marsh habitat and marine habitat. Other factors such as s a l i n i t y l e v e l s , disturbance and vegetation and invertebrate composition would also have an e f f e c t . For example, that American wigeon made l i t t l e use of the marsh at L i t t l e Qualicum R. estuary was probably due to a combination of f a c t o r s , including (1) the r e l a t i v e l y small area of low marsh habitat at the s i t e (a r e s u l t of s i t e topography and channelling of the lower r i v e r ) and (2) disturbance from a nearby road. - 69 -Variations among coastal s i t e s i n average numbers of each species (Table VII) probably also r e s u l t from a combination of f a c t o r s . American wigeon numbers did not d i f f e r greatly among s i t e s . Mallard counts, however, were much higher at L i t t l e Qualicum River and Englishman River estuaries than at Nanoose Creek estuary and Rathtrevor Beach. Both of the former s i t e s contain r e l a t i v e l y large areas of high and middle elevation marsh f l a t s and many channels with sedge stands on the banks. Both are mainly brackish marshes. Nanoose Creek i s more open, with fewer channels and le s s high marsh and i s more s a l i n e (N.K. Dawe, personal communication). Rathtrevor Beach contains no marsh ha b i t a t . Kerwin and Webb (1972) and Chabrek et a l . (1975) concluded that mallards feed more i n fresh and brackish coastal marshes than i n s a l t marshes. There are few data with which to compare these observations, as the few B.C. waterfowl censuses outside the Fraser Delta have combined dabbler species (Hatler 1973, Hay 1976). A la r g e r scale study would be required to i s o l a t e the fa c t o r s that make a s i t e a t t r a c t i v e to mallards. Both s i t e c h a r a c t e r i s t i c s and proximity to other coastal and inland feeding locations should be considered. Disturbance The o v e r a l l e f f e c t of disturbance was to drive ducks either from the s i t e or from the marshes to the marine (usually deep water) zones ( F i g . 11). Thus when using duck abundance as a measure of habitat preference i n estuaries, one should keep i n mind that dabblers may u n d e r u t i l i z e marsh habitat as a r e s u l t of disturbance. Hatler (1973), Tamisier (1976) and - 70 -Owen and Thomas (1979) concluded that disturbance may a f f e c t the d i s t r i -bution of dabblers* Disturbances i n the marshes more frequently a f f e c t e d mallards and those i n marine zones more frequently affected American wigeon. This i s to be expected from the 2 species' r e l a t i v e use of marsh and marine ( F i g s . 6 to 9). Bay areas provided refuges from disturbance; at farm s i t e s , with no comparable habitat, disturbed ducks l e f t the s i t e much more frequently. The l e v e l s of disturbance i n bays as well as i n feeding zones should be considered i n c o a s t a l habitat evaluations. Movement among Feeding Locations i n Relation to Tide Level Campbell (1978) suggested that, although d a i l y routines i n d i s -t r i b u t i o n and feeding i n t e n s i t y are the general r u l e f o r waterfowl, t i d a l routines may take over i n estuaries where feeding i s possible only at c e r t a i n water l e v e l s . Numbers of ducks at coastal s i t e s were not related to time of day (Appendix 1) and there was no evidence of d a i l y routines i n feeding a c t i v i t y . Part 1 Indicated that water l e v e l s on f i e l d s affected dabbler movement between farm and coastal s i t e s . Part 2 indicates that movements among coas t a l s i t e s follow t i d a l routines. Over the "low" to "high" t i d a l range, there was no net movement i n or out of estuaries (Table V I I ) . The shallow bay s i t e , however, was favoured at "low" t i d e s . At "very high" t i d e s , the high marsh at L i t t l e Qualicum River estuary attracted large numbers of dabblers that regularly - 71 -l e f t the s i t e as the tide receded ( F i g . 10). Feeding i n t e n s i t i e s were e s p e c i a l l y high at these locations and t i d e l e v e l s (Table X and F i g . 10). Perhaps i t i s energy-efficient f o r ducks to remain at the estuaries, which provide feeding habitat over most of the t i d a l range, a l t e r n a t e l y r e s t i n g and feeding, while making forays to feeding grounds that become p a r t i c u l a r l y a t t r a c t i v e at c e r t a i n water l e v e l s . These would include: (1) at "very high" t i d e s , estuaries with extensive high marsh f l a t s (mallard e s p e c i a l l y ) , (2) at "low" t i d e s , large marine beaches (American wigeon e s p e c i a l l y ) and (3) flooded f i e l d s (both species). In the "very high" t i d e study, dabbler movement and feeding a c t i v i t y were examined i n r e l a t i o n to small changes i n water l e v e l s ( F i g . i 10). As the t i d e receded and l e s s of the high marsh was a v a i l a b l e f o r feeding, the ducks stopped feeding and flew from the marsh, eit h e r s e t t l i n g i n the bay or leaving the s i t e . Mallards leaving the s i t e flew west and American wigeon leaving the s i t e flew east. This r e s u l t was unexpected, since the 2 species were commonly found together i n mixed f l o c k s . These observations suggest that each species has i t s own regular pattern of movement among s i t e s . Since the departure of ducks from the s i t e was p r e c i s e l y correlated with water l e v e l and not with time of day, i t appears that tide l e v e l i s the key c o n t r o l l i n g f a c t o r . When "very high" tides occurred i n the early morning, unusually large numbers of ducks were already i n the high marsh by f i r s t l i g h t . This suggests that regular, t i d e - r e l a t e d movements also occur at night. - 72 -PART 3: SELECTION OF FOOD ITEMS INTRODUCTION Part 2 indicated that most feeding occurs i n shallow water and that both the estuarine marshes and the l a r g e l y unvegetated marine zones provide important feeding habitat for mallards and American wigeon. Part 3 provides information on what the ducks are eating i n major feeding l o c a -tions of the coastal s i t e s . B e l l r o s e (1976) summarized North American feeding studies of mallards (p. 242) and American wigeon (p. 206). lie concluded that adult mallards are primarily grain and seed-eaters and that American wigeon feed mainly on the leaves and stems of aquatic plants and of c u l t i v a t e d grasses and forage crops. Burgess (1970) and Hughes and Young (1982) concluded that seeds, e s p e c i a l l y the achenes of Carex spp., are the mainstay of wintering mallards' d i e t s i n the Fraser and the Stikine River deltas, r e s p e c t i v e l y . Burgess (1970) reported that American wigeon i n the Fraser Delta eat mainly the leaves of c u l t i v a t e d grasses along with some seeds from t i d a l marshes. - 73 -STUDY AREA AND METHODS Samples were c o l l e c t e d from October, 1980 to A p r i l , 1981 at L i t t l e Qualicum River, Englishman River and Nanoose Cr. estuaries ( F i g . 1), i n marine and marsh zones. Part 2 contains s i t e d e s c r i p t i o n s . Ducks were shot a f t e r they had been observed feeding i n shallow water within the marsh or marine zone f o r at l e a s t 10 minutes. From 1 to 4 ducks were shot at a time. The g u l l e t s (esophagi) were removed immediately and t h e i r contents stored i n 10% formalin. Duck specimens (23 mallards and 40 American wigeon) were frozen and shipped to the B.C. P r o v i n c i a l Museum, V i c t o r i a , B.C. The sampling date, l o c a t i o n and t i d e l e v e l , along with main food items from g u l l e t analyses for each duck or group of ducks shot, are presented i n Table XII. I attempted to obtain samples of the a v a i l a b l e food items at each 2 c o l l e c t i o n s i t e by removing several 0.1 m quadrats of the substrate down to a depth of 10 cm. A f t e r washing and p a r t l y sorting these samples I abandoned the project when I r e a l i z e d how much work was involved i n samp-l i n g a large enough area to estimate the r e l a t i v e abundance of p o t e n t i a l food items. Ducks are f a r more e f f i c i e n t at removing small objects from a tangle of r o t t i n g organic material and mud than are humans. The habitat samples are described q u a l i t a t i v e l y . The contents of each g u l l e t were washed i n a 50 urn sieve and transferred to p e t r i dishes where Inorganic material was removed and food items were sorted under a d i s s e c t i n g microscope. Martin and Barkley - 74 -(1961), Prescott (1970), Hitchcock and Cronquist (1973), Waaland (1977) and Lehmkuhl (1979) were referred to f o r i d e n t i f i c a t i o n of food items. N.K. Dawe (CWS, L i t t l e Qualicum) a s s i s t e d with many i d e n t i f i c a t i o n s . Sorted samples were stored i n 80% ethanol. Volumes of each category and of the t o t a l organic contents of each g u l l e t were measured to the nearest 0.05 ml by water displacement, using a 10 ml graduated c y l i n d e r . C o l l e c t i n g locations were chosen to represent a broad spectrum of heavily-used feeding s i t e s (Table X I I ) . These ranged from the upper marshes at very high tides to the marine beach t i d e l i n e and tidepools at low t i d e s . The data, presented as frequency of occurrence of food items i n various proportions i n duck g u l l e t s , estimate the range of food items and t h e i r r e l a t i v e importance (by importance within each g u l l e t and by f r e -quency of occurrence). C o l l e c t i n g locations were somewhat r e s t r i c t e d both by the p r o v i -sions of my permit and by p r a c t i c a l considerations. In addition, samples are not independent as frequently more than 1 duck was shot at a time. I have not combined samples to c a l c u l a t e aggregate percents as i s generally done with waterfowl food analyses (following Swanson et a l . 1974). Such treatments of the data and subsequent conclusions regarding d i e t s of waterfowl are only v a l i d i f the d i s t r i b u t i o n of sampling e f f o r t r e f l e c t s the d i s t r i b u t i o n of the ducks' feeding e f f o r t . Table XII. Summary of f i e l d notes f o r duck samples. Date No. of Tide S i t e b L o c a t i o n 0 Main food item(s) (1980-81) ducks l e v e l 3 MALLARD MARINE ZONE SAMPLES Jan. 12 2 low Qual beach t i d e l i n e Zostera rhizomes Jan. 14 1 low Qual beach tidepool clams Feb. 9 1 low Qual beach tidepool clams Mar. 12 1 low Qual d e l t a channel filamentous algae Apr. 25 2 mid Qual d e l t a t i d e l i n e mixed green algae MALLARD. MARSH ZONE SAMPLES Nov. 12 2 mid Qual channel edge Carex achenes Nov. 19 1 mid Nan lower f l a t s mixed seeds Dec. 3 1 high Qual mid f l a t s P o t e n t i l l a stems & roots Jan. 6 2 high Eng mid f l a t s mixed seeds & invertebrates Jan. 12 2 high Eng mid f l a t s mixed seeds & invertebrates Mar. 12 1 mid Qual channel edge s n a i l s , mixed seeds Mar. 21 1 very high Qual upper f l a t s s n a i l s , mixed seeds & invertebrates Mar. 21 4 very high Qual upper f l a t s s n a i l s , P o t e n t i l l a stems & roots, mixed seeds & invertebrates Apr. 10 2 mid Nan lower f l a t s s n a i l s , plant fragments, algae 3 see Table V. Qual = L i t t l e Qualicum R. estuary; Nan = Nanoose Cr. estuary; Eng = Englishman R. estuary; see F i g . 1. c see Table VI. - 76 -Table XII (continued) Date No. of Tide S i t e b L o c a t i o n 0 Main food item(s) (1980-81) ducks l e v e l a AMERICAN WIGEON MARINE ZONE SAMPLES Oct. 27 3 low Qual d e l t a t i d e l i n e Ulva, Enteromorpha Nov. 6 3 low Qual beach tidepool Ulva, Enteromorpha Nov. 24 2 low Eng d e l t a t i d e l i n e Ulva, Enteromorpha, Zostera Nov. 28 2 low Eng beach t i d e l i n e Ulva Feb. 2 2 mid Nan beach t i d e l i n e Ulva, Enteromorpha Feb. 14 2 low Eng d e l t a channel filamentous algae Mar. 26 2 mid Qual beach t i d e l i n e mixed green algae Mar. 30 1 low Eng d e l t a channel mixed green algae a see Table V. Qual = L i t t l e Qualicum R. estuary; Nan = Nanoose Cr. estuary; Eng = Englishman R. estuary; see F i g . 1. c see Table VI. - 77 -Table XII (continued) Date No. of Tide S i t e b L o c a t i o n 0 Main food item(s) (1980-81) ducks l e v e l 8 AMERICAN WIGEON MARSH ZONE SAMPLES Oct. 31 1 mid Qual channel edge Agrostis leaves Dec. 5 4 mid Eng channel edge Carex roots Dec. 5 4 high Eng mid f l a t s / channel edge Carex roots & filamentous algae Dec. 10 4 very high Nan upper f l a t s Agrostis seeds Jan. 22 2 mid Nan lower f l a t s filamentous algae Feb. 16 2 mid Nan lower f l a t s filamentous algae Mar. 27 1 high Eng lower f l a t s filamentous algae Apr. 4 1 mid Nan lower f l a t s Plantago leaves Apr. 8 2 high Qual mid f l a t s Agrostis seeds & leaves Apr. 9 2 mid Eng channel edge filamentous algae a see Table V. Qual = L i t t l e Qualicum R. estuary; Nan = Nanoose Cr. estuary; Eng = Englishman R. estuary; see F i g . 1. c see Table VI. - 78 -RESULTS Mean volume of organic g u l l e t contents d i d not vary s i g n i f i c a n t l y among the 4 sample groups (by one-way ANOVA, P<0.01). The o v e r a l l mean was 5.7 ml per g u l l e t (sd=6.3). Food items that made up >1% of organic g u l l e t contents are l i s t e d i n Tables XIII to XVI along with the frequency with which they appeared i n various proportions i n duck g u l l e t s . Major food items for each duck or group of ducks are summarized i n Table XII. A l l items found i n large proportions i n the g u l l e t s were present i n the habitat samples. The bulk of habitat samples from marsh zones con-s i s t e d of dead vegetation and root masses. U n t i l new growth appeared i n mid-February, green vegetation was l i m i t e d to a few grass leaves. Seeds and invertebrates, r e l a t i v e l y scarce i n the habitat samples, were d i s t r i -buted throughout the dead vegetation, s o i l and d e t r i t u s . Filamentous algae i n marine habitat samples were growing on rocks or s h e l l s i n d e l t a areas, whereas i n marsh samples they grew on plants, e s p e c i a l l y S a l i c o r n i a . Growth of filamentous algae was noticeably denser i n the spring samples. Enteromorpha and Ulva occurred mainly i n marine samples as d r i f t fragments, mixed with pieces of brown and red algae and other organic debris. Zostera marina rhizomes appeared i n habitat samples only once. Green Zostera leaves were never abundant. - 79 -Table XIII. G u l l e t contents of 7 mallards shot while feeding i n marine zones. Food item Frequency of occurrence of food items. Categories are percent (by volume) of g u l l e t contents. Data are numbers of ducks. 90-100% 50-90% 10-50% 1-10% Green algae 1. mainly Ulva and Enteromorpha 2. filamentous (Microspora) Red algae Zostera marina rhizomes Vascular plant fragments Polychaetes Bivalves (Mya arenaria and Macoma) 2 1 - 80 -Table XIV. Gullet contents of 16 mallards shot while feeding i n marsh zones. A. Type of food item Frequency of occurrence of food items. Categories are percent (by volume) of g u l l e t contents. Data are numbers of ducks. 90-100% 50-90% 10-50% 1-10% Vegetation Seeds and f r u i t s Invertebrates 1 2 2 6 5 5 4 4 6 3 4 B. Food item 50-100% 30-50% 10-30% 1-10% Vegetation Green algae (Ulva and filamentous) -S a l i c o r n i a shoots -P o t e n t i l i a p a c i f i c a stems and roots 2 Carex shoots and bracts -Plant fragments -Seeds and f r u i t s Chenopodiaceae seeds -P o t e n t i l l a p a c i f i c a seeds -Rubus seeds -Leguminoseae seeds -Plantago seeds and f r u i t parts -1 1 1 2 2 10 3 7 1 1 1 - 81 -Table XIV (continued) Frequency of occurrence of food items. Categories are percent (by volume) of g u l l e t contents. Data are numbers of ducks. B. Food item (continued) 50-100% 30-50% 10-30% 1-10% Compositae seeds - - - 1 T r i g l o c h i n maritimum seeds - 1 1 3 Carex (mainly lyngbyei) achenes 2 - 4 6 Hordeum seeds - 1 2 1 Invertebrates Gastropods (Barleeia acuta) 4 3 5 3 Arachnids - - - 4 Isopods - - - 2 Amphipods - - 1 1 Coleopteran larvae - - - 2 Trichopteran larvae - - 4 4 Lepidopteran larvae - - - 1 Dipteran larvae - - 1 9 Insect eggs and pupae - - - 1 - 82 -Table XV. GuJlet contents of 17 American wigeon shot while feeding i n marine zones. Food item Frequency of occurrence of food items. Categories are percent (by volume) of g u l l e t contents. Data are numbers of ducks• 90-100% 50-90% 10-50% 1-10% Green algae 1. mainly Ulva and Enteromorpha 2. filamentous (mainly Microspora, Cladophora and Monostroma) Zostera marina leaves Vascular plant fragments 11 2 1 1 1 2 - 83 -Table XVI. Gullet contents of 23 American wigeon shot while feeding i n marsh zones. Food item Green algae 1. Ulva 2. filamentous (mainly Cladophora and Monostroma) Brown algae Zostera marina leaves S a l i c o r n i a shoots Carex roots Plantago martima leaves Agrostis leaves Agrostis seeds ( i n heads) T r i g l o c h i n seeds Vascular plant fragments Nematodes Frequency of occurrence of food items. Categories are percent (by volume) of g u l l e t contents. Data are numbers of ducks. 90-100% 50-90% 10-50% 1-10% 2 5 4 1 1 - 84 -DISCUSSION The r e s u l t s indicate that both species s e l e c t a wide v a r i e t y of food types from the estuarine marshes and marine deltas and beaches of southeastern Vancouver Island. This conclusion i s not consistent with the commonly-held view of mallards as seed-eaters and American wigeon as grass-eaters. Green algae were found i n the g u l l e t s of ducks of both species, often as the main constituent, throughout the sampling period, at a l l s i t e s and i n a v a r i e t y of l o c a t i o n s . Algae are c l e a r l y a major food source for American wigeon i n the marine zones and i n lower f l a t s and marsh channels. As mallard foods were more v a r i a b l e and as only 7 mallards were c o l l e c t e d from marine zones, i t i s not c l e a r how Important green algae are i n mallards' d i e t s . It i s c l e a r , however, that mallards do eat algae, sometimes i n large q u a n t i t i e s , i n marine zones and lower marsh f l a t s . Both t h a l l o i d and filamentous algae were consumed. Ducks shot on marine beaches contained mainly Ulva and Enteromorpha, while samples from marsh zones contained mainly filamentous algae; d e l t a samples contained both types. The concentration of feeding ducks i n the d e l t a channels (Table IX) suggests that filamentous algae are preferred over t h a l l o i d green algae when both are a v a i l a b l e . The apparent seasonal s h i f t from t h a l l o i d to filamentous algae i n the American wigeon data (Table XII) i s probably re l a t e d to changes i n a v a i l a b i l i t y . Filamentous algae were sparse i n the winter, and increas-i n g l y abundant from mid-February on. Ulva and Enteromorpha appeared to be - 85 -p l e n t i f u l at a l l seasons. A preference for filamentous over t h a l l o i d species could account for the v i r t u a l abandonment of Rathtrevor Beach by dabblers i n the spring ( F i g . 4). With l i t t l e fresh water input, the s i t e produces almost no filamentous algae; nearby estuaries ( F i g . 1) are p o s s i -bly more a t t r a c t i v e i n the spring. Several authors have observed American wigeon eating Ulva (Munro 1949, Dawe 1980, and Hatler 1973) and several have reported green algae i n g u l l e t or gizzard contents (Lynch 1939 and Vermeer and Levings 1977). How-ever, algae has been regarded as a food eaten by American wigeon only occasionally or when preferred food sources f a i l . Only r a r e l y has algae been mentioned i n connection with mallards (Munro 1943 and Olney 1964). American wigeon i n t h i s study were more opportunistic than i s generally supposed. Part 2 indicated that they feed r e g u l a r l y and inten-s i v e l y i n both marsh and marine areas. Yet, i f t h e i r d i e t were mainly "the stems and l e a f y parts of aquatic plants...and upland grasses and clovers" (Bellrose 1976, p. 206), they would f i n d l i t t l e to s u i t them i n the study area. Green Zostera marina leaves, a recognized marine food of American wigeon (Lynch 1939 and Yocom and K e l l e r 1961), were not abundant i n duck or habitat samples. Other green leaves were sparse i n winter months, and occurred i n only 3 of the g u l l e t samples. However, even when green leaves were abundant, i n the l a t e spring, g u l l e t samples show that American wigeon also selected other food items. At high tides they pulled small Carex r o o t l e t s from the soft mud of channel edges and plucked i n t a c t seed heads from upper marsh areas. At mid tides they gleaned filamentous algae from the stems of S a l i c o r n i a and ate - 86 -green leaves and shoots. At low tides they plucked fragments of green algae from the t i d e l i n e debris and ate the filamentous algae covering rocks and s h e l l s i n d e l t a areas. In the marsh zones, mallards selected a v a r i e t y of foods while' American wigeon concentrated on 1 item at a time. In 18 of 23 marsh American wigeon, a s i n g l e food item formed over 90% of g u l l e t contents, compared with only 3 of the 16 mallard g u l l e t s from marsh zones. Mallards ate s n a i l s , roots and stem bases of P o t e n t i l l a p a c i f i c a , Carex achenes and an assortment of seeds, insect larvae and fragments of vegetation. Seeds, while a major food item, do not dominate mallards' d i e t s to the extent reported i n other c o a s t a l studies (eg. Burgess 1970, Landers et a l . 1976 and Hughes and Young 1982). In marine zones, the small sample of mallards indicates that green algae, clams and eelgrass rhizomes may be important food items. The l a t t e r , however, are probably not abundant i n t e r t i d a l l y , while algae and clams are widespread (Dawe 1980 and Dawe and Lang 1980). Marine invertebrates and insect larvae were frequently consumed i n large quantities by mallards. Invertebrates constituted over half of the g u l l e t contents of 9 of the 23 mallards studied. Snails ( a l l Barleeia  acuta, a small i n t e r t i d a l species) were found i n a l l but 1 of the 16 mallards shot i n a v a r i e t y of marsh areas; insect larvae also occurred frequently, though i n smaller proportions. With the exception of Munro (1943), Olney (1964) and Cronan and H a l l a (1968), studies of non-breeding mallards have concluded that invertebrates are consumed In small quantities only (eg. Burgess 1970, Bellrose 1976, p. 242, and Hughes and Young 1982). - 87 -Breeding mallards, however, may include large proportions of invertebrates i n t h e i r d i e t s (Swanson et a l . 1979). A l l 7 ducks containing over 30% s n a i l s were shot i n March and A p r i l , i n d i c a t i n g that the ducks may have a greater preference f o r invertebrates i n the spring. Variations i n sampling methods may have contributed to the marked differences between the r e s u l t s of t h i s study and those of other coastal dabbler feeding studies. In some studies, the range of food items and the importance of algae and invertebrates could have been underestimated because of bias i n methods of c o l l e c t i n g and analysing samples. • Many studies were based on gizzard analyses. Due to d i f f e r e n t i a l rates of digestion of hard and soft items, use of gizzards results i n a bias toward seeds; soft items such as invertebrates and algae are more rapidly digested (Swanson and Bartonek 1970). Swanson et a l . (1979) con-cluded that t h i s bias resulted i n an underestimation of the importance of invertebrates to breeding dabblers. Although authors of more recent duck studies are c a r e f u l to reduce t h i s bias and the bias r e s u l t i n g from unequal volumes of food i n each duck (Swanson et a l . 1974), the subject of bias i n choice of c o l l e c -t i n g locations i s r a r e l y addressed. C o l l e c t i n g locations i n duck food habit studies are generally not s p e c i f i e d and t h e i r r e l a t i o n s h i p to observed feeding locations i s never discussed. If feeding locations within the study area are f a i r l y homogeneous, t h i s i s not a problem. However, i f the study area i s patchy, samples must include foods selected from the various patches i f conclusions drawn from the data are to be v a l i d for the area as a whole. - 88 -As food passes r a p i d l y through dabblers' g u l l e t s , ducks must be c o l l e c t e d while feeding In a v a r i e t y of l o c a t i o n s . In t h i s study no items were found i n g u l l e t s that were not also a v a i l a b l e i n the sampled habitat; therefore, a g u l l e t sample probably represented the duck's feeding h i s t o r y over not much more than 10 minutes (the minimum time that ducks were observed feeding at each c o l l e c t i n g l o c a t i o n ) . Swanson and Bartonek (1970) found that almost a l l food items force-fed to blue-winged teals (A. discors) had passed through the g u l l e t s within 10 minutes. Due to s a l i n i t y and inundation-period gradients, co a s t a l s i t e s are always patchy. As high and mid-elevation marsh f l a t s of estuaries generally provide more cover f o r observing and for hunting than do lower f l a t s and marine zones, i t i s probable that a bias toward foods a v a i l a b l e i n these areas has been introduced i n t o the l i t e r a t u r e on dabbler feeding habits i n co a s t a l areas. The apparent r a d i c a l differences between the food habits of mallards and American wigeon i n t h i s study and those i n the nearby Fraser Delta may be p a r t l y an a r t e f a c t of sampling methods. The importance of the marine zone i n the Fraser Delta has possibly been underestimated (Part 2). The few ducks c o l l e c t e d i n the marine zone indicate that American wigeon do eat green algae (Vermeer and Levings 1977) and mallards do eat marine invertebrates ( H i r s t and Easthope 1982). Samples from the only major food habits study of the Fraser (Burgess 1970) were mainly gizzard contents, c o l l e c t e d i n unspecified parts of the marshes and a g r i c u l t u r a l lands. Bias inherent i n these methods would tend to underrepresent algae and inverte-brates and to underestimate the range of food items i n the ducks' d i e t s . CONCLUSIONS AND RECOMMENDATIONS 1. Dabblers move between coastal sites and flooded f i e lds . No single site should be considered in i so lat ion, but rather as part of a wetlands com-plex. If an estuary is to be protected as dabbler habitat, other elements of the wetlands complex should be considered for protection. 2. Fluctuations in dabbler numbers along the coastline are related to the degree of flooding on farmlands and to the presence or absence of freezing temperatures. Temperature and flood conditions on nearby f ields should be recorded for each census and considered when interpreting the data. 3. Shallow, non-estuarine bays not previously considered important to dabblers may provide valuable feeding habitat at low tide in f a l l and early winter. 4. Coastal sites are v i t a l during freezing weather when ducks are forced off frozen f i e lds . Duck censuses should be scheduled to include freezing days, both to estimate the numbers of ducks in the region and to catalogue coastal sites that provide c r i t i c a l habitat. 5. F ie lds , when flooding conditions are good, are favoured over coastal s i tes , especially by mallards. Further elimination of flooded fields through drainage w i l l reduce the preferred feeding habitat available to dabblers. 6 . Mallards' and American wigeon's movements both within and among sites are partly dictated by water levels . Each species may have a regular pattern of movement among feeding s i tes , regulated by tides. When - 90 -assessing use of a c o a s t a l s i t e , the ducks must be observed at a l l points of the t i d a l c y c l e . 7. At farm and coastal s i t e s , almost a l l feeding occurs i n or close to shallow water. Available feeding s i t e s s h i f t with weather conditions and with the t i d e s . High marsh zones are popular feeding l o c a t i o n s , e s p e c i a l l y f o r mallards, but are r a r e l y a v a i l a b l e . High biomass of common duck food plants i n t h i s zone does not necessarily t r a n s l a t e i n t o good dabbler habitat. F i e l d s which r a r e l y flood receive l i t t l e use and are not worth maintaining as dabbler feeding grounds. 8. Within shallow water zones, both duck species are opportunistic, feeding on a wide v a r i e t y of food items. In assessing dabbler habitat, I recommend concentrating on factors a f f e c t i n g food a v a i l a b i l i t y : the slope of the land, the extent to which i t floods (at inland s i t e s ) and the shallow water area, both marsh and marine, a v a i l a b l e over the t i d a l cycle at coastal s i t e s . The species composition of marsh vegetation and the d i s -t r i b u t i o n of algae and invertebrates are probably of secondary importance. 9. Marine deltas and beaches are important as refuges from disturbance and as feeding areas, with green algae and marine invertebrates being major food items. These zones form an i n t e g r a l part of estuarine dabbler habitat. 10. Many of the conclusions drawn from Fraser Delta studies do not apply to other B.C. c o a s t a l areas such as those I studied. I question the v a l i d i t y of some of these conclusions for the Fraser system and recommend that they be reexamined. - 91 -REFERENCES Altmann, J . 1974. Observational study of behavior: sampling methods. Behaviour 49(3,4): 227-265. Anderson, D.W. 1975. Population ecology of the mallard. V. Temporal and geographic estimates of s u r v i v a l , recovery, and harvest r a t e s . U.S. Fi s h and W i l d l . Serv. Resour. Publ. 125, Washington. 125 pp. , and K.P. Burnham. 1976. Population ecology of the mallard. VI. The e f f e c t of e x p l o i t a t i o n on s u r v i v a l . U.S. F i s h and W i l d l . Serv. Resour. Publ. 128, Washington. 66 pp. Baldassarre, G.A., and E.G. Bolen. 1984. Field-feeding ecology of waterfowl wintering on the southern high plains of Texas. J . W i l d l . Manage. 48: 63-71. B e l l r o s e , F.C. 1976. Ducks, geese and swans of North America. Stackpole Books, Harrisburg, Pa. 543 pp. , and J.B. Low. 1978. Advances In waterfowl management research. W i l d l . Soc. B u l l . 6: 63-72. Benson, W.A. 1961. An inventory of recreation of the P a c i f i c Coast with s p e c i a l emphasis on waterfowl. Unpubl. rep. Can. W i l d l . Serv., Vancouver, B.C. Bossenmaier, E.F., and W.H. Marshall. 1958. Field - f e e d i n g by waterfowl i n southwestern Manitoba. W i l d l . Monogr. 1. 32 pp. Burgess, T.E. 1970. Foods and habitat of four anatinids wintering on the Fraser Delta t i d a l marshes. M.Sc. Thesis. Univ. B r i t i s h Columbia, Vancouver. 194 pp. Burton, B.A., and R.J. Hudson. 1978. A c t i v i t y budgets of l e s s e r snow geese wintering on the Fraser River estuary, B r i t i s h Columbia. Wildfowl 29: 111-117. Campbell, L.H. 1978. Diurnal and t i d a l behaviour patterns of eiders wintering at L e i t h . Wildfowl 29: 174. Canadian W i l d l i f e Service. 1980. Waterfowl management plan for Canada. F i r s t public d r a f t . Unpubl. manuscript. Can. W i l d l . Serv., Ottawa. 20 pp. Chabrek, R.H., R.K. Yancey, and L. McNease. 1975. Duck usage of management units i n the Louisiana coastal marsh. Proc. Annu. Conf. Southeast Assoc. Game Fis h Comm. 28: 507-516. Cronan, J.M., and B.F. H a l l a . 1968. F a l l and winter foods of Rhode Island waterfowl. Rhode Island Dept. Nat. W i l d l . pamphlet No. 7. 40 pp. - 92 -Dawe, N.K. 1976. Flor a and fauna of the Marshall-Stevenson Unit, Qualicum National W i l d l i f e area. Unpubl. rep. Can. W i l d l . Serv., Qualicum Beach, B.C. 201 pp. 1980. Fl o r a and fauna of the Marshall-Stevenson Unit, Qualicum National W i l d l i f e Area (update to June, 1979). Unpubl. rep. Can. W i l d l . Serv., Qualicum Beach, B.C. 149 pp. , and S.D. Lang. 1980. Fl o r a and fauna of the Nanoose Unit, Qualicum National W i l d l i f e Area. Unpubl. rep. Can. W i l d l . Serv., Qualicum Beach, B.C. 117 pp. , and E.R. White. 1982. Some aspects of the vegetation ecology of the L i t t l e Qualicum River estuary, B r i t i s h Columbia. Can. J . Bot. 60: 1447-1460. Fis h e r i e s and Oceans S c i e n t i f i c Information and Publications Branch. 1981. Canadian t i d e and current tables, V o l . 5: Juan de Fuca and Georgia S t r a i t s . Government of Canada, Ottawa. 73 pp. Goodman, D.C. and H.I. Fisher. 1962. Functional anatomy of the feeding apparatus i n waterfowl Aves: Anatidae. Southern I l l i n o i s Univ. Press, Carbondale. 193 pp. Goss-Custard, J.D., and K. Charman. 1976. Predicting how many wintering waterfowl an area can support. Wildfowl 27: 157-158. Greenwalt, L.A. 1976. A management plan f o r waterfowl. Trans. N. Amer. W i l d l . Nat. Res. Conf. 41: 194-201. Hammack, J . , and G.M. Brown, J r . 1974. Waterfowl and wetlands: toward bioeconomical a n a l y s i s . John Hopkins University Press, Baltimore. 95 pp. Hartman, F.E. 1963. Estuarine wintering habitat for black ducks. J . W i l d l . Manage. 27: 339-343. Hatler, D.F. 1973. An analysis of use, by waterfowl, of t i d e f l a t s i n southern Clayoquot Sound, B r i t i s h Columbia. Unpubl. rep. Can. W i l d l . Serv., Edmonton. 127 pp. Hay, R.B. 1976. An environmental study on the Kitimat region with s p e c i a l reference to the Kitimat River estuary. Unpubl. rep. Can. W i l d l . Serv., Delta, B.C. 85 pp. Heitmeyer, M.E., and L.H. Fredrickson. 1981. Do wetland conditions i n the M i s s i s s i p p i Delta hardwoods influence mallard recruitment? Trans. ' North Am. W i l d l . and Nat. Resour. Conf. 46: 44-57. , and P.A. Vohs, J r . 1984. D i s t r i b u t i o n and habitat use of waterfowl wintering i n Oklahoma. J . W i l d l . Manage. 48: 51-62. - 93 -H i r s t , S.M., and C.A. Easthope. 1981. Use of a g r i c u l t u r a l lands by waterfowl i n southwestern B r i t i s h Columbia. J . W i l d l . Manage. 45: 454-462. Hitchcock, C.L., and A. Cronquist. 1973. F l o r a of the P a c i f i c Northwest. Univ. Washington Press, Se a t t l e . 730 pp. Holland, S.S. 1976. Landforms of B r i t i s h Columbia: a physiographic' o u t l i n e . B u l l . 48, B.C. Dept. Mines and Petroleum Res. V i c t o r i a , B.C. Hughes, J.H., and E.L. Young. 1982. Autumn foods of dabbling ducks i n southeastern Alaska. J . W i l d l . Manage. 46: 259-262. Hunter, R.A., B.A. Pendergast, and K.R. Summers. 1980. A coastal waterfowl habitat inventory i n B r i t i s h Columbia. Unpubl. manuscript. B.C. Min. of Environment, V i c t o r i a , B.C. 19 pp. Johnson, F.A., and F. Montalbano I I I . 1984. Selection of plant communities by wintering waterfowl on Lake Okeechobee, F l o r i d a . J . W i l d l . Manage. 48: 174-178. Joyner, D.E. 1980. Influence of invertebrates on pond s e l e c t i o n by ducks i n Ontario. J . W i l d l . Manage. 44: 701-705. Kerwin, J.A., and L.G. Webb. 1972. Foods of ducks wintering i n coastal South Carolina, 1965-1967. P r o c Annu. Conf. Southeast Assoc. Game Fi s h Comm. 25: 223-245. Krapu, G.L. 1981. The role of nutrient reserves i n mallard reproduction. Auk 98: 29-38. Ladd, W. J r . , J.P. Linduska, and M. Sorenson. 1974. Waterfowl wetlands i n the United States: t h e i r abuse and s a l v a t i o n . In_ M. Smart, ed. International conference on conservation of wetlands and waterfowl. Int. Waterfowl Res. Bureau, Slimbridge, England. Landers, J.L., A.S. Johnson, P.H. Morgan, and W.P. Baldwin. 1976. Duck foods i n managed t i d a l impoundments i n South Carolina. J . W i l d l . Manage. 40: 721-428. Leach, B.A. 1972. The waterfowl of the Fraser Delta, B r i t i s h Columbia. Wildfowl 23: 45-55. Lehmkuhl, D.M. 1979. How to know the aquatic i n s e c t s . Wm. C. Brown Col, Dubuque, Iowa. 168 pp. Lynch, J . J . 1939. Marine algae i n food of Rhode Island waterfowl. Auk 56: 374-380. McMahan, C.A. 1970. Food habits of ducks wintering on Laguna Madre, Texas. J . W i l d l . Manage. 34: 946-949. - 94 -Mack, G.D., and L.D. Flake. 1980. Habitat r e l a t i o n s h i p s of waterfowl broods on South Dakota stock ponds. J . W i l d l . Manage. 44: 695-699. Martin, A.C., and W.D. Barkley. 1961. Seed i d e n t i f i c a t i o n manual. Univ. C a l i f o r n i a Press, Berkeley. 221 pp. , and F.M. Uhler. 1939. Food of game ducks i n the United States and Canada. U.S. Gov. P r i n t i n g O f f i c e , Washington. 308 pp. Munro, J.A. 1943. Studies of waterfowl i h B r i t i s h Columbia: mallard. Can. J . Res. D. 21: 223-260. " 1949. Studies of waterfowl i n B r i t i s h Columbia: baldpate. Can. J . Res. D. 27: 289-307. Olney, P.J.S. 1964. The food of mallard Anas platyrhyncos platyrhyncos c o l l e c t e d from coastal and estuarine areas. Proc. Zool. Soc. Lond. 142: 397-418. Owen, M. 1973. The winter feeding ecology of wigeon at Bridgewater Bay, Somerset. Ibis 115: 237-243. , and G.J. Thomas. 1979. The feeding ecology and conservation of wigeon wintering at the Ouse Washes, England. J . Appl. Ecol. 16: 795-809. , and G. Williams. 1976. Winter d i s t r i b u t i o n and habitat requirements of wigeon i n B r i t a i n . Wildfowl 27: 83-90. Prescott, G.W. 1970. How to know the freshwater algae. Wm. C. Brown Col, Dubuque, Iowa. 348 pp. Sincock, J.L. 1965. Estimating consumption of food by wintering waterfowl populations. Proc. 6th Annu. Conf. Southeast Assoc. Game Fish Comm. 217-221. Summers, K.R., and R.W. Campbell. 1978. Natural h i s t o r y theme study of b i r d and mammal habitats of Canada's P a c i f i c Coast and adjacent coastal waters. Unpubl. rep. Parks Canada, Ottawa. 239 pp. Swanson, G.A., and J.C. Bartonek. 1970. Bias associated with food analysis i n gizzards of blue-winged t e a l . J . W i l d l . Manage. 34: 739-746. , G.L. Krapu, J.C. Bartonek, J.R. Serie, and D.H. Johnson. 1974. Advantages In mathematically weighting waterfowl food habits data. J . W i l d l . Manage. 38: 302-307. , G.L. Krapu, and J.R. Serie. 1979. Foods of laying female dabbling ducks on the breeding grounds. Pages 47-57 in_T.A. Bookhout, ed. Waterfowl and wetlands - An integrated review. Proc. 1977 symp., Madison, W i s e , Northcentral Sect., The W i l d l . Soc. - 95 -, and A.B. Sargeant. 1972. Observation of nighttime feeding behavior of ducks. J . W i l d l . Manage. 36: 959-961. Tamisier, A. 1974. Etho-ecological studies of t e a l wintering i n the Camargue (Rhone Delta, France). Wildfowl 25: 107-117. 1976. Diurnal a c t i v i t i e s of green-winged t e a l and p i n t a i l wintering i n Louisiana. Wildfowl 27: 19-32. Thomas, G. 1976. Habitat usage of wintering ducks at the Ouse Washes, England. Wildfowl 27: 148-152. Vermeer, K., and C D . Levings. 1977. Populations, biomass and food habits of ducks on the Fraser R. Delta i n t e r t i d a l area, B r i t i s h Columbia. Wildfowl 28: 49-60. Waaland, J.R. 1977. Common seaweeds of the P a c i f i c Coast. J . J . Douglas Ltd., N. Vancouver. 120 pp. Wheeler, W.E., and J.R. March. 1979. Ch a r a c t e r i s t i c s of scattered wetlands i n r e l a t i o n to duck production i n southeastern Wisconsin. Tech. B u l l . No. 116. Wisconsin Dept. Nat. Res. White, D.H., and D. James. 1978. D i f f e r e n t i a l use of freshwater environments by wintering waterfowl of coastal Texas. Wilson B u l l . 90: 99-111. Wiens, J.A. 1976. Population responses to patchy environments. Annu. Rev. E c o l . Syst. 7: 81-120. Yocom, C F . 1951. Waterfowl and t h e i r food plants i n Washington. Univ. Washington Press, S e a t t l e . 272 pp. , and M. K e l l e r . 1961. Cor r e l a t i o n of food habits and abundance of waterfowl, Humboldt Bay, C a l i f o r n i a . C a l i f . F i s h and Game 17: 41-53. Zar, J.H. 1974. B i o s t a t i s t i c a l a n a l y s i s . Prentice-Hall Inc., Englewood C l i f f s , N.J. 620 pp. - 96 -APPENDIX 1: PRELIMINARY STUDY, FEBRUARY TO APRIL, 1980 The preliminary study's objectives were: (1) to develop methods for the main study, (2) to assess the degree of day-to-day fl u c t u a t i o n s i n dabbler census data at coastal s i t e s , (3) to test i f time of day i s a fa c t o r i n these f l u c t u a t i o n s , and (4) to see i f dabbler numbers at a farm and a c o a s t a l s i t e were co r r e l a t e d . Counts of dabblers at Nanoose Creek and L i t t l e Qualicum River estuaries were conducted from February 4 to A p r i l 8, 1980. Censuses were done on foot, following set routes and using observation towers and natural vantage points. Counts started at e i t h e r 0900, 1200 or 1500 hours. Duck counts are presented i n Appendix 1, Table 1. In order to test i f duck abundance i s affected by time of day, Kruskall-Wallis tests (non parametric analyses of variance) were performed fo r each species at each s i t e , with data divided according to time of day. A nonparametric t e s t was chosen because the data are highly skewed. The r e s u l t i n g test s t a t i s t i c s (Zar 1974) are, for L i t t l e Qualicum: H=0.324 (mallard) and H=0.756 (American wigeon). None of these values are s i g n i f i -cant (P>0.05). Seasonal f l u c t u a t i o n s i n duck numbers should not a f f e c t these t e s t s since the 3 types of censuses were spread evenly over the study period at each s i t e . Duck abundance at each s i t e , then, does not appear to be affected by time of day. Twelve censuses at L i t t l e Qualicum River estuary were followed immediately by dabbling duck counts at a farm located 3km from the estuary. A log transformation was used to correct f or heteroscedasticity, the - 97 -variances being proportional to the means (Zar 1974). The c o r r e l a t i o n c o e f f i c i e n t s f o r both species were s i g n i f i c a n t l y negative i n two-tailed tests (P<0.01). - 98 -Table 1. Census data, L i t t l e Qualicum River and Nanoose Creek estuaries, 1980. L i t t l e Qualicum Nanoose Date Mallard American Date Mallard American (1980) wigeon (1980) wigeon Feb. 4 157 399 Feb. 6 0 76 Feb. 9 10 4 Feb. 11 14 221 Feb. 15 38 36 Feb. 15 13 305 Feb. 18 27 0 Feb. 19 0 294 Feb. 21 47 10 Feb. 21 0 537 Feb. 22 3 8 Feb. 22 15 404 Feb. 25 25 6 Feb. 27 3 274 Feb. 26 4 2 Feb. 28 0 325 Mar. 1 9 0 Feb. 29 0 308 Mar. 4 45 9 Mar. 3 0 326 Mar. 5 33 7 Mar. 6 4 303 Mar. 8 34 61 Mar. 7 3 295 Mar. 10 62 63 Mar. 12 11 280 Mar. 12 52 50 Mar. 13 5 333 Mar. 15 24 74 Mar. 16 23 347 Mar. 19 109 143 Mar. 17 0 366 Mar. 21 134 129 Mar. 18 3 262 Mar. 22 135 76 Mar. 20 12 335 Mar. 22 135 76 Mar. 20 12 335 Mar. 24 137 192 Mar. 25 33 406 Mar. 25 75 137 Mar. 27 26 265 Mar. 28 239 175 Mar. 31 22 283 Mar. 30 213 149 Apr. 1 6 282 Apr. 5 89 108 Apr. 4 19 309 Apr. 8 119 267 Apr. 5 26 249 Apr. 9 32 300 - 99 -APPENDIX 2. CENSUS DATA FOR EACH SITE Date Mallard numbers at coastal sites'* (1980-81) 1 2 3 4 5 6 7 8 Oct. 6 19 0 194 68 0 0 0 65 10 0 0 31 82 0 0 0 151 15 2 0 82 80 0 0 0 111 24 0 0 0 71 61 0 0 293 Nov. 4 26 0 128 150 0 5 0 126 11 4 0 125 122 0 0 0 168 18 49 0 142 115 12 0 ' 0 82 20 0 0 121 127 0 0 0 69 25 0 0 0 156 4 0 0 106 Dec. 2(F) 22 0 92 121 0 0 0 153 4(F) 89 3 60 261 0 0 0 539 8(F) 29 24 181 490 11 21 0 607 16 24 0 31 164 7 0 0 39 30 2 0 12 133 8 0 0 75 Jan. 6 10 2 32 162 0 0 0 327 13 0 0 5 175 0 0 0 52 20 10 0 0 158 0 0 0 17 27 11 0 0 175 0 0 0 30 Feb.- 3 12 0 8 126 0 0 0 27 10(F) 101 37 0 370 0 18 0 811 11(F) 35 0 11 214 0 20 0 272 17 0 0 0 22 0 0 0 0 24 11 0 0 87 0 0 0 0 Mar. 6 10 0 0 48 17 0 0 80 10 7 0 0 53 0 0 0 28 17 16 0 0 60 5 0 0 184 24 50 0 6 39 0 12 0 155 31 40 0 0 113 27 25 0 49 Apr. 7 24 0 0 67 22 26 0 171 14 0 0 4 85 8 23 0 172 21 7 — 28 12 0 0 14 0 0 0 43 aSee F i g . 1 f or l o c a t i o n of s i t e s F=Freezing day - 100 -APPENDIX 2. CENSUS DATA FOR EACH SITE (continued) Date American wigeon numbers at coas t a l sites* 1 (1980-81) 1 2 3 4 5 6 7 8 Oct. 6 0 0 0 ' 112 0 2 0 177 10 0 0 193 268 0 0 0 149 15 33 0 373 520 0 15 0 281 24 23 0 356 192 17 6 0 388 Nov. 4 82 0 640 608 3 16 0 257 11 260 0 580 564 19 0 19 259 18 174 0 490 392 95 4 0 301 20 304 0 589 67 0 8 0 107 25 134 0 0 94 59 5 0 361 Dec. 2(F) 122 55 250 281 0 0 0 13 4(F) 268 82 44 601 24 0 0 540 8(F) 175 11 373 598 111 0 0 388 16 104 12 0 213 7 0 0 12 30 58 2 8 35 0 0 0 18 Jan. 6 62 67 98 73 0 0 0 34 13 149 28 312 10 57 0 0 56 20 73 0 4 201 0 0 0 14 27 119 15 0 231 0 0 0 50 Feb. 3 199 18 4 301 8 0 4 12 10(F) 287 45 0 271 74 0 51 544 11(F) 220 32 244 288 29 14 2 40 17 175 0 0 27 0 0 0 12 24 268 0 165 23 75 0 0 31 Mar. 6 248 0 195 134 55 0 0 94 10 298 0 12 160 90 5 0 28 17 109 0 0 159 52 9 0 166 24 116 0 0 183 21 8 0 280 31 281 0 0 274 60 19 0 122 Apr. 7 265 0 0 253 26 29 0 223 14 250 0 0 166 51 29 0 192 21 30 28 12 0 0 75 0 0 0 141 aSee F i g . 1 for l o c a t i o n of s i t e s F=Freezing day - 101 -APPENDIX 2. CENSUS DATA FOR EACH SITE (continued) Date Mallard numbers at farm sites'* (1980-81) A B C D E F G H I J Oct. 6 0 0 0 0 0 0 0 0 0 0 10 0 0 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 0 0 0 24 0 0 0 0 0 0 0 0 0 0 Nov. 4 44 10 34 220 0 0 0 0 0 0 11 30 4 27 72 690 0 59 0 28 0 18 49 0 324 247 0 8 115 0 0 0 20 38 10 844 0 0 43 160 0 0 0 25 433 0 115 0 0. 50 8 0 0 0 Dec. 2(F) 35 0 2 30 158 61 6 0 0 0 4(F) 0 0 56 93 0 0 0 0 0 0 8(F) 0 0 73 0 0 0 0 0 0 0 16 135 0 60 70 572 4 31 0 0 0 30 370 47 98 2 386 56 82 5 0 0 Jan. 6 90 10 136 31 80 8 57 0 0 0 13 104 16 68 25 180 5 13 0 0 0 20 130 3 0 146 354 8 2 0 0 0 27 47 3 127 94 77 102 4 0 0 0 Feb. 3 366 33 50 209 0 113 27 0 0 0 10(F) 17 4 4 0 0 4 0 0 0 0 11(F) 0 8 11 248 0 0 0 0 0 0 17 141 17 104 141 374 111 33 0 0 0 24 341 11 89 188 194 29 0 0 17 0 Mar. 6 59 118 77 121 17 38 31 0 0 0 10 108 62 75 56 0 0 14 0 0 0 17 24 46 6 37 0 0 0 0 0 0 24 22 23 4 12 0 0 0 0 0 0 31 23 12 15 0 0 5 2 0 0 0 Apr. 7 28 7 2 33 0 7 0 0 0 0 14 22 14 34 17 0 0 0 0 0 0 21 0 46 6 1 0 0 0 0 0 0 28 0 67 0 6 0 0 0 0 0 0 aSee F i g . 1 for l o c a t i o n of s i t e s F=Freezing day - 102 -APPENDIX 2. CENSUS DATA FOR EACH SITE (continued) Date American wigeon numbers at farm s i t e s (1980-81)  A B C D E F G H I J Oct. 6 0 0 0 0 0 0 0 0 0 0 10 0 0 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 0 0 0 24 2 0 0 0 0 0 0 0 0 0 Nov. 4 314 26 0 0 0 0 0 0 0 0 11 20 8 0 0 333 0 0 0 0 0 18 6 0 303 127 0 0 0 0 0 0 20 48 2 377 0 0 0 5 0 0 0 25 29 6 322 374 0 0 0 0 0 0 Dec. 2(F) 56 0 0 0 260 0 0 0 0 0 4(F) 0 25 92 0 0 0 0 0 0 0 8(F) 0 0 87 0 0 0 0 0 0 0 16 99 230 120 24 828 0 0 0 0 0 30 106 396 170 0 586 16 0 0 0 0 Jan. 6 308 80 173 39 90 0 6 0 0 0 13 130 266 56 10 0 3 0 0 0 0 20 470 6 0 32 175 0 0 0 0 0 27 197 205 99 9 84 0 0 0 0 0 Feb. 3 147 306 38 0 0 7 0 0 0 0 10(F) 74 133 0 0 0 0 0 0 0 0 11(F) 0 179 0 0 0 0 0 0 0 0 17 114 91 50 32 274 60 0 0 0 0 24 178 165 111 74 33 13 0 0 0 0 Mar. 6 31 252 16 102 0 0 0 0 0 0 10 148 370 15 0 0 0 0 0 0 0 17 4 133 22 0 0 0 0 0 0 0 24 15 81 11 0 0 0 0 0 0 0 31 0 57 40 0 0 0 0 0 0 0 Apr. 7 0 89 54 0 0 0 0 0 0 0 14 0 75 33 0 0 0 0 0 0 0 21 0 46 6 0 0 0 0 0 0 0 28 0 17 0 0 0 0 0 0 0 0 aSee F i g . 1 for l o c a t i o n of s i t e s F=Freezing day - 103 -APPENDIX 3. FLOODED AREA ESTIMATES AT FARM SITES Date Flooded area estimates at farm sites* 1 (1980-81) (hectares) A B C D E F G H I J Oct. 6 0 0 0 0 0 0 0 0 0 0 10 0 0 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 0 0 0 24 0 0 0 0 0 0 0 0 0 0 Nov. 4 1.42 7.52 1.26 2.09 0.10 0.10 0 0 0 0 11 1.74 16.27 8.86 4.62 14.23 1.02 1.74 0.22 0.008 0 18 1.29 - 10.30 5.41 3.62 0.10 1.02 1.57 0 0.002 0 20 1.29 10.30 5.41 3.62 0.10 1.10 1.57 0 0 0 25 5.23 12.70 5.41 4.40 0.10 1.26 1.83 0 0 0 Dec. 2(F) 6.89 13.11 7.50 5.77 7.95 2.61 1.83 0 0 0 4(F) 0 0.10 3.45 1.75 0 0.10 0 0 0 0 8(F) 0 0 0.10 0 0 0 0 0 0 0 16 17.26 14.73 6.25 5.77 14.23 1.64 1.44 0 0.003 0 30 24.19 20.34 8.86 6.24 17.95 4.23 1.98 0.28 0.003 0 Jan. 6 10.47 13.78 4.82 2.83 3.28 1.18 1.60 0 0.002 0 13 4.70 9.84 3.35 1.53 1.67 1.05 1.07 0 0 0 20 2.25 11.94 5.69 4.18 11.60 1.28 1.60 0 0.003 0 27 2.25 8.13 4.82 3.35 3.91 1.28 1.07 0 0.002 0 Feb. 3 2.80 5.29 2.58 1.40 0.81 1.00 1.07 0 0 0 10(F) 0.20 0 0.10 0 0 0.17 0 0 0 0 11(F) 0 0.20 0.10 0.38 0 0 0 0 0 0 17 24.79 7.71 6.25 6.79 23.98 4.51 1.71 0.10 0.004 0 24 8.49 8.68 3.91 3.88 16.38 1.54 1.43 0 0.003 0 Mar. 6 3.00 6.70 2.58 3.06 0.10 0.50 1.31 0 0.001 0 10 2.60 5.03 1.25 3.00 0.10 0.20 0.87 0 0 0 17 1.70 4.60 1.00 1.23 0 0.20 0.65 0 0 0 24 0.91 4.21 0.63 0.87 0 0.20 0.35 0 0 0 31 1.39 4.81 1.20 0.62 0 0.35 0.35 0 0 0 Apr. 7 1.40 4.81 1.00 2.26 0 0.15 0.35 0 0 0 14 0.90 5.03 0.63 1.23 0 0.10 0.10 0 0 0 21 0.30 1.57 0.30 0.51 0 0.10 0 0 0 0 28 0.10 0.10 0.10 0.25 0 0.10 0 0 0 0 aSee F i g . 1 f o r l o c a t i o n of s i t e s F=Freezing day 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0096084/manifest

Comment

Related Items