(16 May 2008)
Delayed incubation is the practice whereby a female bird does not begin incubating until at least two and sometimes all her eggs have been laid. The purpose of doing this appears to be to ensure that the chicks all hatch at about the same time and, later, can leave the nest together. This helps avoid the situation where late-developing chicks hold up the rest of the brood, exposing them to possible danger. The development of bird chicks is an astonishingly quick process, with mother nature seeming to want the whole business to go through as quickly as possible. Get the little things into the air fast seems to be her principle!
Delayed incubation is apparently a widespread practice among birds but not, so the wisdom goes, with owls. Almost all references to owl breeding behaviour I've seen say that incubation starts with the first egg, and the writer often goes on to say that the resulting spread of chick ages helps to spread the feeding burden on the parents. Here's a typical example, though the author has another angle on why incubation should start early:
The female [owl] starts to incubate right after the first egg is laid. . . .
The owl habit of starting to incubate as soon as the first egg is laid is an uncommon strategy in the bird world. In most bird species, the female delays starting incubation until her clutch is complete -- until all those three or five or however many eggs are laid. . . . delayed incubation ensures that all the chicks will hatch out at the same time. . . .
It's different with owls, which begin incubation right away. This is a particularly adaptive strategy for owls nesting early in spring in cold climates, where the first-laid eggs would freeze if the female waited before starting to incubate.
That's from page 59 of Cynthia Berger, Owls. Wild Guide series, Stackpole Books, Mechanicsburg, PA, 2005. 144 pp. I found it reproduced on Google Books, but unfortunately the extract stops there and I don't know if Berger gives other reasons why owls might start incubating from the first egg.
Or take Paul Johnsgard, the doyen of American owl authors, in his authoritative overview work North American Owls (second edition 2002). In one of the introductory chapters looking at owls in general he writes:
Incubation behaviour begins with the laying of the first eggs [sic] in most owls, which results in staggered hatching times..." (Ch. 4, Comparative Behaviour, p. 41)
On two of the owls that I've used as examples in this article he says the following. Of the Barred Owl: "Incubation begins with the laying of the first egg, resulting in staggered hatching." (p. 192); and of the Great Horned Owl:
Incubation begins immediately and lasts about 26-35 days, with the longer estimates perhaps reflecting interrupted incubation during cold weather. In one carefully timed case, 2 eggs were laid 3 days apart, and the incubation of each lasted 30 days [cites reference]. In another nest 2 eggs required 34 and 35 days, and a third at least 33 days [cites reference]. (p. 120, italics mine)
The inference of this passage (though it's not spelt out) is that by and large each egg in a given clutch is incubated for about the same time (30 days in one nest and 34 in another), resulting in staggered hatching, and that longer times may be the result of "interrupted incubation" by cold weather rather than of a deliberate strategy by the female to synchronize hatching by delaying incubation of the first egg or eggs. In such a strategy it would be the first egg(s) that showed the longer incubation times. Such a pattern is clearly seen in the tables I present below, where variability of incubation times is the norm and first-laid eggs almost always show the longest times. This the evidence clearly suggests is because the female deliberately does not start serious incubation until the second egg is laid. Whether it's warm or cold or freezing appears to have little or no influence on her behaviour in this respect
The authoritative work on European owls is Heimo Mikkola's Owls of Europe (Poyser 1983). This covers the third species I consider here, the Tawny Owl. In the chapter on that owl he mentions the widely held belief that incubation starts with the first egg, but goes on to make an interesting mention of an early suggestion by H. N. Southern that tawnies practise delayed incubation:
In Wytham Woods, Southern (1970) found an interval of three to four days to be usual between the laying of the first and second eggs. Normally, eggs are laid at intervals of 48 hours, and are incubated for 28-29 days by the female alone. It is always stated that incubation starts with the first egg, but Southern (1970) observed no incubation before the second egg was laid, and he calculated the average incubation time to be 29.7 days for a single egg. The first egg laid was not always the one to hatch first, proving also that incubation may not take place before the second egg is laid." (Mikkola op. cit. pp. 151-2).
I wish I had a copy of Southern's work referred to here as it's frequently cited in the literature. It's almost 90 pages long; here's the full reference: H.N. Southern (1970), The natural control of a population of Tawny Owls Strix aluco. J. Zool. Lond. vol. 162, pp.197-285. I may be wrong, but this sounds like a guy who spent time out with his owls, by night as well as by day, and who in addition to doing the rounds of nestboxes marking eggs or ringing chicks gave time to just sitting there watching the behaviour of the nesting females. It takes patience and dedication! But it seems that he must have seen the behaviour that is now quite commonly noted with owls that can be watched on webcams.
Unfortunately Mikkola's book, unlike Johnsgard's, has no general chapter on breeding behaviour, and the 69 tables at the back of the book give no further clues about brooding behaviour. Nor is there any reference to delayed incubation in the chapters on two closely related species, the Ural Owl and Hume's (Tawny) Owl.
In his Owls of the World (Firefly Books, 2003), James Duncan only makes the somewhat self-contradictory observation that "Owls in the nest are of different ages and sizes because female owls typically start to incubate after the first or second egg is laid." (Under "Nesting" in a section on Breeding Strategies, p. 67.) Here we have a nod in the direction of delayed incubation, but you might be forgiven for thinking that surely a late start to incubation would tend to bunch hatch dates together and so produce chicks of similar age and size. Not only that, but beginning incubation after the first egg was laid would have no effect at all because differences in the chicks' ages would merely reflect the gaps between the times the eggs were laid! Never mind; Duncan continues: "The resulting hatch asynchrony [i.e. spreading of the hatch dates] is an insurance policy for breeding owls against times when not enough prey is available to feed all the young in the nest." This is basically a variation on the argument that hatch asynchrony is a strategy to spread the feeding load, and of course the asynchrony is taken as a fact: female owls organize their laying and incubation behaviour so that chicks have different rather than similar ages and feeding requirements. The reasonable assumption behind this is that it maximizes the chances of a larger number of chicks surviving to the fledging stage.
Well, I wasn't aware of any of this when I started watching Tawny Owls on the nest in 2004, and more seriously in the 2006 nesting season. Even then, when I camped out under a nestbox and monitored a local tawny female using recording equipment and the night-vision setting on a camcorder, I was only aware that at times she seemed remarkably unconcerned about leaving her eggs to go cold for long periods. In fact this experience is probably irrelevant here as it was in the middle of incubation, and the evidence is that hubby wasn't able to bring enough food, forcing her to forage for herself. I think it must have been later, when I picked up the bad habit of watching nesting owls on webcams and following their fortunes on forums, that I began to notice comments on how after a mother owl had laid her first egg she'd go flyabout, leaving it to go cold. Then later, usually after the next egg, she'd settle down and be there almost all the time. Sighs of relief all round -- everyone was happy.
I really became convinced that there was something in it -- that it did happen and that the mother owl did it deliberately -- when I was lucky enough to watch a female Tawny Owl using a nestbox in 2007. This is a nestbox in the UK with a couple of viewing cameras hitched to the web, and the owner kindly let me watch although the website and feed are really intended just for members of the family. That breeding attempt is written up elsewhere on my site (interested readers can try the Cambridgeshire owls). What's important is that this female managed the surprising feat of shepherding three eggs laid about a week apart into a hatching period of just over 24 hours. No accident surely, and so what she accomplished kicks off my data section below.
The second example I've given is the Great Horned Owl female at the California State University at Bakersville campus as this season (2008) she has just done much the same with her own three-egg clutch. She laid over a period of eight days, but by selective brooding brought the eggs to hatch in just three days. Again my interest was aroused because, as we have seen in the quote from Johnsgard (2002) above, Great Horned Owls are believed to begin incubation immediately. I didn't watch her closely at the beginning, but on 6th Feb, when she'd just laid a second egg four days after the first, another watcher reported on the BirdCam forum:
She hasn't been on the egg most of the morning, but I just caught that there is a second egg! Unfortunately, she's only incubating one egg.
I myself saw her at about this time carefully tucking one egg beneath her while leaving the other egg half-exposed in her chest feathers. After the third egg on 10th Feb I checked from time to time and almost never saw her off the eggs -- she had clearly settled into a routine of non-stop incubation. The drawback of this webcam is that it's a daytime-only colour camera, so we don't know what she does at night. Evidence from the Valmont Great Horned Owl infrared webcam at Boulder, CO, is that the brooding female does go out during the night, with a first sortie at dusk (exactly like the CSUB female), but that these sorties are limited to at most 20 minutes.
My third example is the OwlCam Barred Owl pair in Eastern Massachusetts (strictly the Northern Barred Owl, Strix varia varia). The owner of the website where they are featured kept a meticulous diary of events during the five active breeding seasons between 1998 and 2003 (the pair did breed in 1997 but no records are available for that year on the website, and no breeding occurred in 1999). He found himself surprised from the start that the female appeared to neglect the first egg and only begin serious incubation after the second egg was laid. In a typical comment from 1998 he writes:
June spent only about 50% of her time incubating today. . . . I was surprised by this as the temperature was just below freezing. (Entry for 17 March, the day after the first egg was laid.)
In subsequent breeding seasons this established itself as the normal pattern, so much so that the website owner came to realize he should estimate hatch times from the second egg. For example, in the Notes for 20 April 2000 he writes:
I'm slipping my predicted most likely first-hatching date three days (to Wednesday) based on 1998 OwlCam experience and a recent communication from an owl rehabber in Great Britain. It appears that the reference books are incorrect (at least for Barred and Tawny owls) when they say that incubation starts with the laying of the first egg. Even June's apparent part time incubation during the first three days was probably nothing more than guarding the egg.
In an illuminating email to me last year (9/3/07) he expanded on this:
I just finished reading your account of this year’s early nesting [by the Cambridgeshire Owls] and noted one surprise that I discovered with my OwlCam owls. While all of the reference books that I had on Barred Owls indicated that the female started incubating as soon as the first egg was laid, I found that this was not the case. The real incubation did not start until the second egg came (as indicated in my DVD) and I did not start my hatching countdown until that time. When I asked a noted Barred Owl expert about this he said that the field researchers had probably gotten it wrong, as what I was able to see with my side view camera would have looked like incubation to a distant observer. I also heard from an owl breeder in England who said that she saw similar behavior in her Tawny Owls. The female would appear to be incubating the first egg, but when the egg was checked it was cold to the touch. Real incubation did not start until the second egg was laid.
Oh dear, what complications this introduces! For here we have evidence that even if you see an owl sitting on her first egg she may be faking it and not incubating at all! I am immediately reminded of seeing the Bakersville female appearing to incubate one egg while leaving the other, presumably the first, aside. The question arises, does the female keep track of which egg is which, even when she is away, and carefully arrange incubation of each on an individual basis so that all hatch as close together as possible? And how on earth could one find out what was really going on, even with marked eggs, if she can "fake" incubation? You'd have to have a temperature sensor on each egg radioing its signal to a monitor. This is partly relevant here as in my tables I've made the (unavoidable) assumption that, despite the delayed incubation, the first egg laid becomes the first chick out, and so on. But as we have seen from Southern's (1970) work, this is not always the case. Fortunately this does not affect the validity of my conclusion, which is that in all cases recorded below the hatch period was shorter than the laying period, and sometimes considerably so.
So, were these three females of different species being negligent? Were they inexperienced? Do they take some time after laying a first egg to get into the swing of things, to curtail their freedom and resign themselves to that long, tedious 4-5 week stint of almost total confinement? Not least, can the experts really have got it that wrong after at least a century of careful observations?
Well, I can only say that I haven't seen signs that the females here are showing signs of negligence or inexperience. Indeed, the OwlCam DVD presents fascinating evidence that at the chick stage the mother keeps a close track of who's been fed and who needs feeding. It is only a little less apparent that all three females are equally thoughtful and deliberate about what they do in those first days of egg-laying, making it reasonable to conclude that the subsequent outcome -- a compressed hatching period -- is arrived at by design and not by accident. There is no instance in which the hatching period is longer than the laying period, in itself a statistically unlikely outcome if either it is desirable for broods to be markedly staggered or if there is no desired pattern for hatching In fact, over the 19 fertile eggs in seven broods the hatching period is compressed down relative to the hatching period to just less than 40%, from which it is difficult not to conclude that these females were indeed practising delayed incubation. It seems, therefore, that it is more important to the owls to achieve a nearly simultaneous fledge than to spread the burden of feeding.
Case 1: Tawny Owls, Cambridgeshire, England, 2007
The first example is of a Tawny Owl pair that use a nestbox in Peterborough, England. In 2007 the female began laying on 20 January, remarkably early for UK members of this species. Three eggs were laid over a period of about a week and hatched within a little over 24 hours of each other. Her brooding behaviour in that year is not well recorded, but in 2006 she is known to have been in the nest box only sporadically after laying the first egg in a clutch of four.
Data courtesy of Stephen Butler, Tawny Owl Box Watch.
Table 1. Data for Cambridgeshire Tawny Owls, 2007 season.
Data from Tawny Owl Box Watch (www.stephenbutler.co.uk/owl/index.html). For further details see the Cambridgeshire Owls pages on my website. Internal top-mounted camera is IR, allowing night and day observation. An external camera functioned in colour during the day and in b/w at night. The female's brooding behaviour in 2007 is not recorded, but in 2006 she is reported to have "[left] the nest 6 times during the next 2 days [after laying first egg of four], mostly for about an hour, but once for 7.5 hours!". This appears to be clear evidence that incubation of the first egg was delayed.
*In all tables incubation period is calculated making possibly unreliable assumption that eggs hatched in sequence laid.
Case 2: Great Horned Owls, California State University at Bakersville, USA, 2008
This female is well known to owl nestcam watchers, and this is the fifth year she has made use of this site on one of the science faculty buildings. Her daytime activities can be followed on the webcam -- for details see the CSUB thread on the BirdCam forum (there are two urls for the camera, one of which may not work for some).
This year the laying interval was 8 days, and the hatch interval was 3 days.
Table 2. Data for Bakersfield, CA, Great Horned Owls, 2008 season.
Information based partly on my own observations via webcam supplemented by info posted by others on the BirdCam forum and PhotoBucket. Observation is limited by the fact that this (colour) camera operates only in daylight; night time observations, when these owls are most active, cannot be made.
Case 3: The OwlCam Barred Owls, Eastern Massachusetts, USA, 1998-2003
The OwlCam website is a mine of useful information on all aspects of the nesting behaviour of this Northern Barred Owl pair. In particular, the owner noticed the tendency of the female (left in the photo) to leave the first egg unincubated from the second breeding season (1998) and kept meticulous notes on the habit as he was aware that Barred Owls weren't supposed to do it! Delayed incubation was seen in all five seasons during which these owls bred, most notably in the last year, 2003, when three eggs laid over a five-day period were brought to hatch in 1.5 days.
My grateful thanks to Bill Alexander, the owner of the site and producer of the OwlCam DVD, for permission to use his data so extensively and for providing useful supplementary information. To Bill must go the credit for first noticing delayed incubation in owls -- at least, he did so before me!
Table 3A. Data for OwlCam, MA, Northern Barred Owls, 1998 season.
In Tables 3A-3E hatch time is time when chick was completely out of egg shell, where this is known. As in first two tables, incubation periods are calculated from this assuming eggs hatch in sequence laid (unlikely always to be true, but no alternative unless eggs are marked!).
OwlCam owner's comments on brooding behaviour :
17 Mar: "June spent only about 50% of her time incubating today. . . . seen here as she perches in the door while the two-inch egg goes unincubated. She also left the nest area completely for four hours between six and ten PM. I was surprised by this as the temperature was just below freezing."
19 Mar: ". . . much more serious with her incubating in the last couple of days. She spends at least 95% of the day on the egg, but does go out several times during the night. Her departures leave the egg unattended for from thirty minutes to two hours."
20 Mar: "June did not spend as much time away from the nest last night" when 2nd egg was laid.
23 Mar: "June is now spending all of her daylight hours incubating the eggs and is restricting her nighttime outings to less than five minutes."
1 Apr: "She is constantly incubating the eggs through all of these activities."
*Wally "first broke through his shell and started this peeping call last night [19th] at 10pm, but did not break out completely until 11:30 this morning."
**No time given for hatch, but Theodore "peeped" from his shell for 20 hours before mother ate last of shell early evening of 21st.
Table 3B. Data for OwlCam, MA, Northern Barred Owls, 2000 season.
In 1999 the owl pair were frequently present in the area but did not breed. The site owner is certain that this was because the 1998 owlets remained in their parents' territory for much of the first half of 1999.
*24 Mar: "June spent most of her day incubating the egg . . ., but left abruptly at 5pm . . .. The egg went unincubated until she returned five hours later. . . . I'm a bit concerned about the egg being left alone for so long in this cool weather, but she did the same thing with her first egg in 1998. It appears to take more than one egg to hold her attention."
**27 Mar, before 2nd egg: "While June spent less than half of her time incubating her egg on the first two days, she has gradually increased it to nearly full time." Entry for day continues with specific details.
***E.g. 3 Apr: "She has been so devoted to incubating her eggs for the last two days that I've not been able to see them on either camera." and 5 Apr: "June has been even more dedicated to her incubation duties with her two nighttime outings limited to less than five minutes each." Both excerpts from author's supplementary notes.
Rufus broke through shell and started peeping at 8:30pm on 25th but didn't hatch fully until June removed the rest of the shell from his body just after noon on 26th. Emily hatched more quickly, first peeping from shell at dusk on 26th and heard being fed the same night.
Table 3C. Data for OwlCam, MA, Northern Barred Owls, 2001 season.
This is the brood featured in the OwlCam DVD.
*26 Mar: "After laying her first egg on the 23rd, June left the nest at sunset and did not return until the next morning. Even then, she merely stood over the egg and occasionally touched it with her foot . . .. She again left the nest on the second night, but returned 4 hours later just before the temperature dropped below freezing for the first time. She then stayed in the nest for the rest of that night and through the second day, but again did not appear to be incubating the egg. Not until this morning, when she laid her second egg, did she start to take her incubating duties seriously. She spent the entire day incubating her two eggs and did not leave the nest until Ward arrived at 8pm to deliver a meal. She went out to meet him, but returned to her eggs just five minutes later. Her delay in starting incubation of the first egg should result in the two eggs hatching at about the same time -- just as they did last year." An observation that's borne out by the almost simultaneous hatch times of the first two eggs.
**26 Apr: "Emmett . . . started peeping at 9pm last night, but did not emerge from his shell until 6am this morning with a lot of help from June." 2nd owlet (Abigail) already cheeping from egg: "June helped her out of her shell just hours after her brother was hatched." Entry for Abigail hatch is on 27 Apr page.
***Interesting in that this year the incubation period for the last egg was almost as long as for the first despite the female's more persistent presence after laying the 2nd egg.
Table 3D. Data for OwlCam, MA, Northern Barred Owls, 2002 season.
In 2002, unlike the other four years, only two eggs were laid.
*29 Mar: "Two hours after producing the egg, she climbed up into the doorway where she slept in the warm sun for almost an hour before waking up . . . then flew away to perch in a tree about 100 yards from the nest, . . .. As of 11pm, June had not returned . . ., but with the temperature well above freezing, her absence should not harm the egg." Between then and 31st usual pattern of intermittent brooding with lengthy absences (noted in entry for 31 Mar).
**1 Apr: "June is now spending almost all of her time in the nest, .. .. The arrival of the second egg marks the beginning of a six-week period when June will seldom leave the nest." 3 Apr (from Notes): "June is fully into the incubation of her eggs and has still not left them in daylight. . . . total time out of the nest in the last 24 hours has been about 25 minutes."
Cheeping from the first egg began night of 30th and continued until 2nd May, when chick finally emerged and 2nd egg was also cheeping. For a fascinating discussion of this phenomenon, see entry for 2 May.
***Penelope emerged on 3 May "less than 24 hours" after Spencer (entry for 4 May).
A side note. On 16 March 2003 a black squirrel, very rare in the area, made an appearance, and Ward, the male owl, watched it for half an hour. In the entry for that day the author comments "Let's hope that the black squirrel does not prove to be an omen of bad luck." As it turned out, the boisterous 2003 brood were taken by a marten from the unguarded box little more than a couple of weeks after hatching, and the owl pair have not used the nestbox in the four seasons since (to 2007) although they are thought still to be in the area. Hence 2003 is the last year for which data are available.
Table 3E. Data for OwlCam, MA, Northern Barred Owls, 2003 season.
This brood shows the clearest evidence of hatching synchronization, with the first two born "within 2 hours" and the third a day and a half later.
*30 Mar: "Yesterday [29th] was June's last chance to spend time outside the nest and she took full advantage of it. Even though it rained for most of the day and night, she left her first egg alone in the nest as she perched in nearby trees and hooted for Ward to bring her food. She did not return to the nest until just after dawn today when she entered soaking wet and started preening her feathers. Four hours after entering, she sat up . . . and starting making that unique coughing sound that indicates she is about to lay an egg." That was egg no. 2. After laying the second egg "She then started incubating the eggs and has not been outside the nest since [refers to 30th]."
**2 Apr: "She has not left the nest in daylight since laying the second egg two days ago, and has limited her nighttime excursions to less than 15 minutes." No precise lay time is available for the third egg, which was first seen when female stood up at dawn.
***Cheeping from first egg began previous night at 10pm, probably out in early hours as he was seen dry and asking for food at 5.40am (entry for 1 May). Second chick hatched fully "within hours" of first. Third chick "hatched sometime during the night [of 2/3] and was already wheezing for food when the sun came up." (Entry for 3 May)
Table 4. Summary data for the three species of owl.
*Not significant overall, but the first two eggs laid 3 days apart were brought to hatch within "just hours".
Compression calculated as total of hatching intervals (14.25 days) divided by total of laying intervals (38 days).
This article would not have been possible without the data on Barred Owls collected by Bill Alexander of OwlCam, who, as mentioned in the text, noted the phenomenon of delayed incubation before I did. I am very grateful to him for generous permission to make use of his data.
The article is not quite complete as one or two further points of detail need to be considered. I also hope to add data on other owls/breeding attempts as and when they become available.
powered by owls
The female Great Horned Owl at the California State University Bakersfield campus shown here soon after all three of the 2008 brood had hatched.
Delayed incubation in three species of owl
Tawny Owl, Barred Owl and
Great Horned Owl