ONTARIO NOCTURNAL OWL SURVEY 2001 Final Report

Transcription

1 ONTARIO NOCTURNAL OWL SURVEY 2001 Final Report Photo: Bill Allen Bird Studies Canada and Ontario Ministry of Natural Resources Wildlife Assessment Program January 2002 by Debbie Badzinski Bird Studies Canada P.O. Box 160 Port Rowan, Ontario N0E 1M0

2 2001 Ontario Nocturnal Owl Survey ONTARIO NOCTURNAL OWL SURVEY EXECUTIVE SUMMARY The Ontario Nocturnal Owl Survey was initiated in 1995 in response to a need by the Ontario Ministry of Natural Resources to assess populations of representative wildlife species on Crown Lands in Ontario. Bird Studies Canada coordinates the survey in partnership with the Ontario Ministry of Natural Resources Wildlife Assessment Program. The Ontario Nocturnal Owl Survey is a roadside survey that uses playback of owls to elicit vocal responses. In 2001, there was a large increase in both the number of participants and the number of routes surveyed; 172 surveyors completed 198 routes. One hundred and ten routes were surveyed in central Ontario and 88 in northern Ontario. In total, 929 owls of 9 species (excluding 51 individuals of unknown species) were recorded on 174 routes; 24 routes did not record any owls. In central Ontario, owls were encountered on 96 of the 110 routes, with a total of 526 individuals of 9 species. In northern Ontario, owls were encountered on 78 of the 88 routes, with a total of 454 individuals of 9 species. The mean number of owls per survey was 4.8 for central Ontario and 5.2 for northern Ontario. The maximum number of owls recorded on a survey was 18 in central Ontario and 31 in northern Ontario. This year, Boreal Owls showed a significant increase in northern Ontario, and reached their highest levels since the beginning of the survey. Northern Saw-whet Owls showed a similar, but less pronounced increase in both central and northern Ontario. Great Horned Owls also increased significantly from 2000 to 2001 in northern Ontario, but showed little change in central Ontario. Barred Owl and Great Grey Owl populations showed little change. The Ontario Ministry of Natural Resources Wildlife Assessment Program reported an increase in small mammal populations in northern Ontario. It appears that the Boreal Owl and Northern Saw-whet Owl population increase in 2001 was at least partly driven by increased availability of prey. Recommendations for the 2002 survey season include: Implement a random route selection design for new routes Habitat analyses to evaluate whether routes are representative Offer surveyors digital recordings (compact discs) Send surveyors a stop description/location form and an updated route map Create scannable forms for data entry Compare annual indices for Northern Saw-whet Owls with migration monitoring data

3 2001 Ontario Nocturnal Owl Survey 2 INTRODUCTION Owls are considered good indicators of forest health because of their position at the top of the food chain and their dependence on relatively large tracts of forest. However, there is little known about the abundance and population status of most nocturnal owl populations in Canada. Most owl species are wide-ranging, secretive, primarily nocturnal, and occur at relatively low densities (Shepherd 1992). As a result of their unique biology, most owl species, in particular those breeding in boreal forests of Ontario, are not well monitored by existing bird monitoring programs such as the Breeding Bird Survey and Christmas Bird Survey. The Ontario Nocturnal Owl Survey was initiated in 1995 in response to a need by the Ontario Ministry of Natural Resources to assess populations of representative wildlife species on Crown Lands in Ontario. Bird Studies Canada coordinates the survey in partnership with the Ontario Ministry of Natural Resources Wildlife Assessment Program. Owls were chosen as indicator species because of the apparent rarity of some owl species, such as Great Gray Owl and Boreal Owl, the dependence of some owl species on mature forest, and hence their assumed vulnerability to logging practices (Francis and Czerwinski 1995). Broadcast surveys are one of the most effective methods of surveying for owls (Takats et al. 2001). Imitating or broadcasting owl vocalizations invokes vocal responses from many species of owls, although response rate varies seasonally, among species, and with environmental conditions. This survey technique has been used to determine the distribution and population status of several North American owl species and, by recording distance and direction of owls, habitat associations can also be determined (Takats et al. 2001). For these reasons, a roadside playback owl survey protocol was chosen for the Ontario Nocturnal Owl Survey based upon recommendations by Shepherd (1992) and following the design of similar surveys in Manitoba and adjacent Minnesota (Duncan and Duncan 1991, 1993) and in the Red Lake district of Ontario (Gilmore 1994). This report summarizes the results of the survey to date ( ), with emphasis on the 2001 season. In addition, some improvements to survey methodology are recommended and some future research directions suggested. METHODS DEVELOPMENT OF SURVEY PROTOCOL Central Ontario. In central Ontario (south of the 47 th parallel but on the Canadian Shield; see Figure 1) playback protocol is targeted at Northern Saw-whet and Barred Owls. There have been several changes in survey methodology since the survey began in In 1995, the central Ontario protocol consisted of 1 minute of silent listening, followed by 20 seconds of Boreal Owl calls, 1 minute listening, 20 seconds of Barred Owl, 2 minutes listening, 20 seconds of Barred Owl, and 8 minutes of listening for a total of 13 minutes. From , the effects of different types of owl vocalizations and duration of playback on owl response rate were tested. Specifically, we tested whether broadcast of Northern Saw-whet Owl calls would elicit better responses than would Boreal Owl calls, and also whether additional playback of Barred Owls would increase response rate. To do this, the 1995 protocol was

4 2001 Ontario Nocturnal Owl Survey 3 retained on one side of the tape, but beeps were used to break up the 8-minute final period into 2- minute intervals. On the other side, to be played at alternate stops, 20 seconds of Northern Sawwhet Owl calls were substituted for the Boreal Owl, and the final 6 minute period was divided into 3 periods of 20 seconds of Barred Owl calls and 1 min. 40 sec. of silent listening. The results of this study showed that Saw-whet Owls respond equally well to conspecific calls as to Boreal Owl playback. The experiment also showed that Barred Owls typically respond more slowly to playback than do other owls (see also McGarigal and Fraser 1985). Therefore, beginning in 1999, Boreal Owl was used as the first species on the tape, for consistency with the Northern protocol and an additional Barred Owl playback was added at each stop. We also added a second minute of silent listening to the beginning of the survey, as in Northern Ontario. Although additional Barred Owl playback increases response rate, to reduce the length of time that volunteers spent at each stop, the last Barred Owl call was removed from the broadcast tape in This shortened the amount of time spent at each stop from 14 minutes to 12 minutes. Northern Ontario. In northern Ontario (north of the 47 th parallel; see Figure 1), the target species are Great Gray Owl and Boreal Owl. From the protocol at each stop consisted of 1 minute of passive listening, 20 seconds of Boreal Owl calls, one minute of passive listening, 20 seconds of Great Gray Owl calls, and a final listening period, for a total of 3 min. 40 sec. In 1999, an additional minute of silent listening was added to the beginning of the survey, to meet national owl monitoring standards. In 1995, the distance between stops was 0.8 km, and there were 25 stops per route (for a total distance of 19.2 km). To reduce the number of owls that were potentially being counted twice the distance between stops was changed to 1.6 km and the number of stops was reduced to 20, for a total route length of 30.4 km, in Some routes were shorter than 20 stops because of inadequate road lengths. Beginning in 2000 we reduced the minimum number of stops per route to 10 (for a total of 14.4 km); the maximum number of routes remained at the traditional 20 stops (see below for 2000 survey protocol). SURVEY TIMING In order to determine the best time period to conduct the owl surveys, from in central Ontario, and 1995, 1997, and 1998 in northern Ontario, volunteers were asked to survey their route during two different survey windows (March 13 18; April 3-16). In 1999 in central Ontario, and 1996 and 1999 in northern Ontario, each owl survey team was asked to survey their route three times during the following survey windows (March 5-15; April 2-12 April; April 22- May 2). The results of this study showed there would be some loss of survey power with the reduction to one survey window, but the advantages of increased volunteer participation and longer term commitment by volunteers would likely compensate for loss of power (Francis and Whittam 2000). As a result, beginning in 2000, volunteers were asked to survey their route on a single evening within the month of April. We did, however, encourage volunteers to survey within the first two weeks of the period in order to avoid potential problems with noise from frogs and/or spring runoff or melt water problems. Furthermore, if poor weather and/or road conditions were encountered in the first two weeks, volunteers would still have two weeks in which to complete their route.

5 2001 Ontario Nocturnal Owl Survey 4 CURRENT PROTOCOLS Central Ontario. In central Ontario, survey routes consisted of 10 stops spaced 2.0 km apart for a total route length of 18 km. At each stop, volunteers played one side of their broadcast tape, which is 12 minutes long. The tape started with a two-minute silent listening period, divided into two one-minute periods by double-beeps on the tape. Following the silent listening period, a 20- second Boreal Owl broadcast was played, followed by a one-minute listening period. Then the Barred Owl broadcast was played for 20 seconds, followed by a two-minute listening period. The Barred Owl broadcast was repeated three more times with a silent listening period between each broadcast. Northern Ontario. At each stop, volunteers played one side of their broadcast tape, which is 4 min. 40 sec. long. The tape started with a two-minute silent listening period, divided into two one-minute periods by double-beeps on the tape. Following the silent listening period, there were 20 seconds of Boreal Owl calls, one minute of passive listening, 20 seconds of Great Gray Owl calls, and a final listening period, for a total of 4 min. 40 sec. For both central and northern Ontario routes, volunteers were asked to record when the owl was detected (e.g., during first silent listening period, after first call etc.), whether the owl was heard, seen, or both, to discriminate between newly calling owls and duplicate owls (i.e. an owl that had called previously), and to record the estimated direction and distance of each owl detected. This additional information was requested to collect better data on owl locations for assessment of habitat associations. Volunteers were also asked to record the traffic count, background noise level, odometer reading, and time at each stop, as well as any additional comments they felt were pertinent. Routes were surveyed along snow-ploughed secondary roads through forested areas in central and northern Ontario. Most survey teams consisted of two volunteers per route, one listening and one recording. Volunteers drove routes starting at least half an hour after sunset, and stopped at pre-determined intervals to survey owls. They were asked to conduct their surveys on clear calm nights with winds less than force 3 on the Beaufort scale, no precipitation, and temperatures warmer than -15 o C. Surveyors were asked to run each route once on any evening in April, but were encouraged to survey in the first two weeks of the month. Route selection Ideally, routes would have been selected randomly from all suitable secondary roads that were ploughed in winter and had little traffic. Unfortunately, this was initially considered to be too difficult to implement, because inadequate information was available on road suitability to permit us to do this centrally, and it was thought to be difficult to match volunteers with preselected routes. Instead, we relied upon volunteers to select their own routes with coordination centrally to avoid overlap in routes. Some routes in central Ontario were chosen to coincide with pre-selected Red-shouldered Hawk survey routes (Heagy and Francis 1995), while regional OMNR offices selected other routes. The generality of the results of this survey will depend on the degree to which selected routes are representative of the landscape as a whole. Routes near population centres are over-represented,

6 2001 Ontario Nocturnal Owl Survey 5 but this is difficult to avoid with a volunteer-based survey. Because most observers were probably unaware of owl numbers along routes prior to selection, routes should not be overly biased with respect to starting densities. We suspect the bias associated with restricting the survey to roadsides is probably greater than bias due to non-random selection of roads. However, without random route selection, we cannot be certain of this. National owl survey guidelines recommend the use of a randomization strategy (Takats et al. 2001); this will be implemented in 2002 for new routes (see Recommendations and future directions). Once a route was established, we encouraged volunteers to continue to survey the same route, but some routes were abandoned over the years because they were no longer accessible (e.g., no longer ploughed in winter) or because the original volunteers were no longer available to continue to run them. New volunteers were encouraged to adopt any abandoned routes if possible, but were allowed to establish new routes if necessary. Data collection and analysis Data collection. During the survey, volunteers were asked to record all owls seen or heard at each stop, and to keep track of when within each listening period each bird was heard. Participants were also asked to estimate the number of owls at a station that had already been detected at a previous station. The total number of owls on a route was determined by adding all owls that were newly detected at each station. All data are computerized in a Paradox database system. The database includes complete information on the owls recorded at each stop, including precisely which time interval each individual owl was heard calling this allows us to assess the impact of potential changes in protocol, as well as providing a basis for using data from years with slightly different protocols (by restricting analyses to birds detected during comparable parts of the protocol). The number of times a route was surveyed in each year varied throughout the study. From there were two survey windows, and in 1999 there were three windows. Not all routes were surveyed in all windows. Starting in 2000, there was only one survey window. The mean survey date for 2000 was closest to the mean survey date of the second survey window in previous years. Therefore, to estimate population changes among years, the second survey window for each route was used (provided it was completed). If the route was not surveyed in the second window, data from the first window were used. Analysis and summary programs were written in SAS (SAS Institute Inc. 1996). Annual population indices. We calculated annual indices using Poisson regression with a loglink model, treating year and route identifier as class variables. This approach is based on methods developed for analysis of Breeding Bird Survey data (Link and Sauer, 1997, 1998). To compare among years, only owls detected during standardized parts of the protocol and/or adjusted counts for variation in protocol (e.g. differing amounts of playback in central Ontario) were used. As a result, some counts were not integer values, but this did not appear to cause any problems for the SAS program. The issue of over-dispersion, which greatly complicates the analysis, but has been found to be important for some BBS data, was not considered. Most likely, variances have been slightly underestimated, which may over-estimate the statistical significance of changes. For each species, only routes on which the species had been detected at least once

7 2001 Ontario Nocturnal Owl Survey 6 during the survey period were included. Routes on which a species has never been detected do not provide information on trends for that species. This procedure was implemented with the generalized linear model procedure PROC GENMOD (SAS Institute Inc. 1996). Comparing models with and without year effects using a likelihood ratio test tested annual variation in counts. If significant differences were found, contrasts were used to determine which pairs of years differed significantly (P<0.05). Volunteer participation RESULTS In 2001, 172 surveyors participated in the Ontario Nocturnal Owl Survey. In total, 275 routes were assigned to survey teams, of which data were returned for 198 routes (72%). One hundred and ten routes were surveyed in central Ontario and 88 in Northern Ontario. One hundred and fifty-two volunteers surveyed one route each, 15 surveyed two routes each, 4 surveyed 3 routes each and one enthusiastic pair of volunteers surveyed 4 routes (thanks Bob and Joanne Knudsen!). Data from 198 routes from the 2001 survey were included in trend analyses. Of the 198 routes, 143 had been run in at least one previous year, and 55 were new (Figure 1). Of those run before, 17 have been surveyed in all 7 years. The 77 routes that were assigned, but for which data were not returned, included both routes that have been run in previous years as well as new routes that were supposed to have been established in It is likely that some of these were actually run, and the data may be returned at a later date. Owl distribution and abundance In total, 929 owls of 9 species (excluding 51 individuals of unknown species) were recorded on 174 routes; 24 routes did not record any owls (Figures 2, 3 and Table 1). Routes with owls had 1 to 31 individual owls, and 1-5 species. In central Ontario, owls were encountered on 96 of the 110 routes, with a total of 526 individuals of 9 species. In northern Ontario, owls were encountered on 78 of the 88 routes, with a total of 454 individuals of 9 species. The mean number of owls per survey was 4.8 for central Ontario and 5.2 for northern Ontario. The maximum number of owls recorded on a survey was 18 in central Ontario and 31 in northern Ontario.

8 2001 Ontario Nocturnal Owl Survey 7 Table 1 Number of individuals of each owl species and number of routes on which each species was detected during the 2001 Ontario Nocturnal Owl Survey in central and northern Ontario. Central Ontario Northern Ontario Species Individuals Routes Individuals Routes Boreal Owl Northern Saw-whet Owl Barred Owl Great Gray Owl Great Horned Owl Long-eared Owl Eastern Screech Owl Northern Hawk-Owl Short-eared Owl Unknown Owl Survey timing Starting in 2000, a single survey window protocol was implemented. In central Ontario from , there were two survey windows, and in 1999 there were three. In northern Ontario, there were two survey windows in 1995, 1997, and 1998 and three windows in 1996 and In 2001, the mean survey date for central and northern Ontario was 17 April (Table 2). This is almost identical to 2000, when the mean survey dates were 18 April and 17 April for central and northern Ontario respectively. Table 2 Mean survey dates from in central and northern Ontario for each survey window. Starting in 2000, a single survey window protocol was implemented. Central Ontario Northern Ontario Year March 12 April 16 March 14 April March 10 April 14 March 11 April 30 April March 08 April 12 March 12 April March 12 April 13 March 09 April March 09 April 28 April 11 March 09 April 28 April April 17 April April 17 April

9 2001 Ontario Nocturnal Owl Survey 8 Figure 1. Map of Ontario showing boundaries of the study area, and the location of the routes. All routes north of the 47 th parallel use northern Ontario protocol, and all routes south of the 47 th parallel use central Ontario protocol. routes run in 2001 and at least one previous year, new route in 2001, route run in 2000 but not in 2001, routes run only prior to 2000.

10 2001 Ontario Nocturnal Owl Survey 9 Figure 2. Distribution of routes where owls of the four target species (Barred Owl, Great Gray Owl, Northern Saw-whet Owl, Boreal Owl) were reported in 2001 ( ), and routes without owls detected ( ).

11 2001 Ontario Nocturnal Owl Survey 10 Figure 3. Distribution of routes where owls of the non-target species (Eastern Screech Owl, Great Horned Owl, Long-eared Owl, Short-eared Owl, Northern Hawk Owl) were reported in 2001 (, ), and routes without owls detected ( ).

12 2001 Ontario Nocturnal Owl Survey 11 Owl population changes To estimate population change, indices were calculated based on owls recorded during the second survey window each year, or the first window if the route was not surveyed during the second window. This differs from the approach used in reports prior to 2000, and may therefore affect some previous indices. Population indices for owls in central Ontario are shown in Figure 4 and indices for northern Ontario are in Figure 5. Boreal Owl. Annual indices for Boreal Owls in northern Ontario, the only region where adequate numbers were detected for analysis, differed significantly among years (P < ). There was a large decline from 1995 to 1996, followed by a large increase in 1997, and then further declines in 1998 and In 2001, there was a large, statistically significant increase in Boreal Owl population indices. Northern Saw-whet Owl. Northern Saw-whet Owl population indices showed high annual variation in both central and northern Ontario, but the variation was more pronounced in the central study area. In both northern and central Ontario, numbers of Saw-whets showed a large decline from 1995 to 1996 followed by a gradual increase and then a second large decline from In northern Ontario, the population nearly doubled between 1997 and 1998, then remained fairly stable through 1999 and then declined significantly in 2000 to the lowest count since the survey began. A similar pattern was observed in central Ontario, except that the annual population index increased substantially in 1999 to levels that were close to those in 1995, and then declined to their lowest levels ever in Saw-whet Owls increased significantly from in both central and northern Ontario. Barred Owl. In central Ontario, where they are most effectively monitored, Barred Owl populations have remained quite stable with no significant changes from In northern Ontario, Barred Owl population indices showed significant annual variation (P<0.05), but this appears to be due to a relatively high count in 1998, with little change in other years. Great Gray Owl. From , numbers of Great Gray Owls in northern Ontario declined to less than half their 1995 levels. There was a significant decline in 1996 and a significant increase in In 2000 and 2001, Great Gray Owl numbers increased (non-significantly) but have not quite recovered to 1995 levels. Estimates of Great Gray Owl population changes have very low precision, because this species is detected on few routes and rarely detected on the same route twice. Great Horned Owl. Numbers of Great Horned Owls have shown significant annual variation in northern Ontario but have remained relatively stable in central Ontario. In northern Ontario, Great Horned Owls declined significantly from , and then showed a significant increase in This was followed by a significant decline in 2000 and another increase in As with the Great Gray Owl, the survey has low power to detect population changes in this species.

13 2001 Ontario Nocturnal Owl Survey 12 Figure 4. Annual population indices for Barred Owl, Northern Saw-whet Owl and Great Horned Owl, the three owl species that are detected most frequently in central Ontario on the Ontario Nocturnal Owl Survey. Asterisks indicate significant differences between pairs of years: * P < 0.05, ** P < 0.01

14 2001 Ontario Nocturnal Owl Survey 13 Figure 5. Annual population indices for the five species of owls detected most frequently in northern Ontario on the Ontario Nocturnal Owl Survey. Asterisks indicate significant differences between pairs of years: * P < 0.05, ** P < 0.01.

15 2001 Ontario Nocturnal Owl Survey 14 DISCUSSION As a result of an active recruitment campaign, which included press releases, radio interviews and letters to naturalists clubs, the number of routes surveyed and the number of volunteers participating increased considerably in The number of volunteers participating and the number of routes surveyed increased by 31% and 28% respectively, compared to the previous year. This year, Boreal Owls showed a significant increase in northern Ontario, and reached their highest levels since the beginning of the survey. Northern Saw-whet Owls showed a similar, but less pronounced increase in both central and northern Ontario. Great Horned Owls also increased significantly between in northern Ontario, but showed little change in central Ontario. Barred Owl and Great Grey Owl populations showed little change. The Wildlife Assessment Program reported that 2001 was a very high year for small mammal populations in northern Ontario; the number of individual small mammals captured at sampling stations in north-eastern Ontario was more than double the number captured in 2000 (A. Sugar pers. comm.). A similar increase was also noted in Algonquin Park. Park biologists have been tracking small mammal populations over the last 50 years, and found that 2001 was a peak year. So, it appears that the Boreal Owl and Northern Saw-whet Owl population increase in 2001 was at least partly driven by increased availability of prey. Annual fluctuations in owl populations are not yet well understood, especially in Ontario, but presumably occur in response to prey availability (Korpimäki 1985, Sonerud et al. 1988, Houston and Francis 1996). It is not known to what extent these changes are due to changes in numbers of owls (changes in mortality and/or reproductive rates), movement of owls (emigration/immigration) or changes in behaviour (changes in calling rates in response to different reproductive activities). Swengel and Swengel (1997) found that although nocturnal auditory surveys suggested large among-year variation in number of owls detected, diurnal searches did not. Daytime searches failed to detect annual variation in owl abundance, and in years with low numbers of calling owls, diurnal searches located many more owls than did auditory surveys. This suggests that annual variation in numbers of Saw-whets detected in Ontario may be a function of their behavioural inclination to vocalize, as much as the number present. Calling frequency may be related to prey abundance and consequently breeding effort. In years with low prey availability, owls may forego breeding and therefore be less territorial and less likely to vocalize. If prey populations in central Ontario show similar population cycles, as do boreal populations, then this feature could be contributing to the large annual variation found in this survey. A graduate student at Lakehead University has initiated a research project that involves analyzing the standardized mist-net studies of Northern Saw-whet Owls at Thunder Cape Bird Observatory. It would be valuable to compare the results of the Ontario Nocturnal Owl Survey with the Thunder Cape results, to determine whether the population fluctuations documented by the Ontario Nocturnal Owl Survey are related to changes in numbers of migrating owls.

16 2001 Ontario Nocturnal Owl Survey 15 Until recently, there was little known about the habitat preferences of owls in Ontario. The Southcentral Science and Information Section of the Ontario Ministry of Natural Resources has used data collected by the Ontario Nocturnal Owl Survey to create habitat suitability models for Barred Owls (B. Naylor pers. comm.). Location of survey routes and stop locations were superimposed on maps showing the age and composition of the forest. Forest cover around stations at which Barred Owls were detected was compared to stations in which no owls were recorded. Statistical analyses were then conducted to determine the forest types and age classes that owls are found in. Results of this research show that Barred Owls tend to avoid young forest (forests dominated by shrubs and saplings), and use immature, mature, and old growth forest in proportion to its availability. Barred Owls also tend to avoid forests dominated by white pine, red pine, jack pine, or black spruce in favour of forests composed of sugar maple-beech, sugar maple-yellow birch, and black ash-red maple-yellow birch. This research has greatly improved our understanding of habitat use of Barred Owls in Ontario. Hopefully, similar analyses can be performed for other Ontario owl species. It would also be interesting to test for geographic variation in habitat use. Since 2000, volunteer owl surveyors have been collecting detailed information on the distance and direction of owls detected during the survey. Furthermore, we have invested considerable effort into determining precise stop coordinates for all owl routes in Ontario. These data can be used to greatly improve the precision and accuracy of habitat models. RECOMMENDATIONS AND FUTURE DIRECTIONS Random route selection. Starting in 2002, all new routes will be selected using a randomized route selection, to enhance statistical reliability and credibility of the results and to conform to National owl monitoring standards. The goal is to have two owl routes per degree block within the study area. Blocks that do not have roadside access or have unsuitable habitat (i.e. no forest) will be eliminated. For blocks that have less than two routes, new routes will be selected randomly. Ideally, the maximum number of routes per block is 16 (1 route per 1:50,000 map sheet). There are currently a number of blocks that exceed 16, but we will not be eliminating routes. However, if routes become inactive in saturated areas, it will be a lower priority to recruit new surveyors to fill them. When surveyors call Bird Studies Canada to request a new route, we will steer them to the nearest available area and then select their route randomly. Habitat analyses to evaluate effects of non-randomness. In order to evaluate the potential problems of calculating population trends based on routes that were selected non-randomly, a habitat analysis is required. Habitat along owl routes should be compared with habitat along roads in that area as well as habitat in the area as a whole (i.e., not along roadsides). This will tell us whether are owl routes are representative of the landscape. Digital recordings. Starting in 2002, surveyors will have the option of using compact discs or cassette tapes. We would eventually like to phase out the use of cassette tapes because CDs are higher quality and will not degrade from year to year, and therefore we will not have to send out new broadcast recordings every year. The CDs will be created at Bird Studies Canada by copying MP3 files to audio CDs.

17 2001 Ontario Nocturnal Owl Survey 16 Stop descriptions/locations. We will be continuing to collect stop coordinates to improve our ability to assess habitat associations of owls. In 2002, we will be sending owl surveyors an additional data sheet that will be used to record stop coordinates and stop descriptions (this will be optional). We currently have accurate stop coordinates for approximately 75% of active owl routes. We have sent topographic maps (1:50,000) printed from Quo Vadis (a software program that contains scanned images of all the topographic maps for Ontario) to all owl surveyors that have not provided BSC with stop locations. They were asked to mark stop locations on the map and return it to BSC. Scannable Forms. This year, we will be converting the owl data sheets into scannable forms. This will significantly reduce the amount of person hours required for data entry. Database structure. The owl data are currently stored in a Paradox database. These data will be transferred to Microsoft Access for compatibility with data scanning software. Future analyses. Further analyses of the survey results, through comparison with additional data sources, would be valuable to enhance interpretation of the observed apparent population changes. Possible analyses include comparison of annual fluctuations with data on prey availability, if the latter can be obtained, as well as comparisons of population fluctuations on the breeding grounds with migration counts from the Canadian Migration Monitoring Network (especially Thunder Cape, Whitefish Point and Long Point bird observatories). ACKNOWLEDGMENTS This project was funded by the Wildlife Assessment Program of the Ontario Ministry of Natural Resources. Hot chocolate was generously donated by Knechtel s grocery store in Port Rowan, ON. Thanks to Dean Phoenix, Allissa Sugar, Neil Dawson, and George Holborn for their help coordinating the survey, and thanks to Margaret McLaren for helping to develop this program. I thank Jon McCracken for assistance with program management and providing comments on this report, and Ian Richards for assistance with volunteer recruitment and project coordination. Charles Francis provided statistical advice and wrote the original SAS programs that were used to analyze the data. Most importantly, we heartily thank the volunteers who have become too numerous to mention by name, for their participation in the 2001 survey. These people generously donated their time and equipment (tape players, vehicles, flashlights, etc) to venture forth on cold, dark nights to survey owls. Their dedication is making an important contribution to owl conservation in Ontario. LITERATURE CITED Duncan, P. A. and J. R. Duncan Nocturnal audio/playback owl survey of southeastern Manitoba and adjacent Minnesota: 1991 Progress Report. Winnipeg, MB: Manitoba Department of Natural Resources. 8 p. Duncan, P. A. and J. R. Duncan Manitoba s nocturnal owl survey: 1993 Progress Report. Winnipeg, MB: Manitoba Department of Natural Resources. 8 p. Francis, C. M., and M. S. W. Bradstreet Monitoring boreal forest owls in Ontario using tape playback surveys with volunteers. In Proceedings of the Second International Symposium: Biology and Conservation of Owls of the Northern Hemisphere (February 5-9, 1997, Winnipeg, Manitoba).

18 2001 Ontario Nocturnal Owl Survey 17 Francis, C. M. and E. Czerwinski Nocturnal Owl Survey 1995 pilot study. Unpublished report by Long Point Bird Observatory for the Ontario Ministry of Natural Resources. 33 pp. Francis, C. M., and J. S. Dobbyn Nocturnal owl survey 1996 pilot study. Unpublished report by Long Point Bird Observatory for the Ontario Ministry of Natural Resources. 25 pp. Francis, C.M., and B. Whittam Ontario Nocturnal Owl Survey Pilot Study Final Report. Unpublished report by Bird Studies Canada for the Wildlife Assessment Program, Ontario Ministry of Natural Resources. 34 pages. Gilmore, D., and C. MacDonald Northern forest owl survey: Red Lake. Ontario Birds 14: Heagy, A., and C. M. Francis Red-shouldered Hawk and spring woodpecker survey-1995 progress report to Ontario Ministry of Natural Resources. Port Rowan, ON: Long Point Bird Observatory. 47 p. Houston, C. S., and C. M. Francis Survival of Great Horned Owls in relation to the snowshoe hare cycle. Auk 112: Korpimäki, E Rapid tracking of microtine populations by their avian predators: Possible evidence for stabilizing selection. Oikos 45: Lepage, D., C. M. Francis, and V. J. Deschamps Nocturnal Owl Survey 1997 pilot study. Unpublished report by Long Point Bird Observatory for the Ontario Ministry of Natural Resources. 23 pp. Lepage, D., C. M. Francis, and V. J. Deschamps Nocturnal Owl Survey 1998 pilot study final report. By Bird Studies Canada for the Wildlife Assessment Program, Ontario Ministry of Natural Resources. WAP pages. Link, W. A., and J. R. Sauer Estimating equations estimates of trends. Bird Populations 2: Link, W. A., and J. R. Sauer Estimation of population trajectories from count data. Biometrics 53: Link, W. A., and J. R. Sauer Estimating population change from count data: application to the North American Breeding Bird Survey. Ecological Applications 8: McGarigal, K. and J. D. Fraser Barred Owl responses to recorded vocalizations. Condor 87: SAS Institute Inc SAS System for Windows, version SAS Institute Inc., Cary, NC. Shepherd, D Monitoring Ontario's owl populations: a recommendation and report to Ontario Ministry of Natural Resources. Port Rowan, ON: Long Point Bird Observatory. 82 p. Snedecor, G. W.; Cochran, W. G Statistical methods. 6th edition. Iowa State University Press, Ames, Iowa. Sonerud, G. A., R. Solheim, and K. Prestrud, K Dispersal of Tengmalm s Owl Aegolius funereus in relation to prey availability and nesting success. Ornis Scandanavica 19: Swengel, A.B., and S.R. Swengel Auditory surveys for Northern Saw-whet Owls (Aegolius acadius) in southern Wisconsin Pages in Proceedings of the Second International Symposium on the Biology and Conservation of Owls of the northern Hemisphere. February 5-9, Winnipeg, Manitoba. USDA Forest Service General Technical Report NC-190.

19 2001 Ontario Nocturnal Owl Survey 18 Takats, D.L., C.M. Francis, G.L. Holoroyd, J.R. Duncan, K.M. Mazur, R.J. Cannings, W. Harris, D. Holt Guidelines for Nocturnal Owl Monitoring in North America. Beaverhill Bird Observatory and Bird Studies Canada, Edmonton, Alberta. 32 pp.