Cross-species Fostering: An Attempt to Establish Purple Martin colonies by utilizing breeding pairs of Tree Swallows. 2003 – 2004

The following project was carried out in the spring/summer of 2004, with permits from both the Pennsylvania Game Commission (Special Use Permit No. 56-2004) and the U.S. Department of Fish & Wildlife  (Federal Fish and Wildlife Permit Number MB081548-0) The proposal was modified after consultations with the PA Game Commission’s wildlife biologist. See “Amendments to Procedures for Cross-species Fostering Proposal” at the end of the proposal section. The decision was made to carry out the project on a smaller scale (using only four pairs of Tree Swallows) than was originally proposed during the first year. This presentation consists of two parts – the Proposal and the Final Report. Each is about 10 pages in length.


Ken Kostka
American Swallow Conservancy
Pittsburgh, PA

Kirk Piehler, Biologist, Fish & Wildlife Division,
U.S. Army Corps of Engineers (USACE)
Pittsburgh Division

Proposal Background

For the past five years, from 1999-2003, Ken Kostka, in cooperation with the U.S.Army Corps of Engineers (USACE), has attempted to establish Purple Martin colonies at four USACE facilities in Southwestern Pennsylvania, an area of martin scarcity. These efforts include the placement and ongoing management of Purple Martin housing at four locations: Lock & Dam #4 in Natrona, Allegheny Co., PA; Lock & Dam #5 in Freeport, Armstrong Co., PA; Lock & Dam #6 in Clinton, Armstrong Co., PA, and Crooked Creek State Park, Ford City, Armstrong Co., PA. All four locations are ideal Purple Martin habitat – open areas on waterfront property (Hill, 1988).  The first three locations are along the Allegheny River while the third is at Crooked Creek Lake.

The Purple Martin housing offered at these sites consists of both wooden and metal multi-compartmented houses, as well as gourds, both natural and plastic. All martin housing is easily accessible and designed to allow for monitoring and management of nest contents. The housing is kept free of non-native nest-site competitors. House Sparrows (Passer domesticus) and European Starlings (Sturnus vulgaris) are controlled by way of trapping or nest removal. Starlings are also excluded with starling-resistant or starling-proof entrance holes.

To further enhance the chances of attracting martins, the Social Attraction technique (Kostka, 2000) has also been utilized. Decoys have been deployed and martin vocalizations broadcast over outdoor loudspeakers in an attempt to create the illusion that these sites are already colonized, making them more attractive to investigating martins. The broadcast martin vocalizations also act as a beacon, advertising the sites to migrating or cavity-hunting Purple Martins. Artificial nests and mirrors have also been employed to enhance the illusion of colony site activity.

Despite these efforts of five years, no colonies have been established at any of the four sites.

Tree Swallows often attempted to nest in the housing intended for Purple Martins.

Tree Swallow pair copulating

When it became apparent that there was a lack of suitable housing for the Tree Swallows, additional housing was deployed specifically for them to nest in, and their numbers have grown dramatically. At Lock & Dam #4, eight pairs of Tree Swallows bred successfully in 2003, where none had bred in 1998. At Lock & Dam #6, eight pairs bred in 2003, where there was no more than 1 pair in 1998. All but a few of these Tree Swallow nests were constructed in large gourds that were designed to house Purple Martins, but were deployed for the Tree Swallows. These gourds are hung at a height of about three feet from the top of the lock wall, along a wire rope fence that runs the length of the wall, parallel to the river. They are spaced 30-60 feet apart. Other housing intended solely for Purple Martins is also deployed at both sites.


It is proposed that Ken Kostka of the Purple Martin Preservation Alliance, in cooperation with the USACE Wildlife Biologist Kirk Piehler, attempt to establish Purple Martin colonies at two USACE facilities: Allegheny River Lock and Dam #4 in Natrona, Allegheny Co., PA and Allegheny River Lock and Dam #6 in Clinton, Armstrong Co., PA, by using the established Tree Swallow breeding pairs to hatch Purple Martin eggs and fledge the nestlings, some of which are hoped will return in subsequent seasons as breeding adults to help establish Purple Martin colonies at the two sites. Both sites had eight pairs of breeding Tree Swallows in 2003. It is expected that each site will expand to ten breeding pairs in 2004.

Tree Swallows are a good candidate for fostering Purple Martins for several reasons. First, both Tree Swallows and Purple Martins are secondary cavity-nesting members of the family Hirundinidae and have very similar breeding habits and diets (Robertson, 1992 & Brown, 1997). Second, Tree Swallows are already breeding in large numbers at these USACE sites, and these are predominantly ASY adults with prior breeding experience and success. (Kostka, unpub. data) Additionally, the two proposed USACE riverfront sites are fenced and secured facilities, making them safe from vandalism or tampering.

Current Status of Purple Martins in Southwestern Pennsylvania

Purple Martins are scarce in southwestern Pennsylvania, especially along the Allegheny River corridor, with very few or no Purple Martins spotted during breeding bird surveys over the past seven years. (Sauer et al. 2003) According to the Breeding Bird Survey, not a single Purple Martin was counted on any route within a 50 mile radius of the project area in 2001 or 2002. Purple Martins were counted on only 2 of the 24 routes in the years 1996-2000. A total of only 5 Purple Martins were seen on these 24 routes over a period of seven years between 1996 and 2002. While several colony sites exist in this area, these BBS numbers attest to the general scarcity of these swallows in southwestern Pennsylvania.

I looked at all Breeding Bird Survey Routes within a 50 mile radius of project areas for the years 1996 – 2002.

PA-003 Clover, PA-009 Sigel, PA-011 N. Washington, PA-012 N. Pine Grove,
PA-04 Beaver, PA-047 Duvall, PA-048 Pleasant Grove, PA-049 Parkwood,
PA-050 Kittanning, PA-052 Wakena, PA-053 Kaylor, PA-054 Shelocta,
PA-056, Margaret, PA-057 Virginia, PA-059 Martinsburg, PA-088 Jefferson,
PA-154 Mars

Of these 17 routes, only 5 Purple Martins were counted between 1996 and 2002!  That’s a total of 5 martins counted in 119 routes (0.04%).

Additionally, Purple Martins were not listed in the following routes, also within 50 miles of the proposed project area. The BBS site does state “Caution! No species is found on every route….”(Sauer et al. 2003) It seems reasonable to assume that Purple Martins were not listed because they were not seen during surveys.

PA-151  Jeanette,  PA-155- Laurel Hill,  PA-058  Patton,  PA-061 Williamsburg,
PA-101 Harrisville,  PA-189 Mill Run

Each survey route is 24.5 mi long, and consists of 50, three minute roadside counts along the length of the route. (Sauer et al. 2003) Because Purple Martins are birds of open areas near human habitation, which is usually situated near roads, one would expect that martins would be one of the more visible species on survey routes.
According to the Purple Martin Summer Distribution Map (Sauer et al. 2003), which predicts the average number of Purple Martins that could be seen in about 2.5 hours of birdwatching along roadsides by very good birders, the number of Purple Martins predicted was “one or less” for the proposed project area of Allegheny Co. and Armstrong Co., PA. Furthermore, according to the Purple Martin BBS Trend Map, 1966–1996, Purple Martin numbers have undergone a long-term significant decrease in the proposed project area. (Sauer et al. 2003)

History of Purple Martins in Southwestern Pennsylvania

The scarcity of Purple Martins in southwestern Pennsylvania is largely a result of three factors: the continued depredations of non-native European Starlings and House Sparrows, the martin-killing weather of Tropical Storm Agnes in June of 1972, and the lack of interest in the hobby from an increasingly urbanized culture. Tropical Storm Agnes stalled over western Pennsylvania in late June of 1972, in the peak of the Purple Martin breeding season, and 6-10 inches of rain fell on the region from 20-25 June. Since Purple Martins rely exclusively on flying insects for food, and since the continuous rains depleted the air of flying insects, almost all of the Purple Martins in southwestern PA died of starvation, along with their nestlings, during these prolonged rains. Normally, martins would have rebuilt their populations naturally, in part by attracting individuals from outside the devastated area in successive breeding seasons. However, in the martins’ absence, the European Starlings and House Sparrows became entrenched in the unoccupied housing, and in many cases prevented the gradual re-colonization of former sites. Purple Martins are timid when investigating new sites, and are usually repulsed by these non-native species if they have claimed the housing. At sites where a few pairs of breeding

USACE Lock 4, Natrona, PA

martins survived, they were often overwhelmed by starlings or House Sparrows. Finally, after the widespread die-off, many martin “landlords” gave up the hobby and did not pass on the tradition of offering martin housing to younger generations, resulting in a decades-long lack of properly managed housing.

Information gleaned from numerous conversations with long-time residents of the project area, who either managed Purple Martin colony sites or were Purple Martin enthusiasts, suggests that Purple Martins were once fairly common in the project area and throughout southwestern Pennsylvania. The author’s personal conversations with Gerald Presley (deceased), who resided in Brackenridge, Allegheny Co., PA, adjacent to Natrona, PA, related accounts of numerous colony sites in the project area. The author spoke to two of the founders of these colonies prior to their deaths. These colonies were active from the 1940’s through 1972. (Presley, pers. comm.)

Mr. & Mrs. JJ Gleinn, 1437 4th Ave., Natrona Heights, PA
John Mazgaj, 2414 High St., Natrona Heights, PA
Gilson residence, 514 Painter Ave., Natrona Heights, PA
Residence at Idaho Ave, Natrona Heights, PA
John Slivon, 1072 Roup Ave., Brackenridge, PA
Tarentum Community Park, 1st Ave., Tarentum, PA
Mr. Shumaker, 11th Ave., Tarentum, PA

John “Duke” Snyder, the manager of a colony in Butler, PA, has, through local research, determined that there were at least eight colonies within a two mile radius of Butler, Butler Co., PA in 1966. (Snyder, pers. comm.) Roy Bauder, a landlord in Portersville, Butler Co., PA, counted 25-30 colonies in the Portersville, Butler Co., PA area that were active in the from the 1950’s through the mid-1960’s. (Bauder, pers. comm.) Lloyd Getty (deceased) of Saxonburg, PA spoke of numerous colonies in Saxonburg, Butler Co., PA, (Getty, pers. comm.)  C.W. Parker recorded at least nine colonies in Allegheny in 1925. (Todd, 1940)  Presently, there are only two known colonies in Allegheny Co., PA (one at Youghiogheny Country Club in McKeesport and the other at Butlers Golf Course in Elizabeth) and one known colony in Armstrong Co., PA – at Gastown Racetrack near Shelocta.

Factors Involving Establishment of Purple Martin Colonies

The Purple Martins’ social nature adds to the difficulty of establishing new colonies in areas where establishment is already difficult because of the scarcity of Purple Martins. Martins prefer to nest in the company of other martins and usually choose to reside at established sites rather than nest alone at new, untested sites. Subadult females, who must often accept subadult males as mates, usually prefer already established breeding sites, because they have access to extra-pair copulations with genetically superior adult males (Morton, 1990).  These factors help to explain why no new colonies have been established in a multi-county area of southwestern PA, despite concerted efforts to attract them. (Kostka, unpub. data)
Further compounding the difficulty in attracting martins to certain sites in southwestern PA is the abundance of Tree Swallows, which readily accept Purple Martin housing. Interestingly, Tree Swallow numbers appear to be high and increasing in the same areas that martin numbers are low and decreasing. (Sauer et al. 2003) These data suggest that Tree Swallows may be preventing Purple Martins from colonizing some sites in the Tree Swallow breeding range, since they readily nest in housing intended for Purple Martins. The rise in Tree Swallow numbers might be a result of the abundance of housing placed for Eastern Bluebirds (Sialia sialis) as well as Purple Martins, since Tree Swallows often nest in structures intended for these species.

Cross-Fostering in non-Passerine Species

The cross-fostering technique has been attempted in several non-passerine species. Whooping Crane (Grus americana) recovery efforts employed the cross-fostering method using Sandhill Cranes (Grus canadensis) as the foster parents. These efforts were discontinued because of high mortality and a lack of mating success in the fostered cranes. (Ashton, 1991).  In 1979, experimenters used Prairie Falcons (Falco mexicanus) to foster Peregrine Falcons (Falco peregrinus) in South Dakota. Two young were successfully fledged in 1979 in the Black Hills, but efforts were unsuccessful in 1980, and the experiment was discontinued. Hacking later proved successful in aiding Peregrine populations (Ashton, 1991).

Interestingly, cross-fostering has apparently been used to establish White-nest Swiftlet (Aerodramus fuciphagus) colonies. Individuals engaged in the commercial harvest of White-nest Swiftlet nests for bird’s nest soup, a delicacy in the Far East, buy properties in Indonesia that have established colonies of Mossy-nest Swiftlets (Aerodramus vanikorensis) and then place White-nest Swiftlets into their nests. After they mature, the fostered White-nest Swiftlets return and establish a colony. Nest traders claim that a third of all nests now come from these “farms,” but authorities believe that the number is far less and that the establishment of colonies by cross-fostering does not have a high success rate. (MacKay, 1993)

Cross-Fostering in Passerine Species

Tore Slagsvold of the University of Oslo conducted a large-scale cross-fostering experiment in which parents of certain species successfully raised the nestlings of other species in the wild. The four species involved were the Great Tit (Parus major), the Blue Tit (Parus caeruleus), the Coal Tit (Parus Ater), and the Pied Flycatcher (Fidecula hypoleuca). Eggs were exchanged during the incubation period, and all of the host species’ eggs were removed. According to Slagvold, “Great Tits were reared by Blue Tits (155 chicks fledged from 41 nests) and Blue Tits by Great Tits (242 chicks fledged from 41 nests) in 1999-2000. Blue Tits were reared by Coal Tits (Parus Ater) in 2000 (38 chicks fledged from six nests) and Pied Flycatchers were reared by Great Tits and Blue Tits in 1998-1999 (573 chicks fledged from 139 nests)”.

Sexual Imprinting and Cross-Fostering in Passerines

Slagsvold’s research indicates that the impact of sexual imprinting does not seem to be constant across species. Certain species have gone on to breed successfully after being fostered by other species. (Slagsvold, et al. 2002) Cross-fostered Great Tits appeared to be imprinted on the Blue Tit host and failed to pair conspecifically. Cross-fostered Blue Tits, however, had high pairing success.” He goes on to state that “Pied Flycatchers did not seem to be imprinted on their tit hosts at all”. Sexual mis-imprinting occurred in only one of the three species that were cross-fostered. (Slagsvold, 2002) Slagsvold speculated that the lack of sexual imprinting of Pied Flycatchers on their tit hosts may have been due to the fact that the two species have different wintering areas, allowing cross-fostered individuals less time to consolidate the initial sexual imprinting. In addition, the flycatchers were less closely related to the tits than were the other tit species. (Slagsvold, 2002) It should be noted that while Purple Martins and Tree Swallows are of the same family, they are of different genera, and do not overwinter together.

Account of Cross-Fostering in Purple Martins

There is one very old account in the literature of Tree Swallows fostering Purple Martins. The account was documented by J. Warren Jacobs in 1907:

May 31, 1907, Received four Purple Martin eggs from J. Warren Jacobs, Waynesburg, Pa.
June 1, 1907, I find six Tree Swallow eggs in nest. Remove four eggs and place the four martin eggs in nest.
June 6, 1907, Remove the other two Tree Swallow eggs, leaving the four martin eggs only.
June 11, 1907, Swallows sitting on eggs, — OK
June 15, 1907, Eggs not hatched.
June 16, 1907, 7:00 a.m. Find one egg hatched.
June 16, 1907, 6:00 p.m. Find two more eggs hatched.
June 17, 1907, 7:00 a.m. Find the other egg hatched.
June 19, 1907, Visit the young Martins this morning; find them doing well.
June 22, 1907, Visit Martins this morning, find them growing and looking fine. Eyes not open yet and not a feather started. Foster parents feeding them nicely.
June 24, 1907, 6:00 a.m. Visit young Martins. Find them progressing finely. Feathers begin to show under the skin. Eyes just beginning to open. Young birds well cared for.
June 25, 1907, a.m. Find martin eyes about one third open; feathers just begin to protrude through skin.
June 26, 1907, 6:30 a.m. Young Martins eyes open; feathers well started through skin; tail feathers show quite a little. All four young are plump and fat. Foster parents giving them best of care.
June 29, 1907, 7;00 a.m. Visit the Martin house; find no young martins in nest; find one on the ground under house. Could not find the other three. From appearance of young bird found on the ground, should say it has been dead two or three days. (Jacobs, 1907)

While it’s impossible to determine exactly what occurred in this account, it is most likely that the nest was raided by a predator, perhaps a Common Raccoon (Procyon lotor) or a Great Horned owl (Bubo virginianus), the two most common martin nest predators. Although inconclusive, this account clearly demonstrates that breeding Tree Swallows will hatch martin eggs and feed Purple Martin nestlings.

Differences between Species Needing Addressed

Tree Swallow

As previously mentioned, both Tree Swallows and Purple Martins are secondary cavity nesting members of the family Hirundinidae that prefer open areas near water (Robertson, 1992, Hill, 1988) and that feed almost exclusively on flying insects.
There are, however, several issues that need to be addressed when considering Tree Swallows as candidates for fostering of Purple Martin nestlings. First, Tree Swallows are smaller that Purple Martins. Tree Swallows nestlings weigh 20-22 grams (Robertson, 1992), while Purple Martins nestlings weigh 45-55 grams (Brown, 1997). This difference in body mass would be adjusted for by reducing the number of Purple Martin nestlings that the Tree Swallows are assigned to foster by approximately one-half. Instead of 4-6 nestlings, each Tree Swallow pair would only be given two Purple Martin eggs to hatch and therefore only have two nestlings to fledge.

It should be noted that a rather stark size difference does not prevent Brown–Headed Cowbirds (Molothrus ater) from being successfully fostered by a variety of host species. Brown-Headed cowbirds, weighing from 35-45 grams, have been fledged by species as small as creepers, kinglets, and gnatcatchers, weighing approximately 10 grams. (Lowther et al. 1993). Additionally, in Slagsvold’s cross-fostering experiment, Great Tits were successfully fostered by Blue Tits (155 chicks fledged from 41 nests), despite the fact that Great Tits (adult weight ca. 17-20 g) weigh almost twice as much as Blue Tits (adult weight ca. 10-11 g). The author stated that “the [Great Tit] chicks survived well in the nest and afterwards.” (Slagsvold, 1998; Slagsvold and Hansen, 2001).

A second issue needing addressed is that of breeding schedule. Typically, Tree Swallows begin egg-laying two weeks earlier than Purple Martins in Southwestern Pennsylvania. Tree Swallows egg-laying would need to be delayed in order to allow for the appearance of Purple Martin eggs at donor sites. Delaying tactics would include both temporarily blocking the entrance holes of breeding cavities after they are claimed by the Tree Swallows to delay nest-building and intermittently removing nesting material from the nests so as to delay egg-laying without discouraging the pairs that have claimed these cavities. If Tree Swallows begin to egg-lay despite these delaying tactics, their clutch would be allowed to be completed and then removed. Female Tree Swallows often re-clutch in response to the loss of the first clutch (Kuerzi 1941). Purple Martin eggs could then be substituted for the second clutch of Tree Swallow eggs.

PM eggs on left; TS eggs on right

The incubation period is several days shorter for Tree Swallows than for Purple Martins: 14-15 days vs. 15-18 days (Robertson, 1992 & Brown, 1997), but it has been demonstrated that Tree Swallows will continue to incubate and hatch Purple Martin eggs. (Jacobs, 1907) Similarly, the period from hatching to fledging is 6-7 days shorter for Tree Swallows than for Purple Martins: 18-22 days vs. 28-29 days (Robertson, 1992), (Allen & Nice, 1952) but this is not expected to be problematic since the Tree Swallows’ investment in the nestlings should compel them to continue feeding, especially if the nestlingsMartin appear to be uniform and healthy. Cold or unseasonable weather is known to lengthen the time required to fledge nestlings of any species, allowing for flexibility in the length of the nestling period.

Possible Return Rate of Purple Martin Fledglings and other Considerations

A gourd on the Lock wall with Tree Swallow nest

While first year survival and return rates of Purple Martins can vary widely depending on such factors as weather and available housing, first year survival in Purple Martins ranges from 30-50% (Kostka, unpubl. data.), and Purple Martins have exhibited return rates to the natal colony site as high as 20% (Troyer, Kostka, unpub. data) If ten pairs of breeding Purple Martins fledge an average of two Purple Martin nestlings each, two to four martins could be expected return to each natal site as first-year breeders. These martins would act as a catalyst for the establishment of a new colony, attracting and recruiting other martins to the site.  Most SY Purple Martins return to the general area (within 50 miles) of the natal site (Allen and Nice, 1952), so that other returnees could potentially colonize some of the other nearby sites, all managed and located in ideal habitat.

Nestling and first year survival rates might even be higher than normal at the project sites for several reasons. The USACE sites are fenced and lighted areas with a great deal of human activity, reducing the potential for predation.  Nest replacements can almost eliminate the potential for nestling mortality due to nest parasites (Kostka, 1999) as well as allowing for greater fitness of fledglings.  The fostered Purple Martin eggs/young would likely be genetically superior since they are procured from the oldest, earliest-returning, adult (ASY) pairs, which are proven breeders. The removal of Tree Swallow nesting material from multi-compartmented houses and gourd clusters intended solely for Purple Martins will reduce interspecific competition from Tree Swallows, allowing returning Purple Martins to claim and colonize this housing.

Procedure to be followed:

1.  Deploy Tree Swallow/Purple Martin housing (10 gourds per site) at the USACE sites (Lock & Dam #4 in Natrona, Allegheny Co., PA, and Lock & Dam #6 in Clinton, Armstrong Co., PA.) as normal, on or about April 1, 2004.

2.  Allow Tree Swallows to claim the housing, as normal.

3.  Intermittently block cavity entrance holes or remove nesting material to delay nest-building and egg-laying. Deployment of additional housing (gourds) adjacent to the claimed, but blocked housing will help to insure that Tree Swallows do not abandon the site while at the same time delaying nest-building and egg-laying. If eggs are laid despite delaying tactics, clutch will be allowed to be completed and then removed. (Females often re-lay in response to loss of first clutch.)

4.  Allow nest-building and egg-laying to resume after 2 weeks or to coincide with projected availability of Purple Martin eggs at three donor sites.  The first donor colony is a ~50 pair colony at McDanels Launch on the North Shore of Moraine State Park, Portersville, Butler Co., PA. This colony has been managed by Ken Kostka of the Purple Martin Preservation Alliance for the past five years. The second colony is a ~80 pair colony at the private residence of name and address deleted., Grove City, Mercer Co., PA. The third donor colony is the USACE Shenango Lake Recreation area in Mercer, Co., PA. If possible, only 1-2 eggs will be taken from each donor nest, leaving 3-4 eggs in donor nest.

5.  When Tree Swallow clutches have at least 2-3 eggs, remove Tree Swallow eggs and replace with 2 Purple Martin eggs. Any additional Tree
Swallow eggs laid will be removed. If one of the Purple Martin eggs fails to hatch, it will be replaced with a Purple Martin nestling of the
appropriate age.

6.  Continue to monitor Tree Swallow nests. If the fostered martin nestlings become significantly underweight, they will be placed back into the martin nests that they were procured from. Proper weights for nestlings will be determined by average daily weights of Purple Martin nestlings raised by Purple Martins. (Hill, 1994)

7.  Band Purple Martin Nestlings at 14-16 days old, using colored leg bands that allow individuals to be identified with a high power spotting scope. Ken Kostka is a licensed bird bander with both federal and state permits to use colored leg bands with alpha-numerics, allowing for recognition of individual martins.

8.  Continue to manage all USACE sites for Purple Martin attraction in 2005 and 2006, and monitor for return of fostered Purple Martins, which will be individually recognizable since they will have been banded with coded color bands that can be read through a spotting scope.

9.  Repeat procedure in 2005 and 2006 if cross-fostered nestlings fledge successfully in 2004.

We request federal approval to procure 40 Purple Martin eggs and up to 10 nestlings per year from donor sites (listed above) in Pennsylvania in 2004, 2005, and 2006. [This request was modified – see amendments below.] The nestlings would be fostered into nests where one of the eggs failed to hatch due to infertility. These eggs/nestlings are to be placed in 20 Tree Swallow nests at two USACE sites: Lock & Dam #4 in Natrona, Allegheny Co., PA, and Lock & Dam #6 in Clinton, Armstrong Co., PA. We request federal approval to manipulate these Tree Swallow nests at these sites by temporarily blocking entrance holes or removing nesting materials in order to delay egg-laying, as well as permission to remove Tree Swallow eggs. We request permission to remove Tree Swallow nesting material from any housing intended solely for Purple Martins at these sites. We request federal permits granting permission to duplicate these procedures in 2005 and 2006.


Ken Kostka and the USACE have, for the past five years, tried unsuccessfully to establish Purple Martin colonies at ideal sites along the Allegheny River in Southwestern Pennsylvania. At two of these sites, large Tree Swallow colonies have been established. It is proposed that these Tree Swallow pairs be used to cross-foster Purple Martin nestlings. The desired result of this effort is the return of fostered martins to the USACE sites as breeding adults, helping to establish Purple Martin colonies.


Allen, R. W., and M. M. Nice. 1952. A study of the breeding biology of the Purple Martin (Progne subis). Am. Midl. Nat. 47: 606-665.

Ashton, Diane E. and Eileen M. Dowd. 1991. Fragile legacy. Endangered, threatened and rare animals of South Dakota. South Dakota Department of
Game, Fish and Parks, Report No. 91-04. Jamestown, ND: Northern Prairie Wildlife Research Center Home Page.

Brown, C. R. 1997. Purple Martin (Progne subis). In The Birds of North America, No. 287 (A. Poole and F. Gill, eds.). The Academy of Natural
Sciences, Philadelphia, PA, and The American Ornithologists’ Union, Washington, D.C.

Hill, J. R., III, 1988. How to Attract Purple Martins. Purple Martin Update 1(1): 1-3.

Hill, J.R., III, 1994. The growth of nestling Purple Martins. Purple Martin Update 5(3): 1- 9.

Jacobs, J. Warren. 1907. The Purple Martin and houses for its summer home. Gleanings No. 5: 31-35.

Kostka, K. 2000. Social attraction: A new technique for establishing a Purple Martin colony site. Purple Martin Update 9(3): 26-29.

Kostka, Ken, and J Hill. 1999. How and why to do nest replacements for Purple Martins. Purple Martin Update 9(2): 2-5.

Kuerzi, R. G. 1941. Life history of the Tree Swallow. Proc. Linn. Soc. N.Y. 52-53: 1- 52.
Lowther, P.E. 1993. Brown-headed Cowbird (Molothrus ater). In The Birds of North America, No. 47 (A. Poole and F. Gill, eds.). Philadelphia: The
Academy of Natural Sciences; Washington, D.C.: The American Ornithologists’ Union.

MacKay, Jeanine. 1993. Trade and Environment Database. Case Studies. Swifts and Trade. American University. The School of International Service

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

Robertson, R.J., Stutchbury, B.J., and R. R. Cohen. 1992. Tree Swallow. The Birds of North America, No. 11 (A. Poole, P. Stettenheim, and F. Gill,
Eds.). Philadelphia: The Academy of Natural Sciences; Washington, DC: The American Ornithologists’ Union.

Sauer, J. R., J. E. Hines, and J. Fallon. 2003. The North American Breeding Bird Survey, Results and Analysis 1966 – 2002. Version 2003.1, USGS
Patuxent Wildlife Research Center, Laurel, MD

Slagsvold, T. 1998. On the origin and rarity of interspecific nest parasitism in birds. Am. Nat.: 152: 354-367.

Slagsvold, T. & Hansen, B. T. 2001. Sexual imprinting and the origin of obligate brood parasitism in birds. Am Nat. 158: 354-367.

Slagsvold, T., Hansen, B.T., Johannessen, L.E., & Lifjeld, J.T. 2002. Mate Choice and imprinting in birds studied by cross-fostering in the wild. Proc. R.
Soc. Lond. B: 269: 1449-1455.

Todd, Clyde, W.E. Birds of Western Pennsylvania. 1940. University of Pittsburgh Press. pp 375-378.

April 04, 2004

Amendments to Procedures for Cross-species Fostering Proposal

In the first year of this project (2004), the number of Tree Swallows chosen to foster Purple Martins will be limited to four (4) nesting pairs. Therefore, it will only be necessary to procure eight Purple Martin eggs (and up to two nestlings, in case of egg infertility) from the donor sites in 2004. If the Tree Swallows successfully fledge the Purple Martin nestlings in 2004, the number of Tree Swallow pairs utilized to foster Purple Martins may be expanded to the numbers originally proposed for the years 2005 and 2006, if permission is granted by the Pennsylvania Game Commission.

Tree Swallow eggs will be donated to Carnegie Museum of Natural History. Museum personnel will be contacted by April 30, 2004, in order to determine the protocol for transferring the eggs. Every attempt will be made to delay Tree Swallow nest building and egg-laying as long as possible, so that the eggs collected are as fresh as possible and most suitable for preparation and incorporation into a museum collection. Approximately 24 eggs (six per Tree Swallow nest) are expected to be collected.

In the first year of this project (2004) all of the Tree Swallow nesting pairs utilized for cross-fostering will be from the U.S. Army Corps of Engineers Lock and Dam #4 facility in Natrona, Allegheny Co., PA. This site is located in close proximity to the permittee’s residence and will allow for more intensive monitoring of the nests, nestlings, and fledglings, than was initially planned.

The permittee shall monitor the nests at least every third day during incubation, every second day from hatching to age 15 days, and daily from age 16 days through age 40 days. Nestlings will be weighed with a 100 gram Avinet spring scale. If at any time during the nestling stage, the martins are deemed to be significantly underweight, based on the average daily weights graph (Hill, 1994), they will be placed back into the Purple Martin nests and monitored to insure that they are receiving adequate parental care. If it is determined that they are not receiving adequate parental care, they will be taken to a licensed wildlife rehabilitator.

Martin fledglings will be monitored intensively for 10 days post-fledging to insure that they are being adequately cared for by the Tree Swallows. Any martin fledglings that become debilitated as a result of insufficient post-fledging parental care will be taken to a licensed wildlife rehabilitator.

The fostered martins will be located and monitored closely in subsequent years to determine if they mate and breed successfully. The great majority of Purple Martins return to an area within 50 miles of their natal colony site, usually to an already established colony site. All fostered Purple Martin nestlings will be banded with a yellow color band bearing an alpha-numeric code that identifies the individual Purple Martins. These bands can be read with a high-power spotting scope. Permittee is already licensed to use these bands and is proficient in reading them with a spotting scope.


Purple Martin/Tree Swallow Cross Fostering Project Report

Tree Swallow & Purple Martin

Ken Kostka
Purple Martin Preservation Alliance
Pittsburgh, PA

Federal Fish and Wildlife Permit Number MB081548-0
Pennsylvania Game Commission Special Use Permit No. 56-2004

As stated in the project proposal, the purpose of this project was to determine if Purple Martin (Progne subis) colonies could be established by placing Purple Martin eggs into Tree Swallow (Tachycineta bicolor) nests at established Tree Swallow colonies and allowing the Tree Swallows to hatch and raise the Purple Martin nestlings, which would hopefully return to the site in subsequent years to breed. (See original proposal for additional details.) What follows is the chronology of this project.

Abbreviations used: TS=Tree Swallow; PM=Purple Martin; d.o.= days old; g=grams; After banding on 6/14/04 & 6/16/04, the individual martin nestlings are referred to as the alpha-numeric on their yellow color band (i.e., “K360”)

May 16, 2004

Eight (8) Purple Martin eggs were secured from nests at the Purple Martin colony of Dean Kildoo, Grove City, PA. These were the earliest eight eggs laid and were not yet being incubated.  These eggs were transported to the Tree Swallow colony at U.S. Army Corps of Engineers Lock and Dam #4, Natrona, Allegheny Co., PA. This colony consists of 11 pairs of breeding Tree Swallows. Only four nests were used in the first year of the project to determine if the Tree Swallows could successfully hatch the Purple Martin eggs and fledge the nestlings. These four Tree Swallow nests will be referred to as Nest 1, 2, 3, 4.  It should also be noted that the parent at nests 1, 2, 3 are ASY-F, while the female parent at nest 4 is an SY-F. It is also believed that the male parents at nests 1, 2, 3 are also ASY-M, since these are three of the six earliest-nesting pairs.  In several cases, the Tree Swallow eggs were removed over a period of days to avoid alarming the Tree Swallow parents. As stated in the proposal, only two (2) Purple Martin eggs were placed into each Tree Swallow nest to account for the size difference between the two species. All nests were in plastic or natural gourds with access doors that allowed for easy access to the nest.

Nest 1 4 TS eggs present collected 2 TS eggs – inserted 2 PM eggs
Nest 2 6 TS eggs present collected 4 TS eggs – inserted 2 PM eggs
Nest 3 3 TS eggs present collected 3 TS eggs – inserted 2 PM eggs
Nest 4 1 TS eggs present collected 1 TS egg – inserted 2 PM eggs

May 17, 2004

Nest 1 2 PM/3 TS eggs present collected 1 TS eggs
Nest 2 2 PM/3 TS eggs present collected 1 TS eggs
Nest 3 2 PM/0 TS eggs present collected 0 TS eggs
Nest 4 2 PM/1 TS eggs present collected 1 TS eggs

May 19, 2004

Nest 1 TS on nest
Nest 2 2 PM/2 TS eggs present collected 1 TS eggs  (eggs warm)
Nest 3 2 PM/1 TS eggs present collected 1 TS eggs (eggs cold)
Nest 4 2 PM/1 TS eggs present collected 1 TS eggs

May 22, 2004

Nest 1 2 PM/2 TS eggs present collected 1 TS eggs
Nest 2 2 PM/1 TS eggs present collected 1 TS eggs
Nest 3 2 PM/3 TS eggs present collected 1 TS eggs
Nest 4 2 PM/2 TS eggs present collected 1 TS eggs (eggs warm)

May 23, 2004

Nest 1 2 PM/1 TS eggs present collected 1 TS eggs
Nest 2 2 PM/0 TS eggs present collected 0 TS eggs
Nest 3 2 PM/2 TS eggs present collected 1 TS eggs
Nest 4 2 PM/1 TS eggs present collected 1 TS eggs

May 26, 2004

Nest 1 2 PM/1 TS eggs present collected 1 TS eggs
Nest 2 2 PM/0 TS eggs present collected 0 TS eggs
Nest 3 2 PM/1 TS eggs present collected 1 TS eggs
Nest 4 2 PM/0 TS eggs present collected 0 TS eggs

May 28, 2004

Nest 1 2 PM eggs present (warm)
Nest 2 2 PM eggs present (warm)
Nest 3 2 PM eggs present (warm)
Nest 4 2 PM eggs present

May 30, 2004

Nest 1 2 PM eggs present
Nest 2 2 PM eggs present
Nest 3 2 PM eggs present
Nest 4 2 PM eggs present

June 2, 2004

Nest 1 1 PM egg/1PM nestling present (hatching day)
Nest 2 2 PM nestlings present (1 day old)
Nest 3 2 PM eggs present
Nest 4 2 PM eggs present

June 3, 2004

Nest 1 2 PM nestlings present
Nest 2 2 PM nestlings present
Nest 3 2 PM eggs present
Nest 4 2 PM eggs present

June 5, 2004

Nest 1 2 PM nestlings present
Nest 2 2 PM nestlings present
Nest 3 2 PM nestlings present (hatching day or 1 day old)
Nest 4 parent on nest (not checked)

June 6, 2004

Nest 1 2 PM nestlings present
Nest 2 2 PM nestlings present
Nest 3 2 PM nestlings present
Nest 4 1 PM egg/1PM nestling present (hatching day)

June 7, 2004

Nest 1 2 PM nestlings present (4-5 days old)
Nest 2 2 PM nestlings present (5-6 days old)
Nest 3 2 PM nestlings present (2-3 days old)
Nest 4 2 PM nestlings present (0-1 days old)

June 9, 2004

Nest 1 2 PM nestlings present (6-7 days old; 32, 34 grams)
Nest 2 2 PM nestlings present (7-8 days old; 37, 37 grams)
Nest 3 2 PM nestlings present (4-5 days old; 20, 20 grams)
Nest 4 2 PM nestlings present (2-3 days old; 13, 18 grams)

June 14, 2004

Nest 1 2 PM nestlings *banded (11-12 d.o.; K363=46.0 g., K362=53.0 g.)
Nest 2 2 PM nestlings **banded (12-13 d.o.; K360=54.0 g., K361=59.5 g.)
Nest 3 2 PM nestlings present (9-10 days old; 20, 20 grams)
Nest 4 2 PM nestlings present (7-8 days old; 13, 18 grams)
*Color band is yellow K363 on left leg; federal band on right leg is 1781-59963
*Color band is yellow K362 on left leg; federal band on right leg is 1781-59962
**Color band is yellow K360 on left leg; federal band on right leg is 1781-59960
**Color band is yellow K361 on left leg; federal band on right leg is 1781-59961

June 16, 2004

Nest 1 2 PM nestlings present (13-14 d.o.; K363=49.0 g., K362=53.5 g.)
Nest 2 2 PM nestlings present (14-15 d.o.; K360=54.0 g., K361=59.5 g.)
Nest 3 2 PM nestlings *banded (11-12 d.o.; K366=46.0 g., K367=49.5 g.)
Nest 4 2 PM nestlings **banded (9-10 d.o.; K364=50.5 g., K365=45.0 g.)
*Color band is yellow K366 on left leg; federal band on right leg is 1781-59966
*Color band is yellow K367 on left leg; federal band on right leg is 1781-59967
**Color band is yellow K364 on left leg; federal band on right leg is 1781-59964
**Color band is yellow K365 on left leg; federal band on right leg is 1781-59965

June 20, 2004

Nest 1 2 PM nestlings (17-18 d.o.; K363=51.0 g., K362=52.0 g.)
Nest 2 2 PM nestlings (18-19 d.o.; K360=50.0 g., K361=43.0 g.)
Nest 3 2 PM nestlings (15-16 d.o.; K366=58.5 g., K367=57.0 g.)
Nest 4 2 PM nestlings (13-14 d.o.; K364=59.0 g., K365=55.0 g.)

June 22, 2004

Nest 1 2 PM nestlings (19-20 d.o.; K363=52.0 g., K362=53.5 g.)
Nest 2 2 PM nestlings (20-21 d.o.; K360=48.0 g., K361=45.5 g.)
Nest 3 2 PM nestlings (17-18 d.o.; K366=56.0 g., K367=56.0 g.)
Nest 4 2 PM nestlings (15-16 d.o.; K364=55.5 g., K365=53.0 g.)

June 24, 2004

Nest 1 2 PM nestlings (21-22 d.o.; K363=50.0 g., K362=52.5 g.)
Nest 2 2 PM nestlings (22-23 d.o.; K360=46.0 g., K361=43.5 g.)
Nest 3 2 PM nestlings (19-20 d.o.; K366=55.0 g., K367=52.5 g.)
Nest 4 2 PM nestlings (17-18 d.o.; K364=54.5 g., K365=52.5 g.)

June 26, 2004

Nest 1 2 PM nestlings (23-24 d.o.; K363=47.0 g., K362=50.5 g.)
Nest 2 2 PM nestlings (24-25 d.o.; K360=44.0 g., K361=41.5 g.)
Nest 3 2 PM nestlings (21-22 d.o.; K366=49.0 g., K367=46.0 g.)
Nest 4 2 PM nestlings (19-20 d.o.; K364=47.0 g., K365=46.0 g.)

June 28, 2004

Nest 1 2 PM nestlings (25-26 d.o.; K363=42.5 g., K362=47.0 g.)
Nest 2 2 PM nestlings (26-27 d.o.; K360=41.0 g., K361=40.5 g.)
Nest 3 2 PM nestlings (23-24 d.o.; K366=47.5 g., K367=47.5 g.)
Nest 4 2 PM nestlings (21-22 d.o.; K364=43.5 g., K365=41.5 g.)

On the morning of June 28, 2004, the Tree Swallow parents were still in attendance and feeding, and the nestlings appeared to be alert and healthy, but were continuing to lose weight, and several had dropped significantly below the normal weight for a nestlings of their age. Also, the weather had also turned cool and rainy. Therefore, the decision was made to foster all eight nestlings into active Purple Martin nests.

They were placed into nests at the Saxon Golf Course colony, 8.7 miles NNW in Sarver, Butler Co., PA, for two reasons. First, the nesting adults at this colony have been trained for several years to accept supplemental feedings during periods of foul weather. The parents accept large mealworms placed on elevated trays near the housing, and readily feed these mealworms to their nestlings as well. Second, the Saxon site is only eight miles from the permittee’s residence, allowing for close monitoring of the fostered nestlings.

The eight nestlings were placed into three separate nests with martin nestlings of similar age. All three nests had ASY (adult) Purple Martin parents, and no nests exceeded a total of seven nestlings. The additional nestling load could not have been imposed if it were not for the supplemental feeding, which provided the parents an unlimited and readily-available food source. The foster-parents were repeatedly observed carrying mealworms to all three compartments containing the transferred nestlings for the remainder of the day.

The nestling’s weight was also a factor in choosing a foster nest. For example, one of the 26-27 d.o. nestlings (K360) weighed only 41.0 grams, so it was placed into a nest containing 23 d.o. young because it would require at least 3-4 days to regain the weight it needed to fledge successfully. (Martins normally fledge at about 50 grams on day 28.)

The mealworms were made available every day for the next two weeks, well into the post-fledging period. Handling of nestlings over 20 days old is not recommended and must be done with great care to insure against premature fledging. Some of the techniques used when doing nest checks with older nestlings are found in the article Doing nest checks with older nestlings. Fledging-age nestlings had to be kept in a deep 5 gallon bucket after being weighed. Entrance holes were blocked with a rag after the nestlings were placed back into the nest compartment, and the rag was pulled with a string after raising the house and waiting five minutes to allow the nestlings to settle down.
Fostered into compartment 1 (5 nestlings, 22 d.o.)
K361 40.5 g (26-27 d.o.)
K364 44.0 g (21-22 d.o.)

Fostered into compartment 12 (3 nestlings, 25 d.o.)
K362 46.5 g (25-26 d.o.)
K366 47.5 g (23-24 d.o.)OLYMPUS DIGITAL CAMERA
K367 48.0 g (23-24 d.o.)

Fostered into compartment 27 (4 nestlings, 23 d.o.)
K360 41.0 g (26-27 d.o.)
K363 42.5 g (25-26 d.o.)
K365 41.5 g (21-22 d.o.)

June 29, 2004

Large mealworms were again offered in the morning, and continued to be accepted by adults and fed to nestling until mid-day. Weather improved dramatically around 1:00 P.M., and martins no longer took any mealworms from the feeding trays. They were observed repeatedly delivering large insects to nestlings.

Did nest check on WH-12 at Saxon Golf Course and weighed fostered nestlings
K362 48.0 g (26-27 d.o.) +1.5 g
K366 50.0 g (24-25 d.o.) +2.5 g
K367 49.5 g (24-25 d.o.) +1.5 g

June 30, 2004

I again offered mealworms in the morning and afternoon, but no martins landed to take them. Plenty of natural food was being found. Martins were seen arriving with large insects throughout the day.

Did nest check on WH-1 at Saxon Golf Course and weighed fostered nestlings
K361 44.0 g (27-28 d.o.) +3.5 g
K364 47.0 g (22-23 d.o.) +3.0 g

July 1, 2004

Martins observed delivering food throughout the day.

Did nest check on WH-27 at Saxon Golf Course and weighed fostered nestlings.
K360 46.5 g (29-30 d.o.) +5.5 g
K363 48.5 g (28-29 d.o.) +6.0 g
K365 45.5 g. (24-25 d.o.) +4.0 g

Observations continued for the next 6 days, and although it was deemed unsafe to weigh the nestlings after this date (for fear of causing premature fledging), all the fostered martins had gained weight since being fostered into the Purple Martin nests and had apparently been adopted by the Purple Martin parents. Visual observations showed that the nestlings were alert and looked healthy.

July 2, 2004

The nestlings fledged from WH-12.

July 3, 2004

The nestlings fledged from WH-27.

July 5, 2004

The nestlings fledged from WH-1.

July 2-12, 2004

Multiple visual observations of the fledgings (mainly when the parents brought them back to the natal cavity in the evening) indicated that they appeared to be very healthy.

July 27, 2004

One of the fostered martins was seen alive and well in Erie, PA, approximately 100 miles north of the fledging site! Yellow K360 was spotted sitting on the electrical overhead wires of Beach 11 at Presque Isle State Park, a long-established pre-migratory roost, at 7:20 P.M. along with 200 other Purple Martins.  The band was read with a high power Swarovski spotting scope. This martin was again spotted as an SY-F at the Saxon colony on June 7, 2005, and was one of sixteen martins displaced as part of a forced dispersal project.


The Tree Swallow parents did successfully hatch the Purple Martin eggs and raise the martin nestlings until about 20-22 days of age, after which the nestlings began to lose weight at the rate of several grams per day, even though the Tree Swallows continued to feed the nestlings. When it became apparent that the nestlings were dropping below weights at which they could fledge successfully, they were fostered into active Purple Martin nests where they regained weight and fledged successfully.


Purple Martin nestlings normally fledge at a weight of about 45-55 grams on day 27-29. While it is normal for Purple martin nestlings, under normal conditions, to lose up to 1 gram per day in the final week before fledging, the fostered martins in this project lost several grams per day after 20-22 days of age while under the care of the Tree Swallow parents, and dropped significantly below the weight that would allow them to fledge successfully.

Why did the nestlings fail to maintain the weight needed to fledge successfully? There are at least two possible explanations: either the quantity or the quality of the food delivered by the Tree Swallow parents was (or became) inadequate. I believe the first possibility can be eliminated. If the quality of the food being delivered was inadequate, it is unlikely that the nestlings would ever have reached the normal maximum weight of approximately 55 grams at day 18 – something they all achieved.

Although the Tree Swallow parents continued to feed the martin nestlings until they were removed from the nest, it appears that the quantity of food delivered became inadequate. This reduction in the quantity of food delivered to the martin nestlings by their Tree Swallow parents was either instinctive or weather-related. There were below average temperatures on nine of the last ten days of the project, and while this may have impacted the nestlings’ weights, I do not believe it was significant enough to be the primary reason for the rather drastic weight loss since temperatures averaged only about six degrees below normal and there was precipitation on only two of those days.

Additionally, there were seven other Tree Swallow nests at the same site where the Tree Swallow nestlings fledged successfully. Plus, the Tree Swallow foster parents in this project were early-nesting adult pairs, most likely with prior breeding experience. The breeding site is located in ideal habitat on waterfront property where there are usually flying insects available even in cool, rainy weather.

This leaves the possibility that the Tree Swallow foster parents voluntarily withheld or failed to deliver the required amount of food to the nestlings. I believe this is the most probable explanation. Tree Swallow nestlings normally fledge at 18-22 days.  Five of the seven other Tree Swallow pairs breeding at the project site fledged their young in an average of 18 days. I believe that the Tree Swallow foster parents began to deliver less food to the martin nestlings shortly after the period of time at which Tree Swallow nestlings would normally have fledged.


The breeding biology of these two species appears to be too different to allow for successful cross fostering of Purple Martin nestlings by Tree Swallows. Tree Swallows nestlings typically fledge in 17-19 days, whereas Purple Martins require 27-29 days. Tree Swallow pairs will successfully hatch Purple Martin eggs and feed the nestlings adequately until approximately 20-22 days of age. After this point, the Tree Swallow parents apparently fail to deliver the required amount of food, causing the martin nestlings to lose weight rapidly and fall below the 45-55 grams needed to fledge successfully.

Post Script 12/08/05

This project took place over one year ago. In the summer of 2005, a pair of subadult Purple Martins did nest at Lock 4 as a result of a forced dispersal experiment, and I followed their breeding attempt closely. The parents successfully fledged all four young and were still bringing them back to the natal cavity 14 days after fledging. They were seen still feeding their young 10 days post-fledging. My purpose in pointing this out is that the post-fledging care period in martins is lengthy. Even if the Tree Swallow foster parents had managed to fledge the martin nestlings, I doubt they would have been able to provide them with the rather extensive post-fledging care that would be required for survival.