Kaka Count: July 2010

by John Ogden

Was it the third? or was it the fourth? Although it said Kaka Count 4 on the data sheets it was in fact the third, as one was missed out due to ‘other commitments’!

Anyway, the previous two counts had shown a clear decline in the kaka counts between summer and winter, coinciding with an increase in kaka sightings on the mainland (Suzi Phillips – Kakawatchnz). For this reason the Trust decided to do another winter and another summer count, to confirm the supposed ‘off-island’ winter migration. The results for the July count are given in Table 1. This table amalgamates the results of different observers (where there was more than one at a ‘location’) and doesn’t show the relatively high proportion of observers who gave ‘nil’ returns for kaka sightings on the day – seven of the 33 sheets (21%). Indeed the general absence or low activity of kaka in July was noted by most observers and may account for the lesser response to this survey (33 sheets returned) compared to previous ones. But absences are just as important at presences when it comes to statistics, so please return the sheets for the next survey even if there are no kakas to be seen or heard! We are planning one more count in December (this year) to hopefully confirm the trends seen so far.

Table 1. Kaka Count 25/07/10
LOCATION No of data Total AM1 Total PM1 Minimum Maximum
  sheets     at location2 at location3
Allom Bay 1 0 0 0 6
Awana 2 4 7 5 6
Claris 1 0 3 3 6
FitzRoy 2 4 5 3 6
Harataonga 3 5 2 3 3
Kaiaraara Bay Rd 1 3 0 3 3
Kaitoke 1 0 0 0 2
Kaitoke Creek 2 0 1 1 1
Medlands 5 13 14 5 12
Mulberry Grove 1 0 0 0 0
Okiwi 1 2 8 8 8
Okupu 2 0 0 0 1
Puriri Bay 1 9 9 9 10
Rosalie Bay 1 9 5 9 12
Schooner Bay 3 3 3 1 4
Tryphena 2 3 1 3 5
Wairahi 2 2 11 11 16
Whangapara 1 0 0 0 2
Windy Hill 1 4 4 4 4
TOTALS 25/07/10 33 61 73 68 107

(1) Assumes all counts by different observers, AM and PM, were all different birds.
(2) Assumes that all birds at one ‘location’ were the same birds, so the maximum count by any one observer is taken as the minimum for the location.
(3) Uses the maximum estimate for birds reported at the location over the last few weeks.

Another feature of the July data is the variation, especially in the ‘Maximum’ column. This indicates that kaka were concentrated at a few locations, while they were present only in small numbers or absent elsewhere. Thus, in winter the kakas may be concentrated in areas where there are reliable food supplies – such as Peter Speck’s orchard at Rosalie Bay, or in the Medlands valley or the Wairahi. It is interesting too to compare this with the summer count (26/12/07) which, while showing many more birds, showed much less variation between locations.

Table 2 indicates that this July’s count was numerically similar to that made in early September 2008. However, all these results are only samples of the true population size, and it is difficult to know how many birds have not been counted or estimated by the observers. The figures suggest to me that the winter population here is probably in the range 75-175 birds. The Summer population may be up to 300.

Table 2. Comparisons with previous counts
  Number Maximum Minimum Maximum Coefficient Population
  of data number no. based number(3) of  guestimate
  sheets counted(1) on locations(2)   Variation(5) -4
Summer 45 222 141 221 35 250+/-50
Winter 40 136 61 117 32 125+/-50
Winter 33 134 68 107 76 125+/-50

(1) Total of all AM and PM counts by all observers
(2) Total of the minima from all locations (i.e. assumes birds at different locations are different)
(3) Total of maxima for locations.
(4) Assumes that population size must be greater than the minimum number based on locations, and will also probably be greater than the maximum estimated from the data, because some locations have not been observed.
(5) A measure of the variability between locations. Higher % indicates higher variability – i.e. the kaka are more patchily distributed.

The total results from all three counts, and comparisons with kakawatch results from the mainland, suggest that about half the Great Barrier kaka move to the mainland during the winter, leaving in May, and probably returning mostly in September. What determines which birds go, and which stay, we do not know. Powlesland et al. (2009) suggested that kakas breed only in years of abundant food supply. On Great Barrier the kaka which overwinter seem to concentrate on the most reliable food supplies, suggesting that it is a general food shortage which drives the move to the mainland. On the other hand we could see the process in reverse: kaka move to Great Barrier to breed because their main nest predators on the mainland – possums and stoats – are absent from the island (cf. Powlesland et al. 2009).

In conclusion – we know that kaka can be a nuisance or worse, but they are also very unusual parrots, endemic to New Zealand. Their numbers have declined nationally and their range contracted to increasingly disjunct populations. Kaka are put in the ‘Nationally Vulnerable” category of Miskelly et al. (2008). Great Barrier is a stronghold for the North Island, and may well provide recruits to the populations on the Auckland mainland and Coromandel peninsula. Our community-led monitoring on Great Barrier is starting to indicate the significance of our kakas here, and to suggest new lines of research. Thanks to all participants – 101 names occur on the sheets from all 3 counts, so about 10% of the Barrier’s human population has contributed so far! Special thanks to Sue Daly for organising the last count.

References: MisKelly et al. 2008. Conservation status of NZ birds. Notornis 55: 117-135. (www.notornis.org.nz). Powlesland et al. 2009. Breeding biology of the New Zealand kaka (Nestor meridionalis). Notornis 56: 11-33.

See also: http://www.kakawatchnz.org and go to “Research Publications”.