Table Of Content

3 University of Alberta Libra 0 1620 2998665 8 Volume 69 Number 2 June Ft▼ & w [ ***'§£=&T # TJ ^ ■ >/ fpl*. > * i -SJmk v SCI/TECH Fron r on p. 98. Harold Fisher Back cover: Figure 3. Eastern yellow-bellied racer captured near Castle Butte in the Big Muddy Badlands, SK, in May of 2010, thereby extending the known range of the species in the Big Muddy Valley to the north-west. This live eastern yellow- bellied racer is the first one reported in the Big Muddy Valley since 1976. See article by Gardiner et al. on p. 70. Chris Somers Figure 1. Postcard of the whooping crane diorama by R.D. Symons based on Neil Gilmour’s photographs and description of their nesting habitat at Shallow Lake, in the Saskatchewan Museum of Natural History. This exhibit was destroyed in a fire in 1990. (Reproduced with permission from the Royal Saskatchewan Museum.) See article by J.K. Finley on p. 88. Figure 1. Lesser scaup male (left) and female (right) in breeding plumage. See article by G. Hammell on p. 54. Nick Saunders Blue Jay Vol. 69 No. 2 June 2011 53-108 Birds LESSER SCAUP AND RACCOONS: ARE THERE LINKS IN SOUTHWESTERN MANITOBA? Gord Hammell.54 Reptiles EASTERN YELLOW-BELLIED RACER POPULATIONS ON THE CANADIAN PRAIRIES. Laura E. Gardiner, Jessica A. Martino, Ray G. Poulin, Christopher M. Somers... 70 Plants OPHIOGLOSSID FERNS IN MANITOBA: MOONWORTS, GRAPEFERNS AND NORTHERN ADDER’S-TONGUE. Richard J. Staniforth...75 Natural History RETURN OF THE GOLDEN BIRD: THE LAST BREEDING GROUND OF THE WHOOPING CRANE ON THE PRAIRIES. James K. Finley.88 Notes and Letters BEWARE THE WHISKEY JACK. James K. Finley.....95 BOREAL OWL MONITORING IN THE NISBET FOREST: 2010 YIELDS FIRST BREEDING RECORD. Harold Fisher.98 Photo Notes MATING BULLSNAKES. Laura E. Gardiner.100 Poetry GOS. Peter Branson.101 THE LAUGHING GULL. Peter Branson. 102 Mystery Photo THIS ORGANIZATION RECEIVES FUNDING FROM: Saskatchewan LOTTERIES 53 69 (2) June 2011 BIRDS LESSER SCAUP AND RACCOONS: ARE THERE LINKS IN SOUTHWESTERN MANITOBA? Gord Hammell P.O. Box 37, Erickson, MB, ROJ 0P0; E-mail: <[email protected]> ABSTRACT was uncertain given available data. Raccoons (Procyon lotor) invaded the However, there was a detectable shift southern prairie provinces of Canada in hatching chronology to later dates in during the mid-1990s and became a years with wetter summers. It is possible significant predator of waterfowl nests. that the shift in hatching chronology could Nest predation by this novel predator be related to raccoon predation of scaup may alter the reproductive success, and nests and changes in renesting rates thereby the population size of waterfowl among time periods, but it is impossible like the lesser scaup (Aythya affinis). to say for sure given the available data. However, it is difficult to evaluate the Nevertheless, such a shift may have been effects of raccoon predation relative to detrimental to local populations over the other factors that affect scaup populations. years, as later-hatched ducklings have a To investigate possible links between lower probability of recruitment into the raccoons and lesser scaup, a species with breeding population. declining populations, I examined results from studies conducted in southwestern INTRODUCTION Manitoba before (1950’s) and after Lesser scaup (Aythya affinis) are (1970’s) the arrival of the raccoon. medium-sized diving ducks that breed Specifically, I attempted to detect possible in the boreal forests and parklands changes in scaup breeding population from Alaska to Manitoba.1 In breeding size, reproductive success, and hatching plumage, males are black and white and chronology that might be related to the females are dark chocolate brown with arrival of raccoons. I hypothesized that a white face patch (Fig. 1, see inside the post-raccoon period would show: front cover, bottom). Combined North (a) decreasing local scaup breeding American breeding populations of lesser populations due to lower recruitment scaup and similar-looking greater scaup as a result of decreased nest success (A marila) have been declining from highs and female survival and (b) a shift in the in the 1970s of 5-7 million birds to lows of mean hatching date to later in the season 3-4 million in the 2000s (Fig. 2).2 Lesser because of increased nest losses and scaup comprise -90% of the combined more renesting attempts by female scaup. population;1 most of the decline has been Local breeding populations appeared attributed to this species, especially within unchanged before and after raccoon the Canadian western boreal forest, arrival, providing no strong evidence of where most of these birds breed.3 Lesser an impact on numbers of adult birds. The scaup (but not greater scaup) breed in extent of change to reproductive success south-western Manitoba parkland, and 54 UNIVERSil Y LIBRARY Blue Jay UNIVERSITY OF ALBERTA Figure 2. Breeding population estimates, 95% confidence intervals, and North American Waterfowl Plan goal (dashed line) for scaup (lesser and greater) in the traditional survey area (strata 1-18, 20-50, 75-77). long-term but less consistent declines Territories.4-13 The major factor responsible in breeding pairs have also been noted for waterfowl nest failure and female there.34 Several hypotheses have been and duckling mortality on the breeding presented to explain declining lesser grounds is predation,114-16 so predation scaup (hereafter scaup) populations, may be an important factor (but not including changes to food resources necessarily the only factor) in explaining in wintering and migration habitats,57 reduced recruitment. Therefore, changes chemical contaminant exposure,18'11 to the local predator community (size or climate changes affecting boreal forest composition) might contribute to changes wetlands,3 lower female survival,312 and at the local scaup population level.1718 fewer immature birds relative to adults in human harvests (declining age ratio).3 The parkland area of southern However, no clear explanation for the Manitoba experienced a change to the predator community with the invasion decline has been determined. Declining of the raccoon (Procyon lotor) during age ratios in the harvest are suggestive the mid 1900s.19 Raccoons have the of reduced recruitment, and contributing potential to be a significant predator of factors could be any combination of scaup nests, but not females,20 because reduced breeding propensity, reduced they search for prey in the upland and clutch size and renesting rates among emergent vegetation zones adjacent hens, or decreased nest and/or juvenile to ponds,2122 where most scaup nests survival.3 Reduced nest, juvenile, and are found.23-27 Previous studies have female survival have been suggested as reported that raccoons are indeed a proximate causes of decline in local scaup major predator of scaup nests in south¬ populations in south-western Manitoba western Manitoba.2328 In this area, more and in the boreal forest of the Northwest 55 69 (2) June 2011 than 50% of scaup nest over water (nest still some merit in testing my predictions surrounded by water), and more of these to determine whether any of them are nests are successful than are dry-land substantiated. nests.2327 Over-water nests, which may be isolated from dry-land predators for The Consensus Report of the Second some or all of the incubation period, Scaup Workshop indicated several priority presumably would be more vulnerable to research topics, one of which was the raccoons because of their nest-searching establishment of retrospective analyses behaviour. Thus, the arrival of raccoons to examine patterns over time and may have resulted in increased nest space.31 Analyses of existing data were losses, reduced productivity and reduced considered the most immediate research juvenile recruitment. Also, if initial nests need because results are critical for are destroyed, some scaup do attempt forming hypotheses, organizing research second nests (renests) depending on projects, and securing funding. With female age and habitat conditions.28 this priority in mind, and to evaluate the This increased nesting effort resulting novel predator hypothesis, I conducted a from increased initial nest destruction retrospective analysis of existing scaup can lead to greater nest-site exposure breeding pairs, productivity (broods to predators (e.g., fox [Vulpes vulpes] per pair or nest success) and hatching and mink [Mustela vison]) and increased chronology data from studies conducted female mortality.14-29 Scaup exhibit natal in south-western Manitoba prior to the philopatry and breeding-site fidelity,128 1980s, which was prior to the decline so reduced juvenile recruitment and of the continental scaup population. I increased female nest-site mortality focused primarily on the area of Erickson, might result in a smaller local breeding Manitoba ( 50° 30’ N, 99° 55’ W ). My population. Long-term studies of a different main objective was to determine whether species, the canvasback (A. valisineria), in changes in reproductive parameters southern Manitoba have shown that with occurred after the arrival of the raccoon. increasing raccoon numbers, nest losses Erickson is well-suited to this analysis increased and productivity decreased.30 because several historical studies cover The extent of canvasback renesting was appropriate time periods (Rogers study: thought to be closely related to predation. 1957-1960;24’32'33’ Hammell study: 1970- Consequently, for scaup I hypothesized 1972,23 [and unpublished data]; and that the arrival of raccoons as a novel Afton study: 1977-8028). All three studies predator in southern Manitoba affected were conducted at the same locale, and scaup populations at the local level. the raccoon arrival time is known and Based on this hypothesis, I predicted: occurred outside of a drought period, (a) a decrease in the size of the breeding eliminating the severe confounding population over time due to lower juvenile factor of drought on scaup populations recruitment as a result of decreased nest and reproduction.24 In addition, aerial success and female survival, and (b) a photos and discussions with the local shift in mean hatching date to later in the community suggested that agricultural season because of increased nest losses impact (amount of cultivated, hayed, or and more renesting attempts by female pastured land) in the study area was scaup. Importantly, these predictions similar across the different time periods, are not unique to the novel predator thereby controlling for major habitat hypothesis and could apply to any number change as a factor affecting the results. of alternatives based on other factors During the 10-year interval between the affecting reproduction. However, there is earlier studies (1960-1970), raccoons 56 Blue Jay Figure 3. Field study sites (denoted by the small black squares). Shaded area within expanded map represents intensively studied area near Erickson, MB. were noticed for the first time (mid- of south-western Manitoba (Fig. 3) and 1960s) by local residents in the Erickson comprises 259 ha for the 1957-60 study, area. There was no evidence found 680 ha for the 1970-72 study, and 777 that raccoons were a predator of scaup ha for the 1977-80 study. The two former during the 1957-60 study (J.P. Rogers, areas are situated entirely or mostly within pers. comm.). However, they were a the larger area. These block areas were major predator at Erickson by the early most intensively studied, but additional 1970s.23 data were collected from larger lakes and along roadside transects in the Erickson METHODS area of approximately 80 km (1957-60) Study Area and 71 km (1970-72). The topography is The study area for all three time periods rolling with numerous ponds and lakes. is situated in the parkland pothole region The uplands are a mixture of lands sown 69 (2) June 2011 57 to cereal and oilseed crops, hay, pasture, data.28 Care was taken by investigators to and native woodland. The area is well use duckling number and age to ensure described in previous studies.24'27’2834 that sightings of individual broods were not duplicated. It was not possible to Field Methods distinguish which broods were from initial Scaup were counted weekly by nests and which were from re-nests for investigators walking on the entire the earlier (unmarked bird) studies, but I Rogers and Hammell block study areas. assumed that later hatching broods were Estimates of the breeding lesser scaup from renesting attempts.25 population were derived from the average of the two or three counts of pairs seen Local Populations and Nest Success during the pre-laying period (late May To compare changes in population to mid-June). In all studies, scaup nests densities for different sized study areas, were located by observing females flying I re-examined raw data (1970-72) for the to nests and by conducting systematic foot Hammell study (data not available for the searches around the margins of all ponds Afton study) and determined the number on the study area from early June to early of scaup pairs on the same square mile August. Information collected at each (259 ha) used by Rogers. This resulted in site included vegetation characteristics, total pair counts for 1970-72 on the 259- distances to water and dry land edge, ha block area that were >75% of those number of eggs, estimated initiation date, numbers recorded on the larger 680-ha and nest success. area. I also examined raw data from United States Fish and Wildlife Service Broods of scaup are relatively easily (USFWS) annual waterfowl surveys located on rearing ponds, usually for the Erickson area,37 did regression swimming to open areas in the pond analysis for the period 1955-80, and centre when disturbed.23 28 For the Rogers compared the results to those from and Hammell studies, broods were the ground studies. Existing long-term observed in the main study block and population and productivity data from additional numbers were recorded on area other authors were available for south¬ lakes and along the roadside transects. western Manitoba38 and other Prairie- Both authors estimated brood ages by Parkland Region studies,39 40 and were juvenile plumage characteristics.23-24'35'36 compared to Erickson data. To evaluate Hatching dates were determined by nest success at Erickson for the periods backdating from the brood observation before and after the raccoon arrival, I date. For each brood, age class was examined and compared broods/pair estimated and a day-of-year hatch date data for the Rogers and Hammell studies (usually the mid-point of the age class) and nest success data from Afton’s was assigned for the Hammell data. marked bird study. I chose data for Rogers24 conducted weekly surveys author-described non-drought years on his block area, and additional ones only (1957 and 1970-72, 1978-80) in on local lakes and on the roadside just an attempt to reduce drought-induced prior to the expected fledging date of nest success biases,24 which would the earliest hatched ducklings. Hammell most likely mask any raccoon effects. conducted surveys every 3 to 4 days on Again, additional support for this decision the block area and weekly on the roadside is provided by the long-term research transect throughout the entire brood conducted on canvasbacks in south¬ period from late July to late September.23 western Manitoba.30 Here, the effect of Afton provided no hatching chronology the raccoon was most noticeable in wet 58 Blue Jay years, and I assumed that a similar pattern descriptions of spring and summer water might apply to scaup, as most scaup at conditions, dates of first laying attempts, Erickson nest over-water. Nonetheless, and meteorological data41 from Brandon, one might expect some raccoon effect in MB, 63 km south (for 1957-60) and drought years; however, during drought Minnedosa, MB, 25 km south (for 1970- periods, water levels recede drastically 72). Weather data for Minnedosa are and usually all nests become exposed incomplete for the years 1957-60. Again, to predators (including those predators I looked for years with similar climatic normally associated with dry uplands), conditions and avoided drought years in potentially negating the specific effects the analysis because their datasets are of raccoons. very small and drought seriously affects scaup reproduction at Erickson.24 In Hatching Chronology addition, to further account for possible Hammeli estimated brood age differing clutch initiation chronologies, I by juvenile plumage characteristics subtracted from the observed difference developed for scaup.36 Although Rogers in 65%HD the maximum difference aged scaup broods using criteria (6 days) in mean CIDs for first nests developed for juvenile redheads (Athya between years recorded for the same americana),35 he applied an adjustment area from 1977 to 1980.28 Recorded (6 days for each of the eight age classes) CID means could not be compared for to allow for the fact that scaup reach the the two earlier studies, since these data flight stage sooner than redheads (J.R contained unknown degrees of renesting Rogers, pers. comm.). Calculations of influence. possible results obtained using both methods for each age class revealed RESULTS that different ages could be attributed to a Breeding Population, Nest Success, given age class but that these differences and Productivity were small, averaging less than 2 days Erickson (range: + 1 to -4, Hammeli age minus Initial analysis produced equivocal Rogers age) and depended on brood age results for Erickson population trends at observation. Brood age at observation for the period before and after raccoon time is unknown for the Rogers data so arrival. Scaup breeding populations the extent and direction of any bias is were lower after the raccoon arrived, unclear. I assumed that this bias would but this decrease appears to be drought not be significant. Consequently, I did not related and occurred between 1958 and apply any corrections to Rogers’ hatching 1959, before the time of arrival (Table 1). date data and made the assumption that Admittedly, the available data are limited the aging criteria used in the two time and lacking for the years 1961-69, but periods were similar and the data were they do suggest a similar population level comparable. Mean hatch date was not of ~7 pairs/km2 immediately before (1960) provided for the earlier Erickson data, so and after (1970-72) the raccoon arrival, in order to make meaningful comparisons, which occurred in the mid-1960s. This day-of-year of 65% brood hatch (65%HD) result does not support the expectation was selected as a date for reference. of an observed breeding population To conservatively account for possible decrease after raccoon arrival. However, year effects at Erickson (late versus total scaup counted on USFWS annual early, wet versus dry) on first clutch waterfowl and habitat surveys37 near Erickson (1955-80; stratum 40: transect initiation dates (CIDs) and hence the 4, segment 4; transect 6, segments 3 observed 65%HD, I considered author 59 69 (2) June 2011 and 4) showed a downward trend from USFWS data for Erickson, no significant 1955 to 1980 (Fig. 4, simple linear trend in population numbers occurred regression, r2 = 0.199, 24 df, p = 0.02). for the period encompassing the years These data contained very high counts before and after the raccoon arrival (Fig. from 1958-62 (the drought years in 5, r2= 0.003, 19 df, p = 0.8); thus, results the Prairie-Parkland Region),42 and it from both sources for Erickson local is suspected that birds from dried-out population data concur. Therefore, the areas of the prairies immigrated to the expectation that the Erickson population relatively wetter parkland areas during would decline after the raccoon arrival is this period, especially areas like Erickson not supported by the available data. with large permanent lakes (some of which are covered by the USFWS survey Erickson scaup reproductive success segments) and that this influx bolstered (nest success and productivity) data show numbers on the survey segments. This that great variability occurred in years movement of waterfowl to wetter areas before and also in years after the arrival in time of drought is well documented,43-45 of the raccoon (Table 1). Reproductive and a previous study recorded a similar success, evaluated for non-drought increase in scaup numbers at Minnedosa years (see METHODS), was high and during the same drought period.42 By similar in 1957 and 1971 and lower in excluding the drought years from the 1970,72,78,79, and ‘80; that is, lower Table 1. Estimated breeding population (pairs/km2)a and production (broods/pair or percent nest success) of lesser scaup at Erickson, MB, 1957-1960,1970-72, and 1977-80 (see text for data sources). Numbers in parentheses are sample sizes. Raccoons arrived during the mid-1960s. Year Population Production 1957 21 0.50 (27/54) 1958 25 0.05 (3/65) 1959 9 0.08 (2/24)b 1960 7 0.18 (3/17)b 1970 7 0.23 (5/22) 1971 6 0.58 (11/19) 1972 7 0.13 (3/23) 1977 n/a 18% (17)c 1978 n/a 29% (24)c 1979 n/a 40% (47)c 1980 n/a 27% (41 )c aOn Rogers’ block study area (2.6 km2)24 Calculated from the number of ducklings/average brood size cNumber of marked birds in sample 60 Blue Jay