Streaked Horned Lark Conspecific Attraction Feasibility Study

Final Report December 2013

Submitted to: US Fish and Wildlife Service Port of Portland Joint Base Lewis-McChord The Nature Conservancy Metro

Prepared by: Hannah E. Anderson, Adrian Wolf & R. Adam Martin Center for Natural Lands Management

Lark Conspecific Attraction Feasibility Study Ð Final Report 1 Center for Natural Lands Management, [email protected]

Table of Contents

Executive Summary ...... 3 Introduction ...... 4 Background ...... 4 Need ...... 5 Approach ...... 6 Goal and Objectives ...... 8 Methods ...... 9 Site Selection ...... 9 Addressing Threats ...... 10 Habitat Management ...... 10 Conspecific Attraction ...... 12 Results ...... 15 Predator Load ...... 15 Habitat Response ...... 16 Lark Response ...... 17 Discussion ...... 22 Conclusion ...... 24 Acknowledgements ...... 25 Literature Cited ...... 26

Lark Conspecific Attraction Feasibility Study Ð Final Report 2 Center for Natural Lands Management, [email protected]

Executive Summary

The Streaked Horned Lark (Eremophila alpestris strigata, SHLA) was recently listed as Threatened across their range under the federal Endangered Species Act. Population and range reduction, habitat loss, and the loss of ecological processes that created the early successional habitats that larks prefer, have all contributed to the species’ decline. Since most SHLA populations don’t reside on protected lands, techniques that could promote or attract birds to protected habitats could be a vital tool in increasing demographic rates and assisting in species recovery. This feasibility study assessed the utility of using conspecific cues (audio playbacks of songs and calls, as well as three-dimensional decoys) in association with the creation of desired habitat to lure birds from established breeding areas to these newly created habitats. This study measured success at three levels:

1. A male SHLA occurs at a relocation site during any portion of the relocation attempt 2. Both sexes occupy relocation sites during any portion of the attempt 3. Evidence of breeding is documented at the relocation sites

After two years of observations we had 23 detections of SHLA, 20 of which occurred in playback (treatment) plots, suggesting birds respond to conspecific cues. We met success criteria 1 and 2, but larks did not establish breeding territories in treatment plots. Because larks have extremely high site fidelity and suitable habitat may not be limiting, conspecific attraction alone may not be sufficient to attract larks to a new site. However, coupled with careful threat reduction at protected sites and deterrents at occupied sites, conspecific attraction could be a tool to be used with caution in the conservation of SHLA.

Lark Conspecific Attraction Feasibility Study Ð Final Report 3 Center for Natural Lands Management, [email protected]

Introduction Background The conservation of small populations of declining bird species is a particularly difficult challenge to land managers and conservationists (e.g., Sutherland et al. 2005). Declines of bird populations can be multi-faceted and interacting. Factors include habitat-loss, shifts in habitat composition, genetic factors, or behavioral maladaption due to species or populations falling into evolutionary or ecological traps (Schlaepfer et al. 2002).

The Streaked Horned Lark (Eremophila alpestris strigata, SHLA or lark), a genetically distinct subspecies of Horned Lark (Drovetski et al. 2005, Stinson 2005) that historically ranged from the Rogue and Willamette Valleys of Oregon, north to the Georgia Basin of British Columbia and coastal beaches in Washington. Altman (2011) extensively summarized the historic distribution of this subspecies, which was scarce in the northern most part of their range in the Northern Puget Trough and Georgia Basin, and once common and abundant across the South Puget Sound (SPS) and Willamette Valley (WV). However, the prairie lands of both SPS and the WV have been extensively fragmented and degraded (Dunwiddie & Bakker 2011 and references therein), which has contributed to the current range contraction and population collapse of SHLA. Currently, SHLA are only known from scattered populations in the WV, primarily on agricultural sites, dredged material islands of the Columbia River, coastal beaches of southern Washington, SPS prairies, and airports throughout their range. The species has been extirpated from their northern and southernmost ranges. In 2013, the United States Fish and Wildlife Service (USFWS) listed the Streaked Horned Lark as Threatened under the Federal Endangered Species Act (50 CFR part 17).

In addition to habitat loss, the ecological processes such as fire and flood that maintained the early successional habitat conditions that larks prefer have been altered or removed. Only in very few cases do larks occur in natural habitats that are maintained by natural processes. Because nearly all of these historic processes have been halted across the range of this species, SHLA may now associate with the closest surrogate of these historic disturbance regimes. Fire has only been reintroduced in the past 10-15 years, and where it has occurred continuously on military lands in the South Sound, it has been due to uncontrolled wildfires set by exploding ordinance. The Columbia River’s flow is regulated by dams, and accretion of new land on the Columbia River islands or beaches through natural flooding has nearly halted as a result. The modern day surrogates for these landscape processes and the broader landscape features SHLA likely preferred historically are 1) agriculture in the WV, where larks use the underproductive areas of fields or field margins, 2) the placement of dredged material on islands of the Columbia river, and 3) airfields in the SPS and WV, where the act of mowing and spraying, to maintain short statured ground, along with the large flat expanses of field and asphalt that is necessary for aircraft. The vast majority of known occupied sites are primarily used for human purposes (e.g., agriculture, airports, dredged material placement, military training).

Lark Conspecific Attraction Feasibility Study Ð Final Report 4 Center for Natural Lands Management, [email protected]

Need The Streaked Horned Lark population is quite small, with less than 2000 birds estimated (Altman 2011). All known breeding sites in Washington have been monitored consistently for the past several years, and preliminary results indicate a declining population (S. Pearson, pers. comm.). Explanations for decline include low adult and juvenile survival and low fecundity (Camfield et al. 2011, Camfield et al. 2010, Pearson and Hopey 2005). Explanations for the reduced potential reproductive capacity of individual larks include high predation rates (Pearson and Hopey 2004) and low hatchability (Anderson 2010), the latter that is likely driven by inbreeding depression (Drovetski et al. 2005). Population modeling of the Streaked Horned Lark suggests that improvements in adult survival, rather than increasing fecundity or juvenile survival, may have the greatest potential to influence population growth rates (Camfield et al. 2008). Larks are also vulnerable to human-caused disturbance. Although the extent that human-caused disturbances has not been specifically studied, these disturbances likely reduce the potential reproductive capacity of individual larks (lower fecundity). Human-caused disturbance may have effects at every lark life stage, including the wintering grounds, breeding grounds (e.g., nest destruction, nest abandonment and may also alter individual breeding behavior). Larks have high potential to abandon nest sites (Pearson and Hopey 2005) and human-caused disturbance has been linked to nest abandonment in many bird species (Gutzwiller et al. 1998).

The majority of occupied Streaked Horned Lark breeding sites throughout the range have multiple and often conflicting management agendas. Multiple breeding sites are on municipal and international airports (e.g., Olympia, Shelton, Portland International, Corvallis); four sites are on Joint Base Lewis-McChord (Gray Army Airfield, McChord Airfield, Training Area 14, and Range 76), which are subject to military training activities; many are on Columbia River islands subject to dredge material deposition and which receive an unknown level of recreational use (pers. obs.), an unknown number of breeding sites occur on active agricultural lands in the Willamette Valley, and one site is in north Portland (Rivergate) which is slated for development.

While the ultimate cause of decline of the Streaked Horned Lark is not fully understood, direct impacts and disturbance on occupied sites from military training, aircraft, air shows, mowing, dredging, ATV use and development most likely affect the reproductive success and survival of breeding birds (e.g., Pearson and Hopey 2005). Abatement of direct threats to individual birds is a great benefit to this declining species and has been identified as a priority action in a number of conservation plans (e.g., Stinson 2005, JBLM Endangered Species Management Plan 2013). However, threat abatement alone at occupied sites might not be sufficient to stabilize or reverse the decline, because many activities that impact larks at the occupied sites cannot be fully curtailed. For example vegetation management (e.g., mowing) at airports is an aircraft safety requirement. Although the seasonal timing of the mowing may be somewhat altered to reduce impacts to known nests, not all nests locations are known and flightless fledglings are extremely vulnerable. These high human- wildlife conflict areas could thus be functioning as ecological traps. Ecological trap theory suggests that the presence of a trap will drive a local population to extinction, because the animals prefer low quality habitats for reproduction and survival that cannot sustain the

Lark Conspecific Attraction Feasibility Study Ð Final Report 5 Center for Natural Lands Management, [email protected] population (Battin 2004). A behavior that was once positively adaptive, such as selecting high quality habitat, becomes decoupled from these positive fitness effects, and leads to negative fitness consequences (Schlaepfer et al. 2002).

Approach Because larks occur on human-dominated landscapes that may negatively affect them, a conservation tool to aid in their redistribution to areas managed for their conservation may be beneficial. The Streaked Horned Lark prefers early successional habitats, which necessitates adaptive behaviors to prospect and breed at new sites as old sites become heavily vegetated and, therefore, unsuitable. This prospecting and establishing behavior is exhibited in the dynamic landscapes of the Willamette Valley, Columbia River islands, and Washington Coast where larks are known to move and establish territories at unoccupied, yet suitable, sites as they become available and as existing occupied sites become unsuitable (R. Moore, pers. comm., S. Pearson, pers. comm.).

Traditionally, managers and ecologists have thought habitat quality was the main explanation for species occupancy (Hilderbrand et al. 2005). However, more recent research has demonstrated social information and behavioral cues can be important as or more important than environmental cues in habitat selection (Ahlering et al. 2010). Conspecific attraction involves the process through which a given individual uses social cues from either conspecifics, or even heterospecifics (Thomson et al. 2003), to assess the suitability and quality of a habitat. Visual cues include the presence of an individual or groups of individuals, and auditory cues include songs and contact calls.

A wide range of bird species have been documented responding to conspecific cues across disparate phylogenetic groups such as the Atlantic Puffin, Terns, & Storm Petrels (Kress 1997), Common Murre (Parker et al. 2007), Black-capped Vireo (Ward & Schlossberg 2004), Baird’s Sparrow (Ahlering et al. 2006), Black-throated Blue Warbler (Hahn & Silverman 2007), and Bobolink (Nocera et al. 2006), thus demonstrating the utility and potential general applicability of using conspecific cues for avian conservation.

Several aspects of SHLA life history could make response to conspecific cues likely. Ahlering et al. (2010) list three broad categories of life history traits signifying the potential importance of conspecific cues. These include:

1. Clustered spacing of territories either due to coloniality, the aggregation of territories, or patchy distribution. 2. Behavioral traits including extra-pair mating behaviors, large juvenile: adult ratios, low survival, and migratory behavior. 3. A short breeding season or asynchronous breeding.

SHLA meet all three of these suggested requirements. While SHLA are not colonial, they do form aggregations of ‘all-purpose territories’ encompassing foraging habitat and nesting habitat. Thus, even though male SHLA defend their territories from other males, these territories are clustered across the landscape, a curious aspect of their biology first noted Lark Conspecific Attraction Feasibility Study Ð Final Report 6 Center for Natural Lands Management, [email protected] by Bowles (1898). Anderson (2005) found SHLA to have aggregated nesting densities and this may be due the particularly high level of avian predation occurring; though clustering could have either deleterious or advantageous impacts on nesting success (Danchin & Wagner 1997). Research on why birds nest in colonies or aggregations has been varied and fruitful. Hoi & Hoi-Leitner (1997) suggested nesting colonies were initiated by female Bearded-tits, because colonial nesting increased the likelihood of extra-pair copulations, which could increase the likelihood that high-quality females would mate with higher- quality males outside of their original pair-bond, thus increasing offspring fitness. A similar pattern was also found in Purple Martin (Morton et al. 1990). This phenomenon is known as the ‘hidden lek’, and has become one of the most convincing hypotheses addressing the peculiar nature of territorial birds forming aggregations (Fletcher & Miller 2006).

It is believed SHLA form breeding pairs while on the winter grounds (Jewett et al 1953), suggesting some level of monogamy during the breeding season. While extra-pair copulation hasn’t been noted in the SHLA literature, it is suspected to occur. Extra-pair solicitations have been observed multiple times across different populations in the South Sound (A. Wolf pers. comm., R.A. Martin pers. comm.). Martin (personal observation) has also observed multiple male on male visual displays occurring in view of females, with associated male-female chasing post display during the breeding season. These anecdotal observations suggest some level of sexual competition between males during the breeding season for extra-pair copulations. Forth-coming genetic work can address the likelihood of this hypothesis.

The breeding season for SHLA is relatively short, with nest initiations starting in early May and ending in mid to late July. Species that breed asynchronously or sequentially may be more likely to use conspecific cues since more experienced birds would arrive at high quality breeding sites first, followed by less experienced breeders. These less experienced breeders may use the location of early arrivers as cues for quality habitat (Ward & Schlossberg 2004). SHLA demonstrate particularly high site fidelity, with individuals returning to the same breeding territory each year (Anderson, Wolf, Martin pers. obs.), and site-fidelity has been linked to breeding success in multiple studies (see references in Ahlering et al 2010). Further, while infrequent (5 of 16 banded birds) it has been only first year SHLA that have been documented moving to other sites than their natal territory (Pearson et al 2005). Thus, it is likely these first-year birds would use conspecific cues to make habitat choices their first breeding year.

Lastly, while unpredictable habitat conditions may lessen the importance of conspecific cues, if it varies predictably conspecific cues could note adaptive behavior (Doligez et al 2003). Though SHLA is early successional, and historically their habitat patches likely varied spatially year to year following natural disturbance regimes (e.g. post-fire, post- flood) the expected post-disturbance habitat patches likely occurred within predictable landscapes (e.g., Pacific Northwest grasslands, Columbia River beaches and islands). SHLA habitat is likely predictable enough to suggest that conspecific cuing may be an adaptive trait.

Lark Conspecific Attraction Feasibility Study Ð Final Report 7 Center for Natural Lands Management, [email protected]

Goal and Objectives We aim to test the feasibility of using artificial conspecific attraction to lure streaked horned larks from occupied sites that are either threatened with impending development or threatened with disturbance or mortality to nearby sites. The goal of this feasibility experiment is to determine if streaked horned larks respond to conspecific cues by occupying treatment sites.

Our study, while assessing the feasibility of conspecific attraction, is novel in the fact we are not only trying to lure SHLA to unoccupied habitat but are trying to passively move individuals of a specific existing population from one location to another. We are attempting to do this by: 1. Creating suitable SHLA habitat in unoccupied, yet nearby locations from source sites 2. Using site-specific playbacks of conspecific and heterospecific calls and songs, as well as decoys to lure SHLA to these newly created habitats. We are using three criteria to measure success: 1. Male SHLA occur at a relocation site during any portion of the observed study. 2. Both sexes occupy relocation sites during any portion of the observed study 3. Evidence of breeding is documented at the relocation site.

If successful, it may suggest conspecific cues could be used to lure SHLA away from detrimental habitats to those actively created and managed for their conservation. Since human-caused declines in survival rates are potentially the easiest factor to address in SHLA declines, their avoidance and cessation could quickly increase vital demographic rates and hasten the recovery of the species.

Should conspecific attraction prove to be a successful method for attracting individual streaked horned larks to unoccupied sites, conservationists will have the tool available to apply at a larger landscape. This relocation feasibility study is ultimately aimed at benefitting larks by defining a method to ensure larks occupy protected or managed sites, thus reducing proximate causes of decline (such as human-related disturbance) and increasing adult survival.

Lark Conspecific Attraction Feasibility Study Ð Final Report 8 Center for Natural Lands Management, [email protected]

Figure 1. Relocation sites with treatment and Methods control plots. From top to bottom –St. Johns, TA6, TA14. Site Selection

We originally identified four source sites, each with a corresponding pair of 5-acre treatment and control plots at a nearby relocation site. There were three sets on Joint Base Lewis-McChord (JBLM) and one in North Portland: Range 74/76 (R76) – Training Area 6 (TA6); Gray Army Airfield (GAAF) – North Approach; 13th Division Prairie (13th Div) – Training Area 14 (TA14); and Rivergate (RVG) – St. John’s Landfill (STJ). Unfortunately we had to discontinue use of the GAAF set due to accessibility and their potential negative impact to larks. Airfield approach zones are areas where the height of aircraft potentially aligns with the height of flying birds, particularly male larks that may be conducting an aerial display. With the removal of the GAAF plots, three sets of controls and treatments remain, the minimum amount required to be able to use quantitative statistics (Figure 1).

Each source site is a known occupied breeding site for streaked horned larks. Source and relocation sites were selected on a combination of factors including their existing level of threat to the breeding population at the source, their adjacency to suitable relocation sites, and accessibility.

Suitable relocation sites were selected based on their landscape context, their adjacency to source sites, minimal or similar threat to breeding larks than at source sites, minimal threat to other conservation targets (i.e. native prairie), and logistical accessibility. To ensure appropriate landscape context, we situated relocation plots in an open landscape mostly free of trees and shrubs and at least 225 meters from habitat edge, such as forest. 225m is the minimum distance to habitat edge at source sites. All sites were situated within 2.5km of a source site (RVG:STJ=2.2km; R76:TA6=2km; 13th Div:TA14=1.7km). Previous work on the Columbia River suggests that

Lark Conspecific Attraction Feasibility Study Ð Final Report 9 Center for Natural Lands Management, [email protected] adjacency to occupied sites may play an important role in establishment of new sites (Anderson 2011).

While the objective of this study is not necessarily to conduct relocation but only to test its feasibility, the result could be establishment of a population at a new or historic site. Sites were selected with caution so that if a new breeding site was established, we would not be luring larks into a known sink.

Addressing Threats There are threats to breeding larks at all source sites. All JBLM sites receive military training: R76 is in the Artillery Impact Area and receives an unknown level of disturbance including off-road vehicle traffic and unchecked wildfire. 13th Div receives a large amount of military training including helicopter maneuvers, parachute drops, and on-the-ground infantry training. Finally, the RVG site, while relatively free of human disturbance (some dog-walking occurs), is slated for development in the near future - recent warehouse development at the RVG site has reduced available SHLA habitat.

Minimizing threats to breeding larks at relocation sites required site-by-site assessment. The St. Johns Landfill is closed to the public and identifies Streaked Horned Larks as a conservation priority. At the landfill, we were most concerned with predator load and added additional bird deterrent spikes on all nearby wellheads in an attempt to reduce the potential impact of predators at the site. Both TA6 and TA14 will still receive military training, but it is not expected to be an increase over the source site. The added benefit is that neither training area is inside the Artillery Impact Area, and as such will not receive live fire munitions that could start an unchecked wildfire.

Habitat Management We used a variety of management techniques to create SHLA habitat. Management actions included the application of broad-spectrum herbicide, disking, raking, harrowing, and prescribed fire (Table 1). Because each site is unique, it was not possible to standardize the habitat management approach for all plots. For instance, St. Johns is a covered landfill emitting methane gas, therefore prescribed fire is not possible. The JBLM sites all occur on glacial outwash soils and disking this soil type is extremely difficult due to the large amount of rock present. As a result, not all techniques were implemented in all plots. In general, SHLA is associated with sparse, short vegetation with a high percentage of bare ground. The percent cover of bare ground was reported as the strongest predictor of lark use (Pearson & Hopey 2005). JBLM plots target South Puget Sound SHLA habitat characteristics and St. Johns plots target Columbia River SHLA habitat characteristics (Pearson & Hopey 2005, Anderson 2011).

Lark Conspecific Attraction Feasibility Study Ð Final Report 10 Center for Natural Lands Management, [email protected]

Table 1. 2012-2013 Habitat Management Treatment Timeline Site St. Johns Training Area 6 Training Area 14 Habitat Jan - Apr - July - Oct - Jan - Apr - July - Oct - Jan - Apr - July - Oct - Treatment Mar June Sept Dec Mar June Sept Dec Mar June Sept Dec Spray X X X X X Mow X X Disk X X Harrow X X Rake X X X Burn X1 Monitor X X X 1 Due to safety concerns of unexploded ordnance, burning was conducted in TA6-T only.

The variability in management technique is not of concern as streaked horned lark occurs on a variety of habitat types maintained by several different mechanisms throughout its distribution. In south Puget Sound, the lark occurs on remnant glacial outwash prairie habitat, historically maintained by fire. Along the Columbia River, they occur on sandy islands historically maintained by flooding, currently maintained by the placement of dredged material. In the Willamette Valley they occur in sparse agricultural fields. The only two natural processes still in place that maintain lark habitat are wetland drawdowns in the Willamette Valley and beach accretion along the Washington coast. All these site types contain similar structural characteristics of the vegetation and the landscape context is similar. Due to these dissimilarities in the mechanisms that create lark habitats within its natural distribution, we are not concerned that the management techniques are dissimilar. The important piece is that suitable lark habitat structure is achieved.

We conducted qualitative and quantitative assessments of lark habitat quality, and implemented management to stunt vegetation growth, as necessary throughout the year (Figures 2 & 3). Qualitative assessments guided our habitat management activities.

Figure 2. Results of broad-spectrum herbicide Figure 3. Prescribed fire implemented at TA6-T. application. Picture taken April 2012. TA14-C. 2 August 2012.

To quantitatively assess habitat conditions we followed methods described in Anderson (2011), which were modeled on those used by Pearson & Hopey (2005) to allow comparisons between data sets. Data were collected using the point-intercept method

Lark Conspecific Attraction Feasibility Study Ð Final Report 11 Center for Natural Lands Management, [email protected] along four 25meter transects. There were two sets of two transects, each set bisecting at the middle, situated at the left and right center of each plot (Figures 4 & 5).

We used a narrow diameter wooden dowel to sample Figure 4. Vegetation monitoring at each meter along each 25m transect (Figure 6). arrangement. Each transect is 25m Each time a plant or ground cover class touched the situated within 5-acre habitat plots. dowel, it was recorded as a “hit” along the transect. Percent cover of each variable was calculated by dividing the total number of hits by the total number of sampling points along the transects (104). Because structure, and not species composition, is the important driver in predicting lark use we collected information on functional groups, rather than plant species. We collected data on the following habitat components: vegetation (grasses, forbs, shrubs); ground substrate (bare ground, moss, thatch); and vegetation height. We collapsed all vegetation into one functional group category and present it together.

Figure 5. Vegetation monitoring transects at STJ- Figure 6. Vegetation monitoring dowel. STJ-T. T. March 2012 March 2012.

These data were then compared to reported values using a Sorensen (Bray & Curtis) dissimilarity measure. This is a multivariate value, which is the distance between two points in multivariate space, in this instance the ‘ecological similarity’ between habitat values in study plots compared to values reported in Pearson & Hopey (2005). These values are presented as a % similarity, thus smaller values are less similar to those reported in Pearson & Hopey, while larger values are more similar (table 3). Conspecific Attraction In 2011, we engaged a biologist with expertise in avian bioacoustics, Dr. Alison Styring (The Evergreen State College) to record frequency and timing of streaked horned larks at two of our four source sites – Rivergate in north Portland and 13th Division Prairie at JBLM. We used the information and source-specific recordings to mimic occupied sites at the relocation sites (i.e. Rivergate recordings at St. Johns and 13th Div. recordings at JBLM) (Styring 2011). Lark Conspecific Attraction Feasibility Study Ð Final Report 12 Center for Natural Lands Management, [email protected]

During the breeding season (defined as April-August for this study), our playback sequence was a 60 minute loop including 40 minutes of site-specific lark songs and calls, followed by 10 minutes of sympatric species vocalizations (e.g., savannah sparrow, western meadowlark, etc.), followed by 10 minutes of silence. The literature indicated that inclusion of sympatric species and silence was important to ensure the target species does not habituate to the playback. The sequence played from 5:30-9:30am and from 2:00pm to 6:00pm. During the fall migration period (Sept – Oct) in both years, and the spring migration period in 2013 (Feb-Mar) the sequence was modified to remove all lark songs, and played only calls, sympatric species, and silence. The call playback sequence included a variety of lark contact notes and calls and other sympatric species in 10 minute alternating periods of calls and silence. Playback intervals were maintained at 4 hours, but start time in the morning was adjusted to 6:30am (afternoon start time remained at 2:00pm).

The playback sequence was digitally uploaded to a high-quality commercially available audio game-caller (Foxpro NX4) (Figure 7), which was fashioned with a timer and exterior speakers, placed inside a rubber tub to protect them from the weather (Figure 8). Early in the season in 2012 when the playback devices were first deployed we had several logistical problems resulting in intermittent periods when the devices were not operating. By mid May 2012, playback devices were reliably functioning properly.

Figure 7. Foxpro NX4 playback device.

Figure 8. Playback device.

We placed 2 handcrafted three-dimensional decoys of streaked horned larks and one playback device in each treatment plot (TA6-T, TA14-T, STJ-T). One decoy was placed within 2 meters of the playback device and the second was placed approximately 20 meters away (Figures 9 & 10). Several decoys did disappear at St Johns Landfill in 2012. The decoys were recovered on the landfill, but often quite distant from their original location. Wildlife cameras were installed and an immature Bald Eagle and a Coyote were captured on film inspecting the decoy (Figures 11 & 12). To make it more difficult for either animal to lift and remove the decoys, they were modified by welding a metal weight to the bottom (Figures 13 & 14) and burying this weight in the sand. The weighted decoys remained in place throughout the remainder of the study.

Lark Conspecific Attraction Feasibility Study Ð Final Report 13 Center for Natural Lands Management, [email protected]

Figure 9. Lark decoy and playback device Figure 10. Lark decoy and playback device being deployed at St. Johns Landfill. deployed at TA14 on JBLM.

Figure 11. Remote wildlife camera captures Figure 12. Lark decoy inspected by coyote at St. image of immature bald eagle inspecting lark Johns landfill. 17 October 2012 decoy at St. Johns Landfill. 13 June 2012.

Figure 13. Lark decoys are fitted with metal Figure 14. Weighted lark decoy. weights by Metro staff.

Lark Conspecific Attraction Feasibility Study Ð Final Report 14 Center for Natural Lands Management, [email protected]

We monitored lark response with dedicated visits to both control and treatment plots throughout the time that the playbacks and decoys were deployed. Control and treatment plots were monitored at least twice per week. On each visit we conducted a 30 to 50 minute area search recording all bird species detected. When SHLA were detected, we observed them until they left the plot. We recorded sex, age, behavior, weather, temperature, playback status (playing or silent), time, date, and presence of other bird species in the plots.

Plots at St. Johns were also monitored January and February in 2013 and anecdotally in winter months of 2011 and 2012. JBLM plots were not monitored during winter, as SHLA are not in South Puget Sound at this time. If larks were detected in treatment plots, the playbacks and decoys were removed so as to not give the impression that there was not a territory available to be occupied. We used a χ2 test to determine differences between detections between control and treatment plots.

Results Predator Load Over the course of the study period, we detected 7 species of potential SHLA predators at relocation sites. Potential SHLA predators detected at the plots included the Northern Harrier and American Kestrel. The detection frequency for all potential predators (less than 3%) was relatively low besides Killdeer. Also to note, across both years, killdeer were the most common at the St. Johns control plots, and dramatically less frequent in the treatment plot, where the only SHLA detections in 2013 took place (Table 2). There was no consistent difference between treatment and control plots in the numbers or frequency of potential predators, though St. Johns had slightly more species and frequency of occurrence than the South Sound sites. Also note that Northern Harrier was detected more frequently at St. Johns. In the Willamette Valley, Northern Harrier is the primary nest predator. We also have indication that the nest spikes we installed are not always efficient deterrents for this species; we even documented them perching on the well-head with spikes (Figure 15). Figure 15. Northern Harrier perched on well- head with deterrent spikes. St. Johns, May 2013.

Lark Conspecific Attraction Feasibility Study Ð Final Report 15 Center for Natural Lands Management, [email protected]

Table 2. Potential predators detected during plot monitoring visits. Values are % of visits when a given species was seen. Red- Brown- Western American Northern Falcon Year Plot Killdeer tailed headed Meadowlark Crow Harrier spp Hawk Cowbird 2012 STJ-C 13 3 2013 STJ-C 4 2 2 58 2012 STJ-T 10 5 3 2013 STJ-T 7 1 12 6 1 2012 TA14-C 6 3 39 2013 TA14-C 2 7 2 2012 TA14-T 3 2013 TA14-T 2 2012 TA6-C 12 39 3 2013 TA6-C 2 40 2012 TA6-T 3 12 2013 TA6-T 4 32

Habitat Response In general, habitat management resulted in conditions consistent with lark habitat characteristics. While habitat characteristics tended to less dense than those reported by Pearson and Hopey (2005) (table 3), in general habitat management shifted habitat characteristics towards those reported values. Plots became barer, sparser, had shorter vegetation, and a lower frequency of grass and shrubs. Habitat management tended to create habitat more similar to what was found along the Columbia River and Washington coast, compared to habitat found on occupied prairie habitat. Bare ground at the St. Johns control plot were 11% less bare Figure 16. Lark nest and fledglings at Rivergate. then reported territorial and nesting sites, Reprinted from Moore, 2010. and the St. Johns treatment was 15% more bare than the highest range reported. However, bare ground at JBLM on average was 2 to 3 times more bare than documented averages, and significantly higher than reported averages for nesting areas (17.5%, SE 2%) in South Puget Sound. Both vegetation cover and vegetation height were lower than target values, though St. Johns values were more similar to reported values than those at JBLM. However, it is important to know SHLA likely use habitat with characteristics outside what has been reported, since habitat use in the Willamette valley is comprised of extremely bare ground with short statured vegetation. Larks have been known to use dirt fields almost completely denuded of vegetation (R. Moore, pers. comm.) and breeding grounds at the Rivergate source site are extremely bare (Figure 16). The % cover of bare ground in south Puget Sound use areas is the lowest reported from all parts of the range, we do not interpret a higher % cover of bare ground as a barrier to potential use. Lark Conspecific Attraction Feasibility Study Ð Final Report 16 Center for Natural Lands Management, [email protected]

Table 3. Vegetation monitoring results across two years of sampling. St. Johns sites were sampled twice, and JBLM sites were sampled 3 times. Below values are the average ecological similarity (Soreson-Bray Curtis similarity value ± CI95) of conspecific plots to reported habitat characteristics in Pearson & Hopey (2005). These values are presented as a % similarity, thus smaller values are less similar to those reported, while larger values are more similar. The averages of the three strongest reported habitat predictors (SE) are reported along with those recorded for our study. Italicized values are from Pearson & Hopey (2005) for comparison.

Eco. Sim. to S.S. Eco. Sim. To C.R. Site % bare ground vg. height % vg hits reference (%) reference (%)

South Sound 17.5 (2.0) 15.3 (0.7) 85.3 (5.0) TA14-T 37 ± 27 53 ± 36 41.3 (6.3) 4.5 (2.6) 51.5 (8.8) TA14-C 39 ± 28 59 ± 36 55.8 (13.1) 5.2 (2.8) 43.9 (11.4) TA6-T 44 ± 20 64 ± 24 43.3 (5.1) 5.5 (1.9) 47.4 (3.6) TA6-C 43 ± 16 62 ± 21 31.4 (7.1) 5.2 (1.5) 45.5 (5.0) Columbia River 66.7 (6.0) 14.6 (1.8) 33.2 (2.8) STJ-T 18 ± 11 35 ± 15 87.5 (1.0) 1.7 (0.7) 34 (5.7) STJ-C 25 ± 18 40 ± 15 53.8 (1.8) 2.4 (1.1) 40 (13.6)

Lark Response SHLA were observed 23 times over the course of two years of standard observations (April 2012 – October 2013). Overall, SHLA were significantly more likely to be detected in 2 playback treatment plots compared to controls (χ 1= 8.9, p < 0.005). There were 7 observations of larks in our plots in 2012, and 16 in 2013. All but 3 of these detections were in treatment plots (Tables 4 & 5).

Military training and activities did alter survey monitoring and audible detectability at JBLM. For example, on 1 May 2012 at TA14-C, we could not perform surveys due to an airdrop and other military activities. Another possible effect on the plot’s bird activity occurred on 18 May 2012 at TA14-C, when an army van arrived near the plot, and an orange tarp was erected nearby. Military firing at the ranges might have negatively influenced detection of larks due to the intensity of the sonic booms, and other sounds. For instance, on 26 Sept 2012 at TA6, heavy firing NE of the plot made it extremely difficult to hear birds and the tape playback from edge of the plot.

Only one encounter with humans recreating at the plots occurred during the monitoring period. On 16 Oct 2012 at TA14-T, a person was walking 3 dogs in the plot when we arrived to monitor. That same day, a hunter was firing rounds approximately 200m west of TA14-C.

Although we have not statistically assessed the relationship of weather and bird activity, we do not believe that weather conditions or precipitation affected the bird activity during our monitoring visits. We substantiate this claim by providing this observation: during

Lark Conspecific Attraction Feasibility Study Ð Final Report 17 Center for Natural Lands Management, [email protected] intermittent showers on 21 May 2012, monitors recorded a total of 24 individual birds representing five species in the treatment plot at TA6.

We do not suspect that decoys or the playbacks have a negative influence nor deter other species from using the plots. Evidence supporting this claim includes the high detection rate of Savannah Sparrows, and Killdeer in the plots, and the occurrence of several individuals within close proximity to the decoys and playback units (e.g., Savannah Sparrows were detected perched within 1m of the speakers).

During winter months of 2011-2012, mixed flocks of wintering horned larks, including E. a. strigata were anecdotally detected foraging in the treatment plot at St. Johns (Figures 17 & 18). There was no indication that these birds remained into the breeding season. Larks were again detected in January/February 2013 and were documented during regular monitoring.

Figure 17. Horned lark at St. Johns landfill. 11/2011 Figure 18. HOLA mixed flock, including streaked horned lark, at St. Johns landfill. 1/2012.

Table 4. Summary of monitoring effort and number of SHLA and hetero-specifics detected across all plots. Breeding season phenology is defined for this study as April - August.

# SHLA non- # SHLA # breeding season breeding # other Mean Survey Total Survey plot visits detections season species time (min) Time (hrs). detections STJ-C 123 0 0 17 30 60 STJ-T 131 12 5 19 33 71.2 TA14-C 75 0 3 17 41 51 TA14-T 77 0 0 11 42 53 TA6-C 78 0 0 17 40 52 TA6-T 80 0 3 11 42 55

Lark Conspecific Attraction Feasibility Study Ð Final Report 18 Center for Natural Lands Management, [email protected]

Table 5. Lark detections at treatment and control relocation plots documented during regular monitoring Date Site Play- # Sex Behavior2 Comments3 backs indiv 1 4/24/12 TA6-T S 1 F AL, C, F Unbanded, pale individual in plot center, flew south to center portion of plot. 4/28/12 TA6-T n/a 1 M S Male heard upon approach to plot, could have been skylarking. 5/24/12 TA6-T S 1 F F Lark foraging with a mixed flock of EUST and KILL. STHL foraged on moss sporophytes, then flew S and high, lost in clouds. Found in very gravelly bare area and mossy area. 6/1/12 STJ-T P 1 M FL, F Lark landed in plot after 31 minutes Ð extended monitoring time 7/15/12 TA14-C C 1 M FS 80% sure it was a male, behaved like a flushed STHL, got a good look at yellow; maybe a dispersing male? 7/17/12 TA14-C C 2 U FS 2 larks flushed from NW corner, flew >500m N; On way out, heard male singing from scarcely veg'd area N of north road; suitable habitat here for larks, but did not locate birds. 7/19/12 TA14-C C 2 U FS Birds flushed from NW corner of plot, and flew NE towards middle of TA14 1/22/13 STJ-T W 7 U FL, F Some bright yellow. 6-8 flew in. 1/25/13 STJ-T W 12 U F 1/28/13 STJ-T W 7 U F 1/31/13 STJ-T W 3 U F, FL 2/5/13 STJ-T R 1 U F 2/15/13 STJ-T R 2 U F 2/25/13 STJ-T R 2 U F 3/15/13 STJ-T R 1 M F, S 3/18/13 STJ-T R 1 M F Subspecific identity unknown. 3/22/13 STJ-T R 1 M F 3/25/13 STJ-T R 1 M F, FL, S Observed in NW area of plot. It foraged and sang a couple times, then flew up and over to SE area of plot. 3/27/13 STJ-T R 1 M F 4/10/13 STJ-T P 1 M Playbacks and decoys re-deployed 4/8/13, plot monitored on 4/9/13 (no larks detected). One HOLA (subspp undetermined) was interested in decoy. Near playback jumping up at decoy. Playback turned off. 4/11/13 R 1 M F 4/12/13 R 1 M F 4/18/13 R 1 M F, S, C 4/23/13 R 1 M F, S, FL, Sang while flying. FD Codes: 1Playbacks: P = playing, S = silent, R = removed due to presence of HOLA, W = not playing because winter season, C = control plot, no playbacks deployed, 2Behavior: AL = Alert position, C = Call, F = Forage, FS = Flushed, S = Song; 4EUST = European Starling, KILL = Killdeer, SHLA = Streaked Horned Lark

Lark Conspecific Attraction Feasibility Study Ð Final Report 19 Center for Natural Lands Management, [email protected]

In 2012, SHLA detections were split between treatments and control plot sets, with 3 detections at TA14 control plots and 3 at TA6 treatment plots. At TA6, observations of one male and one female respectively occurred during the breeding season just prior to the onset of nesting (24 and 28 April 2012), the third observation of a single female occurred in the middle of the breeding season (24 May 2012), and was seen foraging, and then returning toward the location of the source population (R76). All observations at TA14 occurred from 15 – 19 July 2012 of two unknown sexed birds and a single male. This period coincides with the end of the breeding season, and the two unknown birds could have been young of the year dispersing from nesting areas on TA14. At STJ there was one detection of a male SHLA in the treatment plot on 1 June 2012. Of all the 2012 detections, a breeding behavior was observed only one time (male song on 28 April in TA6).

In 2013 all the detections in the plots occurred at St. Johns, the majority of which (12 of 17) occurred during the winter months. In January and February 2013 SHLA were detected foraging in mixed flocks of multiple HOLA subspecies, which arrived to the treatment plots shortly after they were deployed. In March 2013 a single male individual was repeatedly detected in the treatment plot at STJ. We assume it was the same individual who remained in the plot into the latter part of April. Despite several biologists closely observing this individual, his subspecific identity was undetermined, but could have been merrilii, strigata or potentially some hybrid between the two. The individual had morphological characteristics that could be attributed to either subspecies. The individual was last detected on 23 April 2013, fate unknown.

The animal interacted with the decoy and was captured on video jumping up and touching the decoy with its beak. It would then move a bit away, forage, and return to interact with the decoy. This pattern was repeated four times during the 30-minute survey. Figures 19 & 20 are the still images captured by the observer as well as one from the remote wildlife camera that was set up on the decoys (Figure 21).

Figure 20. HOLA directly right of the playback and Figure 19. HOLA male in STJ treatment plot, to the left of the decoy at STJ treatment plot. 10 subspecific ID undetermined. 10 April 2013. April 2013. This individual was also captured on video jumping up and pecking at decoy.

Lark Conspecific Attraction Feasibility Study Ð Final Report 20 Center for Natural Lands Management, [email protected]

Other Noteworthy Species In addition to SHLA a total of 36 species of birds were found in the plots. Five species comprised the most frequently detected species at both JBLM and St. Johns sites: Savannah Sparrow, American Robin, European Starling, Killdeer, and Western Meadowlark. Nor were there significant patterns in response of birds, whose calls were also within the playbacks (Vesper Sparrow, Savannah Sparrow, Western Bluebird, Western Meadowlark). Though all 3 Vesper Sparrow detections were in the TA6T plots, Figure 21. Remote wildlife camera captures there were not enough samples to assess if this HOLA near decoy. 19 April 2013. was significant or not.

In addition there were several noteworthy and species of conservation interest detected. On 11 May 2012, one Taylor’s Checkerspot, a federally endangered species, was observed. The second noteworthy species was the Common Nighthawk, which was found roosting on the ground on 11 May at TA6-T. On 23 July, we confirmed a Nighthawk nest with one egg and one chick within the plot (Fig. 22), and in 2013 a Nighthawk was flushed from the same nesting area. At both JBLM sites Western Bluebirds were detected. Mountain Bluebirds were detected only at TA14, and Vesper Sparrow and Lazuli Bunting were both detected at the TA6 site.

Figure 22: Common Nighthawk adult (right) flushed from nest of one egg and chick (left), TA6, 17 July 2012.

Lark Conspecific Attraction Feasibility Study Ð Final Report 21 Center for Natural Lands Management, [email protected]

Discussion Conspecific attraction has been noted in many species of birds. Our data suggest this may also be the case for SHLA. While the response to playback treatments was relatively weak, larks were significantly more likely to be detected in plots with conspecific cues than in control plots. We succeeded in meeting two of the three criteria we established for success; both sexes of SHLA were detected at least once during our study period. However, we did not succeed in having SHLA establish a breeding territory at relocation sites.

Larks were detected in plots several times at JBLM in 2012 and never in 2013. Conversely, larks have been detected at St. Johns each winter during the study (2011-2013) and there were several detections during the breeding season (April – August), some of which were accompanied by breeding behaviors (e.g., singing male). While the fate of the male detected in March/April 2013 at St. Johns was undetermined, he may have been a SHLA attempting to breed at the site. He did exhibit breeding behaviors (i.e., singing and flight display) and was maintaining presence in the plot after winter flocks had begun breaking up, potentially indicating the first phases of territory establishment. Three explanatory scenarios and potential fates come to mind regarding this bird. First, let’s assume he was a male E. a. strigata. In that scenario he either was depredated or he returned to another breeding site. Or, we can assume he was a male E. a. merrili, who eventually moved on to his breeding grounds. If he was strigata his presence in early spring indicates he may have been considering the site for breeding. However, the presence of predator perches at the site should be addressed more thoroughly to reduce the potential threat at this site. Instead of spikes, we suggest affixing lengths of flashy ribbon (e.g., “irri-tape”) to deter raptors and other potential predators from perching.

The higher rate of detection at St. Johns may be due to the difference in landscape context and use during the non-breeding season. While source sites were all similar distances from relocation sites, St. John’s and Rivergate are both distinct patches separated by completely unsuitable habitat, while at JBLM relocations sites were surrounded by other potentially suitable habitat. Thus, the patchy context at St. Johns could have increased the likelihood of SHLA detecting and using the created habitat. The fact that larks use St. Johns as wintering foraging habitats, which are the same general characteristics that comprise breeding habitat, could also be contributing to their use of the site during the breeding season.

The same factors that make conspecific attraction likely could also hinder attempts to lure birds using the same cues. The treatment areas, while arising brief interest from birds, could still be undesirable when a higher density of birds are heard or seen close by. If these resident populations are not at population saturation, emigration to habitat further away may not outweigh the benefits of aggregated nesting among the resident location, and density-dependent factors won’t begin to push birds to find new territories.

Social cues at the established sites thus could continually counteract potential influence of our study plots (Betts et al. 2008). Since treatment plots were only composed of 1 decoy “pair” while denser breeding territories were composed of multiple pairs, birds could have interpreted study plots as unsuitable habitat, or more risky than habitat with multiple Lark Conspecific Attraction Feasibility Study Ð Final Report 22 Center for Natural Lands Management, [email protected] pairs, and the low bird density could have been interpreted as a negative social cue (Danchin & Wagner 1997, Hoi & Hoi-Leitner 1997). Related to this, instead of removing playback devices entirely once larks were detected, moving them outside of the treatment plots, and in different locations around the plots is an approach that could minimize the potential ‘competitive effect’ the playbacks might give as well as mimic active breeding pairs in the vicinity.

Also, larks in South Puget Sound are in documented decline and only first-year birds have been documented moving from their natal sites, and never emigration into South Sound. There are few, if any, “excess individuals” that may disperse to establish new sites. Furthermore, habitat availability is not always the limiting factor. For example, in the South Puget Sound there are several apparently suitable unoccupied expanses and even some areas of occupied sites where larks have recently become extirpated. While habitat remains available in these static landscapes, larks may choose to remain in established breeding sites. In addition, the lack of larks documented in the South Sound study areas in 2013 could be a function of the extensive restoration and habitat management that occurred in the fall of 2012. Large prescribed fires at both TA6 and TA14 created large expanses of low stature bare habitat, within the current occupied ranges of SHLA. Larks respond positively to fire, and have been shown to move into burned areas the spring following a fall burn (Pearson et al. 2005). The expansion of suitable habitat within source sites at JBLM likely decreased the probability that birds would choose habitat created in the study plots since the expansion of the number of possible territories at source sites likely increased, and territory size is generally believed to decrease with the increase in habitat quality.

In addition, because larks demonstrate extremely high site-fidelity (Pearson et al. 2008), without sufficient incentive, successfully breeding larks are unlikely to disperse to new breeding sites. In the Willamette Valley, larks have been observed moving from one site to another mid breeding season as habitat becomes unsuitable (R. Moore, pers. comm.). It may be necessary to couple conspecific attraction with attempts to deter larks from high human-wildlife conflict breeding sites through habitat manipulations. Habitat manipulations could include altering the structure and cover of the vegetation to discourage larks from settling and breeding.

Finally, using conspecific attraction as a conservation technique should be considered carefully. Ahlering et al. (2010) cautioned that conspecific attraction may be too risky for use with threatened or endangered species, given that managers often do not understand species-specific habitat requirements, nor are the ultimate causes of decline addressed. However, they also noted that conspecific attraction may be less risky than forced movements of wild or captive reared animals. They acknowledged that the relocation technique is most appropriate for species with small or declining populations, but cautioned that decisions need to be made carefully. They also noted that the most prominent gap in our knowledge of conspecific attraction in songbirds was our lack of information about the consequences of attracting birds by such manipulations. For example, Ward et al. (2011) found that nest predation increased following successful

Lark Conspecific Attraction Feasibility Study Ð Final Report 23 Center for Natural Lands Management, [email protected] attraction of Foster’s and Least Terns to new nesting locations, and suggested that a predator control program be part of a conspecific attraction conservation tool.

Conclusion In conclusion, our study indicated that SHLA do respond to conspecific cues at some level, but other factors influence the animal’s process for selecting breeding habitat. Brief visits of SHLA shortly after the deployment of playbacks suggest the birds at least noticed these treatments and investigated the habitats. Lark life history traits and habitat preferences generally meet the criteria outlined by Ahlering et al. (2010), although their extreme site tenacity implies that the artificial conspecific cues will likely attract failed breeders and juveniles, unless active measures are taken to deter mature breeders from existing sites. Conspecific attraction might be an appropriate tool to attract Streaked Horned Larks away from high human-wildlife conflict areas into protected habitat, but attracting larks with conspecific cues should not be implemented as a conservation strategy unless the ultimate and proximate causes of the species decline are addressed at the attraction site. Although this technique may be risky for use with a threatened species, conspecific attraction is less risky than forcibly translocating wild or captive reared animals.

Lark Conspecific Attraction Feasibility Study Ð Final Report 24 Center for Natural Lands Management, [email protected]

Acknowledgements This project would not have happened without the dedication and support of numerous individuals and entities. US Fish and Wildlife Service, Port of Portland, Joint Base Lewis- McChord, and Metro have provided significant cash and in-kind support for the project. JBLM, Metro, and Port of Portland have provided access to their sites and helped implement the project. Thank you to Kim Flotlin and Cat Brown, USFWS, and to David Clouse and Jim Lynch at JBLM for their support. At the Port of Portland, thanks to Marla Harrison and Dana Green for their support and to Carrie Butler for quick access to Rivergate when needed (which included some very early hours) and to the support of Paul Vandenberg, Elaine Stewart, and Therese Mitchel at Metro. I am particularly grateful to Adrian Wolf, CNLM, for his bird knowledge, field experience, and willingness to help through all phases of the project. Adrian conducted a large proportion of field monitoring, was an excellent partner for talking through project ideas and changes, and helped tremendously in the production of this report. Thanks also to Adam Martin, Jerrmaine Treadwell, Maureen Puffer, and Tim Leque for their help monitoring for lark response at JBLM and data entry. Adam also conducted the statistical analyses and drafted an early version of the final report. Therese Mitchell at Metro, conducted the majority of lark field monitoring at St. Johns, all the habitat management, assisted with vegetation monitoring, and managed wildlife cameras and playback devices at St. Johns. Alsion Styring, The Evergreen State College, recorded source specific songs, and Greg Falxa, Cascadia Research Collective, was instrumental in getting the playback devices in working order. Thanks to the CNLM and JBLM prescribed fire crew, particularly Bob Wilkin and Mason McKinley, who made the fire happen in plots at TA6. Huge gratitude to Mark Roth, CNLM, Todd Zuchowski, JBLM, and Rod Gilbert, JBLM for their knowledge and expertise in prairie habitat management, as well as implementing the habitat treatments at JBLM. Finally, thank you to The Nature Conservancy of Washington, especially Misha Henshaw and Jim Evans, for their unwavering support of CNLM in general, and this project in particular.

Lark Conspecific Attraction Feasibility Study Ð Final Report 25 Center for Natural Lands Management, [email protected]

Literature Cited Ahlering, M.A., D.H. Johnson, J. Faaborg. 2006. Conspecific attraction in a grassland bird, the Baird's Sparrow. Journal of Field Ornithology, 77(4), 365-371.

Ahlering, M. A., Arlt, D., Betts, M. G., Fletcher Jr, R. J., Nocera, J. J., & Ward, M. P. (2010). Research needs and recommendations for the use of conspecific-attraction methods in the conservation of migratory songbirds. The Condor, 112(2), 252-264.

Altman, B. 2011. Historical and current distribution and populations of bird species in prairie-oak habitats in the Pacific Northwest. Northwest Science, 85(2), 194-222.

Anderson, H. E. 2005. Nest predation of the Streaked Horned Lark (Eremophila alpestris strigata) on lowland Puget prairie remnants, Washington State. The Evergreen State College, Olympia, Washington

-----. 2011. Columbia River Streaked Horned Lark Restoration Trial. Final Report to USFWS. Center for Natural Lands Management, Olympia, WA.

Anderson, J. K. 2010. Comparing endangered streaked horned lark (Eremophila alpestris strigata) fecundity to other grassland birds (Doctoral dissertation, Evergreen State College).

Betts, M. G. et al. 2008. Social information trumps vegetation structure in breeding-site selection by a migrant songbird. Proceeedings of the Royal Society B: Biological Sciences 275.1648: 2257-2263.

Bonham, C.D. 1989. Measurements for terrestrial vegetation. John Wiley and Sons, New York, NY.

Bowles, J.H. 1898. Notes on the Streaked Horned Lark. Osprey 3:53-54.

Camfield, A. F., S.F. Pearson & K. Martin. 2010. Life history variation between high and low elevation subspecies of horned larks Eremophila spp. Journal of avian biology, 41(3), 273-281.

----- 2011. A Demographic Model to Evaluate Population Declines in the Endangered Streaked Horned Lark. Avian Conservation & Ecology, 6(2).

Danchin, E., & R. H. Wagner. 1997. The evolution of coloniality: the emergence of new perspectives. Trends in Ecology & Evolution 12.9 (1997): 342-347.

Davis, J. M., & Stamps, J. A. (2004). The effect of natal experience on habitat preferences. Trends in Ecology & Evolution, 19(8), 411-416.

Lark Conspecific Attraction Feasibility Study Ð Final Report 26 Center for Natural Lands Management, [email protected]

Davis, J. M. 2008. Patterns of variation in the influence of natal experience on habitat choice. The Quarterly review of biology, 83(4), 363-380.

Drovetski, S.V., S.F. Pearson, and S. Rohwer. 2005. Implications of mitrochondrial DNA diversity on the conservation status of the streaked horned lark Eremophila alpestris strigata. Conservation Genetics 6:875-883.

Doligez, B., Cadet, C., Danchin, E., & Boulinier, T. 2003. When to use public information for breeding habitat selection? The role of environmental predictability and density dependence. Animal Behaviour, 66(5), 973-988.

Dunwiddie, P. W., & J.D. Bakker. 2011. The future of restoration and management of prairie- oak ecosystems in the Pacific Northwest.Northwest Science, 85(2), 83-92.

Fletcher Jr, R. J., & C.W. Miller. (2006). On the evolution of hidden leks and the implications for reproductive and habitat selection behaviours. Animal Behaviour, 71(5), 1247- 1251.

Gutzwiller, K. J., H. A. Marcum, H. B. Harvey, J. D. Roth and S. H. Anderson. 1998. Bird tolerance to human intrusion in Wyoming montane forests. Condor 100:519-527.

Hahn B.A., E.D. Silverman. 2007. Managing breeding forest songbirds with conspecific song playbacks. Animal Conservation 10:436-441.

Hilderbrand, R. H., A. C. Watts, & A.M. Randle. 2005. The myths of restoration ecology. Ecology and Society, 10(1), 19.

Hoi, H., & Hoi-Leitner, M. 1997. An alternative route to coloniality in the bearded tit: females pursue extra-pair fertilizations. Behavioral Ecology, 8(2), 113-119.

Jewett, S. G., W.P. Taylor, W.T. Shaw, & J.W. Aldrich. 1953. Birds of Washington state (p. 767). Seattle: University of Washington Press.

Kress, S.W. 1997. Using animal behavior for conservation: case studies in seabird restoration from the Maine coast, USA. Journal of the Yamashina Institute for Ornithology, 29(1), 1-26.

Morton, E.S., L. Forman, M Braun. 1990. Extrapair fertilization and the evolution of colonial breeding in Purple Martins. Auk 107: 275-283.

McCune B., J. B. Grace. 2002. Analysis of Ecological Communities. MjM Software Designs, Gleneden Beach, Oregon. Pp 143-146.

Lark Conspecific Attraction Feasibility Study Ð Final Report 27 Center for Natural Lands Management, [email protected]

Nocera, J. J., G.J. Forbes, & L.-A. Girdaldeau. 2006. Inadvertent social information in breeding site selection of natal dispersing birds. Proceedings of the Royal Society of London B 273: 349-355.

Parker, M. W., S.W. Kress, R.T. Golightly, H.R. Carter, E.B. Parsons, S.E. Schubel, ... & S.M. Wisely. 2007. Assessment of social attraction techniques used to restore a Common Murre colony in central California. Waterbirds, 30(1), 17-28.

Pearson, S.F. and M. Hopey. 2005. Streaked horned lark nest success, habitat selection, and habitat enhancement experiments for Puget lowlands, coastal Washington and Columbia River islands. Natural Areas Program Report 2005-1. Washington Dept. of Natural Resources. Olympia, WA.

Pearson, S.F., A.F. Camfield, & K. Martin. 2008. Streaked horned lark (Eremophila alpestris strigata) fecundity, survival, population growth and site fidelity: Research progress report. Washington Department of Fish and Wildlife, Wildlife Science Division, Olympia, WA.

Schlaepfer, M. A., M.C. Runge, & P.W. Sherman. 2002. Ecological and evolutionary traps. Trends in Ecology & Evolution, 17(10), 474-480.

Stamps J. A., & Swaisgood, R. R. 2007. Someplace like home: experience, habitat selection and conservation biology. Applied Animal Behaviour Science, 102(3), 392-409.

Stinson, D.W. 2005. Washington State Status Report for the Mazama Pocket Gopher, Streaked Horned Lark, and Taylor’s Checkerspot. Washington Department of Fish and Wildlife, Olympia 129+xii pp.

Styring, A. 2011. Streaked Horned Lark Vocal Attractions Study. Final Report submitted to Center for Natural Lands Management. The Evergreen State College, Olympia, WA.

Sutherland, W. J., I. Newton, and R. E. Green. 2005. Bird ecology and Conservation: A Handbook of Techniques. Oxford University Press. 386 pp.

Thomson, R. L., J.T. Forsman, & M. Mönkkönen. 2003. Positive interactions between migrant and resident birds: testing the heterospecific attraction hypothesis. Oecologia, 134(3), 431-438.

United States Department of Interior. Fish and Wildlife Service. Endangered and Threatened Wildlife and Plants; Determination of Endangered Status for the Taylor’s Checkerspot Butterfly and Threatened Status for the Streaked Horned Lark; Final Rule. Federal Register October 2013, 78:192.

Ward, M. P., & S. Schlossberg. 2004. Conspecific attraction and the conservation of territorial songbirds. Conservation Biology, 18(2), 519-525.

Lark Conspecific Attraction Feasibility Study Ð Final Report 28 Center for Natural Lands Management, [email protected]

Lark Conspecific Attraction Feasibility Study Ð Final Report 29 Center for Natural Lands Management, [email protected]