Anything you build on a large scale or with intense passion invites chaos.
– Francis Ford Coppola
Little and Saunders’s Tern represent one of the biggest identification challenges of the Middle East, especially since most people visit the area in non-breeding season, when all the birds are in winter plumage. In that context, the moult has risen as the main key feature to tell these species apart, but even that is sometimes hard to understand.
During our recent trip to Oman, we had the chance to take a close look at a big flock of little terns that turned out to be mainly Saunders’s Tern Sternula saundersi, with only a few safe Little among them. The flock was resting at Khawr Dhurf, a place that never disappoints, and it was composed by over 300 individuals, most (>90%) in non-breeding plumage but several actively moulting. All the individuals in breeding plumage were identified as Little on the basis of rump and tail colour and head pattern, except for one.
I took the chance to study the moult state of Saunders’s Tern at this time of the year, but first I needed to identify them all so I wasn’t getting any winter-plumage Little Tern influence. What a challenge: all the key features weren’t as obvious as in summer plumage or even as in winter time, when the moult is arrested and you can theoretically trust the number of black (worn) primaries. Most of the plumage features included in the literature have been questioned at some point; mainly because of the existence of tricky birds caught at a breeding site in Israel but also because pretty much every detailed analysis of a vagrant Saunders’s has shown some degree of mismatch between characters.
- Head pattern: In summer plumage, Little shows more white above the eye, with a triangular-shaped white entrance from the front that usually reaches the end of the eye. In the other hand, Saunders’s doesn’t show white above the eye. The white area is restricted to the front and only reaches the beginning of the eye forming a square 90-degree angle.
- Tail and rump: Saunders’s shows a grey rump and tail, concolorous with the rest of the upperparts, while Little Tern shows some contrast between these areas. This feature has been questioned by these birds caught in Israel and there seems to be some overlap between the two species, maybe involving different age classes.
- Bill colour: Saunders’s shows a more extensive dark tip, which can reach around 30% of the bill. The dark tip is more restricted in Little, some times virtually absent.
- Leg colour: Little is meant to show brighter yellow legs in summer plumage, whereas Saunders’s gives a dirtier impression due to an overall more olive-green leg colour.
All these features are pretty much gone in wintertime and we are left with the moult, which is the actual focus of this post. The moult in the little tern complex is one of the funniest within the Aves class: complete moult cycles that look partial, apparently random suspensions, age-dependant temporal delays… All sorts of pieces of a jigsaw puzzle, impossible to make fit at a glance. We’ll focus on primary moult, since it’s the easiest to assess in field photos.
In a nutshell, post-breeding moult consists in a complete moult of primaries, starting from P1. However, as soon as this moult wave has reached the mid-hand, another wave starts from P1. When this second wave has reached the mid-hand, yet another wave starts from P1. Thanks to Martí’s sketches, I’ve made a gif to show how this progresses:
When the next breeding season comes, the bird arrests its moult and we then can see two obvious generations of primaries with a very strong contrast: some black and worn outer primaries and some fresh and white inner primaries. These moult cycles aren’t partial but arrested complete: the bird will resume as soon as breeding season is over to suspend again later in the season. Hence, there isn’t a difference between moult extensions in Saunders’s and Little; both perform an endless series of complete moult cycles. The difference is in moult timing: Saunders’s moults later, so every moult suspension catches the birds with a higher number of black and worn outer primaries. The traditional threshold used to tell these species apart in winter is 4: >4 black primaries = Saunders’s, <4 = Little. Easy-peasy?
The photo below can be considered a random sample of the flock, including 45 assessable individuals. There are at least 3 birds with an apparent white tail and rump, contrasting with the back, numbers #18, #17 and #38. These birds have been excluded from the analysis for being too good candidates for Little. I acknowledge there might be some more Little Terns showing a grey rump, but I decided to follow the traditional key features for the sake of being able to do something. #36 was finally excluded from the analysis because its moult can’t actually be assessed. It’s worth noting however that the state of the moult of these individuals is within the variation of the flock, as shown in the moult analysis.
I counted the primaries belonging to each moult wave in every individual in which I could assess all 10 primaries. I then plotted the data in two different but related ways: one shows the number of primaries belonging to each wave and the other the last primary to be moulted in each wave.
This second plot seems to confirm the 4 as a good threshold, with most birds showing a moult limit somewhere between P4 and P6 and therefore having left a minimum of 4 old primaries during the winter. All 3 waves are visible in most birds, although the third wave is just incipient, with most birds having moulted P1 and P2 or doing so now. It feels impossible for the second wave to make it to the wingtip before the breeding season suspension, and just as hard for the third wave to reach somewhere near P7. Therefore, the black outer primaries we see in breeding season must belong still to the first wave, the rest of the primaries being a combination of second and third wave feathers.
Is there any difference between the expected and the observed moult stage then? Just one, which doesn’t have many implications in the identification: the synchrony between moult waves. The second plot shows how there are only between 2 and 4 primaries of difference between the second and the third wave, so the third wave starts before the second wave has reached the midhand. Indeed the third wave was active in many more birds than the second wave.
To sum it up, this is the situation I found now, averaging the sample of the flock:
How will this evolve later into the breeding season? Most photos of Saunders’s Terns in breeding season I’ve seen show a very strong contrast between P7 and P8 (having left 3 old primaries) and then there are two kinds: most show entirely uniform inner primaries and a few others show a subtle contrast somewhere around P4.
So here’s a possible explanation: outer primaries in breeding season invariably still belong to the first wave (just as I saw them now) and then you get:
1- Birds in which the rest of the primaries belong to the third wave, which has overtaken the second wave creating a very strong (almost black vs. white) contrast, usually between P6 and P7 or P7 and P8 (>=3 black and old primaries). This fits with the fact that in some birds the second wave hadn’t resumed, while the third was already active. An example here.
2- Birds in which there are 2 generations of inner primaries, but both fairly new. The contrast is subtle but does exist. See the photos above by Mike Barth.
The first strategy seems to be commoner than the second in Saunders’s and it would be interesting to know what happens in Little. There’s something still missing here: the delay in Saunders’s moult timing is real, but it doesn’t explain the whole thing: in the end there’s only 1-2 primaries of difference. In Saunders’s, the second wave stops earlier in autumn, leaving a higher number of outer black primaries, but then they have more time than enough to resume and reach as far as Little do, but they don’t. It’s not a matter of unsuitable conditions to moult, indeed they start with the third wave, but for whatever reason the second wave doesn’t advance much during late winter and spring. The same might happen in Little though, otherwise we’d not see any black primary in spring. All in all, I’ve reached two not-so-ambitious conclusions:
- The delay in Saunders’s Tern primary moult timing only affects the second wave and only happens in summer-autumn (post-breeding stage), but can still be visible in spring because not much happens in early spring.
- The inner primaries are moulted much quicker than the outer, probably because they are cheaper to grow. This fact creates an asynchrony between waves: despite, by the start of the breeding season, the third wave is probably equally extensive in Saunders’s as it is in Little and can potentially happen at the same time, the second wave doesn’t advance as quick, so the number of black primaries in breeding season is still somewhat useful.