We examined the fitnotion of rictal bristles on Willow Flycatchers by testing whether their removal or the placement of small pieces of transparent tape on them would reduce the ability of captive birds to capture live house flies. Neither operation adversely affected the birds' ability to capture prey, indicating that the rictal bristles do not aid in prey capture. Other experiments with bird specimens placed in a wind tunnel demonstrated that rictal bristles may protect the eyes from food items the bird is trying to capture. Particles released in front of the bird's open mouth and blown back towards its head struck an eye more frequently after the rictal bristles had been removed. Much speculation has centered on the function of rietal bristles, stiff whisker-like feathers arising from the rictus and the margin of the leathered skin behind the horny covering of the upper mandible. Some authors have suggested that rictal bristles may serve as mechanoreceptors (Wallace 1955, Pettingill 1970) or chemoreceptors (Jany 1955). Dyer (1976) proposed that in birds that feed on large and potentially dangerous insects, such as hees and acridid grasshoppers, these bristles protect the eyes from the preys' legs and stingers. Other authors, noting that rictal bristles are common among insectivorous birds, have proposed their utility in capturing prey, possibly serving as an insect net or funnel (Welty 1962, Van Tyne and Berger 1976). Even if true, this cannot be their sole filnction, however, because rictal bristles are also found on many noninsectiwlrous birds (Stettenheim 1973). Evaluation of these hypotheses has been hampered by the lack of experimental data. Lederer (1972), however, used high speed photography to show that the Great Crested Flycatcher (Myiarchus crinitus), Eastern Phoebe (Sayornis phoebe), and Eastern Wood Pewee (Contopus virens) usually catch flesh flies (Sarcophaga bullata) with their bill tips. From this, he suggested that rictal bristles perform no filnction in prey capture. However, when a bird misses its prey with its bill tip, rictal bristles may still serve to deflect the missed prey into the bill. They may also allow the bird a second strike opportunity by ilnpeding the insect's escape or serving as mechanoreceptors that signal the missed prey's new location. Another hypothesis, owing to the rictal bristles' location between the eyes and mouth, is that these feathers may prevent escaping prey or parts of captured prey from hitting the bird's face and eyes. We examined whether rictal bristles aid the Willow Flycatcher (Empidonax traillii) in prey capture or help protect the eyes. In our experiments, we manipnlated the rictal bristles on specimens and observed whether there was any change in the birds' ability to capture prey or in the numher of objects striking the eyes.

METHODS Ten Willow Flycatchers were captured with mist-nets during July and August 1977 and 1978 at the Hudson Biological Reserve near Pullman, Washington. Each bird was housed separately in a 1.5 x 1.8 x 2.5 m screened cage and fed ad libitum a variety of live in- sects. Lighting dnring all tests was by overhead flu- orescent and incandescent lights, which provided ap- proximately 250 lux. All tests were conducted 5 to 15 days aier a bird's capture. Each bird was observed at a distance of 2 to 3 m as it captured live house flies (Musca domestica), first during a control period and then during one or two experimental periods. The first experimental period lasted one day while the control and the second ex- perimental period lasted two to three days. To guard against the possibility that capturing and handling the birds during the experimental procedures might have been an important variable, all birds were similarly captured and handled prior to the control period. We assumed that small objects placed on the tips of the rictal bristles to increase their length, weight, and wind resistance might temporarily disrupt their filnc- tioning if the bristles aid in prey capture by serving as mechanoreeeptors. To test this, we removed each bird tom its cage at the beginning of the first experimental period and folded a 2 x 2 mm piece of transparent ad- hesive tape over the end of a bristle. Two randomly selected rietal bristles on each side of each bird's head TABLE 1. Capture success ratio of flycatchers when their rictal bristles were left intact, taped or removed. Control Bristles taped Bristles removed Number Success Number Stlecess X Number Success X Specimen attempts ratio attempts ratio (vs. control) attempts ratio (vs. control) Sally-gleaning attempts A 40 0.85 41 0.80 0.06 47 0.77 0.51 B 35 0.89 19 0.89 0.14 45 0.82 0.22 C 60 0.82 43 0.79 0.01 55 0.91 1.35 D 50 0.56 -- -- -- 48 0.77 3.98* E 75 0.87 -- -- -- 91 0.84 0.12 Aerial hawking attempts A 24 0.92 21 0.81 0.38 24 0.92 0.27 B 18 0.89 10 0.90 0.30 23 0.78 0.23 C 51 0.78 49 0.78 0.002 44 0.77 0.01 D 72 0.65 -- -- -- 25 0.92 5.33* E 39 0.69 -- -- -- 38 0.89 3.64 * P < 0.05. TABLE 2. Effect of taping or removing the rictal bristles on the ratio of capture attempts with multiple bill snaps to those with a single snap. Control Bristles taped Bristles removed Number Double Nnmber Donble X Number Double X 2 Specimen attempts snap ratio attempts snap ratio (vs. control) attempts smp ratio (vs. control) A 61 0.05 65 0.09 0.35 75 0.05 0.08 B 48 0.07 18 0.06 0.22 63 0.09 0.08 C 79 0.08 62 0.07 0.10 68 0.10 0.08 were taped, after which the bird was returned to its cage. If rictal bristles aid in catching prey, removing or cutting them off should also reduce a flycatcher's abil- ity to catch insects. We tested this in the second ex- periment by cutting off the bristles at the skin line so that they no longer projected above the contour feath- ers. Three birds were tested during the first experi- mental period, and these same individuals plus two others were then tested during the second experimen- tal period. During the control and experimental periods, we re- corded the number of attempts each bird made to cap- ture flies by either aerial hawking or sally-gleaning and the proportion of successful attempts (capture success ratio). The ratio is a conservative estimate of capture success. House flies were used as prey because their small size, speed and maneuverability tested the fly- catchers' capturing ability, and emphasized any de- crease in this ability. Each bird's capture success ratio during each experimental period was compared to the ratio obtained during the control period using a 2 x 2 contingency table corrected for continuity. A flycatcher sometimes made an audible snap as it rapidly closed its bill during a capture attempt. It usu- ally snapped only once during a capture attempt but occasionally made two or more snaps as it struck re- peatedly at the same fly. The ratio of multiple snaps to single snaps was determined for three of the birds dur- ing each of the control and experimental periods, and then compared using a 2 x 2 contingency table cor- rected for continnity. To test whether rictal bristles shield the eyes from items a bird is trying to eat we placed six preserved flycatchers, fixed with their mouths opened 1-2 era, in a wind tunnel and then sinmltaneously released 10 to 20 pieces of wood or plastic 0.2 to 2.0 mm in length from a fixed position 1-3 em in front of each bird's mouth. The birds were held in position by a rod that exited from the rear of the bird's body and connected to the back of the wind tnnneh The wind speed was 5 m/s. The directionality of the rictal bristles was not changed or controlled on any specimens. To measure the frequency of items striking the eyes, we put ad- hesive discs over the eyes so that any striking ohjeets would adhere to them. The eye disc protruded 1-2 em above the surthee of the f'aee. The number of adhering particles was then counted. Each specimen was tested in this manner 10 times and the results totalled for each specimen. The results for each specimen were then statistically compared using the ehi-sqnare test to de- termine whether the number of particles striking each side of the bird's face significantly differed tom one another. The rietal bristles on one side of the specimen were then cut off and the experiment repeated. These data were again totalled for each specimen and statis- tically analyzed. RESULTS Small pieces of tape on the ends of the rictal bristles apparently did not disrupt the birds' ability to capture flies because there was little change in capture success ratio for aerial hawking or sally-gleaning attempts by any of the birds (Table 1). Removing the bristles also did not reduce the ability of any of the birds to capture flies (Table 1). In fact, one bird had a significantly higher capture success ratio after its rietal bristles were cut off. The ratio of capture attempts in which the birds nmde multiple snaps to those in which it made only a single snap did not change significantly during the control and experimental periods for any of the birds (Table 2). This indicates that neither cutting off nor taping rietal bristles altered the birds' ability to rapidly snap again at missed prey. Before the bristles were removed, the number of par- tides adhering to the adhesive discs overlying the birds' eyes differed significantly in only one of the six specimens (Table 3). After the bristles were cut one side of each bird's face, however, particles hit the discs on that side significantly more often in all six specimens (Table 3). TABLE 3. Ratio of particles adhering to eye discs before and after the rictal bristles were removed i}om one side of the bird's head (experimental side). Before bristles removed After bristles removed Ratio Ratio Nmnber experimental to Ntlmber experimental to Specimen particles control side X 2 particles control side AA 251 1.11 0.67 362 1.81 29.9'* BB 98 0.92 0.16 123 1.73 8.85** CC 143 1.13 0.57 166 1.96 17.57'* DD 141 1.52 5.96* 136 1.61 7.52** EE 181 1.03 0.05 275 1.55 12.66'* FF 183 0.87 0.92 194 1.49 7.44** Total 997 1.08 1.37 1,256 1.68 81.53'* * P 0.05. ** P 0.01. DISCUSSION Lederer (1972) shoved that the Great Crested Fly- catcher, Eastern Phoebe and Eastern Wood Pevee capture flesh flies with their bill tips so that the rictal bristles could not [nction to hmnel insects into the mouth. Our results, using a different species of fly- catcher, different prey species, and different experi- mental conditions, agree with his findings that rictal bristles do not function as insect funnels and further failed to support any of the other prey-capture hypoth- eses. Taping several ricta] bristles to increase their length, weight and wind resistance had no ef{ct on prey-capture success. While this suggests that rictal bristles do not aid in prey captnre by serving as mechanoreceptors, it is possible that the taping simply did not disrupt their functioning as mechanoreceptors. Hovever, regardless of hov rictal bristles function, their removal would be expected to reduce the pro- portion of successful capture attempts if they aided in prey capture. No significant reduction was evident for any of the experimental birds. Likewise, if the rictal bristles serve to increase the birds' ability to restrike rapidly at missed prey, removing them should also have decreased the frequency of multiple snaps, but no decrease was evident. Rictal bristles might prevent food items from striking the eyes, as might happen when the prey is missed or breaks apart on capture. The results of the wind tunnel experiment support this possibility. The mean distance between individual bristles in 40 specimens of Willow Flycatcher that we examined was 1 min. Hence, rictal bristles in this species probably are most efficient at stopping particles larger than 1 mm in diameter. These bristles may also help keep smaller particles out of the eyes by diverting the air flow away from the face. The bristles' location, around the rictus instead of the eye, seems tinusual if a primary fimction is to pro- tect the eye. Some birds, indeed, do have bristles around the eyes that apparently serve as eyelashes (Stettenheim 1973). Why then should flycatchers and other birds have them (inly on the rictus? Probably in any bird the number, size, and location of bristles to protect the eyes is determined by evolutionary pres- sure restilting from the conflicting needs to protect the eyes and yet obtain an adequate field of vision. For birds that are likely to be hit by particles during normal flight or by particles coming ti'om any direction, bris- tles immediately around the eyes are the only vay to protect the eyes, In Willov Flycatchers, hovever, dan- ger to their eyes probably arises chiefly from attempts to capture prey that can move unpredictably in escap- ing or are likely to break apart on capture. The location of bristles on the rictus should block any prey parts coming from the bill and still allov a clear forward field of view and in all other directions except dovn- vard. For instance, when chasing prey, the rietal bris- tles start to block the bird's view only after the bill tip reaches the insect. Besides protecting the eyes, the rictal bristles may protect the feathers on the face from becoming soiled by {Bod items. Possibly the short bristles along the lov- er bills of many flycatchers serve a similar filnction. While our results {hiled to support any of the prey- capture hypotheses, they do not and cannot prove the null hypothesis--that rictal bristles do not aid in prey- capture. Furthermore, our findings do not exclude the possibility that rictal bristles aid in prey capture in oth- er birds or even in other tyrant flycatchers, oving to the latters' diverse food habits and capture techniques (Fitzpatrick 1980). We can tinderstand the filnctions of rictal bristles only after experimental studies have been conducted on many bird species using different types of prey. ACKNOWLEDGMENTS We thank D. O. Conover for her help in conducting these experiments and preparing the manuscript. Crit- icism of an earlier draft by D. Aylor and R. A. Conover improved the vork. This research was funded by grants from the Frank M. Chapman Memorial Fund of the American Museum of Natural History and the So- ciety of Sigma Xi. LITERATURE CITED DYER, M. 1976. On the function of rictal bristles, vith reference to Nigerian birds. Nigerian Ornithol. Soc. Bull. 12:45-48. FITZPATRICK, S. W. 1980. Foraging behavior of neo- tropical tyrant flycatchers. Condor 82:43-57. J^N3/4, E. 1955. Schnabelborsten als Triger des junsten Sinnesorgans der Vbgel. Proc. XI Int. Ornithol. Congr. (1954):275-278. LEDERER, R. J, 1972. The role of avian rietal bristles. Wilson Bull. 84:193-197, PETTINGILL, O. S., JR. 1970. Ornithology in laboratory and field. Fourth ed. Burgess Publ., Minneapolis, MN. STETTENHEIM, P. R. 1973, The bristles of birds. Liv- ing Bird 12:201-234. VAN TYNE, J., AND A. J. BERGER. 1976. Fundamentals of ornithology. Second ed. John Wiley and Sons, Nev York. W^L[^CE, G. J. 1955. An introduction to ornithology. Macmillan, New York. WELT3/4, J. C. 1962. The life of birds. W. B. Saunders, Philadelphia. Department of Zoology, Washington State Universittt, Pullman, Washington 99164. Present address of first author: Department of Ecologtt and Climatology, The Connecticut Agricultural Experiment Station, Box 1106, New Haven, Connecticut 06504. Accepted for publication 2 May 1980.