Maps used by volunteers, and graphs showing the results of predator control.
Predator control
This shows the grid of bait lines, running in straight lines 100m apart. Bait stations are placed every 50m in most areas, but a 100m spacing is being trialled in some newer blocks. Cordons of stoat traps, all of which are kill-traps, also run through the area, and are being re-baited two-weekly all year (and sometimes once a week, at the time of peak stoat activity November to March).
(Click on image to see an enlarged view)
"Owned" bait lines
Some volunteers have taken on responsibility for the maintenance and baiting of particular lines for at least one year. Having these lines pre-allocated ("owned") assists the organisation of each baiting cycle.
(Click on image to see an enlarged view)
Rodent monitoring lines
The effectiveness of rodent control is measured at intervals of around two months, from soon after baiting starts through to late summer. Tunnels with a peanut butter bait in the centre are placed in the localities shown above. Ink is applied to pads next to the bait, and as the rodent exits its footprints are picked up on blotting paper pieces that it has to cross.
A total of 180 tunnels are in place, in random positions, so that a variety of habitats are sampled. These are compared with 30 tunnels at three locations outside the Ark.
The protocol is to set the baits and papers on a day when there is little rain, and the papers are collected next day. Typically the incidence of rat footprints is around 2-5% in the Ark, and 70%+ outside.
(Click on image to see an enlarged view)
Monitoring results within the Ark: Before baiting and after
Predator control started early in 2003. The reduction in rat numbers has been spectacular - since 2007 the usual result of monitoring immediately after the Spring baiting is such that we would expect to find only one or two visited tunnels out of the 150 placed throughout the Ark. In March, around six months after the last baiting, and at the time of the natural peak for rat populations, the numbers are about 3 times higher, but still generally only 5 -8%.
The initial increase in mouse numbers was attributed to them flourishing in the absence of rats, but since 2006 their numbers have also dropped sharply - this was unexpected, but has been sustained and it now seems clear that the combination of broudifacoum and a 100m by 50m grid of bait stations is sufficient to significantly reduce mouse numbers. Click the forward arrow above to see the results for after baiting, only, which allows a more sensitive vertical axis.
(Click on image to see an enlarged view)
Monitoring results in Ark: after baiting started
This graph shows the same results as the previous page, but with the pre-baiting results excluded. The smaller vertical axis is now able to show variations more sensitively.
Monitoring results outside the Ark
The comparison between this and the results from within the Ark is striking confirmation of the effectiveness of the rodent control. Some of the variability shown above does not have an obvious explanation, but some can be explained in terms of the normal cycle of rat populations, where there is a large decrease in numbers during the winter, and a steady build up during the summer to a peak in late summer and autumn.
(Click on image to see an enlarged view)
Mustelid trap catch 2006-2007
The majority of catches are found on the periphery of the Ark, but there is also a significant cluster in the central area around the Auckland City Walk. Stoats can move large distances in a day, and it is likely that these ones came via the golf course, attracted by the population of rabbits there.
Mustelid trap catch 2007-2008
A similar pattern to the previous 12 months results is shown here - the majority of catches are made by the cordon of traps on the boundary, but the central area abutting the golf course is also active.
(Click on image to see an enlarged view)
Mustelid trap catch 2008-2009
Numbers are down this year (67 stoats, 5 weasels, 5 ferrets). Again the majority are around the periphery of the forest, as the golf course is an internal open space (with a thriving population of rabbits to attract mustelids).
For the first time there were no mustelids in the central lower Andersons Track area, but unfortunately a cluster of five can be seen in another internal area nearby ie the lower part of the Upper Kauri track. The importance of multiple traps is clear, if animals that travel so fast and so far are to be adjacent to a trap when they pause.
Mustelid trap catch 2009-2010
The catch for the 12mths ending June 2010 was 78 stoats, 15 weasels, 5 ferrets, 26 hedgehogs, and 2 possums. Invasion from the periphery, as before, particularly from the farmland at the north-western corner, and heading up the Waitakere Stream valley to the golf course and beyond.
Mustelid catch November and December 2011
Every year, early Summer sees an explosion in the number of invading mustelids, as the females leave their nest and take their litter with them. The pattern of catches is of clusters in the same area, reflecting the movement of these groups.
Stoat catch by month 2004-2008
The parameter shown is catch per trap - the number of traps has risen steadily throughout the period, so catch per trap corrects for the effect of this increase on actual numbers caught.
There is a consistent increase in numbers in mid-summer, when the female stoats come off their nests with their brood of around eight offspring. In two years (04-05 and 06-07) this produced a very large peak in December/January, but in the other two years there was a much less marked increase, spread evenly from December to February. It is not clear why there was this change in pattern - in 05-06 there was a bigger catch in the preceding Autumn, which could explain lower breeding in Spring, however 04-05 and 06-07 had similar patterns for most of the year but very different results in December/January.
Traps are checked and re-baited less often in winter and more often in summer - for most of them the frequency is two-weekly and weekly, respectively. The bait used for most is rabbit meat, with a cracked egg as well (especially important in summer, when meat rapidly deteriorates), but for some a rabbit paste is used.
All traps used are kill-traps, either Fenns or DOC 200's.
Rat catch January to April 2010
Although the Ark control of rats is achieved by the baiting programme, the by-catch of rats in mustelid traps is valuable as it provides information on the distribution of the rat population.
The period shown above starts two months after the completion of the 2009 baiting programme, but it is not until April that rats can be detected in the Ark itself, along with the expected periphery catch. The one exception is the valley below the dam, and this area will be treated more intensively in the 2010 baiting programme.
Rat catch May to August 2010
The first baiting of the 2010 programme was in May. Rats in the internal areas are seen in May, as in the preceding month, but by June rats are only evident on the periphery.
Whitehead survey 2011
In 2011 volunteers spread out to walk circuits at several widely spaced locations throughout the Waitakeres. Following a standard protocol, they stopped at intervals to listen for the song of whiteheads. Whiteheads have had two releases, both in the Ark area, but they are known to be highly mobile birds who disperse widely. Laurence Béchet has used Quantum GIS software to produce this map which shows the location of these circuits and monitoring spots.
Whitehead Survey 2012
Here are the results of the 2012 survey. Map produced, using Quantum GIS software, by Laurence Bechet, with assistance from Cameron Wilson.
Rat catch Winter 2012
This shows rat presence between the end of the Autumn baiting in May, and the start of the Spring-Summer baiting. See the next map for rat presence following this baiting.
Rat catch early 2013
There is evidence of post-baiting population in two sections of the internal part of the Ark, one in the north and one in the south. These are indicated by red markers ie persistence is interpreted as March catch, or monitoring tunnel prints - a catch in February which did not repeat in March can be regarded as showing response to baiting. Comparing this with the previous map, the northern area is new, but the southern one is possibly a continuation from mid-2012.