Pacific Seabird Group
DEDICATED TO THE STUDY AND CONSERVATION OF PACIFIC SEABIRDS AND THEIR
ENVIRONMENT
Dr. Julia K. Parrish ñ PSG Chair
School of Fisheries & Zoology Dept. Box 355020, University of Washington , Seattle, WA 98195-5020
jparrish@u.washington.edu
Dr. William J. Sydeman ñ PSG Chair Elect
Point Reyes Bird Observatory,
Craig S. Harrison ñ PSG Vice Chair, Conservation
4001 N. 9th St, Number 180, Arlington, VA 22203
charrison@hunton.com
To: Marbled Murrelet Inland Surveyors and Managers
From: Dr. Julia K. Parrish, Chair ñ Pacific Seabird Group (PSG)
Date: 10 April 2000
Re: Updates in the 1994 Inland Survey Protocol
The attached document is an update of the 1994 Inland Survey Protocol. This document clarifies many aspects of the protocol and provides additional supporting evidence for many of the recommendations. It also incorporates the information that was distributed in annual letters from PSG since 1994, consolidating this information in one place. This 2000 protocol supercedes all previous protocols.
What is New: In general, more detail has been added throughout. A more comprehensive review of the biological significance of murrelet behaviors has been added. General Surveys have been removed, as this type of survey was not reliable for documenting occupancy of a site. Radar surveys and how they can compliment audio-visual surveys have been described. Guidance on how to define survey areas, sites, and stations is better detailed. The field data form has been reorganized for clarity.
What Remains Unchanged: The number of survey visits has not changed from those established in 1992 and 1995: 4 visits in each of two consecutive years to establish presence, and 10 visits to establish occupancy. We continue to advise that you consult with regulatory agencies to resolve specific questions on survey effort. Data collected since 1994 currently are being analyzed to determine if this level of effort will change, and how survey visits should be distributed among years. When this analysis is complete, a revised protocol will be distributed for peer scientific review and adoption by the PSG. This revision is anticipated in Spring of 2001.
Approximately 70 people participated in discussions of this update during the last two Inland Survey Protocol Subcommittee meetings, held February 1999 and 2000. Thus, this document has benefited from a variety of perspectives, from those conducting surveys to federal and state regulators. As always, the survey protocol will continue to evolve as new information on murrelet ecology is brought to light.
METHODS FOR SURVEYING MARBLED MURRELETS IN FORESTS:
AN UPDATE TO THE PROTOCOL FOR LAND MANAGEMENT AND
RESEARCH
by the Pacific Seabird Group
Marbled Murrelet Technical Committee
Complied and edited by:
Diane M. Evans
William P. Ritchie
S. Kim Nelson
Elena Kuo-Harrison
Peter Harrison
Thomas E. Hamer
15 April 2000
In addition to the compilers/editors, the following members of the Pacific
Seabird Groupís Marbled Murrelet Technical Committee contributed to writing this
update:
Janet Anthony, Brian Cooper, Danielle Prenzlow Escene, Lee Folliard, Ron LeValley,
Sherri Miller, Sharon Dechesne-Mansiere, Steve Singer
Others who submitted comments and suggestions for the final product were: Gayle Anderson, Dave Buchholz, Steven Courtney, Eric Cummins, Kim Flotlin, Elizabeth Gayner, John Grettenberger, Shelley Hall, Scott Horton, Dave Huber, John Hunter, Paul Jones, Linda Long, Sarah Madsen, Douglas Meekins, Ken Ostrom, Phyllis Reed, Lynn Roberts, Clint Smith, Amanda Wilson, Mike Wilson
The 1994 Protocol on which this update was based was compiled by C. J. Ralph, S. Kim Nelson, Margaret Shaughnessy, Sherri Miller, and Tom Hamer. We gratefully acknowledge their efforts in providing the original ideas and some of the text that appears in this update.
This document was assembled by Diane Evans, Coordinator, Pacific Seabird Group Inland Survey Protocol Subcommittee.
TABLE OF CONTENTS
INTRODUCTION 1
PURPOSE AND OBJECTIVES 1
BACKGROUND ON THE SPECIES 2
METHODS 3
LITERATURE CITED 22
APPENDIX A Description of Eggs 33
APPENDIX B Observer Training Protocol 36
APPENDIX C Hearing Examination 41
APPENDIX D Forest Bird and Mammal Species Potentially Misidentified as Marbled Murrelets, and Potential Predators 43
APPENDIX E Marbled Murrelet Vocalizations 45
APPENDIX F Data Form and Instructions 47
APPENDIX G Use of Radar for Marbled Murrelet Surveys 58
LIST OF FIGURES
Figure 1. A 4-ha (10-acre) timber harvest area is located in the middle of a 122-ha (300-acre) stand of potential habitat. The survey area (timber harvest area and potential habitat within one-quarter mile of the harvest boundary) includes 83 ha (206 acres). 26
Figure 2. A 2-ha (5-acre) timber harvest area is located on the edge of a 101-ha (250-acre) stand of potential habitat. A one-quarter mile boundary around the proposed harvest area includes 26 ha (65 acres) of potential habitat. This captures a small portion of the entire contiguous habitat. The additional contiguous habitat that extends beyond the one-quarter mile boundary should also be considered for surveys. 27
Figure 3. A large survey area of 138 ha (340 acres) divided into three survey sites, with an example of a naming convention to uniquely identify survey area, sites, and stations. 28
Figure 4. The survey area (timber harvest area and potential habitat within 1/4 mile of the sale boundary) includes 85 ha (211 acres). Eleven stations (approximately one station per 8 ha [19 acres]) are needed to survey this area because of limited visibility and steep and complex terrain. 29
Figure 5. Two stations are needed to survey this 12-ha (30-acre) site that has a long and narrow shape. 30
Figure 6. Less than 6 ha (15 acres) of this 12-ha (30-acre) site would be surveyed with this survey station. At least one additional station would be needed to survey this site adequately. 30
Figure 7. Example of using one large brushed trail to access many survey stations. 31
Figure 8. Decision tree to evaluate stands for occupancy of Marbled Murrelets. 32
METHODS FOR SURVEYING MARBLED MURRELETS IN FORESTS:
AN UPDATE TO THE PROTOCOL FOR LAND MANAGEMENT AND
RESEARCH
by the Pacific Seabird Group
Marbled Murrelet Technical Committee
INTRODUCTION
In North America, Marbled Murrelets (Brachyramphus marmoratus) occur from Alaska south to Santa Cruz, California, and winter as far as Baja, Mexico. This seabird is closely associated with old-growth and mature forests for nesting (papers in Ralph et al. 1995), and declines have been attributed in part to loss or modification of forest habitat (USFWS 1997). To be effective in maintaining adequate nesting sites, forest land managers need to determine murrelet inland distributions and patterns of habitat use.
The Pacific Seabird Group (PSG), a professional scientific organization, has taken a lead role in coordinating and promoting research on murrelets. PSG surveys for Marbled Murrelets have been conducted in forests since 1992, following protocols put forth in Ralph and Nelson (1992) and Ralph et al. (1993, 1994). These protocols were designed to provide researchers and land managers with standardized techniques to detect murrelets in forests. From inland surveys conducted over the past seven years and from research directed at various aspects of this speciesí breeding ecology, much new information has been synthesized on nesting behavior, activity patterns, and habitat use.
This document is an update of the 1994 protocol. It draws on new information to improve the methods used to survey for Marbled Murrelets in forests. It provides supporting documentation for many of the recommendations, and clarifies some aspects of the protocolís use and application. The number of survey visits has not changed from those established in 1992 and 1995: 4 visits in each of two consecutive years to establish presence, and 10 visits over two years to establish occupancy. This guide is intended as a working document -- as new information is learned, the protocol will be further revised.
PURPOSE AND OBJECTIVES
The objectives of this protocol are to provide scientifically-based methods for biologists, managers, and researchers to: (1) document the presence or probable absence of murrelets in a forest stand at the time of surveys; (2) interpret the biological significance of behaviors observed to evaluate how murrelets are using stands; (3) further classify ëpresenceí stands as ëoccupiedí; (4) identify the geographic distribution of the murrelet; and (5) provide consistency in surveys among land managers. This protocol is based on analyses of previous survey data to provide a statistically-reliable approach to classifying sites. While applicable in Washington, Oregon, and California, the described methods may require modification for use in British Columbia and Alaska, and may not be applicable during years with abnormal oceanographic conditions, such as an El NiÒo year.
BACKGROUND
The Marbled Murrelet occurs only in North America. What was previously referred to as the Asian race of the Marbled Murrelet is now a separate species, the Long-billed Murrelet (Brachyramphus perdix). The Marbled Murrelet is state-listed as endangered in California and threatened in Oregon and Washington (Nelson and Sealy 1995). It also is listed as threatened in British Columbia. In September 1992, the U.S. Fish and Wildlife Service listed Marbled Murrelets as federally threatened in Washington, Oregon, and California (USFWS 1997). The species currently is not listed in Alaska.
In the Pacific Northwest, detections of Marbled Murrelets are closely associated with older-aged coniferous forests or younger forests with structural elements similar to old-growth, such as remnant old-growth trees or younger trees with platforms created by deformities or dwarf mistletoe infestations (papers in Ralph et al. 1995, Grenier and Nelson 1995, Nelson and Wilson 1999). Unlike most members of the family Alcidae, Marbled Murrelets most often nest in trees. An exception includes areas of southcentral and western Alaska and northern British Columbia, where nests have been found on the ground in forested and non-forested areas (e.g., Simons 1980). As of 1997, at least 130 tree nests had been located (Nelson 1997); the total after the 1999 field season is approximately 270 nests (K. Nelson pers. comm.). From these sites and additional data gathered over the past 15 years, it is apparent that murrelets nest in old-growth and mature coniferous forests throughout most of their range (see papers in Nelson and Sealy 1995 and Ralph et al. 1995).
Few murrelet surveys were conducted in forested sites before 1984. Methods for conducting surveys from a fixed location were initially evaluated and modified through research in Oregon and California (Nelson 1989, Paton and Ralph 1990). PSG ëprotocol surveysí (Ralph and Nelson 1992) have been conducted since 1992 on federal, state, and private forest lands. The protocol was updated in 1993 and 1994 (Ralph et al. 1993, 1994), and additional modifications were distributed in annual letters from the PSG Marbled Murrelet Technical Committee (8 March 1995, 24 April 1996, 24 April 1997, and 11 March 1998).
The 1992 protocol established the guideline of 4 visits per site in each of 2 consecutive years to determine presence/probable absence. The 1995 update letter presented the results of the preliminary analysis of ëpí and ëqí, which identified the need for 10 visits to determine occupancy with 95% probability, but did not specify how to distribute visits across years. The April 1996 and April 1997 letters recommended that the 4 visits to determine presence/probable absence and the 10 visits to determine non-occupancy be conducted in each of 2 consecutive and normal years. The March 1998 letter recognized the lack of data analyses to support the recommendation that surveys be conducted in each of 2 consecutive years, and clarified that analyses to address this issue were continuing. These analyses have not been completed at the time of this update. When they are completed, revised guidance on the number and distribution of survey visits will be provided. This current PSG protocol recognizes 4 visits in each of 2 years to determine presence, and 10 visits total for occupancy.
METHODS
Definitions
For purposes of this handbook, the following definitions apply.
In North America, Marbled Murrelet nests have been found in mature and old-growth habitat and in younger (60-80 years) forests that have trees with dwarf mistletoe or other deformations or structures that provide a nest platform (Nelson 1997, Nelson and Wilson 1999). Douglas-fir, coast redwood, western hemlock, western red cedar, yellow cedar, mountain hemlock, and Sitka spruce predominate nest stands found to date (Hamer and Nelson 1995a). Therefore, potential habitat is habitat that should be surveyed for murrelets, defined as (1) mature (with or without an old-growth component) and old-growth coniferous forests; and (2) younger coniferous forests that have platforms. A platform is a relatively flat surface at least 10 cm (4 in) in diameter and 10 m (33 ft) high in the live crown of a coniferous tree. Platforms can be created by a wide bare branch, moss or lichen covering a branch, mistletoe, witches brooms, other deformities, or structures such as old squirrel nests. It is important to note that murrelets can use small patches of habitat within larger areas of unsuitable habitat (Nelson and Wilson 1999). They can also nest in larger, residual trees that are found in low densities, sometimes less than one tree per acre (Grenier and Nelson 1995, Ralph et al. 1995). The presence of platforms is the most important stand characteristic for predicting murrelet presence in an area (Hamer et al. 1994). Platform presence is more important than tree size, which alone is not a good indicator of platform abundance. Therefore, any forested area with a residual tree component, small patches of residual trees, or one or more platforms is potential murrelet habitat. Contiguous potential habitat is that which contains no gaps wider than 100 m (328 ft).
A survey is the process of determining murrelet presence, probable absence, and occupancy of a site by visiting it on the ground and observing for murrelets. The survey area is the entire area that is being surveyed. For example, it could be an isolated 4-ha (10-acre) stand of potential habitat or 142 ha (350 acres) of a 202-ha (500-acre) stand of potential habitat. Large survey areas should be divided into survey sites, which contain one or more survey stations (see p. 10 for a more complete discussion of survey sites). A survey station is the location where the observer stands when conducting a survey visit. A survey visit is a single morning's survey. The survey period is the 2-hour period in which a survey visit is conducted; it begins 45 minutes before official sunrise and continues at least 75 minutes after sunrise, except in Alaska (see p. 17).
The unit of measure for surveys is the detection of a single bird or group of birds, defined as the sighting or hearing of one or more birds acting in a similar manner and initially occurring at the same time. Sequential detections are distinguished by a break of five seconds or more. For example, a bird flying around in a circle overhead for three minutes calling continuously would be counted as a single detection. If that bird stopped calling and was out of sight for more than five seconds and then started to call or was seen again in a different area, the observer should count it as two detections. This is because the observer would be uncertain if it was the same bird or a different bird calling. When a group is observed and then splits into two groups, the observation is treated as a single detection. If two groups are spotted separately and then coalesce, the groups should be counted as two detections.
The following definitions apply to sites that have been surveyed for murrelet activity. These definitions are detailed on p. 6, ëClassification of Sitesí. A site with murrelet presence is a site of potential habitat where there has been at least one murrelet detection. Presence sites include occupied sites. An occupied site is where murrelets have been observed exhibiting subcanopy behaviors, which are behaviors that occur at or below the forest canopy and that strongly indicate that the site has some importance for breeding. Occupied sites include nest sites. A nest site is a site with an active nest or evidence of a nest, including eggs, eggshell fragments, or a downy chick.
Significance of Murrelet Behaviors
Murrelet nests are extremely difficult to find; therefore, a set of behavioral criteria has been established to determine if potential habitat is likely to be occupied by murrelets. These behaviors have been documented at active nest sites and can be observed during an audio-visual survey, providing the rationale to use them as indicators of occupancy. In addition, these behaviors have been associated with purposes other than attending an active nest, suggesting that the stand has some importance for breeding.
Flight
Marbled Murrelet flight is direct and involves rapid, often continuous wing beats. Flight speeds average 73-136 km/h (45-85 mi/h) and can reach maximum speeds of 158 km/h (98 mi/h) (Hamer et al. 1995, Burger 1997, Cooper and Blaha 1999). Murrelets generally fly at higher altitudes en route between nesting and foraging areas, and fly lower at or near nests.
Subcanopy flights. Subcanopy flights include those below, through, into, or out of the forest canopy within or adjacent to potential habitat. Flight below the canopy is most commonly observed during the breeding season (O'Donnell et al. 1995). Adults flying to nests approach from below the forest canopy, often along a route of gaps among overstory trees or other natural ëcorridorsí (Nelson and Peck 1995, Singer et al. 1995). Nesting birds can consistently use the same flight path within a season, although each bird of a pair may have different paths, and arrival paths may differ from departure paths. Thus, birds flying along the same route on successive days could indicate nesting. In addition to direct flights to nests, murrelets can engage in ëfly-bysí before and after visits to the nest, where a nesting bird flies past the nest tree below the canopy at nest height. ëFly-bysí occurred during the incubation and nestling periods in California (Singer et. al. 1995), but also have been observed at nests after nesting was completed (K. Nelson pers. comm.). While an observer may not be aware of a nest, these flights lend support for the association of subcanopy flights with nesting. Subcanopy flights are often nonvocal, but can include wing-beat sounds. Observations of subcanopy behaviors usually consist of 1 to 2 birds.
Some flights that are observed below the canopy, and thus technically ësubcanopyí behaviors, are not indications of occupancy. For example, murrelets en route to nesting areas in the Santa Cruz Mountains flew quite low (just above the top of riparian hardwood trees) when following stream channels inland, particularly on foggy or heavily overcast mornings (S. Singer pers. comm.). In addition, low-flying birds have been observed in steep canyons or crossing ridge lines in non-habitat areas (S. Singer pers. comm., C. Smith pers. comm.). In general, if subcanopy flights are seen in areas lacking potential habitat, they are not an indication of occupancy. If there is any question about the significance of subcanopy behaviors, please consult with your regulatory agency for advice.
Landings. In addition to landing at active nests, murrelets have been observed landing in trees near known nests throughout the breeding season, or in trees with nests that were active in a previous year (Nelson and Peck 1995). Murrelets also land in trees prior to egg laying, presumably to inspect potential nest sites (Nelson and Hamer 1995) and to copulate (D. Buchholz pers. comm.). Landings also may indicate territorial behavior, resting or roosting (Naslund 1993b).
Circling and above-canopy flights. Circling and other above-canopy flights, such as dives, indicate possible occupancy of a site (reviewed by Nelson and Hamer 1995). These behaviors are a red flag that should prompt additional survey effort to observe subcanopy activity. Shallow or steep dives that originate above the canopy but terminate below canopy have been observed more frequently (67%) near known nest trees. These ëjet divesí may function to maintain pair bonds or be used in territorial defense (Nelson 1997). Circling is common over some nest sites (Nelson and Peck 1995; Hamer, Ralph, unpubl. data), and generally includes > 2 murrelets. Circles can be small (~10-20 m radius) or greater than 1 km-radius (0.62 mi). Occasionally, observers note ëpartial circles,í or birds following a curving flight path. In most cases, it is likely that the birds were circling, but limited visibility prevented the observer from seeing complete circles. Upon leaving a nest, such as after an incubation exchange or fish delivery, breeding birds may join with other murrelets over the nest site before departing for the ocean (Nelson 1997), but nesting birds at three active nests in the Caren Range of B.C. were not observed to circle before returning to the ocean (P. Jones pers. comm.), nor was circling from other birds observed over this stand. Murrelets also have been seen circling over young or non-forest habitats. However, in most cases these areas have been near or adjacent to old-growth trees (T. Hamer unpubl. data, Nelson unpubl. data). When evaluating the significance of circling behavior, the height of the bird(s) above the canopy, frequency of circling, and distance from potential habitat should be considered. We recommend that in all cases where circling is observed, additional surveys be conducted to determine occupancy.
Vocalizations
Murrelet vocalizations are described in Appendix E. Interpreting the association of calling (an audio detection) with the status of a site is difficult. The most audible call, the ëKeerí call, is heard at nest sites, while flying, and at sea (Nelson 1997). Vocalizations at the nest generally are soft and not readily audible from the ground, but are given frequently by both adults during incubation exchanges and chicks during feedings. Loud calls from the nest are more rare. Nevertheless, loud calls were heard from seven nests in Oregon while birds attended a chick or egg, or prior to egg laying (Nelson and Peck 1995, A. Wilson pers. comm.). In the Caren Range of B.C. (P. Jones pers. comm.) and in California (Singer et al. 1995), however, no loud calls were recorded during approaches or exits from active nests. Calls that emanate from one location within the survey site may be a less ambiguous indication of nesting activity than calling in general. Many ëKeerí calls are from birds heading to the local area, but some are from birds in flight traveling beyond the site being surveyed. As social interactions increase, calling also increases, and an increase in calling in late summer may be related to subadults and nonbreeders visiting forest stands (Nelson 1997).
Of the nonvocal sounds heard during surveys, ëjet soundsí are associated with dives (described under Circling), which can be associated with nest sites.
Classification of Sites
The behaviors described above lead to three classifications of sites.
Probable absence. A site of potential habitat where no murrelets were detected after the requisite number of surveys.
Presence. A site of potential habitat where murrelets were detected, but subcanopy behaviors were not observed. Additional survey effort is required at areas with birds present to determine whether or not a site is occupied. Presence sites include those with:
Occupied site . An occupied site is defined as a site where at least one of the following subcanopy behaviors or conditions occurs:
Occupied sites include nest sites, but an occupied site also can be used for purposes other than nesting that are essential for the complete life history of the bird (Nelson 1997). For example, courtship displays in other alcids can take place near, but not at, the breeding site. Murrelets have been observed landing in unsuitable trees in unsuitable habitat contiguous with or near suitable habitat in Oregon and British Columbia (K. Nelson pers. comm.). These landings generally involve more than one murrelet and the birds remain standing in these young trees for a period of time. Thus, the places where birds nest and where they engage in courtship or other breeding-related activities might not be in the exact same area or stand, but these areas are just as important as nesting sites for their life history.
During surveys, the behaviors observed should be clearly documented. When appropriate, narrative also should be provided to include additional detail and insight into reported detections.
Inland Limit for Surveys
The following data document the extent of the inland range as currently known. These are not intended as strict guidelines by state. Some regions within states might not support murrelet activity as far inland as the maximum limits listed below suggest (e.g., Dillingham et al. 1995, Witt 1998a). These data are provided as guidelines when planning surveys, particularly if the intent of inland surveys is to encompass all areas potentially used by Marbled Murrelets. Consult with your regulatory agency if you are unsure how far inland to survey in your region.
Most nests in North America have been located within 30 km (19 mi) of the coast, with the exception of Alaska, where they have been found only up to 10 km (6 mi) inland (Nelson 1997, Whitworth et al. in press). It is important to note, however, that nest searches have been conducted in more restricted areas than surveys, so the farthest inland detection should be used as your guideline for conducting surveys. In British Columbia the most distant nest from the coast was 35 km (22 mi) inland (Lougheed 1999), but a grounded fledgling with an egg tooth was reported 101 km (63 mi) inland (Rodway et al. 1992). The majority of occupied sites in Washington have been discovered within 63 km (39 mi) of marine waters. The most distant nest in Washington was 35 km (22 mi) from the coast, and farthest inland occupied site was 84 km (52 mi) in the foothills of the northern Cascade Mountains. Distribution may be more limited in Cowlitz County of southwestern Washington, as all of the more than 100 Marbled Murrelet surveys conducted beyond about 32 km (20 mi) inland have had negative results. However, an anecdotal murrelet observation was reported by Ritchie and Rodrick (2000) near Mount Adams at a distance greater than 96 km (60 mi).
Marbled Murrelets have been reported at distances of 105 km (65 mi) to 129 km (80 mi) inland at several locations in the Oregon Cascade Mountains, although nesting behaviors were not observed (Nelson 1997). The farthest nest from the Oregon marine coast has been located 49 km (30 mi) inland in the south-central Coast Range (Witt 1998b), and the farthest inland occupied site was at 61 km (38 mi; Nelson 1997). Murrelet distribution may be more limited in the Siskiyou Mountains, where the farthest inland detection was at 51 km (32 mi; Dillingham et al. 1995). In northern California, the most distant nest was 29 km (18 mi) inland, while the most distant confirmed occupied site was 40 km (25 mi). Marbled Murrelets also have been detected 59 km (37 mi) inland near Happy Camp, but more than 200 survey sites within 45 to 72 km (30 to 45 mi) of the ocean on the Six Rivers National Forest farther to the south yielded no detections (Hunter et al. 1998). In the Santa Cruz Mountains of central California, the most distant nest was found 18 km (11 mi) from the coast.
Habitat Assessment
Identifying where murrelet surveys should be conducted is a critical first step in the process. Failure to identify potential habitat, and thus ëclearí an area for management activities, could have a substantial negative impact on the population. A habitat assessment is an on-the-ground evaluation of the habitat within an area of proposed management activity. We are not attempting to define habitat here, given the large regional variation, but instead we describe the procedure in general terms. A habitat assessment cannot be completed from maps and aerial photos alone. It should include a ëwalk-thoughí of the entire project area, looking specifically for the presence of platforms or, in younger-aged areas, for small patches of habitat or remnant large trees. By definition (p. 3), large-diameter trees do not have to be present for an area to contain potential habitat. Moss cover or deformities can create platforms on smaller-diameter limbs. Alternatively, moss does not have to be present within the canopy, as murrelets can nest on duff platforms (Hamer and Nelson 1995a). Perceived lack of flight access for murrelets into an area should not eliminate that area for consideration. Stands on >20% slope often create natural access due to the layering of canopy trees, and streams create natural flyways (Hamer et al. 1994). Aspect has not been identified as a limiting factor for murrelet nests (Hamer and Nelson 1995a). In summary, any area with a residual large tree component, small patches of potential habitat, or suitable nest platforms should be evaluated for the need for surveys.
Choosing a Survey Type
The previous protocol (Ralph et al. 1994) described two types of surveys, General and Intensive, which were designed to address different objectives of murrelet surveys. General Surveys are no longer recommended for timber surveys or for research, as they were not designed to document probable absence. This protocol adds Radar Surveys as an option for very specific and limited objectives. Please note that Radar Surveys may not be used in place of Intensive Surveys for determining occupancy or probable absence.
Intensive Survey. Intensive surveys are designed to determine probable absence or presence of a specific site, help determine occupancy, monitor murrelet activity levels at specific sites (e.g., for a pre-harvest inspection), locate nests, and establish murrelet use patterns. When conducting an Intensive Survey, the observer visits only one station per morning. Intensive surveys are recommended for all proposed timber harvest and management activities. They can also be used for surveys in rugged or roadless areas.
Radar Survey. Radar surveys employ a stationary marine radar system to detect and track murrelets in flight. Radar surveys cannot determine occupancy, but can often be used to identify stands where occupancy is a possibility. Because it is likely that radar can reliably determine presence of birds in a shorter period than the current audio-visual protocol in some areas (Cooper and Blaha 1999), radar surveys can be used as a ëcoarse filterí to quickly and accurately determine whether murrelets are present near, or adjacent to, a forest stand. For the purposes of this protocol, radar surveys can be applied to document presence and help identify where follow-up efforts of intensive surveys for determining occupancy would be most effective. To apply the radar technique in addition to the standard audio-visual ground survey technique, it is necessary to consult with the appropriate state and federal agencies. A rigorous sampling design will need to be approved by these agencies. The applications of, and limitations to, radar surveys are detailed in Appendix G.
Procedures for Intensive Surveys
Intensive Surveys incorporate a three-step process:
(1) Design the survey, including habitat assessment, defining the survey area, and establishing survey sites and stations.
(2) Conduct survey visits in accordance with the protocol to determine if murrelets are present at the site.
(3) If murrelets are present at the site, increase the number of survey visits to determine if the site is occupied. This is necessary because subcanopy behaviors are often more difficult to detect.
Additional surveys could be conducted at occupied sites to locate nests or attempt to determine the birdsí spatial and temporal use patterns throughout the entire stand. This would require extensive efforts with numerous people conducting simultaneous surveys. If biologists are interested in verifying nesting within the stand, PSG has developed a protocol that assists observers with nest verification: "Techniques for finding tree nests of the Marbled Murrelet" (Naslund and Hamer 1994).
If, during any phase, behavior is found that strongly indicates nesting (i.e., any of the subcanopy behaviors, with exceptions noted on p. 6), then the site would be considered to be an occupied site.
Defining survey area and sites
Survey area. The minimum area surveyed should be the potential habitat that falls within the proposed project area and within one-quarter mile (402 m) of the project area boundary that is contiguous with the project area (Figure 1). The intent of the one-quarter mile guideline is to increase the likelihood that all of a contiguous block of potential habitat is surveyed, not just that portion that lies within the project boundary. For example, a proposed project boundary might bisect a contiguous block of potential habitat. By defining the survey area as one-quarter mile beyond the project boundary, the entire block of contiguous habitat is more likely to be included. The hypothesis that contiguous habitat is important is based on the following observations on the nesting behavior of murrelets and alcids in general:
(1) Although Marbled Murrelets nest solitarily, more than one pair of birds are usually found in a single, contiguous forest (Nelson and Peck 1995). The interaction of murrelets in a single stand seems important for social and breeding purposes.
(2) As two or more pairs of murrelets might nest asynchronously in a stand (or perhaps even renest), you may have murrelets nesting at different times - and therefore different places - in the same stand in the same year.
(3) Over several years, murrelets might use more than one nest tree or use different parts of a stand for nesting (Nelson 1997). Murrelets exhibit high nest site fidelity, with some stands supporting 20+ years of murrelet use (Divoky and Horton 1995). A few nest trees have been used in consecutive years (Singer et al. 1995, Nelson 1997, Manley 1999); however, most are not, suggesting that breeding birds may move elsewhere within a stand in successive years or may not nest every year.
When a project is planned in a large expanse of potential habitat, surveying the entire contiguous block may allow a more thorough evaluation of the potential impacts to portions of the habitat that are greater than one-quarter mile from the project boundary. For example, in many situations the potential habitat occurs in a long, linear configuration. When the project area is at the edge of this large block, even a one-quarter mile boundary may not include the entire stand of potential habitat (Figure 2). This was the intent of the guideline in the previous protocol that the survey area should include contiguous habitat within oneñquarter mile or 51 ha (125 acres), whichever was greater. This allowed for a larger portion of the potential habitat to be surveyed when a relatively small portion occurred within the one-quarter mile zone. It also provided a limit to the survey area when the contiguous potential habitat extended over a large landscape. We recommend that the one-quarter mile zone define the minimum survey area. The portion of a contiguous stand that extends beyond that also should be considered for surveys. For projects that have completed the first year of surveys under the 125 acre guideline, the second year of surveys should be completed within that design.
Other potential habitat within the one-quarter mile delineation that is discontinuous with the project area may also need to be surveyed if disturbance is a concern. Disturbance is a regulatory issue; consult with your regulatory agency for guidance.
Following are two examples of determining the survey area. The first involves a 122-ha (300-acre) stand of potential habitat, with a planned harvest of 4 ha (10 acres) located in the center of the stand (Figure 1). A 402-m (one-quarter mile) area around the edge of the 10-acre harvest would include 83 ha (206 acres). The second example involves a 101-ha (250-acre) stand of potential habitat, with a planned harvest of 2 ha (5 acres) located on the edge (Figure 2). A 402-m (one-quarter mile) area around the boundary of the 2-ha harvest would encompass 26 ha (65 acres) of potential habitat. The remaining contiguous potential habitat could be surveyed to better evaluate potential impacts.
The survey area should be defined by the occurrence of potential habitat. It should not include large expanses of unsuitable habitat, but this should be ascertained by visiting the area on the ground to determine the best way to delineate it. Potential habitat that is separated from other potential habitat by more than 100 m (i.e., surrounded by unsuitable habitat), should be delineated as its own survey area. This 100-m guideline should be applied when defining the area, not at the scale of scattered individual remnant trees or patches. In places where remnant trees are scattered equally throughout younger forest, the contiguous potential habitat should be delineated by forest that contains this combination of young and remnant trees. If a large expanse of young forest without remnant trees is adjacent to the potential habitat, it should not be included in the survey area boundary.
Survey site. A survey site is the unit by which surveys are designed and carried out, and the unit to which the requisite number of visits applies. We recommend limiting the size of the site to 49 ha (120 acres). When the survey area is small (< ~49 ha), the site encompasses the entire survey area. In this case, the terms ësurvey siteí and ësurvey areaí are interchangeable, and the protocol applies equally. More typically, survey areas are large (> 49 ha), and should be divided into sites (Figure 3). Some flexibility is allowed in exceeding the 49-ha (120-acre) site guideline, but sampling intensity and coverage are compromised when the site exceeds 55-57 ha (135-140 acres).
The survey site boundary should not be confused with the management project or survey area boundaries. Because the survey area, by definition, is contiguous potential habitat, the classification of probable absence, presence, or occupied determined at the site applies to the survey area. For example, if a large block of contiguous potential habitat is divided into three survey sites, and one of those three sites yields subcanopy detections, the entire survey area is considered occupied, not just that one site, because all the sites form one large piece of contiguous habitat.
The site contains > 1 survey stations which are laid out together and which collectively are surveyed to determine the status of the site and the survey area. For the site, every station must be visited at least once and a minimum of 4 survey visits total must be conducted per year to determine presence. For example, if a site contains less than 4 stations, more than one visit must be made to one or more of the stations (see Number of Survey Visits, p. 18). If the site contains more than 4 stations, the site will receive more than 4 visits per year. Individual survey sites within the same survey area may be visited on the same or consecutive days, but survey visits to determine presence within a survey site should be separated by a minimum of 6 and a maximum of 30 days (see Distribution of Visits, p. 16).
It is critical that each site be identified by a unique name or number and legal description that will identify that particular site over the years. Furthermore, the boundary of the site must be clearly delineated on a topographic map or aerial photo. Stations within sites also must have unique identifiers, but in addition, all stations within a site must share the same site name. It must be unquestionably clear which stations belong to a site, as there is no other way of determining if the site was surveyed with the requisite number of visits. Multiple sites within a survey area should share the same area name. Figure 3 illustrates one example of a naming convention, which uses alpha-numeric codes in a hierarchical fashion to identify stations, sites, and areas.
Survey stations and their placement
Survey station placement is one of the most crucial aspects of survey implementation. Marbled Murrelets can be difficult to detect in and around their breeding areas, in part due to their small size, rapid flight, cryptic plumage and crepuscular behaviors. Where the likelihood of detecting murrelet activity is low, such as where a small number of birds are nesting due to small stand size or extreme distance to marine waters, good station placement is imperative if murrelet use of the stand is to be correctly classified. OíDonnell (1995) reviewed the effects of station placement on the number of murrelet detections. He states that the number of visual sightings of murrelets is strongly influenced by the location of the observer. Furthermore, sensible placement of survey stations can serve to largely control site characteristics that may influence the observerís ability to hear and see murrelets.
There are three steps involved in station layout. The first step is to determine adequate coverage and establish preliminary station locations. This can be accomplished by overlaying circular mylar disks on aerial photos or topo maps. This is detailed below under ëNumber of Survey Stationsí and ëA Simple Technique for Delineating Site Boundaries and Determining Station Locationí. The maps or photos are used to identify topography; openings or gaps in the canopy; patchiness of habitat; and natural and artificially-created flight corridors such as streams, lakes, rivers, meadows, avalanche chutes, landslides, paths, and roads. Local knowledge of the area is helpful, but not essential, at the initial design stage.
The next step is to locate the stations on the ground and refine their placement based on site-specific factors. This may help to identify openings that were not evident on aerial photographs, or identify potential sources of localized noise disturbance. Because of the high proportion of audio detections during most surveys, placing stations near sources of loud noises, such as busy roads, is less optimum than a quieter location covering the same area. This step also may identify patches with the most suitable murrelet nesting habitat, such as areas with the highest density of potential nest structures. On-site review allows these locations to be factored into the survey design. Other considerations when placing stations include the growth and foliation of adjacent vegetation, increase in snow melt runoff when locating stations early in the spring, and the viewing window. Chances of detecting murrelets flying silently are increased dramatically if the birds are viewed against a light or bright sky as a background, which silhouettes the birds in the early dawn light.
A third step is not always necessary, but often overlooked. This involves the addition of new or supplementary stations which may or may not conform to the minimum requirements stated in the protocol. These additional stations can serve to better survey a stand in a difficult setting. They also can be added after surveys have begun, where detections indicate potential activity in a portion of the survey area receiving minimal coverage under the existing survey design. For example, once presence has been detected and the objective is to determine occupancy, supplementary stations can be added to augment the data previously collected. Such a station could be one that affords a good view of the target stand but is greater than 50 m from its edge. Stations could also be surveyed in tandem, with one observer placed adjacent to a stream that has good visibility but limited hearing, and a second observer at a station with quiet conditions. Note that two stations surveyed in tandem counts as only one protocol visit for the site.
Station Effective Area. The distance at which observers conducting audio-visual surveys detect murrelets determines the effective area of a survey station, and thus the number of stations needed to cover the survey site. Previous data (Ralph et al. 1994) suggested that observers generally see birds only within 100 m (328 ft) or hear birds within 200 m (656 ft). Observers can detect birds at greater distances, but many are missed at these distances and classifying behavior is more difficult. At some locations, visibility is restricted and subcanopy behaviors can only be seen at distances less than 100 m. A study on the Olympic Peninsula used radar to measure detection distances and found a steep, steady drop in the number of murrelets beyond 100 m (even without accounting for the fact that sampling area increased with distance from observer): 36 (39.1%) occurred £ 100 m from the observers, 22 (23.9%) occurred 101ñ200 m from the observers, and detections continued to drop with distance from observer (Cooper and Blaha 1999). The use of radar in recent studies has demonstrated that observers could miss a large number of murrelets in some areas. Concurrent radar and audio-visual surveys in the Santa Cruz Mountains and on the Olympic Peninsula found that ground observers missed 71-100% and 85-93%, respectively, of the murrelets detected on radar, even when provided with the birdsí bearing and travel direction by the radar operator in the California study (Cooper and Blaha 1999, Singer and Hamer 1999). Missed observations aside, this audio-visual survey protocol is built on ground observersí effective detection distances. Until additional data and more complete analyses suggest otherwise, this protocol uses 200 meters as the maximum detection distance, and thus defines station effective area as a 200-m radius circle centered on the survey station.
Based on a stationís effective area, up to 12 ha (30 acres; roughly equivalent to the area of a 200-m radius circle) can be surveyed from a single survey station under ideal circumstances. This is the maximum coverage for a single station. In many cases, each station will cover less. The number of hectares surveyed by a station will vary with topography, stand shape, canopy closure, understory denseness, location of the survey area, and station placement. An area with closed canopy, limited visibility and/or steep terrain with many drainages will need many more survey stations than is expected based on acreage only (Figure 4). For example, on the Olympic Experimental State Forest in Washington, average station density was 1 per 7 ha (17 acres) because of the presence of streams, ridges and steep slopes (Horton and Harrison 1996).
Effect of topography and stand shape. In a square stand with a slope of 0 o, one survey station will cover 12 ha. As the slope steepens, the number of stations required to effectively survey the area increases (Figure 4). This is because the 12-ha estimate of murrelet detectability is based on the horizontal distance one can see or hear a Marbled Murrelet (see above), and slope distance is not equivalent to horizontal distance. An estimate of average slope of a stand can be determined using stereoscopic analysis or from measurements on the ground; horizontal distance can then be determined from standard slope distance conversion tables. The best way to determine the number of stations needed in each stand is to use the Technique for Delineating Site Boundaries and Determining Station Placement (see below).
Stand shape also will influence the number of survey stations. If the stand is a perfect square or circle, you probably will not have to worry about stand shape. However, if your stand is rectangular or oddly shaped, you will need to make sure you cover the entire area with survey stations. For example, if you have a perfectly flat (no slope), 12-ha (30-acre) stand that is very long and narrow, one station will not adequately cover the entire stand (Figure 5).
A general rule of thumb is that your stations should be located throughout the site. Station placement should incorporate topographic features and cover every hectare of a given site, no matter the size. Stations that are located up-slope from the survey site, such as along a ridge with the survey site in a valley below, may offer a broad, sweeping view of the entire site but provide very limited chances to observe murrelets that are accessing the site from an elevation below the ridge top. If your site includes a ridgetop, mid-ridge and river bottom, you must make sure that your stations effectively survey (not necessarily be placed in) the ridgetop, mid-ridge, and river bottom. If your site is only 12 ha, but is long and narrow, you will need to place a station on each end of the site at a minimum (Figure 5). The additional number of stations required will depend on slope. Remember that if a station is placed on the edge of a site, you may be surveying less than 12 ha of that site (Figure 6).
Location with respect to openings. Generally, murrelets remain unseen to the observer; 80% of detections from two sources in Washington (n = 8376) were audio, compared with 13% visual and 7% both seen and heard (WDFW interagency database). Rates of audio detections were similar in California and Oregon (Paton and Ralph 1988, Nelson 1989). However, behaviors indicating occupancy are derived almost exclusively from visual observations. Therefore, stations should be located so that the observer has an unobstructed view of the sky. Whenever possible, stations should be placed in forest clearings, on quiet roads, at the edge of the site, or in or adjacent to rivers or streams. Murrelets often use stream or river corridors as flight paths to access nest sites. Streams create noise disturbance, but the increased opportunity to observe occupied behaviors outweighs the negative aspects of noise. However, stations should be located no farther than 50 m (164 ft) from the edge of the site being surveyed (e.g., see Figure 6). A common error is inadequate survey coverage of interior portions of survey sites. In many cases survey stations are placed along roads or adjacent to the edge of the target site because of easy access and better visibility, but generally the entire site can not be surveyed adequately if all stations are located around the perimeter. Stations must also be located within the site so that the entire site has survey coverage. Even if well-placed openings are not available in a site, station coverage should not be compromised. The number of stations in a site should not be decreased just because openings are not available or are not well-placed (the number may, however, need to be increased). Ultimately, some stations may need to be set in areas without a good view of the sky. Additionally, it is important to consider that the silhouette of a dark bird flying directly overhead against the light-colored sky is easier to see than a bird flying against a dark background when viewed from the top of a ridge or high point.
When there are few clearings within a site, such as in areas with closed canopies or steep complex terrain, visibility will be restricted and the detection of subcanopy behaviors will be very limited. To make up for a lack in visibility, we recommend that station coverage and density be increased in these sites. Surveyors should consult with their wildlife resource agency for direction in these cases.
Location with respect to potential habitat. In many younger-aged stands, potential nesting habitat often is located in small patches (micro-sites) separated by areas of unsuitable habitat. In some cases, patches containing the most likely nesting habitat may be ineffectively covered even though the site is being surveyed to the specifications of the protocol. Interpretations of what is potential (or likely) habitat differs, and the complete range of conditions murrelets use for nesting is still not known. In cases where habitat quality varies throughout the survey site (specifically, where larger residual trees containing suitable platforms are spaced at regular or irregular intervals within a site that contains no other potential platforms), survey stations should be strategically placed to cover the most likely nesting habitat within a site, as long as stations remain distributed throughout all potential habitat within the site. In cases where portions of the site do not contain potential nesting habitat (i.e., no platforms), survey stations should be placed to cover the potential habitat within the site. If more than one survey visit is required to some of the stations within the site to meet protocol, the additional surveys should be conducted at those stations with the best habitat; or a combination of best habitat, visibility, and proximity to previous detection.
Modifying station placement. To maximize the observer's chance of seeing birds, he/she may move up to 50 m (164 ft) from the station during the survey visit. The new location should be less than a one minute walk away, and the observer should note time and direction of movement. In subsequent visits to an area, additional stations can be established to obtain visual observations. For example, if birds were heard in a nearby gully during a survey, the observer can set up one or more additional station(s) in the gully to increase the probability of observing subcanopy behavior. It is important to assign a unique identification to any new stations, including those which have been moved more than 50 m. If an observer thinks that there is a good chance of observing murrelets at a particular station, additional surveys visits can be made to that station. However, all potential habitat within the survey site must be surveyed.
Summary. The following bullets summarize the most important points about survey station placement:
Intensive surveys can be labor intensive. For areas that are difficult to access because they have steep slopes, cliffs, thick brush, or are long distances from roads, it may facilitate the survey effort if one or more wide trails are brushed through the stand (this should be done outside of the nesting season if power equipment is used). These trails can serve as access points to several stations (Figure 7). In extreme cases, it may be necessary to camp out, hiking to the station before dark the evening before the survey visit.
A Simple Technique for Delineating Site Boundaries and Determining Station Location
The use of aerial photos and a stereoscope can be used to delineate site boundaries, and to locate canopy gaps, road landings, and other suitable locations from which to survey. The 3-D image you get from stereoscopic analysis of 1:12,000 (or other size) aerial photography is helpful for identifying topographic features and determining the appropriate number of stations. Once you have a pair of aerial photos set up under your stereoscope, locations for potential survey stations can be determined by looking for gaps and other open locations, and using a mylar measuring template with 30, 20 and 15 acre circular areas (see Figure 8) to estimate ground (horizontal) distance based on topography. Gaps or open location stations should be selected first, and then determine the locations for other survey stations by using the mylar template. Your 12-ha (30-acre) template should be made to the scale of the aerial photo being used. Survey sites (< ~49 ha [120 acres]) can then be delineated by circumscribing the area covered by a set of adjoining stations.
Station locations, site boundaries, and the number of stations per site should be finalized after field review. Remember to consider the growth and foliation of adjacent vegetation and increase in snow melt runoff when locating stations early in the spring. Stations should be marked with uniquely-numbered flagging and stakes. Station locations should then be marked on orthophotos or topo maps using photo interpretation and/or measured distances and azimuths from field notes, or using global positioning systems (GPS). UTM coordinates should then be determined for each survey station.
Outer ring: 12 ha (30 acres; 195-m radius)
Middle ring: 8 ha (20 acres; 160-m radius)
Inner ring : 6 ha (15 acres; 138-m radius)
When to survey
Time of year. Although nesting sites are primarily used during the breeding season, Marbled Murrelets have been observed at some inland sites during all months of the year (Carter and Erickson 1992, Cross 1992, Naslund 1993a, OíDonnell et al. 1995). However, these areas are most effectively surveyed during the spring and summer, when activity levels are greater and attendance is more consistent and longer in duration. Murrelet activity increases to moderate intensity during spring, possibly during incubation, and reaches a peak level from early July to early August in California, Oregon, and Washington (OíDonnell et al. 1995, W. Ritchie pers. comm.). This increase in activity in July might be associated with nesting birds, but also could be attributed to nonbreeders prospecting for future nest sites (OíDonnell et al. 1995, Nelson and Peck 1995, Jodice 1998, Whitworth et al. in press). The number of detections decreases markedly after this peak, presumably because many birds have completed their nesting activities and begun a flightless molt at sea.
Based on past survey data and current knowledge, surveys for management applications should be conducted during the following periods: 15 April to 5 August in California (Carter and Erickson 1988, OíDonnell et al. 1995); 1 May to 5 August in Oregon, Washington, and British Columbia; and 15 May to 5 August in southeastern and southcentral Alaska (Kuletz et al. 1994, but see Brown et al. 1999 for potentially earlier start in southeastern Alaska). These dates bracket a substantial portion of the incubation period into the nestling period, based on chronologies identified by Hamer and Nelson (1995b), but should not be confused with breeding seasons for these areas (see below). For the purposes of researching breeding ecology or monitoring nest sites, surveys could be initiated at least two weeks earlier and extended at least two to three weeks beyond the periods recommended above.
The breeding season is defined by the earliest known nesting and latest known fledging dates, and is used by regulatory agencies to avoid adverse effects to the species. The breeding season extends 24 March ñ 15 September in California, and 1 April ñ 15 September in Oregon and Washington. Thus, the survey period misses some nesting activity, and potentially some opportunities to determine occupancy at a site. For example, 13 of 26 (50%) nests in California were active after, and 4 (15%) before, the survey season (Hamer and Nelson 1995b). Of 22 nests documented in Oregon, 7 (32%) were active after, and 1 before, the survey season (K. Nelson unpubl. data). This trend is consistent with data from Washington, where 33% of 9 nests were active after the survey season, and a combined total of 44% were active during either the pre- or post-survey season (W. Ritchie unpubl. data).
Murrelet visitation to nesting areas during the non-breeding season may be important in forming or maintaining pair bonds, retention of nest sites, and for selecting future nest sites (Naslund 1993a, Nelson 1997). At two sites in northern California, calling frequency (mean number of calls per detection) was greater during winter than spring and summer, although the duration of detections was shorter (OíDonnell et al. 1995). However, birds are also more likely to be absent during winter, leading to incorrect probable absence determinations (Brown et al. 1999). Therefore, while winter surveys may be helpful for determining site presence in some areas, they cannot be counted towards surveys required in a given year.
Distribution of visits. Surveys should be conducted at regular intervals over the entire survey season. Past survey efforts have shown that detection levels can fluctuate greatly at the same survey area, or even the same station, throughout the breeding season (Manley et al. 1992, Rodway et al. 1993, Kuletz et al. 1995, Jodice 1998). Surveys should be scheduled to begin within the first two to three weeks of the survey season. To help maintain even distribution, surveyors should aim for a minimum of 6 and a maximum of 30 days between survey visits when the survey objective is to determine presence or probable absence. We recommend that when increased effort is needed to determine occupancy, survey visits be spaced as evenly as possible throughout the breeding season, aiming for no more than 2 of the requisite visits occurring within 6 days of each other.
Most areas show an increase in detections in July, but the timing of peak detections varies year to year. For example, in Washington, peaks have occurred from 24 June to mid August (W. Ritchie pers. comm.). Given that an objective of this protocol is to detect murrelets if they are present, survey visits should adequately cover this time of increased activity. We recommend that half of the total annual visits to each site be conducted from the last week of June on, with at least half of these visits conducted within the peak period of activity (generally the first three weeks of July in British Columbia, Washington, Oregon, and California). For example, if 4 visits are required, 2 would be scheduled before the last week of June and 2 would be after the last week in June. Of these latter 2, 1 would occur within the first three weeks of July. In Alaska, at least half of the recommended annual visits should be conducted at each survey site after 1 July, with at least half of these visits corresponding to the peak of activity during the last three weeks of July. Adherence to this schedule, as closely as possible, will increase the quality of surveys and result in a more accurate determination of activity.
Time of day. As in earlier versions of this protocol, the survey period in California, Oregon, Washington, and British Columbia is defined as the 2 hours from 45 minutes before to 75 minutes after official sunrise or for 15 minutes after the last detection, whichever is longer. In southeastern Alaska, surveys should begin at least 60 minutes before sunrise (Brown et al. 1999), and surveys should begin 90 minutes before official sunrise in southcentral Alaska (Kuletz et al. 1994). Exceptions to this timing are detailed below under ëEnvironmental conditions affecting surveysí. By following these guidelines, some survey visits will last longer than 75 minutes after sunrise, especially on overcast days or days with heavy fog when detections generally continue longer. However, if a survey has unquestionably determined occupancy during the regular 2-hour survey period, staying longer is not necessary, although it could yield additional supporting detections.
More recent radar surveys in Washington consistently have detected murrelets in stands earlier than 45 minutes before sunrise. An average of 20% of radar detections occurred before the official survey start time at stands on the Olympic Peninsula (Cooper and Blaha 1999). However, because many of these targets were silent and would not have been detected on a PSG audio-visual survey in the near-dark conditions, surveys will continue to begin 45-90 minutes before sunrise depending on regional location (see above). It also should be recognized that opportunities to observe occupied behaviors can occur after the survey period, particularly during chick rearing. For example, while there appears to be an initial wave of fish deliveries to chicks right at sunrise (in low light conditions when an observer is less likely to detect them), second feedings occurred on average 54 minutes after sunrise (SE 9.6, n = 40 observations) and as late as 225 minutes post-sunrise (Nelson and Hamer 1995b). Similarly, at three nests in British Columbia observed in three different years, 63% of 104 feedings occurred more than one hour after sunrise (P. Jones pers. comm.). Later arrival times generally were associated with cloudy mornings. Thus, additional opportunities to observe an occupied behavior (i.e., adults flying into a stand to deliver fish) occur after the end of the survey period, particularly during the height of chick rearing.
Use the Nautical Almanac to determine sunrise times for your area. Do not rely on tide tables, local newspapers, or television stations because they can vary up to 15 minutes from official sunrise. Sunrise tables can be obtained from the U.S. Naval Observatory at web site http://aa.usno.navy.mil/AA/data.
Marbled Murrelets also can be detected inland during the evening. Radar surveys generally find consistent but lower volume of flights inland compared with morning surveys (B. Cooper pers. comm.). Evening audio-visual surveys could be useful in determining presence or occupied behavior, but are not recommended as part of this protocol because they would not count toward determining probable absence.
Environmental conditions affecting surveys. The effects of environmental conditions on murrelet surveys are twofold. They affect (1) the timing, duration, and intensity of murrelet activity; and (2) the ability of observers to detect the birds audibly and/or visually.
Murrelet activity at inland sites begins later, lasts longer, and is often more intense on mornings with overcast conditions, fog, drizzle, or rain than on mornings with clear conditions (Hamer and Cummins 1990, Manley et al. 1992, Naslund 1993b, Rodway et al. 1993, Nelson and Peck 1995). On the Olympic Peninsula, radar observations indicated that the timing of murrelet movements averaged 10 minutes later on overcast or foggy mornings (Cooper and Blaha 1999). If rainy, cloudy, or foggy conditions exist at the end of the regular 2-hour survey period, observers who continue to survey for an additional 30 minutes might detect possible late activity. Cloudy conditions are defined as greater than 75% cloud cover. Foggy conditions are defined as a cloud ceiling lower than the height of the tallest trees at the site or by low fog which decreases horizontal visibility to less than 100 m.
The conditions described above also potentially limit an observerís ability to detect murrelets audibly or visually. Rain and wind can make it difficult to hear murrelets calling. Low cloud ceilings or thick fog make it difficult to see murrelets. We recommend that if conditions that limit murrelet detectability, including heavy rain, hail, strong wind, logging activity, vehicle traffic, or loud aircraft, exist for more than 10% (12 minutes) of the survey period, the survey be rescheduled and repeated again on another morning soon after, unless occupied behaviors are detected on that morning. These conditions also include a cloud ceiling lower than the height of the tallest trees at the site or low fog which decreases horizontal visibility to less than 100 m. Because murrelets might still be detected during these conditions, if the surveyor is already at the station, the survey should not be terminated (even if it will be repeated), and the observer should remain for the duration of the 2-hr period unless heavy rain or strong wind threaten his/her safety. Note that the 12-minute limitation refers to 12 minutes of interrupted observations (continuous or discontinuous) once the survey is underway, and does not allow for the survey to begin late or end early.
Number of survey visits
The following data and recommendations are unchanged from previous protocols. The original analyses that determined 4 visits for presence and 10 for occupancy were not designed to address how those visits should be distributed across years, and they did not include no-detection sites. The analysis currently underway, which also incorporates more recent survey data, will address these issues.
However, there is evidence that one year of intensive surveys is inadequate to determine probable absence, presence, or occupancy at a site. Nelson (unpubl. data) found that murrelets occupied several stands in year one, were absent in year two, and occupied the stands again in year three. This finding, coupled with the variation between survey visits and the potential for between-year variation due to changes in ocean conditions and food supplies (e.g., El NiÒo Southern Oscillation), all indicate that probable absence of murrelets from a site is difficult to determine.
In addition to year to year variation in detection rates at sites, the number of detections, especially subcanopy behaviors, are affected by limited visibility. Ralph, O'Donnell, and Miller (unpubl. data) analyzed data from two areas with moderate (85%) and open (60%) canopy cover and relatively low detection rates of five or fewer detections per morning. They found that it would require between two and five survey visits to determine, with a 95% certainty level, that murrelets were probably absent from the area.
Based on the potential for variation in detection levels at survey areas, intensive surveys should be conducted for at least two consecutive years. To establish presence or probable absence of murrelets at a site, each survey station should be visited at least once per year or a minimum of four survey visits per year to each survey site, whichever is the greatest number of visits.
The two consecutive years should include at least one year with normal levels of breeding to account for years in which low breeding effort may result in a reduced number of detections in otherwise occupied stands. We have no definitive basis to determine a normal year at this time, but will address this topic when revised guidance on number and distribution of survey visits is distributed.
The number of visits per station will vary with the number of stations established at survey sites. If one to three stations are established, divide the number of visits among stations so the survey effort equals four visits per year for two consecutive years (i.e., 1 station = 4 visits per year; 2 stations = 2 visits per station per year; 3 stations = 1 visit to 2 of the stations and 2 visits to the third station per year). Additional visits should be conducted at the station(s) of highest quality (i.e. those with the greatest number of detections, the best view of the sky or stand, and/or in habitat with the highest potential). If four or more stations are established, a minimum of one visit per station per year for two consecutive years is needed to determine presence or probable absence. If murrelets are present at a survey site but subcanopy behaviors have not been observed, at least 10 visits total are needed to determine occupancy.
For purposes of evaluating survey effort needed at a stand, refer to the decision tree (Figure 8). The following example demonstrates the use of the branching choices. There is an isolated 10-ha (25-acre) site of older forest where surveys are to be conducted to assess murrelet activity. One or two stations are established in the site, depending upon topography and station location, and one survey visit is conducted. If no murrelets are detected, or if no subcanopy behaviors are seen, the site must be surveyed again, alternating visits between stations if more than one station is needed. As the decision tree illustrates, a minimum of 8 survey visits (4 each year for 2 consecutive years) with no detections will be required to determine probable absence of murrelets from the site. However, if murrelets are detected on the first survey visit (or any subsequent visit), another path is followed in the tree. If the detection(s) include(s) observations of subcanopy behaviors, the site is determined to be occupied, and no further survey visits are required. However, depending on the objective of the surveys, you may choose to continue surveys at the site. If murrelets are detected, but no subcanopy behaviors are observed, a second survey visit will be necessary and the decision process begins again. Caution: a ëpresenceí detection near the end of the second survey year might require additional years of surveys to determine if the site is occupied. Increased survey effort should begin immediately following the documentation of presence to avoid adding additional years to the survey effort.
How Long Are Surveys Good For?
The detection of occupied behaviors in forests implies that the area serves as a breeding location for murrelets. As a breeding area, murrelets may nest there every year, in alternate years, or once in several years (e.g., Manley 1999). The extent of colonization, recolonization, or abandonment of nest areas is unknown. However, recent observations of murrelets in 70-100 year-old forests regenerated from heavy timber harvest in Mendocino County, CA, and in northwestern Oregon may indicate colonization as the habitat has matured to suitability (R. LeValley pers. comm., D. Buchholz pers. comm.), although it is not known if birds merely moved from an adjacent, contiguous site or from a longer distance away. In addition, Marbled Murrelets are believed to have strong breeding site fidelity (DeSanto and Nelson 1995, Divoky and Horton 1995). Forest patches, nest trees and nest cups have been reused in subsequent years (by the same or different birds), and murrelets have been observed landing in a previously-used nest tree in a year when it was not used for nesting (Nelson and Peck 1995, Singer et al. 1995, Hamer and Meekins 1999, Manley 1999). Repeated use of forest stands suggests that these sites play a role in supporting reproduction.
Although it is possible for murrelet presence/probable absence in forest stands to change through time, we have no data from which we can recommend how long after surveys are completed that the results of those surveys remain valid. Murrelet surveys reflect the breeding status of sites for the time period during which surveys were conducted. We believe that occupied stands should be treated as occupied indefinitely. For probable absence sites, if a significant time lag (>5 years) occurs between the completion of protocol surveys and the implementation of activities that would modify suitable habitat, additional surveys may be appropriate to support the results of previous surveys. Consult with regulatory or evaluating agencies regarding these issues.
Data Collection
Training
Training is recommended for observers conducting surveys on most forest birds (Kepler and Scott 1981), and Marbled Murrelets are no exception. Intensive training and annual review and evaluation in detecting and identifying Marbled Murrelets and their vocalizations is strongly recommended, as most murrelets remain unseen to the observer. It is recommended that an intensive instructional period with a minimum of four training mornings be implemented. Training should be conducted at a site with high activity levels to expose trainees to a wide range of vocalizations and activity levels during the morning. Trainees should be provided a tape with the full range of vocalizations of known murrelet calls, and be able to compare them with similar calls of other species, such as American Robin, Varied Thrush, Evening Grosbeak, and Red-shouldered Hawk. They must also become familiar with the common call groups (e.g., ëKeerí group, ëGroaní group, ëWhistleí group) described further in Nelson (1997) and Dechesne (1998). We recommend that all trainees have their hearing tested by a professional, and have adequate vision. See Emlen and DeJong (1981) and Ramsey and Scott (1981) for discussions on counting birds and variable hearing abilities. See Appendices B-E for more details on training, evaluations, hearing tests, confusing species, and vocalizations.
Data quality
In addition to each agency or entity housing their own survey data, data also are voluntarily submitted to state or regional clearing houses. Thus, it is essential that the data be accurate. We recommend that data quality be assessed at several levels, beginning with the supervising field biologist of the field crews. Supervisors should have field experience with murrelet surveys, and should review all data sheets to help assure that the data meets the highest quality possible. This review should ensure that: correct and consistent site and station identifiers were used, the survey visit started on time, observations were not disrupted for more than 12 minutes total, detections were accurately recorded, and occupied detections were accurately defined. We also recommend that relatively inexperienced surveyors (i.e., with only 1-2 seasons conducting surveys) not be responsible for delineating survey sites and designing station layout.
Equipment needed
Equipment for surveying should include: a clipboard, pencil, data forms, wrist watch, a light source (i.e., a headlamp or a flashlight), binoculars, compass, and a permanent marker and colored flagging for marking the locations of survey stations. A tape recorder is strongly recommended for all surveys and is extremely useful in areas of high activity. Use of a tape recorder helps to prevent missing detections, as the observer can continuously watch the sky, and allows one to record simultaneous detections.
Reporting observations
We recommend that data collected during survey visits be recorded on the data sheet provided in Appendix F. This data form was revised by the Washington Department of Fish and Wildlife and differs from that provided in the 1994 Protocol.
Detailed information on murrelet behavior needs to be recorded with each observation. Include in the notes section of the form information on the location of the bird's flight (over drainage, ridge, etc.), unusual behaviors or interactions, and details on subcanopy behaviors (e.g., ëbird flew between two trees and then headed up the Drift Creek drainageí).
Observations of birds landing in trees, and chicks or eggshells on the forest floor, should be reported immediately to interested scientists and responsible wildlife agencies in your area so that active nests can be searched for.
LITERATURE CITED
Brown, M., J. G. Doerr, J. Fowler, A. Russell, and P. J. Walsh. 1999. Marbled Murrelet activity patterns and survey efficiency at inland sites in southeastern Alaska. Northwestern Naturalist 80:44-50.
Burger, A. E. 1997. Behavior and numbers of Marbled Murrelets measured with radar. Journal of Field Ornithology 68: 208-223.
Carter, H. R. and R. A. Erickson. 1988. Population status and conservation problems of the marbled murrelet in California, 1892-1987. Report to the Nongame Bird and Mammal Section, California Department of Fish and Game, Sacramento, CA, Job II.B.2. 74 pp.
Carter, H. R. and R. A. Erickson. 1992. Status and conservation of the marbled murrelet in California, 1892-1987. Pp. 92-108 in H. R. Carter and M. L. Morrison, eds. Status and conservation of the marbled murrelet in North America. Proc. West. Found. Vertebr. Zool. 5.
Cooper, B. A., and R. J. Blaha. 1999. Development of radar as a tool to increase the accuracy and efficiency of inland surveys for marbled murrelets. Unpubl. report prepared for Olympic Natural Resources Center, Forks, WA, by ABR, Inc., Forest Grove, OR. 48 pp.
Cross, J. A. 1992. Over winter surveys for the marbled murrelet east of Verlot, Washington 1990-1991 and 1991-1992. Unpubl. Rep. to Wash. Dept. Wildl., Olympia, Wash. 10 p.
Dechesne, S. B. C. 1998. Vocalizations of the marbled murrelet (Brachyramphus marmoratus): vocal repertoire and individuality. M.Sc. Thesis. University of Victoria, Victoria, B.C.
DeSanto, T. L. and S. K. Nelson. 1995. Comparative reproductive ecology of auks (family alcidae) with emphasis on the marbled murrelet. Pages 33-47 in C. J. Ralph, G. L. Hunt Jr., M. G. Raphael, and J. F. Piatt, eds. Ecology and conservation of the marbled murrelet. U.S. Dept. Agric., For. Serv. Gen. Tech. Report PSW-GTR-152, Albany, CA.
Dillingham, C. P., R. C. Miller, and L. O. Webb. 1995. Marbled murrelet distribution in the Siskiyou National Forest of southwestern Oregon. Northwestern Naturalist 76:33-39.
Divoky, G. J., and M. Horton. 1995. Breeding and natal dispersal, nest habitat loss, and implications for marbled murrelet populations. Pp. 83-87 in Ralph, C. J., G. L. Hunt, Jr., M. G. Raphael, and J. F. Piatt, tech. eds. Ecology and conservation of the marbled murrelet. U.S. Dept. Agric., For. Serv. Gen. Tech. Report PSW-GTR-152, Albany, CA.
Emlen, J. T. and M. J. DeJong. 1981. The application of song detection threshold distance to census operations. Pages 364-352 in C. J. Ralph and J. M. Scott, eds. Estimating numbers of terrestrial birds. Studies in Avian Biology No. 6, Cooper Ornithological Society.
Grenier, J. J., and S. K. Nelson. 1995. Marbled murrelet habitat associations in Oregon. Pp. 191-201 in Ralph, C. J., G. L. Hunt, Jr., M. G. Raphael, and J. F. Piatt, tech. eds. Ecology and conservation of the marbled murrelet. U.S. Dept. Agric., For. Serv. Gen. Tech. Report PSW-GTR-152, Albany, CA.
Hamer, T. E. and E. B. Cummins. 1990. Forest habitat relationships of marbled murrelets in northwestern Washington. Unpublished report, Wildlife Management Division, Nongame Program, Washington Department of Wildlife, Olympia, WA. 57 pp.
Hamer T. E. and D. J. Meekins. 1999. Marbled Murrelet nest site selection in relation to habitat characteristics in western Washington. Report to the U.S. Fish and Wildlife Service, North Pacific Coast Ecoregion, Olympia, WA. 28pp.
Hamer, T. E., and S. K. Nelson. 1995a. Characteristics of marbled murrelet nest trees and nesting stands. Pp. 69-82 in Ralph, C. J., G. L. Hunt, Jr., M. G. Raphael, and J. F. Piatt, tech. eds. Ecology and conservation of the marbled murrelet. U.S. Dept. Agric., For. Serv. Gen. Tech. Report PSW-GTR-152, Albany, CA.
Hamer, T. E., and S. K. Nelson. 1995b. Nesting chronology of the marbled murrelet. Pp. 49-56 in Ralph, C. J., G. L. Hunt, Jr., M. G. Raphael, and J. F. Piatt, tech. eds. Ecology and conservation of the marbled murrelet. U.S. Dept. Agric., For. Serv. Gen. Tech. Report PSW-GTR-152, Albany, CA.
Hamer, T. E., B. A. Cooper, and C. J. Ralph. 1995. Use of radar to study the movements of Marbled Murrelets at inland sites. Northwestern Naturalist 76: 73-78.
Hamer, T. E., W. P. Ritchie, E. B. Cummins, and C. W. Turley. 1994. Forest habitat relationships of marbled murrelets in western Washington. Unpublished report, Wildlife Management Division, Nongame Program, Washington Department of Wildlife, Olympia, Washington. 51 pp.
Horton, S. and P. Harrison. 1996. Olympic Experimental Forest marbled murrelet surveys: 1996 Annual Report. Washington Dept. Natural Res., Forks, WA.
Hunter, J. E., K. N. Schmidt, H. B. Stauffer, S. L. Miller, C. J. Ralph, and L. Roberts. 1998. Status of the marbled murrelet in the inner north coast ranges of California. Northwestern Naturalist 79:92-103.
Jodice, P. G. R. 1998. Behavioral ecology of marbled murrelets (Brachyramphus marmoratus) in forest and marine ecosystems of Oregon. PhD. Dissertation. Oregon State University, Corvallis, OR.
Kepler, C. B. and J. M Scott. 1981. Reducing bird count variability by training observers. Studies in Avian Biology 6: 366-371.
Kuletz, K. J., N. L. Naslund and D. K. Marks. 1994. Identification of marbled murrelet nesting habitat in the Exxon Valdez oil spill zone. Exxon Valdez Oil Spill Restoration Project Final Report (Restoration Project R15). U. S. Fish and Wildlife Service, Anchorage, AK. 70 pp.
Kuletz, K. J., D. K. Marks, N. L. Naslund, N. J. Goodson, and M. B. Cody. 1995. Inland habitat suitability for marbled murrelets in southeastern Alaska. Pages 141-149 in C. J. Ralph, G. L. Hunt Jr., M. G. Raphael, and J. F. Piatt, eds. Ecology and conservation of the marbled murrelet. U.S. Dept. Agric., For. Serv. Gen. Tech. Report PSW-GTR-152, Albany, CA.
Lougheed 1999. The characteristics of 23 marbled murrelet nests in BC located by radio telemetry. Abstract. Pacific Seabirds 26(1): 52.
Manley, I. A. 1999. Re-use of nest trees by marbled murrelets. Abstract. Pacific Seabirds 26 (1): 39.
Manley, I. A., R. Shortt, and A. E. Burger. 1992. Marbled murrelet activity patterns in the Carmanah Valley on the southwest coast of Vancouver Island. Pages 71-75 in K. Vermeer, R. W. Butler, and K. H. Morgan, eds. The ecology, status and conservation of marine and shoreline birds on the west coast of Vancouver Island. Occasional Paper Number 75, Canadian Wildlife Service, Environment Canada, Ottawa, ON.
Naslund, N. L. 1993a. Why do marbled murrelets attend oldñgrowth forest nesting areas year-round? Auk 110:594-602.
Naslund, N. L. 1993b. Breeding biology and seasonal activity patterns of marbled murrelets (Brachyramphus marmoratus) nesting in an old-growth forest. M.S. Thesis, University of California, Santa Cruz. 146 pp.
Naslund, N. L., and T. E. Hamer. 1994 Techniques for finding tree nests of the marbled murrelet. Pacific Seabird Group Marbled Murrelet Technical Committee, U.S. Fish and Wildlife Service, Migratory Bird Management, 1011 E. Tudor Road, Anchorage, Alaska 99503.
Nelson, S. K. 1989. Development of inventory techniques for surveying marbled murrelets (Brachyramphus marmoratus) in the central Oregon Coast Range. Report to the Nongame Program, Oregon Department Fish and Wildlife, P.O. Box 59, Portland, OR, Publ. No. 88-6-01.
Nelson, S. K. 1997. Marbled Murrelet (Brachyramphus marmoratus). In The Birds of North America, No. 276 (A. Poole and F. Gills, eds.). The Academy of Natural Sciences, Philadelphia, PA, and the American Ornithologistsí Union, Washington, D.C.
Nelson, S. K. and T. E. Hamer. 1995. Nesting biology and behavior of the marbled murrelet. Pp. 57-67 in Ralph, C. J., G. L. Hunt, Jr., M. G. Raphael, and J. F. Piatt, tech. eds. Ecology and conservation of the marbled murrelet. U.S. Dept. Agric., For. Serv. Gen. Tech. Report PSW-GTR-152, Albany, CA.
Nelson, S. K., and R. W. Peck. 1995. Behavior of marbled murrelets at nine nest sites in Oregon. Northwestern Naturalist 76: 43-53.
Nelson, S. K., and S. G. Sealy, eds. 1995. Biology of the marbled murrelet: inland and at sea. Northwestern Naturalist 76.
Nelson, S. K. and A. K. Wilson. 1999. Marbled Murrelet habitat characteristics of state lands in western Oregon. Unpubl. 1997/98 Annual Rep., OR Coop. Fish and Wildlife Research Unit, Oregon State Univ., Department of Fisheries and Wildlife, Corvallis, OR. 108 pp.
OíDonnell, B. P. 1995. A review of the effects of station placement and observer bias in detections of marbled murrelets in forest stands. Pp. 139-140 in Ralph, C. J., G. L. Hunt, Jr., M. G. Raphael, and J. F. Piatt, tech. eds. Ecology and conservation of the marbled murrelet. U.S. Dept. Agric., For. Serv. Gen. Tech. Report PSW-GTR-152, Albany, CA.
OíDonnell, B. P., N. L. Naslund, and C. J. Ralph. 1995. Patterns of seasonal variation of activity of marbled murrelets in forested stands. Pp. 117-128 in Ralph, C. J., G. L. Hunt, Jr., M. G. Raphael, and J. F. Piatt, tech. eds. Ecology and conservation of the marbled murrelet. U.S. Dept. Agric., For. Serv. Gen. Tech. Report PSW-GTR-152, Albany, CA.
Paton, P. W. C., and C. J. Ralph. 1988. Geographic distribution of the marbled murrelet in California at inland sites during the 1988 breeding season. Unpublished report, California Department of Fish and Game, Sacramento. 35 pp.
Paton, P. W. C. and C. J. Ralph. 1990. Distribution of the marbled murrelet at inland sites in California. Northwestern Naturalist 71:72-84.
Ralph, C. J., and S. K. Nelson. 1992. Methods of surveying marbled murrelets at inland forest sites. Pacific Seabird Group, Marbled Murrelet Technical Committee. May 1992. 21 pp.
Ralph, C. J., S. K. Nelson, M. M. Shaughnessy, and S. L. Miller, compilers. 1993. Methods for surveying marbled murrelets in forests. Pacific Seabird Group, Marbled Murrelet Technical Committee, Technical Paper #1. March 1993. 47 pp.
Ralph, C. J., S. K. Nelson, M. M. Shaughnessy, S. L. Miller, and T. E. Hamer. 1994. Methods for surveying marbled murrelets in forests: a protocol for land management and research. Pacific Seabird Group, Marbled Murrelet Technical Committee. March 1994. 51 pp.
Ralph, C. J., G. L. Hunt, Jr., M. G. Raphael, and J. F. Piatt, tech. eds. 1995. Ecology and conservation of the marbled murrelet. U.S. Dept. Agric., For. Serv. Gen. Tech. Report PSW-GTR-152, Albany, CA.
Ramsey, F. L. and J. M. Scott. 1981. Tests of hearing ability. Pages 341-345 in C. J. Ralph and J. M. Scott (eds.). Estimating numbers of terrestrial birds. Studies in Avian Biology No. 6, Cooper Ornithological Society.
Ritchie, W. P., and E. Rodrick. 2000. Marbled murrelet Brachyramphus marmoratus. Pp. 4.1-4.23 in Priority Habitats and Species Recommendations, Vol. IV: Birds. Wash. Dept. Fish and Wildl., Olympia, Wash.
Rodway, M. S., H. R. Carter, S. G. Sealy, and R. W. Campbell. 1992. Status of the marbled murrelet in British Columbia. Pp. 17-41 in H. R. Carter and M. L. Morrison, eds. Status and conservation of the marbled murrelet in North America. Proc. West. Found. Vertebr. Zool. 5.
Rodway, M. S., H. M. Regehr, and J.-P. L. Savard. 1993. Activity patterns of marbled murrelets in old-growth forest in the Queen Charlotte Islands, British Columbia. Condor 95:831-848.
Simons, T. R. 1980. Discovery of a ground-nesting marbled murrelet. Condor 82:1-9.
Singer, S. W. and T. E. Hamer. 1999. Gazos Creek marbled murrelet monitoring program ñ 1999 Annual Report. Unpubl. report for the Sempervirens Fund and the Apex Houston Trustee Council, prepared by Steven W. Singer Environmental and Ecological Services, Santa Cruz, CA.
Singer, S. W., D. L. Suddjian, and S. A. Singer. 1995. Fledging behavior, flight patterns, and forest characteristics at marbled murrelet tree nests in California. Northwestern Naturalist 76: 54-62.
U.S. Fish and Wildlife Service. 1997. Recovery plan for the threatened marbled murrelet (Brachyramphus marmoratus) in Washington, Oregon, and California. U.S. Dept. Interior, Fish and Wildlife Service Region 1, Portland, OR.
Whitworth, D. L., S. K. Nelson, S. H. Newman, G. B. van Vliet, and W. P. Smith. In press. Foraging distances of radio-marked Marbled Murrelets from inland areas in southeast Alaska. Condor.
Witt J. W. 1998a. Distribution of the marbled murrelet in southwestern Oregon. Northwest Science 72:96-101.
Witt, J. W.. 1998b. Notes on activity and characteristics of an inland marbled murrelet nest site in Douglas County, Oregon. Northwestern Naturalist 79:27-32.
Figure 1. A 4-ha (10-acre) timber harvest area is located in the middle of a 122-ha (300-acre) stand of potential habitat. The survey area (timber harvest area and potential habitat within one-quarter mile of the harvest boundary) includes 83 ha (206 acres).
Figure 2. A 2-ha (5-acre) timber harvest area is located on the edge of a 101-ha (250-acre) stand of potential habitat. A one-quarter mile boundary around the proposed harvest area includes 26 ha (65 acres) of potential habitat. This captures a small portion of the entire contiguous habitat. The additional contiguous habitat that extends beyond the one-quarter mile boundary should also be considered for surveys.
Figure 3. A large survey area of 138 ha (340 acres) divided into three survey sites, and an example of a naming convention to uniquely identify survey area, sites, and stations.
Figure 4. The survey area (timber harvest area and potential habitat within 1/4 mile of the sale boundary) includes 85 ha (211 acres). Eleven stations (approximately one station per 8 ha [19 acres]) are needed to survey this area because of limited visibility and steep and complex terrain.
Figure 5. Two stations are needed to survey this 12-ha (30-acre) site that has a long and narrow shape.
Figure 6. Less than 6 ha (15 acres) of this 12-ha (30-acre) site would be surveyed with this survey station. At least one additional station would be needed to survey this site adequately.
Figure 7. Example of using one large brushed trail to access many survey stations.
Figure 8. Decision tree to evaluate stands for occupancy of Marbled Murrelets.
APPENDIX A
DESCRIPTION OF MURRELET EGGS AND EGGSHELL FRAGMENTS
Compiled by Steve Singer
Size and Shape
Marbled murrelet eggs are subelliptical in shape with sizes ranging from 57.0 ñ 63.0 mm in length and 35.0 ñ 39.5 mm in width (as reviewed in Nelson 1997). Egg mass ranged from 36 ñ 41 g in five records reviewed by Nelson (1997). One reported measurement of eggshell thickness was 0.21 mm at the waist (Kiff 1981). Surface texture is usually non-glossy.
Color and Markings
Egg background color can be described in general terms as olive-green, lime green, or greenish-yellow and more precisely corresponds to Munsell colors of 2.5 GY 8/3, 2.5 GY 8/4, 7.5 Y 8/4, 7.5 Y 8.5/4, and rarely, 6.5 GY 8/3 (see Table A-1). Eggs are variably marked with irregular spots and splotches that are brownish, blackish, grayish, purplish, or sepia-like in color (Figure A-1). Spots and splotches may be as long as 8 mm in their longest dimension (Becking 1991), although most are smaller than 2 mm in diameter.
It is not yet known if there is any geographic variation in egg color or markings. Some published descriptions have failed to match eggshell color with known color standards, thereby limiting their usefulness. Those that have done so have used Ridgway (1912), Smithe (1974, 1975, 1976), or the Munsell Book of Color (Anonymous 1976). Of these color standards, only the latter has enough described colors to provide an exact match for all egg colors based on unfaded color swatches. A comparison of different color standards used to describe Marbled Murrelet eggshell colors is provided in Table A-1.
Figure A-1. Nearly intact side of large Marbled Murrelet eggshell fragment, Big Basin Redwoods State Park, 1993. Photo by S. Singer.
Table A-1. Background Marbled Murrelet egg color as defined by different color notation standards.
|
Ridgeway (1912) |
Smithe (1974, 1975, 1976) |
Munsell Book of Color (Anon. 1976) |
|
"pale glass green" |
No equivalent |
2.5 GY 8/3 |
|
"pale chalcedony yellow" |
No equivalent |
7.5 Y 8/4 7.5 Y 8.5/4 |
|
"pale dull green-yellow" |
No equivalent, but somewhat lighter that #59 "lime green" and more yellow than #162 D "opaline green" |
2.5 GY 8/4 2.5 GY 8.5/4 |
|
"pale turtle green" |
#162 D "opaline green" |
6.5 GY 8/3 |
Sources: descriptive articles in References, also unpublished data.
Recommendations on Describing Eggshell Fragments
Eggshell fragments are often found in murrelet nests or on the ground below. Their condition can be useful in determining the fate of the nest if not otherwise known. Researchers should collect the following information:
Eggshells should be donated to museums and scientists with the proper permits in your area. It is not permissible to keep eggshell fragments without the proper federal and state permits.
Copies of the Munsell Book of Color are available in the library of any college or university with an Arts Department or can be ordered from Gretag MacBeth, 617 Little Britain Road, New Windsor, New York, 12553. Their web site is at http://munsell.com.
References
Anonymous 1976. Munsell book of color. Gretag MacBeth, New Windsor, NY.
Becking, R. W. 1991. Eggshell fragments of the Marbled Murrelet (Brachyramphus marmoratus) in San Mateo County, CA. Northwestern Naturalist 72(2): 75-76.
Day, R. H., K. L. Oakley, and D. R. Barnard. 1983. Nest sites and eggs of Kittlitzís and Marbled Murrelets. Condor 85: 265ñ273.
Hirsch, K. V., D. A. Woodby, and L. B. Astheimer. 1981. Growth of a nestling Marbled Murrelet. Condor 83: 264-265.
Kiff, L. F. 1981. Eggs of the Marbled Murrelet. Wilson Bulletin 93: 400-403.
Nelson, S. K. 1997. Marbled Murrelet (Brachyramphus marmoratus). In the Birds of North America No. 276 (A. Poole and F. Gill, eds.). The Academy of Natural Sciences, Philadelphia, PA. and the American Ornithologistís Union, Washington, D.C.
Pratt, H. D. and J. P. OíNeill. 1976. Naturalistís color guide [Book Review]. Auk 93(2): 404-406.
Preston, F. W. 1953. Shapes of birds eggs. Auk 70(2): 160-182.
Reed, P. and C. Wood. 1991. Marbled Murrelet chick and eggshell fragments from inland Washington. Northwestern Naturalist 72: 77-78.
Ridgway, R. 1912. Color standards and color nomenclature. Published by Author. Washington, D.C.
Sealy, S. G. 1975. Egg size of murrelets. Condor 77: 500-501.
Simons, T. R. 1980. Discovery of a ground-nesting Marbled Murrelet. Condor 82: 1-9.
Singer, S. W., N. L. Naslund, S. A. Singer, and C. J. Ralph. 1991. Discovery and observations of two tree nests of the Marbled Murrelet. Condor 93: 330-339.
Singer, S. W., D. L. Suddjian, and S. A. Singer. 1995. Fledging behavior, flight patterns, and habitat characteristics of Marbled Murrelet tree nests in California. Pp 54-62 in S. K. Nelson and S. G. Sealy, eds. Biology of the Marbled Murrelet Inland and At Sea. Northwestern Naturalist 76.
Smithe, F. B. 1974. Naturalistís Color Guide Supplement. American Museum of Natural History, New York, NY.
. 1975. Naturalistís Color Guide. American Museum of Natural History, New York, NY.
. 1981. Naturalistís Color Guide ñ Part III. American Museum of Natural History, New York, NY.
APPENDIX B
MARBLED MURRELET OBSERVER TRAINING PROTOCOL
Compiled by Sherri L. Miller1, C. John Ralph, and Ron LeValley
Introduction
Presented here is a protocol to train and evaluate potential observers. The training program helps the trainees to develop their ability to see and hear murrelets in the forest and to accurately record observations on a data form. The evaluation process provides a method for determining if an individual's abilities will yield reliable and dependable survey data.
Training for first and second year murrelet surveyors should include all of the following steps: (1) a hearing test (see Appendix C); (2) a seminar on murrelet biology and forest survey protocol; (3) field training, with a minimum of three survey mornings, from a qualified instructor within or outside of your agency or organization; and (4) a field exam with a qualified evaluator in their geographic area. Trainees should take the field exam after they understand the protocol and are proficient in survey techniques. Once a trainee passes the field exam, they are qualified to conduct murrelet forest surveys.
After two years of survey experience that includes murrelet detections on multiple surveys, training in subsequent years should include steps (3) and (4) as listed above, except with two to three survey mornings in the field recommended prior to the field exam. Surveyors who do not perform murrelet surveys regularly should also include step (2) in their annual evaluation. It is important that surveyors refamiliarize themselves each year with the calls and techniques needed to conduct accurate murrelet surveys. We also recommend that to help maintain their skills, surveyors who do not encounter murrelets during the season should visit a site with moderate activity levels at least one time during the season. This mid-season refresher would best take place during late June or very early July to prepare a surveyor for the increased activity levels documented in July.
Observer Qualifications
Our experience indicates that most individuals with adequate sight and hearing skills are capable of being trained to recognize Marbled Murrelets following the PSG protocol. However, the quality and reliability of observations is greatly enhanced if surveyors possess basic bird identification skills, or, preferably, begin with the ability to identify by sight and sound the common birds of the survey areas. Surveys taking place at sites with low or zero murrelet abundance require a higher degree of competence and documentation (Hunter and LeValley 1996). Given the expense of sorting out false positive detections, land managers should be willing to expend the effort to insure that the data gathered are of the highest quality possible.
Seminar
A seminar on the biology of the Marbled Murrelet should include the following: a summary of murrelet biology, including species description, breeding chronology, flight behavior, habits, habitat and nest site description, and a summary of potential threats to the bird. A slide show or video including pictures of adults, juveniles, chicks, eggs, eggshell fragments, and some habitats used by murrelets should be included. The seminar can provide information regarding the legal history and current status of the species. Questions from the audience should be answered regarding all aspects of the biology of the species.
The importance of adequate training and preparation for the evaluation should be emphasized at the seminar. Proper training will not only help observers to pass the evaluation, but also will improve the quality of data collected throughout the season.
The survey protocol should be reviewed and information presented on where to survey (potential habitat), when to survey (dates and times of survey), how to establish the survey site, and the number of survey stations required. Examples of how to establish survey stations should be presented with a discussion of where to place stations at the site.
A complete description of how to record and interpret data and bird behavior should be included in all aspects of the training and reviewed annually.
Field Training
Field training should always be conducted at an area of high murrelet activity, preferably well in excess of 25 detections per morning. If this level of activity is not found in the local area, it is recommended that trainees be transported to an area of high murrelet activity.
Before the first day of training, it will be helpful to the trainees to read and become familiar with the PSG survey protocol (Evans et al. 2000) and to listen to the Marbled Murrelet vocalization cassette tape with accompanying descriptions. An outline of the daily objectives for the training and equipment for surveys should be obtained from the instructor prior to the training session. Equipment needed for the training includes: a tape recorder, at least one blank cassette tape, binoculars, a compass, a watch, and 5-10 blank survey data forms.
Outline of field training schedule
Day 1.--The first day of the field training begins 15 minutes before the survey time, at an area that will not disturb nesting murrelets. Trainees can listen to the murrelet vocalization tape while the instructor identifies the types of murrelet calls. The instructor should discuss calls from other species which may cause confusion. The training tape can include some of these calls of other species.
At the survey training site, trainees observe and listen for murrelets while the instructor points out the birds and calls. The instructor can discuss several topics. These include: (1) murrelet behaviors in the forest and the importance of behaviors in identifying occupied sites; (2) the data sheet, including the types of data taken and priorities when recording information; (3) observation and recording techniques; and (4) calls of other birds which can cause misidentification problems.
The use of the Notes section on the data form should be encouraged. Interpretation of survey results will be enhanced by narrative notes that clarify the data. The instructor should discuss the importance of using binoculars to identify some species which can be confused with murrelets (e.g., swallows, swifts). It also should be noted that since the use of binoculars during a survey can cause the surveyor to miss murrelets, their use should be limited to species verification. The instructor should encourage trainees to ask questions throughout the session and during a discussion period following the survey.
During the last portion of the survey period, the instructor can record a few detections to demonstrate recording methods. At the close of the session, trainees are asked to practice recording before the next day's session by observing birds of any species flying overhead. By recording these birds, they will become more familiar with the data required and the order of recording the information.
A classroom session on this day can be used to explain details of recording observations on the data forms and mapping detections, and to show videos and/or a slides.
Day 2.--On this day, trainees practice identifying murrelet calls and observing behaviors during visual detections. The instructor should measure and mark distances and tree heights at the training site to help trainees sharpen their skills for estimating distance to, and height of, the birds. A 50-m or 100-m tape can be used after the session to further help with distance estimates.
The instructor should record a few detections on a tape recorder, play back the recording, discuss the data with the trainees, and answer questions. Trainees can then record detections on their own while receiving assistance from the instructor. At the end of the morning's session, trainees should transcribe a portion of their data with the assistance of the instructor. This is an excellent way to see what data they are missing or recording incorrectly. Again, we suggest that trainees spend some time before the next session observing and recording birds of other species. Estimating height and distance also can be practiced on other birds.
Day 3.--All trainees can conduct a complete survey on this day, as the instructor circulates between trainees, helping with comments on accuracy and technique. At the end of the survey, tapes are transcribed, and any questions on data are clarified by the instructor. Trainees should be familiar with the techniques for conducting and recording a murrelet survey by the end of this day. It is helpful for the trainer to record and transcribe a segment of the morningís activity for comparison to the trainees transcriptions.
Day 4.--If the instructor also is a qualified evaluator, the simultaneous survey described below will be conducted on this day of the training.
Evaluation Survey (Field Exam)
When training is completed, an evaluating agency or organization should be contacted and arrangements made for an evaluation survey. Evaluation is based on the results of a simultaneous survey conducted by the trainees and an evaluator. The number of participants per evaluation will be determined in part by the size of the site. More importantly, the evaluator must be able to watch the participants and their reactions to birds to assess their ability. We recommend that group size be limited to 10 trainees per evaluator whenever possible, with a maximum of 12 trainees per evaluator.
Participants should arrive at the site early enough to allow time for instructions and still begin the survey at the appropriate time. During the survey, trainees are positioned approximately 5-8 meters apart. This helps ensure that observers have essentially the same viewing field, such that similar numbers of birds can be detected by all observers, but reduces the likelihood that they will cue in on detections by watching the evaluator or other observers. Watches should be synchronized or a time check recorded on the tape recorders at the beginning of the survey. The evaluator may call out a time check during the survey, at which time all surveyors record the time on their tape.
In periods of low activity during the evaluation survey, the evaluator can record calls of other species, recording the same type of information as for a murrelet. These observations can then be checked against the trainees' recorded data to determine whether species are being correctly identified. The evaluator should record at least 10 to 20 observations of other species.
At the end of the survey, the evaluator and trainees should go to a location where the data are transcribed under the direct supervision of the evaluator. After the data are transcribed, all of the data sheets are turned into the evaluator, who tallies the results for each participant.
Evaluation of Survey Results
To evaluate the results, we suggest that, for each 10-minute period of the survey, the number of detections of murrelets be tallied according to the following six categories: number heard; number seen; total number of detections; number of detections with occupied behaviors; number detected within 200 meters of the participant; and those detected at greater distances. If birds are both heard and seen, they are tallied once in each of the first two categories.
Each surveyor's results for three categories -- number heard, number seen, and number of occupied detections and/or total detections -- are compared with the evaluatorís observations for each 10-minute period and for the entire survey. Two criteria of success are described below, one for sites with many birds, making it likely that some birds would be missed during peak activity, and another criteria for sites with fewer detections.
To ensure consistency with previous evaluations, it is best that the tallied data be reviewed again by another experienced person for interpretation and evaluation of the results. For example, in California, one instructor reviews all of the results obtained under each evaluator, and then these decisions are reviewed again by a representative of the California Division of Forestry and California Department of Fish and Game.
Sites with many birds
At a site where the evaluator records 35 or more detections, reasonable measures of success for a surveyor are the following: if the participant records at least 60% of the number of observations in two of the three categories, and at least 50% in the remaining category, it can be considered that the participant has sufficient skill to determine the presence of murrelets in a forest stand. These figures are based on our experience in training people and comparing with expert surveyors. In these comparisons, the expert surveyors always detected in excess of 70% of the best observer. Further, we feel that any person detecting more than 60% of the birds in a stand with 35 or more detections would be unlikely to overlook so many birds that a nesting stand would be misclassified as "unoccupied".
Sites with fewer birds
If the survey site generally has fewer murrelet detections and the evaluator records less than 35 detections during the simultaneous survey, a different measure of success can be used. A participant should record at least 70% of the number of observations in two of the first three categories, and at least 60% in the remaining category.
At the discretion of the instructor and evaluator, the criteria listed above may be relaxed for trainees that meet one or more of the following conditions: (1) birds not heard by the trainees are usually in excess of 200 m from the evaluator; (2) occupied behaviors are in excess of 80% of the standard and agree with the evaluator; and (3) missed detections occur during very busy (greater than 10 detections) 10-minute periods when the presence of many birds may make individual detections difficult to define.
Qualified Instructors and Evaluators
Instructors and evaluators should be highly-qualified field ornithologists very familiar with not only murrelets, but also all other bird species (especially their calls and songs) at evaluation sites. An Evaluator should have recently performed at least one season of murrelet surveys at sites with many detections or several seasons at lower activity sites. An Instructor is a senior-level evaluator who should have a minimum of three yearsí survey experience from a variety of survey situations (both high and low detection sites) and in a range of forest stand types. Instructors should have a demonstrated ability to teach and interpret the survey protocol. As such, an instructor must be knowledgeable in the areas of murrelet ecology, general habitat associations, protocol interpretation, survey design, and regional management and regulatory requirements.
Instructors are responsible for the first three days of the training sessions as outlined above. The final evaluation should be performed by an evaluator (who may also be an instructor) who has proven current abilities to survey consistently within 10% of other evaluators.
Instructors and Evaluators should complete an extensive refamiliarization session annually, and an annual hearing test should be done. The annual session should include a complete review of changes in survey protocol, new information suggesting alternative interpretation of survey data, and an update from local regulatory agency staff. It also should include simultaneous surveys with other instructor/evaluators. Consistent results between the evaluators during simultaneous surveys must be achieved before outside evaluations begin. The level of consistency between the surveyors should be within 10%. A potential evaluator should spend at least 5 mornings conducting simultaneous surveys with a qualified and experienced evaluator, and obtain the same 10% consistency.
Follow-up Surveys
Follow-up surveys should be conducted by crew leaders with the trainees at their assigned survey sites after the initial training and evaluation. These surveys help to identify deficiencies in survey technique which may develop once observers are conducting field work. Two types of follow-up surveys should be conducted: (1) at low-use sites, within 1-2 weeks after certification; and (2) a mid-season survey at any site with detections, especially for those who have not seen or heard murrelets during the early part of the survey season.Follow-up surveys at low-use sites are important to verify that observers are (1) identifying single murrelets in areas with few observations; and (2) not confusing murrelet calls with the calls of other forest birds in their survey areas. Because most training and certification are done at high-use sites, it is imperative that crew leaders verify that observers know how to accurately conduct surveys at low-use sites. It is recommended that these follow-up surveys take place for 1-2 days at sites with an average of 10-20 detections per morning. The crew leader should conduct a simultaneous survey, similar to the initial evaluation, to identify how the observers would benefit from additional instruction. If no low-use sites are available in your area, high-use sites can be used. In this case, the survey period could be split between the periods of peak murrelet abundance and the non-peak times. The crew leader could then focus on the non-peak times and compare the numbers and types of observations recorded by the observers.
Mid-season training should occur during late June or very early July, and should include 1-2 days of surveys at low- or high-use sites. Crew leaders should review the survey protocols and reevaluate the observersí survey skills. This also is an important time to answer questions that have developed over the survey season and to revitalize crew morale.
Literature Cited
Hunter, John E., and Ron LeValley. 1996. Improving the reliability of marbled murrelet surveys in low abundance areas. Pacific Seabirds 23(1): 3-4.
APPENDIX C
HEARING EXAMINATION for MARBLED MURRELET FOREST SURVEYS
Compiled by William Ritchie and S. Kim Nelson
Introduction
Given that a larger proportion of Marbled Murrelet detections are audible rather than visual, normal levels of hearing are required of all Marbled Murrelet survey personnel. Surveyors should have their hearing tested by a certified audiologist or physician prior to conducting protocol surveys. These standard tests are available at a reasonable price and offered at any clinic with an audiologist on staff. Most employers will reimburse employees for the cost of testing. Some large firms and agencies employ their own audiologist.
In addition to having normal hearing, or corrected hearing that meets the test criteria, a prospective murrelet surveyor must attend a recognized training program and demonstrate a proficiency in their ability to conduct protocol Marbled Murrelet surveys under the guidelines set forth by the Pacific Seabird Group. Individuals also should have normal or corrected vision.
Examination Procedure
Prior to testing, the patient will be asked to answer questions pertaining to their medical background and exposure to sources of loud noise. Audiologists use specialized equipment calibrated to provide diagnostic pure-tone audiometric testing. An audiometer provides a measure of a personís ability to hear sounds of different frequencies and intensities. These tests are typically performed in sound-treated examination rooms in order to obtain accurate results. The results of the testing should report the patientís hearing thresholds at sound frequencies within the normal range of human hearing, between 250 Hz and 8,000 Hz. Upon completion, the audiologist or physician should provide the patient with an audiogram and confirmation of normal hearing ability. An audiogram represents the hearing thresholds in decibels (dB), and can be displayed graphically or as a list of values.
It is recommended that a hearing test be conducted prior to hiring individuals for murrelet survey work. Producing the results of an acceptable hearing test should be a condition of hire for everyone expected to conduct protocol murrelet forest surveys. Results of the hearing test must also be reviewed by the training evaluator before a surveyor can qualify as proficient. A personís hearing should be tested at a minimum of once every two years, or more frequently if they have been exposed to any loud noise. The Occupational Safety and Health Association (OSHA) defines loud noise to be of an intensity >85 dB for >8 hours in duration (e.g., small aircraft flights, chainsaw, gunshots, loud music, etc.). This is roughly equivalent to a situation where a normal level of conversation within three feet begins to become difficult to discern due to the intensity of the noise. When assessing previous exposure to loud noise, one should consider that as noise intensity levels increase, the duration time of exposure before reaching the critical threshold will decrease. OSHA recommends annual testing whenever an individual is exposed to these conditions.
Evaluating Test Results
A review of the audiogram is necessary to determine if an individual has an acceptable level of hearing to conduct murrelet surveys. Marbled Murrelet vocalizations and sounds associated with flight range between 2,000 Hz and 5,000 Hz. In order for a person to have an acceptable hearing test, they should have good hearing at all frequencies, but especially in this range. The American Medical Association and OSHA define good hearing as 0 to 25 dB in both ears. This means that at all frequencies within the normal range of human hearing, an individualís hearing thresholds should be 25 dB or less. Individuals with good hearing, or corrected hearing that meets the definition of ëgoodí hearing, are qualified to conduct protocol murrelet forest surveys upon successful completion of an approved survey training program.
Marginal hearing is defined as 0 to 25 dB in one ear, and a level not to exceed 60 dB in the other ear. If an individual with marginal hearing can demonstrate proficiency in their ability to detect Marbled Murrelets during the survey evaluation, given their impaired hearing, they can conduct protocol murrelet forest surveys. This determination is made at the discretion of the training evaluator. Evaluators must be assured of the surveyorís ability to identify murrelet vocalizations at distances greater than 200 meters (600 ft), and their ability to discern correct detection and flight directions.
Poor hearing is defined as greater than 25 dB in both ears. Individuals with poor hearing, including those who meet the definition of poor hearing with corrective devices, are not qualified to conduct protocol murrelet forest surveys.
APPENDIX D
FOREST BIRD AND MAMMAL SPECIES POTENTIALLY MISIDENTIFIED AS MARBLED MURRELETS
AND
POTENTIAL MURRELET PREDATORS
The following species have been identified as sources of potential confusion if present during a Marbled Murrelet forest survey. They may be misidentified by sight, sound, or both, by an inexperienced observer. Observers should be able to identify the species on this list to ensure the accuracy of the survey data reported. Marbled Murrelet flight is characterized by rapid, constant wing beats. See Appendix E for a detailed description of murrelet auditory sounds. Species are identified by Common Name/A.O.U. code (birds).
Potentially Misidentified
Heard and Seen
Common Nighthawk (CONI)
Varied Thrush (VATH)
American Robin (AMRO)
European Starling (EUST)
Heard
Killdeer (KILL)
Bald Eagle (BAEA)
Red-shouldered Hawk (RSHA)
Red-tailed Hawk (RTHA)
Osprey (OSPR)
Northern Flicker (NOFL)
Red-breasted Sapsucker (RBSA)
Hairy Woodpecker (HAWO)
Olive-sided Flycatcher (OSFL)
Western Wood-Pewee (WWPE)
Stellerís Jay (STJA)
Gray Jay (GRJA)
Swainson's Thrush (SWTH)
Hermit Thrush (HETH)
Hutton's Vireo (HUVI)
Black-headed Grosbeak (BHGR)
Song Sparrow (SOSP)
Western Tanager (WETA)
Evening Grosbeak (EVGR)
Mammal: Douglas squirrel
Seen
Wood Duck (WODU)
Harlequin Duck (HADU)
Common Merganser (COME)
Spotted Sandpiper (SPSA)
Band-tailed Pigeon (BTPI)
Mourning Dove (MODO)
Black Swift (BLSW)
Vaux's Swift (VASW)
Tree Swallow (TRSW)
Violet-green Swallow (VGSW)
American Dipper (AMDI)
Mammal: Bat spp.
Potential Marbled Murrelet Predators
The following is a list of potential predators of adult Marbled Murrelets or their nests (eggs or young). The presence of these predators during the survey should be noted at the bottom of the last page on the Survey Activity Table form.Bald Eagle (BAEA)
Sharp-shinned Hawk (SSHA)
Cooper's Hawk (COHA)
Northern Goshawk (NOGO)
Red-shouldered Hawk (RSHA)
Peregrine Falcon (PEFA)
Great Horned Owl (GHOW)
Barred Owl (BAOW)
Northern Spotted Owl (SPOW)
Steller's Jay (STJA)
Gray Jay (GRJA)
American Crow (AMCR)
Common Raven (CORA)
Douglas squirrel
Deer mouse
Keenís mouse
Northern flying squirrel
Townsendís chipmunk
Bushy-tailed woodrat
APPENDIX E
MARBLED MURRELET VOCALIZATIONS
Reviewed by William Ritchie
Familiarity with murrelet vocalizations is essential for anyone planning to conduct a protocol survey. The majority of murrelet detections are auditory in nature (Hamer and Cummins 1990, Nelson 1990, Paton and Ralph 1988), especially at interior forest survey stations with limited visibility. Marbled Murrelet vocalization recordings are currently being collected and analyzed to characterize the different calls. Presently there are four recognized vocalization categories: (1) "Keer" calls, (2) Whistle calls, (3) Groan/grunt calls (formally known as alternate calls), and (4) Fledgling begging calls (Nelson and Peck 1995, Nelson 1997, Nelson et al., in prep.). These categories of vocalizations can include a variety of variable call combinations. To date there have been no identified sexual differences, call functions, or geographic variability in murrelet vocalizations. However, in time we may be able to associate vocalizations with behavior.
The most distinctive and commonly heard vocalization is the "Keer" call. The frequency range for this call is 2000 to 5000 Hz, with a mean frequency of about 3500 Hz. There are typically 2 or 3 elements to the "Keer" call, with the initial note of the call reaching a maximum at 5000 Hz (see sonagram in Nelson 1997). "Keer" calls are intermediate in length at about 300-350 milliseconds. This call can be described as a piercing, high pitched "gull-like" call that phonetically sounds like "Keer-Keer". Whistle calls generally consist of a short broadband initial segment followed by a narrow-band mid-frequency note of longer duration than the "Keer", and without the repeating series of calls. This type includes the whistle-like "Kee", single note calls similar to the initial segment of a "Keer" call, and the "soft-que" call, a long plaintive sounding ("eeeh-eeeh") whistle. Groan/grunt (alternate) calls can be heard frequently at inland sites, at sea, or while the adults are present at the nest during feeding visits. These vocalizations are similar to the raspy, nasal-sounding calls given by other alcids at breeding colonies. Many times a groan call is part of a "Keer" call sequence, or given in reply to another vocalizing murrelet. Adults bringing fish to the nestling often give a muted grunt call sounding like "rrUH-rrUH". The fledgling food begging call is a continuous series of soft, high-pitched "peep"s. This behavior has been observed to occur when an adult arrives at or near the nest to feed the chick. In most cases vocalizations at the nest are not audible from the ground.
There are two additional auditory detections that may be heard at inland sites. These are not vocalizations, but sounds produced by air passing over the feathers of a murrelet in flight. The jet sound can be heard when a murrelet is in a steep descent or when it is ascending following such a dive. This loud, slightly wavering, whooshing sound is a bit like a jet plane rapidly passing overhead. It is rarely heard and often occurs near or above nesting areas. The sound of the murrelet wing beat has a wide frequency range, resulting in a rapidly alternating sound. These sounds have been described as similar to that of a rope being twirled rapidly in the air or a hand saw blade being shaken (Nelson 1997). Though the detection of murrelet wing sounds is often associated with below canopy flight, it also originates from murrelets flying above the canopy. If wing-beat sounds are detected during a survey without any visual sighting, additional surveys would be necessary to determine if the site is occupied.
Once a surveyor learns the basic calls, they should develop their ability to identify similar-sounding passerine vocalizations. This will help identify murrelet calls at sites with background noise and differentiate distant murrelet calls from other similar-sounding calls.
References
Hamer, T. E. and E. B. Cummins. 1990. Forest habitat relationships of marbled murrelets in northwestern Washington. Unpublished report, Wildlife Management Division, Nongame Program, Washington Department of Wildlife, Olympia, WA. 57 pp.
Nelson, S. K. 1990. Distribution of the Marbled murrelet in western Oregon. Report to the Nongame Program, Oregon Department of Fish and Wildlife, P.O. Box 59, Portland, OR, Publ. No. 89-9-02.
Nelson, S. K. 1997. Marbled Murrelet (Brachyramphus marmoratus). In The Birds of North America, No. 276 (A. Poole and F. Gills, eds.). The Academy of Natural Sciences, Philadelphia, PA, and the American Ornithologistsí Union, Washington, D.C.
Nelson, S. K., and R. W. Peck. 1995. Behavior of marbled murrelets at nine nest sites in Oregon. Northwestern Naturalist 76: 43-53.
Paton, P. W. C., and C. J. Ralph. 1988. Geographic distribution of the marbled murrelet in California at inland sites during the 1988 breeding season. Unpublished report, California Department of Fish and Game, Sacramento. 35 pp.
APPENDIX F
DATA FORM AND INSTRUCTIONS FOR ITS COMPLETION
Cover Page (page 1)
Item #
1 Page Number of the total number of pages of data for the survey. This includes Cover Page, Survey Activity Table page(s), and Map page(s).
2 Survey Visit to Protocol: Circle Y (Yes) or N (No) to indicate if the survey was conducted following the guidelines of the Pacific Seabird Group protocol. Include the initials of the person who is making this statement, often the crew or project leader. To answer this question will involve a review of the survey visit by someone affiliated with the survey effort, who should check the survey form for compliance with the protocol, and possibly speak with the observer. The review is not to be done by the observer. An affirmative response does not necessarily imply that the entire survey effort was acceptable or that regulating or evaluating agencies will find the survey to be valid.
3 Total Detections: Total number of murrelet detections recorded during a survey visit. All detections should be assigned a detection number (Detect. #), including un-mappable detections. No other species observations should be included in this count.
4 Other Species of Concern: Circle Y (Yes) or N (No) to indicate if other species of concern were observed; refer to your state or provincial Fish and Wildlife agency Species of Concern list. Record details of observation(s) at the end of the last page of the Survey Activity Table.
5 Month, Day, Year: Date of survey visit. Use 2 digits for Month and Day, and four digits for Year (e.g., May 10, 2000 = 05/ 10/ 2000).
6 Area Name: Name of survey area being surveyed.
7 Site Name and Number: Site name and number from which survey visit is conducted. Each survey site should have a unique number or alphanumeric identifier.
8 Station Number: Station number from which survey visit is conducted. Each survey station should have a unique numeric identifier relative to a survey site.
9 Station Location: Location of station where survey visit was conducted. Several mapping coordinate systems are in use throughout the range of the murrelet. Township, Range, Section is used in WA and OR. UTM is available anywhere with GPS.
T,R,S - Record township, range, meridian [circle either E (east) or W (west)], section, sixteenth section ("Q, Q"), of quarter section ("Q"). Use 2 digits for each T,R,S value (e.g., T 09 N, R 06 W, S 10, NW QQ of SE Q).
UTM - Enter the UTM zone, and the Easting (x) and Northing (y) coordinates from a USGS, or equivalent (for B.C.), topographic map if the T, R, S system is not used in your area, or if a GPS is used. Indicate the source used to determine the station location (e.g., type of map or GPS). If a GPS is used, indicate whether the coordinates are differentially corrected or what the error value (FOM) was when the position was taken, and what map datum (e.g., NAD 27 CONUS, WGS 84, etc.) the unit was set for.
10 Observer Name(s): First name, middle initial, and last name of the observer(s).
11 Initials: Initials of observersí full name.
12 Affiliation: Agency, tribe, or company name.
13 Phone: Agency, tribe, or company telephone number including area code. This should be a contact who can be reached during and after the survey season in the event that questions arise regarding the survey data.
14 Station Elevation: Using a USGS 7.5 minute or 15 minute topographic map, or a properly calibrated altimeter or GPS, record the station elevation. An equivalent topographic map may be used for B.C. Indicate whether the value is in feet or meters.
15 Position on Slope: Select the code that best describes the station's position on slope. Codes: B = Canyon bottom or coastal plain, L = Lower 1/3, M = Middle 1/3, U = Upper 1/3, R = Ridgetop.
16 Station Placement: Circle whether survey station is located Inside or Outside the survey site. Stations on the survey site boundary are considered Inside.
17 Distance from Survey Site Boundary: This measurement applies only to Outside stations. Indicate distance from the survey station to the survey site boundary. Stations are generally located <50 meters (164 feet) from the edge of the survey site boundary.
18 Units of Measure for All Horizontal Distances: Indicate the units used for all horizontal distances reported on the survey form (e.g., meters, feet, yards, etc.). This will include distances to survey site boundary for stations outside of the site and for closest distances to birds. The units must be consistent throughout the survey visit.
19 Station Canopy Cover: Select the canopy cover class code that best describes overhead canopy cover at the survey station. Codes: 1 = 0 -25%, 2 = 26 -50%, 3 = 51 -75%, 4 = 76 -100%. This can be derived as an ocular estimate of the area immediately adjacent (approx. 25 m radius) to the survey station, or an actual measurement using a densiometer or other device. This data can be useful in determining the viewability from a station.
ENVIRONMENTAL CONDITIONS: Record conditions as observed at the survey station at the beginning and end of the survey visit; note other significant changes in conditions as they occur throughout the survey visit.
20 Sunrise Time: Official sunrise time derived from The Nautical Almanac tables based on the date of the survey visit and geographic area. Add 1 hour for daylight-saving time! Use 4-digit "24 Hour Time"( e.g., 5:18 A.M. = 0518, or 6:30 P.M. = 1830). Copies of these tables may be available for your specific geographic area from your regulatory agency or Marbled Murrelet Survey Training Instructor.
e.g., Geographic Area of Survey (WA) Sunrise/Sunset Table
Clallam, western Jefferson Co. Tatoosh Island, WA
Grays Harbor, Lewis, Thurston Co. Olympia, WA
Pacific, Wahkiakum Co. Astoria, OR
San Juan, Skagit, Whatcom Co. Vancouver, B.C.
21 Source or Table: Indicate the Sunrise/Sunset table or source reference used to determine the survey times. Enter appropriate code if applicable.
22 Begin Survey Time: Actual time survey visit is started using "24 Hour Time" described above. A morning visit should begin at least 45 minutes before official sunrise. If a survey visit actually begins later, also note number of minutes late (e.g., "5 min. late").
23 End Survey Time: Actual time survey visit is completed using "24 Hour Time" described above. A morning visit generally ends 75 minute after official sunrise; more time is added depending on whether murrelet detections occur at the end of a visit, and may be extended if overcast conditions with rain and fog are present at the end of the standard survey period.
24 Temperature at Sunrise: Record temperature at official sunrise time. Indicate whether Celsius (C) or Fahrenheit (F). Be sure the thermometer is placed above the ground when taking the temperature.
25 Temperature at End of Survey: Record temperature at the end of the survey visit. Indicate whether Celsius (C) or Fahrenheit (F). Be sure the thermometer is placed above the ground when taking the temperature.
26 Time: Record times in 4-digit "24 Hour Time". Enter time when survey visit began and ended and indicate "Begin Survey " and "End Survey " in the Notes column. Also enter the time when significant weather or environmental conditions occur that affect murrelet detectability from the station.
27 Vertical Viewing:
a Ceiling: This is the height of the primary cloud/fog layer relative to the canopy of the survey site as viewed from the station. Record the appropriate code: UL = Unlimited (clear); HI = > 2.0 canopy height; MID = > 1.25 to < 2.0 canopy height; LO = < 1.25 canopy height; U = Unknown; cannot see adequately to describe due to station placement.
b Cloud Cover: Select the class code that best describes the amount of overhead cloud cover visible from the station. This is an ocular estimate.
Codes: 0 = 0% (clear sky; no cloud cover); 1 = about 33% of sky covered; 2 = about 66% of sky covered; 3 = 100% of sky covered; U = Unknown; cannot see adequately to describe conditions due to station placement.
c Visibility to 2 Canopy: From the survey station, note whether vertical visibility is unimpaired to 2 canopy heights. Codes: Y = Yes; N = No; U = Unknown; cannot see adequately to describe conditions due to station placement.
28 Horizontal Visibility to 100 m: From the survey station, note whether horizontal visibility is unimpaired within 100 m (328 ft). Codes: Y =Yes; N = No; U = Unknown; cannot see adequately to describe conditions due to station placement.
29 Audibility to 200 m: From the survey station, note whether audibility is unimpaired within a 200 m (656 ft) radius. Codes: Y = Yes; N = No
30 Precipitation: Select the appropriate codes to indicate precipitation intensity at the survey site as observed from the station. List only one code per column. Use the following codes in each of the type columns:
Rain: N = None; L = Light (mist, drizzle, soft rain); M = Moderate (obscuring rain); H = Heavy (intense rain).
Fog: N = None; L = Light (translucent haze, thin fog); M = Moderate (obscuring fog); H = Heavy (dense fog).
Other: For other precipitation conditions use the following type and intensity codes: N = None; HL = Light Hail, HM = Obscuring hail, HH = Intense hail; SL = Snow flurry, SM = Obscuring snows, SH = Intense snow storms, Blizzard.
31 Wind: Record the wind speed based on the Beaufort Wind Scale. Observe the effects of wind conditions on trees and vegetation visible at ground level at the station and record the appropriate code (0= <1 mph, calm; 1= 1-3 mph, leaves barely move; 2= 4-7 mph, leaves rustle and small twigs move; 3= 8-12 mph, leaves and small twigs in constant motion; 4= 13-18 mph, small branches move; 5= 19-24 mph, large branches and small trees start to sway; 6= 25-31 mph, large branches in constant motion; 7= 32-38 mph, whole trees move; 8= 39-46 mph, twigs and small branches break).
32 Noise: Record the appropriate code(s) to indicate noise conditions that affect ability to hear clearly within a 200 m (656 ft) radius: N = None; A = Airplane; B = Bird song/calls; C = Creek or other water drainage; M = Machinery (logging, mining, road construction, etc.); P = Precipitation (rain/hail); T = Tree drip; V = Vehicle (trucks, cars, etc.); W = Wind; O = Other (explain in Notes column). List more than one if applicable.
33 Notes: Record "Begin Survey" and "End Survey" to correspond to appropriate times recorded. Note any other pertinent information that can help to better describe or explain the conditions during the survey visit.
SHADED AREA AT BOTTOM OF PAGE FOR STATE OR PROVINCIAL FISH AND WILDLIFE AGENCY USE ONLY
Survey Activity Data Page
Item #
1 Detections - Page Total: Enter the total number of murrelet detections; every detection should have a detection number (detect. #), including un-mappable detections. This is the total number of detections per single-sided page.
2 Page Number of the total number of pages.
3 Initials: Initials of observersí full name.
4 Month, Day, Year: Date of survey visit. Use 2 digits for Month, Day, and four digits for Year (e.g., May 10, 2000 = 05/ 10/ 2000).
5 Area Name: Name of survey area being surveyed.
6 Site Name and Number: Site name and number from which survey visit is conducted. Each survey site should have a unique number or alphanumeric identifier.
7 Station Number: Station number from which survey visit is conducted. Each survey station should have a unique numeric identifier relative to a survey site.
8 Data Reference Number: State or provincial Fish and Wildlife agency use only. Used for identifying and tracking individual survey visits.
9 Units of Measure: Indicate measurement used for Closest Distance to Bird. Circle either U.S. or Metric.
SURVEY ACTIVITY: Record details of murrelet detections in this table. A detection is defined as the visual or auditory observation of one or more murrelets acting together in a similar manner and initially occurring at the same time.
10 Status and I/O: State or provincial Fish and Wildlife agency use only. Used for detection status coding and identifying bird location relative to survey site boundary.
11 Detection #: Each separate murrelet detection is sequentially numbered one per line as it occurred throughout the survey visit. When mapping the detections, use the detection numbers to cross reference the corresponding line entry. Number only the prioritized subgroup if a group of birds split, because the whole occurrence is considered one detection. Line out the Detect. # column for all associated subgroups. See the Survey Activity section above.
12 Detection Time: Record the time in 4-digit "24 Hour Time" when a murrelet detection occurred. Be sure to record time when survey visit began and ended, and indicate "Begin Survey" and "Ended Survey" in the Notes column on the corresponding lines.
13 Initial Detection Direction: Record the direction where the murrelet is first detected relative to the observer. The direction is recorded at a minimum of 45 degree increments (e.g., N = North; SW = Southwest; E = East).
14 Type: Record the detection type using the following codes: H = Heard only (auditory sound(s) with no visual observation); S = Seen only (visual observation with no auditory sounds); B = Both Seen and Heard (visual observation with accompanying auditory sounds).
15 Auditory Information: There are numerous call types that have been assembled into call groups based on their sounds. It is felt that these call groups can be precisely identified by properly trained observers, while more specific call types are best left to sonographic analysis (see Nelson 1997 and Dechesne 1998). Review cassette tapes of Marbled Murrelet vocalizations and other auditory sounds to assist with identification. Tapes of other forest bird calls/songs that may have similar sounding notes should also be reviewed periodically.
Vocal Series (vocalizations): Record auditory sounds using the codes listed below. In general, the most common call groups are the Keer group (keers, keheers, and quacks) = "K"; the Groan group (longer, variable groans formerly known as alternate calls) = "G"; and the Whistle group (longer, variable whistle) = "O". Birds most often grade their calls between two of these groups within a series or bout of calling.
Identify the call group detected using the appropriate code or codes for the type of call heard at the start and end of the detection. The detection may consist of one call type, or a vocal series that grades between two groups. Should the calls grade between two groups, identify the start and end points of the gradient, e.g., "K-G". Record the number of calls heard from 1-5. When more than 5 calls are heard in the same detection, record "M" for multiple. Indicate Yes or No to record if overlapping calls (OL) are heard as part of the detection.
Other (non vocal sounds): In addition to the vocal sounds described above, there are two other auditory sounds attributed to marbled murrelets. These non vocal sounds are Wing sounds or wingbeats = "W"and Jet sounds associated with aerial or power dives = "J". Record all types heard for each detection.
16 # of Birds Seen: Enter the number of birds visually observed.
17 Behavior: Record the behavior type of the bird(s) according to the following codes:
C = Bird(s) seen circling over the forest at > 1.0 canopy height. This behavior includes flight paths that deviate from a straight line, such as full, quarter, and half circles, angular turns, etc.
B = Bird(s) seen circling at or below the forest canopy, i.e., < 1.0 canopy height. This behavior includes flight paths that deviate from a straight line, such as full, quarter and half circles, angular turns, etc.
F = Bird(s) seen flying in a straight flight path over the forest at > 1.0 canopy height.
T = Bird(s) seen flying through in a straight flight path at or below the forest canopy, i.e., < 1.0 canopy height.
L = Bird(s) seen landing in, perching, or departing from a tree. This is a rare event.
S = Bird(s) heard emitting > 3 calls from a fixed point in a tree within 100 m (328 ft) of observer. This is a very rare and unusual event.
U = Bird(s) behavior unknown, i.e., bird(s) seen but behavior not identified, or canopy height not quantified, or detection was heard only and was not stationary.
18 Initial Flight Direction: This is the direction that the murrelets are seen heading when initially detected, i.e., the direction the birds are traveling when first detected. This information allows for accurate mapping of visual detections, and compliments the Bird Depart Direction data. Enter direction in a minimum of 45 degree increments (e.g., N = North; SW = Southwest, etc.).
19 Bird Height: This is determined from visual observations only. Enter an estimate of bird height in decimal units based on bird location relative to the height of the forest canopy, i.e., the tallest trees observable from the survey station. The height of the tallest observable tree is equivalent to a unit of 1.0 canopy height. If a bird was seen flying halfway beneath the height of the tallest observable tree, the bird height is "0.5 canopy heights." A bird seen flying over the canopy at one quarter the height of the tallest tree observed is at "1.25 canopy heights."
20 Closest Distance to Bird(s) Seen: Record the closest horizontal distance from observer to the murrelet(s). A bird flying directly overhead is equivalent to a horizontal distance of zero. Distances are recorded only for visual detections. Most visual detections are within 100 meters (328 feet). Indicate units of measurement at top of the column.
21 Bird Depart Direction: The direction the murrelet was last detected heading, i.e., the direction the bird(s) was traveling when last detected. Enter direction in a minimum of 45 degree increments (e.g., N = North; SW = Southwest, etc.).
22 Final Detection Direction: The final direction the murrelet was detected relative to the observer. The direction is recorded at a minimum of 45 degree increments (e.g., S = South; NE = Northeast; W = West).
23 Notes: Additional information which can help to concisely describe and map a detection is entered here. For example: groups of birds that split or join other birds; unusual observed behavior; flight path directional information ("circled clockwise" or "counter clockwise").
MAPPING MURRELET DETECTIONS
To each survey form, attach a copy of a registered aerial photo, orthophoto (1:12,000), or a USGS or equivalent (for B.C.) topographic map showing the area and site surveyed. Be certain to indicate the corrected scale if the original scale was enlarged or reduced on a photocopier. Delineate the site boundary, and identify the observer station location using a circle with a dot in the center (u). If plotting detections on aerial or orthophoto maps, use a topographic map to aid in determining the correct location to plot the detections with respect to the terrain.
Plot the murrelet detections using the directional information, Behavior, and Closest Distance to Bird(s) data from the Survey Activity Table. Indicate the murrelet flight path and behavior (circling, straight flight path, stationary, etc.) relative to the station location using the symbols below. On 1:12,000 scale orthophoto maps, 1 mm = 12 m (39 ft); on 7.5 minute topographic maps the scale is 1:24,000, so 1 mm = 24 m (79 ft).
Audible detection: A dashed line with arrow head (ñ ñ ñ ñ ! ) indicates an audibly tracked flight path.
Visual detection: A solid line with arrow head (ññññ! ) indicates a visually observed flight path.
Stationary or Unknown Bird Depart/Final Direction: A triangle with a dot in the center ( ) indicates a stationary detection, or a visual or audible detection without a Bird Depart or Final Direction.
APPENDIX G
USE OF RADAR FOR MARBLED MURRELET SURVEYS
Compiled by Brian A. Cooper and Tom E. Hamer
Introduction
The current ground-based Inland Forest Survey Protocol for Marbled Murrelets depends on the use of audio-visual cues to detect birds in flight. Collecting biological information on murrelets this way is difficult, because of the low light conditions during their dawn and dusk peaks in inland activity and their small size, cryptic coloration, and rapid flight speed (Hamer et al. 1995). Further, because 85% of the murrelet detections are auditory (Paton et al. 1990), it is difficult to determine with accuracy the number of birds that actually are flying over a particular area. Ornithological radar, which does not have this auditory bias, has been used successfully to study Marbled Murrelets in both the Pacific Northwest and Alaska (Cooper 1993; Hamer et al. 1995; Burger 1997, 1999; Cooper and Blaha 1998, 1999; Cooper et al. 1997, 1998, 1999a, 1999b, 1999c; Hamer and Meekins 1998a, 1998b; Lougheed 1998, Singer and Hamer 1999a, 1999b; Manley and Conroy 2000). Radar techniques also have been used to study other avian species for nearly five decades (Eastwood 1967), and marine radar recently has been used to study other nocturnally-active seabirds (Day and Cooper 1995, Cooper and Day 1998, Bertram et al. 1999).
The way radar works is that pulses of electromagnetic radiation are sent out and then the radar is reversed for receiving the echoes that are reflected back from an object (e.g., a bird, plane, or ship). The radio waves travel at the speed of light and the distance to the object is related to the time lapse between transmission and reception of the echo. The distance that an object can be detected depends on many factors, including the area of the radar cross-section of the object, and the wavelength and power output of the radar. For birds, this distance may vary from a few hundred meters for the smallest marine radars to over 150 km in the case of long-range surveillance radars.
The intent of this appendix is to provide information on the uses and limitations of ornithological radar for Marbled Murrelet surveys. This document is not meant to be an exhaustive discussion or set of survey protocol guidelines, but rather a starting point to inform others of its potential uses. If one does wish to apply this technique, it first will be necessary to get approval of your study plans from the appropriate state and federal agencies.
Uses of Radar for Marbled Murrelet Surveys and Research
The major uses of radar for murrelet surveys and research include: (1) determining if murrelets are present in an area; (2) locating "hotspots" of activity over an area; (3) providing an index of abundance for a drainage; and (4) for population monitoring. Radar studies indicate that audio-visual observers detect an average of 7ñ15% of all Marbled Murrelets within 200 m during intensive murrelet surveys, although the percent detected varied widely among sites and among days within a station (Cooper and Blaha 1998, 1999). Further, approximately 11% of the murrelets that are detected on intensive surveys are birds passing over the stand of interest on their way to another area. Although radar will not work at all stands because certain terrain types preclude its use, results of Hamer et al. (1995) and Cooper and Blaha (1998, 1999) suggest that radar could be used as a 'coarse filter' to quickly and accurately determine whether murrelets are present near, or in the area adjacent to, a forest stand. Cooper and Blaha (1999) detected murrelets on 100% of the surveys with radar (Mean ± sd = 1.0 ± 0.0 days to determine presence, n = 14 sites). In contrast, it took from 1 to 5+ survey days to detect murrelets at the same sites with audio-visual techniques (5+ days means that presence was not detected within the 5 survey days at 2 of the 14 sites).
Because most birds during ground surveys are detected by auditory means, only limited information can be collected on bird flight behavior, flight direction, and flight path, and no information can be collected on relative abundance. Further, the distance of birds from the audio-visual observer is estimated. Radar can supply information on the murrelets' flight path and flight behavior, flight direction of targets to the nearest degree, number of targets, and the distance from the radar to the target to the nearest meter (Hamer et al. 1995). All of this information is critical in determining where birds are headed, which forest stands are likely being used, and the relative abundance of birds in the area. Thus, the quality and usefulness of the survey information collected by radar is much higher than data produced by the ground survey protocol. Radar also might improve survey efficiency because it reliably samples a much larger area (up to a 1400 m radius) than audio-visual observers (up to 200 m radius).
The available data suggest that radar has great potential for quickly determining presence and probable absence of murrelets in a suitable area, but we cannot yet recommend the number of visits or years that would be required to determine presence or probable absence with statistical certainty. If birds were found, however, audio-visual surveys still would be necessary to determine if the stand was 'occupied' by nesting murrelets. Because radar energy cannot penetrate forest vegetation, it generally cannot be used to determine whether a stand is occupied by murrelets (in some cases it may be possible, if circling behavior above the canopy is observed). Forest vegetation, hills, and ridges show up as ground clutter or solid echoes on the radar screen, preventing detection of birds in these areas. Given suitable survey locations, however, radar can reliably determine presence in a shorter period than the current audio-visual protocol (Cooper and Blaha 1999). Information from radar surveys also can be used to locate 'hotspots' over a larger stand in which to focus audio-visual observations. Further, the radar method could improve the accuracy of the protocol by detecting presence of birds at low-use sites where murrelets might be missed completely by audio-visual observers. Survey accuracy also can be improved because radar often can help determine whether murrelets that are flying over the survey area actually are in transit to another area.
In addition to surveying Marbled Murrelets at the stand level, ornithological radar can be used to obtain an index of abundance for Marbled Murrelets on a drainage scale (Cooper 1993; Hamer et al. 1995; Burger 1997; Cooper et al. 1998, 1999a, 1999b, 1999c; Hamer and Meekins 1998a; Singer and Hamer 1999a). Conservation biologists can use this type of landscape information to prioritize lands for potential habitat acquisition efforts. Careful selection of sampling locations in appropriate drainages and adequate sampling intensity during the breeding season is essential for this type of application, to ensure that a large proportion of the birds using a particular drainage are detected.
Because radar-based counts have low among-day variability, radar sampling also may be well suited for long-term population monitoring. Statistical power analyses suggest that radar-based sampling can produce results in a timely, cost-effective fashion (Cooper et al. 1999b, 1999c).
Limitations of Radar
The major limitations of the radar technique are: (1) it cannot determine occupancy (and sometimes presence) because birds flying near or within the canopy are shielded from the radar and missed; (2) it cannot be used at all sites because of topographic and physiographic constraints; (3) species identification errors are possible; and (4) X-band radar cannot be used during rain conditions (but can be used during drizzle or fog conditions). Fortunately, there are methods that will minimize the impact of many of these limitations. Perhaps the greatest limitation of the radar is that it cannot be used at all locations. For instance, radar cannot be used in areas closely surrounded by tall trees that block the radar beam. Use of a lift-equipped radar can help minimize that problem. With a 10.5-m lift-equipped radar, it was possible to use radar at 56% of 50 randomly chosen murrelet stands in the northwestern Olympic Peninsula of Washington (Cooper and Blaha 1999). Radar observations would only have been possible at 30% of these sites without the lift.
In most cases, radar needs road access to transport the system to the site, although radar systems have been transported by helicopter to some rugged coastal sites in British Columbia, where the radar has been placed on a tripod near the beach for monitoring at stream estuaries and canyon entrances (Burger 1997, 1999). In addition, a boat-based radar has been successfully used to observed murrelets from a boat at coastal areas in Alaska and British Columbia (Cooper 1993, Manley and Conroy 2000).
Some topographic situations are not conducive to use of radar. Whenever energy is reflected from the ground, surrounding vegetation, and other objects that surround the radar unit, a ground-clutter echo appears on the display screen. Because ground clutter can obscure bird targets, it should be minimized by elevating the forward edge of the antenna and/or by using a ground-clutter reduction screen (described in Cooper et al. 1991). The antennae of the radar can also be hinged so that it can be raised or lowered at will to reduce ground clutter (Singer and Hamer 1999a). Ground clutter also can be reduced by positioning radar in locations that are surrounded closely by trees or low hills. These objects act as radar fences that shield the radar from low-lying objects farther away from the lab. Using radar fences, only a small amount of ground clutter appears in the center of the display screen, creating ideal conditions for detecting avian targets. For further discussion of radar fences, see Eastwood (1967), Williams et al. (1972), and Skolnik (1980).
Radar works as line-of-sight, such that birds flying in 'radar shadows' (ground clutter) behind trees or hills will not be detected. The impact of 'shadow zones' can be reduced by selecting sites that minimize the size, location, and orientation of shadow zones.
Another limitation of radar is that one does not know exactly how many murrelets are associated with a particular radar target. One or more birds that are flying close together on the same flight path can appear as one echo on the radar monitor. However, one can apply a correction factor to the total number of targets by using the average flock size of targets observed visually.
It is possible that Marbled Murrelets observed entering one watershed could nest in an adjacent watershed (Rodway et al. 1993). For some types of studies, this bias would not be a concern, but for studies that require an index of abundance for a particular drainage, it may be necessary to monitor both drainages. To determine whether Marbled Murrelets were flying between drainages, it might be possible to conduct surveys on ridges or passes between drainages (Singer and Hamer 1999a).
Murrelets primarily are identified by their flight speed, which tends to be greater than most other species. There are individual sites, however, that have large numbers of problematic species, like Band-tailed Pigeons or waterfowl, that can fly at speeds above the 64 km/h (40 mi/hr) threshold that has been used to select for Marbled Murrelets. We stress that concurrent audio-visual observations (at radar lab) and radar observations be made, at least initially at each site, to assess the relative abundance of potentially confounding species and to help filter out non-murrelets from the radar database (Hamer et al. 1995, Cooper et al. 1999b, 1999c; Singer and Hamer 1999b). For radar studies with the objective of determining presence or probable absence, even one error in identification can be critical, so it may be beneficial to always make concurrent audio-visual and radar observations in those instances.
Data Collection
Radar observations should be made only by trained observers, skilled in use of radar, interpretation of radar signals, and in locating appropriate sampling sites. The location of appropriate sampling sites is the skill that requires the most expertise and is most important. If a poor sampling location is chosen, it will decrease the chances of detecting murrelets at that site. Each radar site should be analyzed for its ability to detect murrelets within the drainage. This can be accomplished by making a map of the radar screen with location of ground clutter, shadow zones, streams, and stand boundaries (if applicable). The amount and location of effective sampling area can then be quantified. Preparation of this map involves tracing the radar screen at a site and adding layers delineating ground clutter and shadow zones where low-flying birds would not be detected. The shadow zones are drawn based on a visual assessment of all clutter-free zones on the screen. Mapping exercises should be completed for each site so that data collected from these sites can be properly interpreted and assessed at a later date.
Recommendations
If radar-based sampling is to be used for survey, inventory, or monitoring purposes, we make the following recommendations, based on the results of several radar studies to date (Cooper 1993; Hamer et al. 1995; Burger 1997; Cooper and Blaha 1998, 1999; Cooper et al. 1998, 1999a, 1999b, 1999c; Singer and Hamer 1999a, 1999b):
For inventory and monitoring purposes, one also should:
For inventory purposes, sample at locations that funnel birds into a small, discrete area or plan on deploying more than one radar so that the entire width of a watershed is sampled. The use of radar is less restrictive for monitoring purposes than it is for inventory purposes, because population monitoring measures temporal trends of consistently-collected data and does not require that a large proportion of individuals in a drainage be counted. Thus, for monitoring, it is possible to use sites where there is low day-to-day variation in counts, but where one does not sample most of the birds in that particular drainage.
Radar Equipment
All of the radar surveys of Marbled Murrelets to date have used an X-band marine radar system. We recommend using a 10-kW radar system with a magnetron in good working order. Over time, the magnetron wears out, which makes the unit less sensitive and thus less useful for detecting murrelets. Full descriptions of mobile radar systems can be found in Gauthreaux (1985a, 1985b) and Cooper et al. (1991).
Radar References
Burger, A. E. 1997. Behavior and numbers of Marbled Murrelets measured with radar. Journal of Field Ornithology 68: 208-223.
Burger, A. E. 1999. Radar inventory of Marbled Murrelets in Clayoquot Sound, 1996ñ1998. Unpubl. report prepared for Forest Renewal BC and Ministry of Environment, Lands, and Parks, Nanaimo, BC, by Alan Burger Consulting, Victoria, BC. 44 pp.
Bertram, D. F., L. Cowen, and A. E. Burger. 1999. Use of radar for monitoring colonial burrow-nesting seabirds. J. Field Ornithology 70: 145ñ157.
Cooper, B. A. 1993. Feasibility of using boat-based marine radar to study Marbled Murrelets in Kenai Fjords National Park, Alaska, 1993. Unpubl. report prepared for U.S. Fish and Wildlife Service, Anchorage, AK, by Alaska Biological Research, Inc., Fairbanks, AK. 19 pp.
Cooper, B. A., and R. J. Blaha. 1998. Evaluation of the Marbled Murrelet Inland Forest Survey Protocol with radar. Unpubl. report prepared for Olympic Natural Resources Center, Forks, WA, by ABR, Inc., Forest Grove, OR. 54 pp.
Cooper, B. A., and R. J. Blaha. 1999. Development of radar as a tool to increase the accuracy and efficiency of inland surveys for marbled murrelets. Unpubl. report prepared for Olympic Natural Resources Center, Forks, WA, by ABR, Inc., Forest Grove, OR. 48 pp.
Cooper, B. A., and R. H. Day. 1998. Summer behavior and mortality of Dark-rumped Petrels and Newell's Shearwaters at power lines on Kauai. Colonial Waterbirds 21: 11-19.
Cooper, B. A., and R. J. Ritchie. 1995. Visual and radar studies of the altitude of bird migration in east-central Alaska. Journal of Field Ornithology 66: 590-606.
Cooper, B. A., R. H. Day, R. J. Ritchie, and C. L. Cranor. 1991. An improved marine radar system for studies of bird migration. Journal of Field Ornithology 62: 367-377.
Cooper, B. A., P. Henson, and M. G. Raphael. 1997. Use of radar for the study of rare birds. Endangered Species Bulletin XXII (6): 8-9.
Cooper, B. A., P. Henson, and G. Miller. 1998. Use of ornithological radar for population monitoring of Marbled Murrelets in Oregon. Unpubl. report prepared for U.S. Fish and Wildlife Service, Portland, OR, by ABR, Inc., Forest Grove, OR. 44 pp.
Cooper, B. A., R. J. Blaha, H. Stabins, and D. R. Herter. 1999a. Radar surveys of marbled murrelets in the upper Green River Drainage, Washington, 1999. Unpubl. report for Plum Creek Timber Company, Seattle, WA, by ABR, Inc., Forest Grove, OR. 25 pp.
Cooper, B. A., M. G. Raphael, and D. M. Evans. 1999b. Radar studies of Marbled Murrelets on the Olympic Peninsula, Washington, 1996-1999. Unpubl. report prepared for USDA Forest Service, Olympia, WA, by ABR, Inc., Forest Grove, OR. 62 pp.
Cooper, B. A., C. Strong, and N. Bentivoglio. 1999c. Radar-based monitoring of Marbled Murrelets in Oregon. Unpubl. report prepared for U.S. Fish and Wildlife Service, Portland, OR, by ABR, Inc., Forest Grove, OR. 47 pp.
Day, R. H., and B. A. Cooper. 1995. Patterns of movement of Dark-rumped Petrels and Newell's Shearwaters on Kauai. Condor 97: 1011-1027.
Eastwood, E. 1967. Radar ornithology. Methuen and Co., Ltd., London, United Kingdom. 278 pp.
Gauthreaux, S. A., Jr. 1985a. Radar, electro-optical, and visual methods of studying bird flight near transmission lines. Unpubl. Final Report prepared for Electric Power Research Institute, Palo Alto, CA, by Clemson University, Clemson, SC. 76 pp.
Gauthreaux, S. A., Jr. 1985b. An avian mobile research laboratory: hawk migration studies. Pp. 339-346 in M. Harwood, ed. Proceedings of Hawk Migration Conference IV. Hawk Migration Association of North America, Washington, CT.
Hamer, T. E., and D. J. Meekins. 1998a. Use of radar to determine the presence or absence of Marbled Murrelets in the Green River Watershed, Washington. Unpubl. report prepared for Plum Creek, Inc., Seattle, WA, by Hamer Environmental L.P., Mount Vernon, WA. 14 pp.
Hamer, T. E., and D. J. Meekins. 1998b. Use of radar to monitor Marbled Murrelets at inland sites in the North Cascades of Washington. Unpubl. report prepared for U.S. Forest Service, Mount Lake Terrace, WA, by Hamer Environmental L.P., Mount Vernon, WA. 16 pp.
Hamer, T. E., B. A. Cooper, and C. J. Ralph. 1995. Use of radar to study the movements of Marbled Murrelets at inland sites. Northwestern Naturalist 76: 73-78.
Lougheed, L. W. 1998. Monitoring marbled murrelets with radar in Desolation Sound, British Columbia. Abstract. Pacific Seabirds 25(1): 35.
Manley, I., and C. Conroy. 2000. Radar surveys of marbled murrelets on the northwest coast of Vancouver Island. Poster presented at the 27th Annual Meeting of the Pacific Seabird Group, 23-26 February 200, Napa, California.
Paton, P. W. C., C. J. Ralph, H. R. Carter, and S. K. Nelson. 1990. Surveying marbled murrelets at inland forested sites: a guide. USDA For. Serv. Gen. Tech. Rep. PSW-120. 9 pp.
Rodway, M. S., H. M. Regehr, and J. L Savard. 1993. Activity levels of marbled murrelets in different inland habitats in the Queen Charlotte Islands, British Columbia. Can. J. Zool. 71:977-984.
Singer, S. W., and T. E. Hamer. 1999a. Use of radar to monitor marbled murrelets in the Santa Cruz mountains, California. Abstract. Pacific Seabirds 26(1): 45.
Singer, S. W., and T. E. Hamer. 1999b. Gazos Creek Marbled Murrelet monitoring programñ1999 annual report. Unpubl. report for the Sempevirens Fund and the Apex Houston Trustee Council by Steven W. Singer Environmental and Ecological Services, Santa Cruz, CA.
Skolnik, M. I. 1980. Introduction to radar systems. McGraw-Hill, New York, NY. 581 pp.
Williams, T. C., J. Settel, P. O'Mahoney, and J. M. Williams. 1972. An ornithological radar. American Birds 26: 555-557.