The Seabird 2000 Survey

This year saw the third census of all breeding seabirds in Britain and Ireland, following on from Operation Seafarer in 1969/70 and the Seabird Colony Register census in 1985/87. The organisers of the survey asked us to supply details for the Observatory and for nearby Mew Island and Big Copeland.

Work began early in the year as Black Guillemots are best counted in early morning between late March and early May. The method requires that all adult plumaged birds within 300 m of the shore, or on shore, be counted on a reasonably still day with a calm sea. Our count was made on the 19th April 2000 when 72 birds were recorded.

Except for Manx Shearwaters, the other species were surveyed on the 3rd June when there were five occupied sites for Fulmar, 14 apparently occupied nests for Common Gull, 540 individual Lesser Black-backed Gulls, 290 Herring Gulls and two Great Black-backed Gulls.

So far, so good, but what about the Manx Shearwaters? The other species are reasonably straightforward to census, but we have struggled for years to come to a satisfactory estimate of the size of our breeding colony, mostly based around a combination of numbers of chicks and adults ringed each year and the substantial number of retraps. Chick ringing peaked in 1996 when 711 were ringed. Studies elsewhere suggest that about 70 % of pairs raise a chick to fledging age, so the minimum at that time must have been around 1,000 pairs. Obviously, we cannot assume that we catch and ring all the chicks and this is where experience and judgement must be brought to bear. Some would claim we catch up to 90 % of chicks leaving the Island, while others have suggested the fraction could be as low as one half. This puts the colony size between 1,100 and 2,000 pairs. The number of chicks ringed each year has fallen since 1996 to 655, 314, 350 and 450 up to the year 2000. There have been a number of potential explanations put forward for this decline, but colony growth may also have been checked during this period when less chicks were ringed.

The new census technique

The new census method is relatively time-consuming, strenuous work, but quite straightforward and very exciting! In summary, the method involves:

• sketch areas occupied by shearwater burrows on a map and measure the area as accurately as possible;
  
• divide the colony into equally sized quadrats (we used 6 metres square plots), count and number the quadrats;
  
• using random selection, pick the quadrats to be sampled. (We used a technique called stratified random sampling where each of our sub-colonies was treated separately and the number of quadrats sampled in each was determined by the area of the sub-colony. This technique improves the precision if the density of occupation of sub-colonies varies.);
  
• a random number generator was used to select the quadrats to be sampled;
  
• using a marker stick to identify a datum point for each sub-colony, we measured 'down and across' to locate the quadrat to be sampled;
  
• the survey instructions suggest that 30 is a reasonable number of quadrats to be sampled;
  
• a cane was placed in the centre of the quadrat and a rope 2.96 m long attached to the cane. The rope had another cane tied at the other end and, having marked the start point, we simply 'walked' the cane at the end of the rope to make a circle with area 27.5 square metres;
  
• the number of burrows present inside the circle was counted and each one was tested for the presence of Manx Shearwaters;
  
• from the number of occupied burrows per area of quadrat sampled we could then calculate the number expected to be present in the whole sub-colony.
  

So, how did we check whether any shearwaters were present in the burrows? By the ingenious method of playing a tape recording of a male shearwater at the entrance to each burrow for a maximum of 15 seconds. If a male shearwater is present in the burrow it will almost always reply to the tape, while females never respond. The method has been tested at other colonies where the burrows could be opened to check for occupancy. The researchers who developed the method found a correction factor of 1.98. That is, if 50 birds answered the tape then there were actually 50 x 1.98 = 99 pairs present. The survey must be done during the day in mid-to-late incubation period (late May to early June), but before any chicks have hatched. At this time it is reasonable to expect that incubation will be shared equally by male and female parents, though only one will be present during the day while the other is off feeding. Daytime sampling also helps to avoid counting non-breeding immature or prospecting birds which are present in lower numbers during the day. The instructions recommend using a correction factor of 2 (i.e. assume 2 breeding pairs for every reply to the tape).

It was with some excitement and anticipation that Kerry Leonard and I approached our first burrow, tape in hand. Would anything happen? Not the first, not the second, nor the third – oh dear! But then, at the fourth attempt, the tape had hardly started when it was almost drowned out by a loud and instantaneous response from deep inside the burrow – success! More and more 'hits' followed, sometimes only a distant and brief cackle of response but, far more commonly, an instantaneous and loud reply. It is tempting to think our males didn't like the Scottish accent of the bird on our tape. Buoyed up by our success, and despite some heavy rain, we pressed on with the work, assisted at times by Ian McKee and Fiona Maitland.

After two days of hard slog we had sampled 50 quadrats in 14 of our sub-colonies (only North Garden, Radar and East Side were omitted). The results are shown in Table 1.

The calculation for a sub-colony, for example Well Path, is as follows:

And this is repeated for each of the sub-colonies sampled.

Table 1 – Summary of results from the Seabird 2000 census on the Observatory

Discussion

The results were astonishing, 2,819 breeding pairs is about twice our highest previous estimate.

We sampled 50 quadrats or 1,375 square metres which contained between them 211 burrows. Of these 211 burrows we had replies to the tape from 56. The whole colony occupies approximately 35,400 square metres (8.5 acres), perhaps one fifth to one quarter of the area of the Island above the rocky shoreline.

Averaging over the whole area of the sub-colonies we see the number of burrows is:
        211 ´ 35,400 ¸ 1375 = 5432.

Using the correction factor of 2, the 56 replies correspond to 112 breeding pairs present. The total number of pairs therefore is:
        112 ´ 35,400 ¸1375 = 2883.

The result is slightly different from the sum of all the sub-colony estimates in Table 2, this is due to averaging out the density of occupied burrows over the whole colony area. The figure of 2,819 pairs remains the best estimate.

It is obvious from the table that this survey method cannot be accurate at the sub-colony level, we know there are breeding birds in Gully, Lower Gully and West Warren. But the apparent lack of birds in these sub-colonies is balanced by over-estimates at others. Cliff, North Landing, North Point and Alley have, according to the survey, almost a pair in every burrow. So we cannot take the data for an individual sub-colony as representative, too few samples were taken to ensure an accurate result. It is simply chance that no birds were recorded in Gully, Lower Gully and West Warren, and similarly 'too many' were recorded in other sub-colonies.

Although it would be very time consuming, it would be most interesting to check every burrow in every sub-colony. We did try this for Lower Shore where it is relatively easy to establish the borders of the occupied area and to locate all the burrows. The narrow strip below the cliff was divided into eight sectors, marked off with string and each burrow was checked. The results are shown in Table 2.

The results suggest there are 41 males present and, therefore 82 breeding pairs in total. (By chance this agrees well with the estimate from the main survey of 84 pairs.) But the main point of this exercise was to convince the doubting Thomas's among us who pointed out that chick ringing in Lower Shore was typically around 5 to 15 per year (maximum 34 in 1995), so there couldn't possibly be 80-odd pairs, could there?

So the result of 2,819 pairs stands as the most accurate estimate of breeding pairs on the Observatory. When account is taken of the presence of immatures and non-breeding adults there could easily be over 8,000 individuals associated with the Observatory colony.

Big Copeland

What then of the Big Isle? In recent years, relatively little attention has been given to the possibility that significant numbers of Manx Shearwaters may be present there. Reports from people staying overnight on the Island suggested no great evidence that more than perhaps a few hundred pairs are present.

Unfortunately, in the time available, and taking account of the much larger size of Big Copeland, it was not possible to carry out the same kind of detailed survey there. However, Kerry Leonard and Ian McKee did spend a full day walking around the edge of the Island, making rough estimates of the size of areas containing burrows and checking a small sample of burrows with the taped calls. The areas they checked are shown on the map (Figure 1) where they are identified by letter codes. The results shown in Table 3 can be no more than an indicator that more work is urgently needed.

The indication from this initial survey is that Big Copeland could hold perhaps 3,000 to 5,000 pairs. It may easily hold a similar number of birds as the Observatory.

It is all the more important therefore that we take urgent steps to investigate the Big Island colony. A survey there would have two important goals: how many breeding pairs are there; and what, if any, is the degree of exchange with the Observatory colony? We have previously reported that only around 14 % of birds ringed as chicks are ever seen again. How many of our 'missing' birds are actually breeding on Big Copeland? What is the rate of exchange between the two Islands and how does it vary with the age and sex of the birds concerned?

It seems the more we find out about these birds, the less we know! There is work here to occupy our shearwater enthusiasts for many years to come.

| Home |