Ageing Northern Saw-whet Owls (Aegolius acadicus) from Remigial Molt Patterns

Migrant Northern Saw-whet Owls (Aegolius acadicus) are common in much of the United States and Canada, with concentrations around the Great Lakes (Rasmussen et al. 2008). Adult and juvenile Northern Saw-whet Owls can be distinguished by the presence or absence of multiple generations of flight feathers (Evans and Rosenfield 1987, Pyle 1997a). Adult Northern Saw-whet Owls exhibit a pattern of remigial molt with up to three generations of feathers (Evans and Rosenfield 1987, Pyle 1997a). Remigial molt typically occurs from July to September, and is almost always incomplete (Evans and Rosenfield 1987, Slack 1992, Rasmussen et al. 2008). Therefore, we used molt patterns of known-age birds to determine the most prevalent molt pattern for different age classes.

Previous molt studies have been limited by both their small sample size and by their analysis of only one wing per bird, which can be misleading because molt is often asymmetrical in owls (Evans and Rosenfield 1987, Slack 1992). Although Evans and Rosenfield's findings have essentially been adopted and used in many studies, they have not been tested or modified to increase age-classification rates. Here we describe two basic molt patterns that can be used to distinguish SY and TY Northern Saw-whet Owls, and we quantify the effectiveness of these criteria to correctly age owls.

Methods

We pooled recapture data from three Midwestern banding stations: Linwood Springs Research Station in Wisconsin (44°28.233′N, 89°38.267′W), Whitefish Point Bird Observatory in Michigan (46°46.267′N, 84°57.250′W), and Hawk Ridge Bird Observatory in Minnesota (46°50.750′N, 92°2.000′W). Banders at all three stations captured Northern Saw-whet Owls during either fall or spring migration. Additionally, Whitefish Point Bird Observatory captured Northern Saw-whet Owls during the summer. All three stations used Northern Saw-whet Owl callers that broadcast the male “toot” call, which greatly increases capture rates (Erdman and Brinker 1997). Additionally, Whitefish Point Bird Observatory used a caller broadcasting a female “tssst” call (Neri et al. 2018). The Wisconsin site used data collected during netting operations from 1988 through 2013, the Michigan site from 1998 through 2013, and the Minnesota site from 1973 through 2008.

Owls were captured in mist-nets, banded, and measured according to each station's protocol. Banders recorded the molt patterns of both wings on every owl, due to possible variability between wings (Slack 1992). Only owls originally banded as HY birds and recaptured in a subsequent molt cycle, i.e., known-age owls, were used in this analysis. Banders determined molt patterns by looking at the dorsal side of the wing under incandescent or fluorescent lighting. Ultraviolet lighting on the ventral side of the wing (Weidensaul et al. 2011) was used only intermittently and as a secondary technique to verify patterns seen under normal lighting on the dorsal side. We classified feathers as new, old, and very old, referring to the three different generations that can be present in a given plumage. New flight feathers were the most recent generation, with the darkest color tones and showing the least amount of wear. Old flight feathers, which were 1 yr older than new feathers, had mid-range color tones due to fading and wear. Very old flight feathers (hereafter, VO), those 2 yr older than new feathers, exhibited the most color loss due to fading and showed the most wear. We abbreviate the primaries 1 through 10 as P1, P2, P3, etc., and secondaries 1 through 12 as S1, S2, S3, etc.

Results

Of the more than 50,000 Northern Saw-whet Owls banded in all years at the three stations, 722 were recaptured after at least one molt cycle (i.e., July and August) had occurred. Of those recaptured birds, only two (0.28%; 95% CI: 0.0–0.7%) showed a single generation of feathers on both wings. This indicates that birds with only one generation of remiges were >99% likely to be a hatch-year bird (HY). Of the 722 recaptured birds, 286 were initially banded while exhibiting no visible molt and were classified as HY. All others were omitted from our analysis of molt patterns. Of these 286, 155 (54%) were second-year birds (SY), 72 (25%) were third-year (TY), 39 (14%) were fourth-year (4Y), 16 (6%) were fifth-year (5Y), three (1%) were sixth-year (6Y), and one (0.3%) was eighth-year (8Y).

Second-year Molt Patterns. Of the 155 SY birds, 128 (83%) had the previously published (Evans and Rosefield 1987) SY molt pattern, which consists of a solid block of old feathers in the middle of at least one wing. However, we considered both wings and found that SY patterns can be more varied. Based on our observations, we established the following criteria. An SY pattern consists of a solid block of old feathers on both wings, with or without one skip on one or both wings (Fig. 1). A skip is defined as one new feather within a block of old feathers. This old feather block always includes at least S1 and P1 in both wings, and consists of at least three old feathers. It can be as small as P2–S1 or as large as P8–S10 and is not necessarily symmetrical between wings. One skip may be present in either the primaries or secondaries but not both, and no VO feathers are present. Pyle (1997b) suggested HY Northern Saw-whet Owls may have narrower outer primaries and more pointed center rectrices than adults, but we found no differences in the shape or pattern of juvenile and adult remiges. One hundred and fifty-six birds matched our criteria for the SY age class (Fig. 1), of which 155 were known SY (99.4%; 95% CI: 98.1–100.61%). One 4Y bird incorrectly fit our SY molt pattern criteria.

View Figure

• Download as Powerpoint
• Download Figure

Third-year Molt Patterns. Evans and Rosenfield (1987) suggested a TY age class, but had a very small sample size. Based on Evans and Rosenfield's description of two outer blocks of old feathers on one wing, only seven birds from our sample fit into this class. With our much larger sample size and our examination of both wings, we found greater variation. Based on these observations, we established the following criteria. A TY pattern consists of two blocks of old feathers on the wing, with or without a VO block positioned discretely between the two old blocks. A block is defined as >1 feather and the pattern is not necessarily symmetrical between wings. Among the primaries, the old block of feathers never includes P1, P2 or P3. Among the secondaries, the old block of feathers never includes S1 or S2. If a VO block is present, there are no skips in that block and it is located at the center of the wing. This block is discrete and never includes feathers more distal than P4 or more proximal than S4. All old and VO blocks are discrete with new feathers in between. The number of remiges in each block can vary greatly among individuals in both SY and TY patterns. Of the 72 TY birds in our sample, 52 (72.2%; 95% CI: 61.9–82.6%) correctly matched our TY criteria, and 20 (28%) did not fit and were classified as ASY. From our sample of 286 total birds, 54 fit the criteria for the TY pattern; 52 of these were correctly classified as TY birds and the remaining two were 5Y birds (Fig. 2).

View Figure

• Download as Powerpoint
• Download Figure

Discussion

It is unusual that a species the size of a Northern Saw-whet Owl goes through an incomplete adult prebasic molt. We found that feathers retained for two or three years were the most centrally located remiges on the wing. Because molting feathers requires a substantial amount of energy (Rohwer et al. 2009), we hypothesize those remiges receive the least amount of wear and not molting them as often allows the owl to conserve energy.

Other studies of molt in large raptors have used sequencing of specific flight feathers to determine age (Solheim 2012). However, in the genus Aegolius, we found that analyzing the overall pattern on both wings was more useful for ageing, due to the substantial individual variation in molt patterns.

Of the total sample of 286 known-age birds, 210 (73.4%) matched one of our two proposed molt patterns. Of those 210, 207 (98.6%) were correctly aged (Table 1). Given our large sample size, we are confident that our criteria for the SY and TY age classes are extremely accurate for ageing Northern Saw-whet Owls. Our TY criteria did fail to distinguish all the TY birds in our known-age sample; 20 of the 72 TY (27.7%) birds did not match this pattern and were classified as ASY. However, this pattern correctly assigned a TY age class to 52 of 54 (96.3%) birds in our sample. We caution that our TY criteria are accurate when a bird fits the pattern, but in our work they distinguished only 72 % of the TY birds and classified the remaining TY birds as ASY.

Table 1. Age classes assigned for all 286 Northern Saw-whet Owls using the criteria for each molt pattern. Age class abbreviations: SY, second-year; TY, third-year; 4Y, fourth-year; 5Y, fifth-year; 6Y, sixth-year; 8Y, eighth-year.

View Fullscreen

The fourth-year birds exhibited variable molt patterns, making it difficult to develop a reliable 4Y pattern. We found that 4Y molt patterns resembled the SY pattern, having a large block of old remiges in the center of the wing, but generally with new remiges at P1 or S1. We hypothesize that Northern Saw-whet Owls alternate between some versions of SY and TY patterns throughout the rest of their lives, with the patterns growing more variable as they age. Indeed, we did observe this in the few individuals that were recaptured in multiple years.