Owl Guards:  Preferences and Breeding Success at an Established Colony

Ken Kostka
Purple Martin Preservation Alliance
http://www.purple-martin.org
Pittsburgh, PA
9/21/05

 
Introduction:
Placing owl guards on martin housing has become a commonly practiced management technique. These guards help to prevent owls and other large raptors from getting close to the entrance hole of Purple Martin nesting cavities and extracting the martins.  Some landlords are reluctant to install owl guards in the belief that the barriers might discourage martins by impeding their entrance and exit of the entrance hole area. This study was intended to determine whether martins are more likely to choose compartments with or without owl guards in T-14 martin houses.

 
Methods:
  I have used owl guards for the past two seasons at the Saxon Golf Course colony near Saxonburg, PA. These guards consist of ˝” square strips of wood placed 3 inches apart and attached vertically to the porches of each section of the T-14’s.  (see photo) The colony had 100% occupancy (28 pairs) in 2004 when all compartments had owl guards.  I wanted to see how frequently martins would choose to nest in compartments without guards if given the choice, and to see which compartments would be chosen first - guarded or unguarded.  Therefore, I removed guards from half (14 compartments) of  the two T-14’s (28 compartments total) prior to their arrival for the 2005 season. There were an equal number of guarded and unguarded compartments at all heights and facing all compass directions. Both houses were wooden T-14’s, each containing 14 identical compartments. All entrance holes were the Conley Excluder. 

 Two factors could have biased the results: First, there is a high degree of site fidelity among Purple Martins and many returning martins had either nested in or had fledged from the owl-guarded compartments in 2004.  Since they had become accustomed to them, they may have come to ignore or prefer them, even though they may have initially “disliked” them and considered them an obstacle.  Second, there is also a significant amount of cavity fidelity among Purple Martins (Kostka, unpub. data), and returning breeders may have chosen the same cavity regardless of whether it still had guards.  For these reasons, I do not intend to draw any hard conclusions with regard to true preference for or against owl guards in cavity-hunting martins. True preference results could only be established by testing the preference of first year breeders, and would have to take into account whether the test subjects fledged from guarded and unguarded cavities, since this could affect their choice as breeding adults.  My main objective was to determine if the martins, overall, considered the guards to be a significant obstacle or inconvenience.  If they did, then it seems logical to conclude that more martins would tend to choose the unguarded cavities first.  

     I used clutch initiation date to determine which compartments were chosen first and presumably preferred.  Only 20 of 28 compartments were used in this experiment because I blocked the last eight compartments claimed as part of a Forced Dispersal Project that was undertaken in early June. There is no CID (clutch initiation date) for these compartments since no eggs were laid before they were closed, but all had nearly completed nests.

 Data: [terms for data table: WH-X = compartment #; 0X/XX/05 = clutch initiation date; OG= Owl Guards present; NO = No Owl guards present; XF = # of young fledged)

WH-02            05/13/05       OG      3F
WH-08            05/13/05       OG      5F
WH-18            05/14/05       OG      5F
WH-14            05/15/05       NO      6F
WH-20            05/16/05       OG      4F
WH-12            05/17/05       NO      3F
WH-11            05/25/05       NO      4F
WH-07            05/25/05       NO      1F
WH-28            05/26/05       OG      6F
WH-19            05/29/05       OG      4F
WH-01            05/31/05       OG      2F
WH-25            05/31/05       OG      4F
WH-27            06/01/05       OG      4F
WH-24            06/03/05       NO      4F
WH-10            06/04/05       OG      4F
WH-13            06/05/05       NO      0F
WH-06            06/05/05       NO      2F
WH-03            06/06/05       OG      0F
WH-21            06/06/05       OG      4F
WH-22            06/07/05       NO      2F

 Results:  Four of the first five compartments (80%) and nine of the first 13 compartments (69%) claimed had owl guards.  12 of the 14 guarded compartments (86%) and 8 of the 14 unguarded compartments (57%) were among the first 20 chosen.  A total of 45 nestlings fledged from the guarded compartments, and 22 nestlings fledged from unguarded compartments. An average of 3.75 nestlings fledged from guarded compartments and 2.75 young fledged from unguarded compartments.

 Discussion:  Since most of the compartments claimed, both initially and overall, had owl guards, it is unlikely that the martins at this colony considered the guards a highly significant obstruction. Furthermore, the pairs nesting in guarded compartments fledged an average of one extra nestling.   I did remove starling nests (and trapped the starlings) from WH-21 and WH-22 on 04/18/08; these were unguarded compartments and certainly prevented martins from claiming them.  But this occurred early in the season, and I doubt it significantly affected the results. I lost 13 nestlings (found dead) that were 20+ days old, and  found a large owl feather under a tree about 150 ft. from the housing, making me suspect I was experiencing owl predation (owls capturing one or both parents and causing the young to starve); therefore, I eventually replaced all the owl guards in mid-July, with many nestlings still unfledged and many fledglings still returning to the box at night.  This also may have affected the compared reproductive success results.