Content uploaded by Carolyn M King
Author content
All content in this area was uploaded by Carolyn M King on Jul 03, 2014
Content may be subject to copyright.
New Zealand Journal of Zoology, 2001,
Vol.
28: 57-78
0301^223/01/2801-0057 $7.00/0 © The Royal Society of New Zealand 200157
Population biology of the ship rat and Norway rat
in Pureora Forest Park, 1983-87
J. G. INNES
Landcare Research
Private Bag 3127
Hamilton, New Zealand
C. M. KING
Department of Biological Sciences
University of Waikato
Private Bag 3105
Hamilton, New Zealand
M. FLUX
230 Belmont Hill Road
Lower Hutt, New Zealand
M. O. KIMBERLEY
New Zealand Forest Research Institute
Private Bag 3020
Rotorua, New Zealand
Abstract Populations of ship rats {Rattus rattus)
and Norway rats
(R.
nofvegicus) were sampled over
the five years 1983-87 at Pureora Forest Park, by
Fenn and rat kill-traps every three months. Fenn and
rat traps recorded similar capture rates in compara-
ble habitats, although Fenns caught more heavy and
fewer young
rats.
Ship rats (n = 1793 collected) were
more abundant, heavier and larger in native forest,
regardless of logging history, than in exotic forest
of any age. Young ship rats (age classes 1-3) were
most abundant in unlogged interior native forest, and
in autumn and winter after summer and autumn
breeding. Capture rates declined after peaking in
1985,
probably due to reduced recruitment of young
rats following lower pregnancy rates in adult fe-
males.
The irregular annual seasonal cycle of
repro-
duction and abundance observed at Pureora is the
same as that described for non-commensal ship rat
Z00035
Received 20 June 2000; accepted 4 October 2000
populations elsewhere in New Zealand and the
world. Thirty five of 43 Norway rats collected came
from a single trap by the Waipapa Stream, apparently
set near a permanent colony. Pregnant female Nor-
way rats were trapped in every season, suggesting
year-round breeding. This implies that both species
can recover rapidly after control operations con-
ducted at any time of
year,
but especially in spring
and summer. Future research should include manipu-
lative exploration of factors limiting ship rat abun-
dance and Norway rat distribution.
Keywords rats; Rattus rattus; Rattus norvegicus;
New Zealand; Pureora Forest Park; native podocarp
forest; exotic forest; measurements; colour morphs;
population structure; age; reproduction; recruitment
INTRODUCTION
Pureora Forest Park comprises a large tract of
indigenous forest on the volcanic plateau of the
central North Island of New Zealand. From
November 1982 to November 1987 we studied the
introduced small mammals resident in the Park,
which at that time also contained extensive exotic
forests inside its legal boundaries (King et
al.
1996c).
The research was intended to provide background
information relevant to the control of predators of
declining endemic forest fauna such as the kokako
(Aves; Callaeas cinerea wilsoni Gmelin). We in-
cluded in our survey all three mustelids present in
New Zealand (Mustela erminea, M. nivalis and M.
furo),
both European rats (Rattus rattus and
R.
nor-
vegicus),
and the feral cat (Felis catus), hedgehog
(Erinaceus europaeus) and house mouse (Mus
musculus).
The Norway rat is the largest, and was apparently
the first, of the European rats to reach New Zealand,
aboard European and American sailing ships from
1769 onwards (Moors 1990). It soon became
widespread and abundant on the mainland, but since
about the 1860s it has been replaced throughout the
North Island by the ship rat (Atkinson 1973), much
58New Zealand Journal of
Zoology,
2001, Vol. 28
North Island
KEYPureora
4 Trapline /Park H.6.
Road
Exotic plantationv
Logged native forest
Unlogged native forest
SOUTH
BLOCK2 kmMt Pureora :
Fig. 1 Map of the study area.
Pureora Forest Park boundaries are
shown
as
they were during
the
field
work. The Forest Park Headquar-
ters includes the Visitor Centre and
the meteorological station. The
former settlement at Barryville is
now deserted except for the old
sawmill (closed). Traplines are
identified by their codes (see text):
trap FU16, on the bank of the
Waipawa Stream, sampled a local
population of Norway rats.
as the Norway rat had replaced the smaller kiore (R.
exulans) that had arrived with Polynesian voyagers
(Atkinson & Moller 1990), at a date still being
debated (Holdaway 1999). The common rat at
Pureora today is certainly the ship rat. By
comparison, few Norway rats were collected during
our study, but our research on them is significant
because it is the first to document a mainland non-
commensal population. In commensal habitats such
as rubbish tips and farm buildings, Norway rats are
still more common than ship rats (Moors 1990).
King et al. (1996c) described the field data on
abundance, distribution and habitat preferences of
the eight species regularly monitored. Ship rats were
captured both in Fenn traps set for mustelids and in
rat traps set for rodents, and both techniques showed
that they were common and widespread. Ship rats
were more abundant in native forest, regardless of
logging history, than in exotic forest of any age. In
native forest they were equally abundant on forest
edges and in the interior, but catch rates were highest
on warmer steeper sites and lowest in early success-
ional sites. They were virtually absent from young
(4-9 years) exotic plantations, but present in older
stands which had woody understories of native
fruiting species. Ship rats were trapped most
Innes et al.—Population biology of ship and Norway rats59
frequently in autumn and winter and least so in
summer. Their abundance peaked at up to 20 rats per
100 TN (trap-nights; a trap-night is
1
trap set for 1
night) in 1985 in all forest types. Figures are cor-
rected for sprung traps after Nelson & Clark
,
1973).
By contrast, Norway rats were rare, caught only in
Fenn traps, and only in native forest.
This is the last of the three companion papers to
King et al. (1996c), which describe the results of
systematic necropsy of the trapped animals. The
previous two dealt with mustelids and feral cats
(King et al. 1996a), and feral house mice (King et
al.
1996b). The aim of this paper is to record the
measurements, population structure and repro-
duction of ship and Norway rats at Pureora in relation
to habitat, season and year. The results described in
this paper replace the preliminary figures quoted by
Innes (1990).
Previous studies of the biology of ship and
Norway rats were summarised by Innes (1990) and
Moors (1990) respectively. Long term studies of ship
rat demography initiated in the 1970s by the former
Department of Scientific and Industrial Research in
the Orongorongo Valley, Wellington and on Mt
Misery, Nelson remain largely unpublished, except
for the preliminary accounts given by Brockie (1992)
and Wilson etal. (1998).
STUDY SITES
At the time of our study, Pureora State Forest Park
occupied 75 000 ha of the ranges west of Lake
Taupo. Our three study sites were all within 12 km
of Pureora Village, at altitudes ranging from 550 to
700 m above sea level (Fig. 1). Two were in logged
and unlogged native podocarp-broadleaved forest,
and the third was in exotic forest (mostly Pinus
radiatd).
For further details on the study areas and
field routines, see King et al. (1996c).
In logged native forest, we set rodent traplines to
sample both the narrow strip (6-12 m) of dense
cover along a road edge (line RL1) and the forest
interior
up
to 500 m from the nearest road (line RL2).
A third rodent trapline (RU) sampled unlogged,
unroaded native forest. In the exotic forest we set a
rodent trap line in a 724 ha area of young Pinus
radiata plantation established in 1978 (line RE),
adjacent to the Pikiariki Ecological Area, but not in
the older plantations east and south of
it.
However
the latter were sampled by
a
Fenn trapline (FE2), and
a Fenn line also sampled the young exotics (FE1);
both were set along
roads.
A roadside Fenn line (FL)
sampled logged native forest at Ngaroma in the north
of the study area, and the final Fenn line (FU) crossed
the unlogged native forest of the Waipapa Ecological
Area. The "roads" were single-lane gravel tracks
carrying about 0-10 vehicles per day.
METHODS
Rodent traplines
Rodent traplines all had 36 trapping stations at 50 m
intervals (total length 1.8 km). One each of the
"Supreme Eziset" rat and mouse traps was set at each
site,
baited with peanut butter and rolled oats,
according to the standard method established by B.
M. Fitzgerald (Fitzgerald & Karl 1979; Innes 1990).
After a pilot trapping session in November 1982, all
lines were set and inspected daily for three days
during four trapping sessions
a
year, in the last weeks
of February, May, August and November of 1983-
87 inclusive. One line
(RL1,
logged forest road edge)
was closed after February 1985 (Fig. 2). The wooden
bases of all traps were soaked in linseed oil before
first
use,
and the springs were oiled periodically. We
inspected all traps daily and recorded their condition,
according to the routines described by Cunningham
& Moors (1983).
Fenn traplines
The 122 steel Fenn traps were set singly, in wooden
tunnels to protect non-target species (King & Edgar
1977),
baited with fish-based catfood, and inspected
daily . After 1984 we fixed two horizontal wires
across the tunnel entrances in an effort to reduce
interference by possums (Trichosurus vulpecula).
Fenn traps were all spaced at 300 m intervals and set
in the last weeks of January, April, July and October
for 1982-87 inclusive, but the lines were of variable
length depending on the extent of suitable habitat
available. For further details see King etal. (1996c).
Laboratory procedures
All animals caught were returned to the laboratory
frozen, and later examined for a standard list of
physical attributes. The number of rats available for
examination or dissection was always less than the
total trapped, because some were scavenged in traps
or damaged during storage, transit or processing.
Norway rats were put through the same procedures
as ship rats.
We collected fleas (by brushing the fur, and
inspecting the plastic collecting bag) only from rats
killed during the winter of 1987 (the Fenn trapping
60New Zealand Journal of
Zoology,
2001, Vol. 28
RL1
RL2
Fig. 2 Density indices (rats cap-
tured perl00 trap-nights) for ship
rats through the five years, in rat
traps (prefix R) and Fenn traps
(prefix F). A Captures in logged
native forest, along traplines FL
(road edge), and RL2 (forest inte-
rior);
B
captures in unlogged forest
interioralong traplines
RU
and FU;
C captures in exotic forest along
traplines RE and FEl, and FE2 (in
older plantations).
Oct. 86Oct. 87
AFU
~T 1 T~T~ "T—T ~1 r"~l 1 1
T~"T~T—T
Oct. 82 Oct. 83 Oct. 84 Oct. 85 Oct. 86 Oct. 87
C
D) 6
J2
2
en
0 J
' <>FE2
Oct. 82 Oct. 83 Oct. 84 Oct. 85 Oct. 86 Oct. 87
Innes et al.—Population biology of ship and Norway rats61
in July, and the rodent trapping in August). Fleas
found were stored in 70% alcohol, and later passed
to B. M. Fitzgerald for identification.
Measurements and morphs
We recorded whole body weight, paunched weight
(after removal of stomach), total length and tail
length, omitting rats that had been chewed in the trap
or were missing tails or feet. In analyses of weight
we excluded the pregnant females, and in analyses
of length we excluded any rat with a damaged spine
or tail. We classified each ship rat as representing
the "rattus", "frugivorus" or "alexandrinus" morph;
for colour illustrations of
each,
see King (1990).
Age determination
We classified all rats of both species into one of the
seven age-classes described by Karnoukhova (1971)
based on molar toothwear, as illustrated in Innes
(1990).
This method gives consistent results open to
comparison with other New Zealand studies, but it
has never been calibrated against known age animals
in New Zealand. Moller & Tilley's (1986) recom-
mended modifications of the method for Norway rats
were published after most of our laboratory work
was complete.
Criteria for assessing reproductive activity
The condition of the vagina (perforate or imper-
forate) is an unreliable indicator of reproductive
condition, so was merely recorded without further
analysis.
We took active pregnancies and lactations to
define present breeding, although the number of
actively breeding females is thereby underestimated.
The gestation period of
ship
rats is 20-22 days, and
of the Norway rat, 22-24 days (Brooks & Rowe
1987),
but the embryos are not visible to the naked
eye for about the first week in either
species.
Females
that had bred at some time in their lives, though not
necessarily during the sample period in which they
were collected, were defined as those that were
pregnant, lactating, or with uterine scars. Uterus
condition was classified into one of three standard
categories of relative thickness. Those classed as
"thread" could be either immature or mature but
quiescent. Enlarged uteri associated with ovarian
activity were classed as "string" or "cord".
Litter size was estimated as the mean number of
viable embryos (excluding resorptions) per pregnant
female. The traditional criterion indicating breeding
condition in males, the position of the testes (scrotal
or abdominal), is unreliable, so we determined male
breeding condition mainly from whether or not the
tubules in the epididymides were visible. We also
tried a new method, based on the length and width
of the testes in mm. By assuming that width and
depth were equal, we derived a rough estimate of
testis volume (length x width x depth = volume in
mm3).
We defined recruitment as the addition of young
rats (age classes 1-3) to the trapped samples. The
category "young" refers to chronological age as
reflected by tooth wear, not to reproductive maturity.
Statistical analysis
The statistical analysis was conducted using means
calculated for each trapline x trapping session. The
statistical packages SAS and GENSTAT were used
to perform a multi-factor analysis of variance
(ANOVA) for each of the variables of interest (age,
sex ratio, weight, length, reproductive variables etc).
The factors included in the ANOVA model were
trapline, year and season. The data from the pilot
trapping session in November 1982 were excluded
from analyses involving comparisons between years
and seasons, since they represented only one season
of that year. These procedures are able to accom-
modate unbalanced sample sizes and can test for
differences in each variable whilst controlling for all
the other variables appropriate to each comparison.
Because the samples were large, most of the
variables could be analysed adequately with PROC
GLM. Percentage variables were better handled by
generalised linear models with logit link function and
binomial error function, and by using deviance ratios
to test the significance of each factor. In all tables,
significant differences were assumed if P
<
0.05.
Because ship rats were abundant, and frequently
caught both in rat and in Fenn traps, we were able
to compare population parameters derived both from
rodent traplines and from Fenn traplines. The two
types of traps sampled the same areas, but they were
set at different spacings, with different baits and in
different months. The mechanism of the Fenn trap
is also heavier than that of the Supreme rat traps,
which might affect the weight distribution of the
captures, so we analysed the results given by the two
trap types separately.
Because each of the main habitats of interest was
represented by a single trap line without replication,
we could not make any true tests of difference be-
tween habitats. Instead, the line x year interaction
term was used as the error term for testing for dif-
ferences between lines. This procedure will detect
any differences between habitats that remained
62New Zealand Journal of Zoology, 2001, Vol. 28
consistent over the five years covered by the study.
Similarly, season was tested against the season x
year interaction. Other factors were tested against the
residual error. Least significant differences (LSDs)
were used to detect significant (at a = 0.05) differ-
ences between adjusted means. The raw data are
available on request from MOK.
RESULTS
Ship rats
Distribution of captures
A total of 1793 ship rats was collected, mostly from
indigenous forest (Table
1).
Only seven of 179 traps
of both types set in native forest caught no ship rats;
60 traps caught
1
-3,
and 14 traps caught 20 or more
each. At the broad scale on which we sampled, the
rats were virtually ubiquitous, although the skewed
frequency distribution of trap success suggests that
rat density was locally variable (King et al. 1996c).
Comparisons between catches in Fenn versus rat traps
More ship rats were caught in Fenn traps than
Supreme rat traps (Table 1), since the total number
of corrected trap nights recorded by Fenns was much
higher. However, when the two trap types were set
in comparable habitats, the capture rates they
recorded were similar (Fig. 2A, C).
Rats caught in Fenn traps were significantly larger
and heavier than those from rat traps (Table 2),
probably because Fenns have a stronger trigger
mechanism. Fenns also caught proportionately fewer
young rats (age classes 1-3), of which none was from
age class
1
(Table 2). A logistic regression was used
to test the relative importance of weight and age to
explain these results, since the two parameters are
correlated. Each parameter fitted individually
affected the catch (for age as a linear function, P
=
0.0001;
for weight fitted by class, P = 0.0001) but
in combination both were still significant (for age,
P
---
0.0005; for weight, P
=
0.005).
Overall, 47% of the rats collected were males
(Table
3).
Gender ratios of the samples from the two
trap types were the same, even though males are
significantly heavier than females (Table 4). The
minimum weight threshold required to set off a Fenn
trap appears to be (usually) above the weight of the
smallest rats (age class 1, averaging
41
g, Table 4).
We had previously shown (King et al. 1996c) a
significant decline in numbers of captures of
rats
in
Fenn traps through each 10 day trapping session,
usually to about half the initial catch by the tenth day.
Here we examined the necropsy data to see if there
were any differences in dominance (as indexed by
size and age) or breeding history between the rats
caught at the beginning and end of
the
10 day Fenn
trapping sessions. The capture records for rat traps
were too few, and the trapping periods too short, to
do the same for them.
There were no significant correlations between
the percentages of males with visible tubules, the
Table 1 Numbers of ship and Norway rats captured, and trapping effort (CTN: corrected trap nights, corrected for
unavailable traps; Nelson and Clark 1973), listed by habitat and trap type (from King et al. 1996c). The capture totals
include scavenged or severed remains, provided that another individual of that speeies with the same part missing was
not later captured in the same area. A 'trap-night' is one trap set for one night.
Ship rat
Norway rat
Total CTN
Trap
type
Mouse
Rat
Fenn
Mouse
Rat
Fenn
Mouse
Rat
Fenn
Unlogged
native forest
interior
1
106
473
0
0
35
2160
2157
6281
Old
Logged
native forest
interior
1
161
-
0
0
-
1969
2007
0
Habitat
exotic
Logged
native forest
roadedge
1
91
658
0
0
8
977
991
7746
plantation
(planted
before 1966)
„
-
299
_
—
0
0
0
8166
Young exotic
plantation
(planted 1978)
0
2
0
0
0
0
2085
2156
2079
Totals
3
360
1430
0
0
43
7191
7311
24 272
Innes et al.—Population biology of ship and Norway rats63
mean number of uterine scars in females, or the
distributions of age class, weight or total length with
trapnight (numbered 1-10) for rats caught in Fenns.
However there was a significant (P
=
0.026) positive
correlation between trap-night number and the
percent of the catch comprising females which had
bred or were pregnant or lactating; that is, breeding
females tended to be caught later in the trapping
period.
Because the two trap types sampled the popu-
lations differently, we have adjusted all the results
for trap type or else present them separately in the
analyses that follow.
Population structure
Rat traps set in unlogged interior native forest (line
RU) consistently caught proportionally more young
rats (age classes 1-3) than traps in logged interior
Table
2
Effect
of
trap type (Fenns compared with
rat
traps)
on the
catch
of
ship rats.
Whole body weight
Paunched weight
Total length
Tail length
%
young
(age
classes
1-3)
%
male
Distribution
of
age classes:
Class
1
Class
2
Class
3
Class
4
Class
5
Class
6
Class
7
Fenn
N
1
149
1148
1050
1074
1256
1256
0
31
253
585
332
49
6
traps
mean
134g
127
g
366
mm
193
mm
19
%
46%
0
2.5%
20.1
%
46.6%
26.4%
3.9%
0.5%
N
307
306
318
319
331
331
8
16
87
141
70
5
1
Rat traps
mean
124
g
1
18
g
354
mm
188
mm
34%
46
%
2.4%
4.9%
26.5%
43.0%
21.3%
1.5%,
0.3%
P-valuc
0.002
0.002
<
0.0001
0.0002
<0.0001
0.38
0.0001
0.0351
0.0151
0.2716
0.0694
0.0522
0.6743
Table
3 Age and
gender of ship rats
by
trapline, season
and
year.
For
each factor, values
in a
column followed
by the
same letter
do not
differ significantly
at oc= 0.05.
Habitat
&
trap type
Season
Year
Total
RU
FU
RL2
RL1
Fl.
FE2
Spring
Summer
Autumn
Winter
1983
1984
1985
1986
1987
"A
n
92
411
157
82
588
257
473
298
353
463
298
259
451
300
279
1587
i young (class
mean
41
28
24
35
19
21
19
16
43
33
26
37
37
16
21
25
1-3)
a
abc
be
ab
c
c
a
a
b
b
ab
a
a
b
b
n
92
411
157
82
588
257
473
298
353
463
298
259
451
300
279
1587
%
male
mean
37
43
45
57
50
45
43
51
56
45
54
51
48
41
50
47
a
a
a
a
a
a
c
ab
a
be
a
ab
ab
b
ab
64New Zealand Journal of
Zoology,
2001, Vol. 28
native forest (line RL2: Fig. 3, Table 3), and this
difference was also apparent although not significant
with the catch in Fenns (FU compared with FL: Fig.
4,
Table 3).
Young rats were more abundant (from a third to
almost
a
half of the total catch) in autumn and winter,
after recruitment from summer and autumn breeding,
but they still averaged about one-sixth of the catch
in spring and summer (Table 3). However, the
youngest rats of age classes 1-2 could appear in the
traps at any season, indicating year-round breeding.
Proportionately more young rats were trapped in
1984 and 1985. when numbers were high (Fig. 2)
relative to other years. During the following (post-
peak) year, from October 1985 to July 1986,
significantly fewer young rats were trapped than at
the same time in other years, although no fewer
females were pregnant, nor males sexually active, at
that time (Table 5).
Gender ratio did not vary significantly between
lines,
seasons or age classes, although fewer of the
trapped samples were males in 1986 than in other
years.
Measurements
Rats collected in the Fenn traps on line FE2, set in
the older exotic plantations, were slightly shorter and
lighter than those collected in Fenns set in native
forest (Table
4;
means adjusted for sex and
age).
The
differences in length were significant in comparison
Table
4
Weight and length of
ship
rats by habitat, season, year, age class and gender. For each factor, values in any
column followed by the same letter do not differ significantly at
cx=
0.05.
Line
RU
FU
RL2
RLI
FL
FE2
Season
Spring
Summer
Autumn
Winter
Year
1983
1984
1985
1986
1987
Age
1
"
2
3
4
5
6
7
Sex
Male
Female
Total
Whole body
n
83
368
145
76
527
238
447
236
303
451
253
234
426
266
258
8
45
318
648
362
50
6
824
613
1437
weight
(g)
mean
127a
132
a
133
a
1
18
ab
134 a
128
b
128
b
133
ab
127a
127ab
126
b
125
ab
129
a
138a
127
ab
41
a
79
b
111
c
134
d
148e
152
c
l50ed
141
b
121
a
132
Paunched
n
83
367
145
75
527
237
446
235
303
450
250
234
426
266
258
8
45
317
646
362
50
6
822
612
1434
weight
(g)
mean
120ab
126 a
126 a
113
b
127a
121
ab
122
a
126 a
121
a
121
a
119b
119b
123
b
131a
121
b
39 a
75
b
106
c
127d
140
e
143
e
139
ed
134
b
1
14
a
126
Total
n
85
339
151
79
488
213
401
257
304
393
263
228
391
242
231
8
40
295
619
344
45
4
732
623
1355
length
(mm)
mean
361
ab
366
a
363
ab
347
c
365
a
357
be
362
a
361
a
357
a
361
a
354
c
356
be
364
ab
371
a
357
be
246
a
318b
346
c
366
d
374
e
378
e
380
ed
368
b
358
a
364
Tail
n
86
349
151
79
494
215
405
262
310
397
265
230
397
246
236
8
40
299
630
347
45
5
738
636
1374
length
(mm)
mean
191
ab
195
a
192
a
186
b
193
a
187b
192
a
190
a
190 a
193 a
188
b
188
b
192ab
196
a
190
b
130
a
170
b
184 c
194
d
197
e
197
ed
189
edc
194
a
192
a
193
Innes et al.—Population biology of ship and Norway rats65
Fig. 3 Age distributions of ship
rats caught in rat traps, by season,
in unlogged native forest interior
(RU) and in logged native forest
interior (RL2).
I i | i i
10
RURL2
r
P<
W
1
2
3
4
5
6
7
MM
n = 36
1)
E
F
1
2
3
4
5
6
7
S3
an = 21
<
1
2
3
4
5
6
7
a
n = 15
••
31
'"TT
20 30 40
Per cent
i'
111
5060
1
2
3
4
5
6
7
3
iH
i
•
i i |
n = 42
11111
•
111
g
•
1111 1111
111
•
i
1n
=
31
n
= 49
11 11111 i i
35
10 2030 40
Per cent
50 60
with lines FU and
FL,
and the difference in paunched
weight was nearly so for FU. Neither mean weight
nor length differed between seasons, although rats
trapped in 1985 and 1986 were larger than rats from
other years. Males were on average longer and
heavier than females.
Mice collected from Pureora during the same
study showed significant seasonal variations in the
inter-relationships between age, whole body weight
and head-body length (King et
al.
1996b). Mice of a
given age were significantly larger in summer than
at any other season. There was no such seasonal
variation in ship rats. Both weight and length
increased significantly with age class up to class 4,
after which neither parameter increased. In fact, rats
in age class 7 regressed in size, so that they were not
heavier or longer than rats in age class 4 (Table 4,
Fig. 5).
66New Zealand Journal of
Zoology,
2001, Vol. 28
FU
1
2
3
E 4
6
7
1
2
^ 3
0)
I4
6
K
6
7
1
2
^ 3
s
I 4
^5
6
7
]
11111
<
11
>
11111111111
n
=
111
=
147
111 II i
n = 69
11 "n I II II I ir | i |
n = 76
11
'''
• 1
• i ' M
n =119
0 10 20 30 40 50 60
Per cent
1
2
3
4
5
6
7
ill
I
1 1 1 1
FL
]
mm n
= 152
1 1 1 1
[
1 1 < 1
|
1 1 1 1
|
1
I
1
' I
1 1 < 1
|
Fig. 4 Age distribution of ship
rats caught in Fenn traps, by sea-
son, in unloggcd native forest inte-
rior^
FU)
and
in
logged native forest
roadside (FL).
1
2
3
4
5
6
7
I
11
a
i
111111
•
111
> >
n = 118
I
• • • • I
•
• • •
1
2
3
4
5
6
7
m
1
IK?!
fill
n= 143
n
= 175
0 10 20 30 40 50 60
Per cent
Colour morphs
Of 1587 individuals classified for colour morph,
79%
were the brown backed, white bellied "fru-
givorus" morph,
13%
were the dark "rattus" morph,
and 8% were the brown-backed, grey-bellied
"alexandrinus" morph. In the North Island, the
"alexandrinus" morph had previously been recorded
only from Northland, although it is the commonest
morph in South Island samples (Dowding & Murphy
1994;
Innes 1990; Smith 1986), and is the typical
morph in UK (Corbet & Harris 1991).
The proportion of Pureora rats classified as
"rattus"
morph did not vary significantly between
seasons, years or genders. Fewer "rattus", and
more "frugivorus" were collected from the older
exotic plantation (line FE2) compared with other
habitats. The "rattus" morph rats tended to be
lighter (mean weight 130 g, n = 208) than
"frugivorus" rats (134 g, n = 1288), although not
significantly so, and proportionately fewer
"rattus"
were caught in Fenn traps (12% of 1181)
than in rat traps (16% of 295). Proportionately
Innes et al.—Population biology of ship and Norway rats67
Fig. 5 Body weight, paunched
weight, total length and tail length
of ship rats by age class (see Table
4).
—•— Body weight
—•— Paunched weight
••• -Total length
•••Tail lenqth
Age class
Table 5 Age, weight, length and reproductive status ot
ship rats trapped in the period of declining numbers (Oc-
tober 1985 to July 1986) compared with samples taken in
the same period in other years.
Po
(Oct
% young
% male
Weight
Paunched weight
Length
Tail length
"o perforate
"o pregnant
No.
embryos
No.
scars
% preg orlact
% with scars
% with scrotal testes
% with visible tubules
Testes volume
st-peak year
85-July 86)
17.5
50
137
130
372
198
98.6
19.9
5.8
8.7
36.8
59.3
71.7
80.6
2566
Other years
30.6**
50
128**
121 **
358**
190**
97.5
16.2
5.2
8.6
29.4 +
48.9 +
59.6**
72.1 *
2381 +
+ difference significant at a = 0.1; * significant at a
0.05;
** significant at a = 0.01.
more "frugivorus" than "rattus" morph rats were
trapped in exotic forests.
Reproduction
Mean measures of several parameters describing the
female breeding cycle are shown in relation to
habitat, season and year in Table 6.
The proportion of females with perforate vaginae
increased rapidly from age classes 1-2 (60-70%) to
age classes 3-7 (98-100%), so this parameter is not
informative. No indicator of female reproductive
activity (% pregnant; mean number of live embryos;
% with uterine scars; mean number of uterine scars;
% pregnant; % lactating; % pregnant females with
resorbing embryos) varied significantly with habitat.
However, as expected from recruitment data,
pregnant rats were trapped most often in summer and
autumn. Fewer pregnancies were recorded in spring,
and only six from winter during the 5 years of
trapping, all from lines FU or RU in 1983 or 1984.
The mean number of embryos per female varied
slightly with season, from 4.2 in spring to 5.9 in
winter, although not significantly. A histogram of the
number of uterine scars borne by 346 females (Fig.
6) shows peaks at 5-6, 10-11 and 18 scars, corres-
ponding to 1, 2 and 3 litters respectively, but with
considerable spread. Most pregnant females trapped
in 1983 had resorbing embryos, significantly more
than in other years.
No females in age class
1,
and few in age class 2,
were pregnant, lactating or had uterine scars when
trapped, but most class 4 females were pregnant in
summer and most class 5 females in autumn. There
were no significant differences between age classes
in the proportion of females pregnant or lactating in
winter or spring. Most age class 4 females trapped
in spring had not yet bred at all, whereas most
females of age classes 5-7 trapped in spring and
winter had bred in the previous summer or autumn.
In all seasons, litter size was no greater in older
females than young ones, but significantly more
females of age class 5-7 had uterine scars (Table 6).
As expected, most males had scrotal testes and
visible tubules in the cauda epididymus in spring,
68New Zealand Journal of
Zoology,
2001, Vol. 28
Fig. 6 Frequency distributions
of uterine scars in female ship rats.
The range 1-8 scars peaking at 6
represents females that have had
only a single litter; the range 8-14
peaking at 10-1 1, those that have
had two litters, and the range of
14-23 peaking at 18, those that
have had three litters.
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Number of uterine scars
Table 6 Parameters of reproduction in female ship rats, by line, season, year and age class. For each factor, values in
any column followed by the same letter do not differ significantly at a = 0.05.
Line
RU
FU
RL2
RL1
FL
FE2
Season
Spring
Summer
Autumn
Winter
Year
1983
1984
1985
1986
1987
Age
1
2
3
4
5
6
7
Total
% perforate
n
78
52
122
200
258
33
200
147
187
209
155
125
206
122
135
5
19
161
337
187
29
5
743
mean
100 a
98 a
98 a
89 a
100
a
99 a
98 a
98 a
97 a
98 a
97 a
98 a
98 a
98 a
98 a
60
68 b
99 a
100
a
100 a
100
a
100
99
% pref
n
77
52
122
200
258
32
199
148
186
208
155
122
207
122
135
4
20
162
335
186
29
5
741
jnant
mean
20 a
19a
15
a
11
a
20 a
17a
9c
36 a
21 b
2c
25 a
16abc
9c
21 ab
13
be
0
5ab
11
b
19a
19a
14
ab
0
16
No.
embryos
where
n
9
6
13
23
35
3
14
39
32
4
16
17
18
24
14
0
0
13
52
21
3
0
89
present
mean
3.74 c
4.60 be
5.43 abc
7.30 a
5.15 abc
5.77 ab
4.27 a
5.59 a
5.62 a
5.90 a
4.25 b
5.92 a
5.49 ab
5.70 a
5.37 ab
-
5.23 a
5.17a
5.33 a
5.33 a
-
5.20
No.
:
scars
where present
n
30
18
58
102
115
17
72
67
99
102
75
61
94
50
60
0
0
33
146
137
20
4
340
mean
8.79 a
8.73 a
8.75 a
9.64 a
8.21 a
7.90 a
8.87 a
6.87 a
8.82 a
10.07 a
7.01 a
8.90 a
8.98 a
7.81 a
10.59 a
-
5.79 b
7.14b
10.25 a
10.55 a
16.25
8.60
% breeding
n
71
50
116
187
242
29
190
131
171
203
155
107
189
117
127
4
20
155
314
172
25
5
695
mean
29 a
33 a
24
a
40 a
29 a
31 a
16b
56 a
44 a
6b
34 a
31 a
24 a
38 a
24 a
0
5c
18c
27 b
38 a
20 be
0
26
% with
n
66
45
108
182
237
31
170
123
174
202
148
103
193
94
131
4
20
155
302
162
21
5
669
scars
mean
42 a
57 a
46 a
50 a
47
a
57 a
40 a
53 a
56 a
51 a
49 a
57 a
48 a
50
a
47 a
0
Od
21 c
48 b
85 a
95
a
80
51
Innes et al.—Population biology of ship and Norway rats69
summer and autumn (Table 7). However, even in
winter, when few females were pregnant, most males
still had visible tubules. In nearly half of them the
testes were still scrotal, although the volume of the
testes regressed significantly in autumn and winter.
There was no significant difference in testis
volume between the age classes, but most males
whose testes remained scrotal in winter were age
class 5 or older. By spring a majority of age class 4
males'
testes were scrotal, and by summer this was
true of age class 3 males.
Norway rats
Distribution of captures
A total of 43 Norway rats was collected (31 males,
11 females), but only in Fenn traps in native forest
(Table 1). Thirty five of them came from a single trap
on a stream bank at the edge of the unlogged forest
of the Waipapa Ecological Reserve (FU16, see Fig.
1).
This trap produced a regular crop of 4-10 Norway
rats of both sexes per year from an apparently
permanent colony nearby (Table 8), as well as 13
ship rats and three stoats. The rest of the Norway rats
came from eight sites scattered through the logged
forest along the Ngaroma Road, each yielding only
a single rat.
There were no significant differences in the
distributions of captures by sex, season and year,
probably because the samples were too small to
detect any.
Age and measurements
The frequency distribution of age classes (Table 9)
approximated a normal curve, peaking at class 6. We
do not know what chronological age this represents.
As in all rodents, body measurements varied
significantly with age class. Table 9 presents least
Table 7 Parameters of reproduction in male ship rats, by line, season, year and age class. For each factor, values in
any column followed by the same letter do not differ significantly at a = 0.05.
Line
RU
FU
RL2
RL1
FL
FE2
Season
Spring
Summer
Autumn
Winter
Year
1983
1984
1985
1986
1987
Age
1
2
3
4
5
6
7
Total
% with scrotal
n
78
39
131
199
322
48
258
147
162
250
140
133
241
174
129
3
26
177
376
208
25
2
817
testes
mean
51 a
62 a
69 a
59 a
63 a
62 a
68 ab
75 a
60 b
41 c
55 b
52 b
60 b
77 a
63 b
33
12
d
38 c
66 b
80 a
68 ab
50
62
% with
n
79
40
133
207
325
48
270
147
162
253
140
133
241
177
141
3
26
179
383
214
25
2
832
visible tubules
mean
65 a
71 a
78 a
68 a
78 a
77 a
92 a
81 a
56 b
61 b
65 a
75 a
76 a
80 a
67 a
0
8c
48 b
86 a
92 a
92 a
100
77
Testes
n
79
40
133
206
318
48
270
138
162
254
133
132
241
175
143
3
26
179
378
211
25
2
824
volume (ce)
mean
2.19b
2.34 b
2.54 b
2.14b
2.67 a
2.66 a
2.73 a
2.87 a
1.89 b
2.19b
2.44 a
2.41 a
2.23 a
2.70 a
2.33 a
0.13
0.37 d
1.50 c
2.66 b
3.24 a
3.53 a
2.49
2.50
70New Zealand Journal of Zoology, 2001, Vol. 28
squares estimates calculated by the weighted means
model, in order to compensate for the uneven
numbers of rats in each group. Such a small data set
allows few conclusions, except that Norway rats
appear to be still growing up to and including age
class 6, and that the reduction in weight but not in
linear measurements in the oldest rats of both sexes
suggests a drop in condition.
Reproduction
Six of the eleven females were pregnant. Three were
caught in spring, and one was also lactating; the other
three represented each of the other seasons, one
apiece, but none was also lactating. One fully mature
female from each habitat caught in autumn/winter
was lactating but not pregnant. One from each habitat
in spring had bred before, although neither was
pregnant or lactating then. To find such even
representation of seasonal activity in such a small
sample suggests that the breeding season in Norway
rats at Pureora is, like that of ship rats, not well-
defined.
Three females collected from trap FU16 had 6, 8
and 8 live embryos of 2-15 mm crown-rump length,
none resorbing, and one had 7 embryos in unspeci-
fied condition. One female from the logged forest
had 7 embryos, 4 live plus 3 resorbing, and another
had 6 embryos, all resorbing. In 8 females checked
Table 8 Seasonal and annual distribution of captures of Norway rats.
Spring
Summer
Autumn
Winter
Totals
1983
1984
1985
1986
1987
Totals
Line FU
Male
10
7
7
5
29
4
9
5
5
6
29
Numbers
unlogged forest
Female
3
1
2
0
6
0
1
3
1
1
6
of rats caught
Line FL
Male
1
1
1
0
3
1
0
1
I
0
3
logged forest
Female
2
1
0
2
5
0
4
1
0
0
5
Table 9 Body measurements ± SE and sample size (n) of Norway rats by age class.
Age class
Males
3
4
5
6
7
F-ratio
All ages (n = 30)
Females
3
4
5
6
All ages(n= 10)
Whole body weight (g)
132 ±27 (2)
213 ±20 (5)
209± 16(10)
248
±
19(11)
233 ± 15(2)
2.19 (P= 0.10)
220 (range 104-345)
151 (1)
162 ±6 (2)
198+15(6)
152(1)
181 (range 150-230)
Total length (mm)
320 ±30 (2)
368+ 11 (5)
359 ±9 (8)
370 ±9 (9)
352+ 18(2)
1.57 (P= 0.22)
361 (290^18)
314(1)
341 ±6(2)
342 ± 12(6)
339(9)
339(312-384)
Tail length (mm)
149 ±20 (2)
163 ±4 (5)
160 ±4 (9)
164 ±4 (9)
160 ± 11 (2)
0.61 (P
=
0.66)
161 (130-180)
140(1)
158 ±6 (2)
155
±8 (6)
150(1)
154(121-175)
Innes et al.—Population biology of ship and Norway rats71
for uterine scars, five had none and the others (all in
age class 5) had 14, 15 and 20 visible scars. So far
as they go, these data suggest that litter size in non-
commensal Norway rats is about 6-8 in both
habitats.
The testes of most (27 of
30,
with 2 unclassified)
adult males contained visible tubules. The majority
had fully descended into the scrotal position in all
seasons except winter, but there were always some
in the abdominal position in any season, even with
visible tubules. The only males without visible
tubules were a juvenile (age class 3) caught in spring,
and two mature adults caught in winter. One of these
two had visible tubules in abdominal testes, and the
other had non-visible tubules in testes that were still
scrotal. It is impossible to say from these data
whether or not males remain fertile in winter, or
whether scrotal position reliably indicates breeding
condition in Norway rats.
Fleas
The fleas identified from rodents sampled during
winter 1987 are listed in Table 10. In such a small
collection, it is remarkable that the most common
flea on each species was different: Nosopsyllus
fasciatus dominated on ship rats, Pygiopsvlla hoplia
on Norway rats, and Leptopsyllus segnis on mice.
One female ship rat carried all three species (ten N.
fasciatus, one
P.
hoplia and one
L.
segnis), but most
rat carcases yielded only one or two fleas of a single
species.
DISCUSSION
This study was originally planned to survey the
species distribution in, and monitor the population
variations of, the small mammal fauna of podocarp
and exotic forest of Pureora Forest Park, with
especial reference to providing information to
managers about potential predators of
kokako.
We
used the same techniques developed previously in
beech forest (King 1983) in order to be able to
compare the faunas of small mammals living in these
different habitats. The beech forest traplines
collected mainly stoats and mice, plus a few ship rats
mostly in Fenn traps (to 1.7 C/100 CTN) in certain
years only (King & Moller 1997). In sharp contrast,
at Pureora the same methods collected eight species
of small mammals, including huge numbers of ship
rats (to 20 C/100 TN (Fig. 2A). Furthermore, the
large samples of ship rats from the mixed broad-
leaved podocarp forests of Pureora were distributed
relatively evenly between seasons, years and sites.
Smaller samples were collected from closed-canopy
exotic forest, but practically none from a young
plantation.
Both Daniel (1978) and Brockie (1992:158)
tabulated the abundance of ship rats in different parts
of New Zealand as determined by trap catch.
Unfortunately, most such comparisons are weak,
because not all studies used the same trap spacings,
layouts, seasonal schedule and bait. Nevertheless it
is clear that few ship rats inhabit dairy farmland or
pure beech forest. Numbers were highest (82
captures per 100TN) on Rosa I., off Stewart Is.
(Hickson et al. 1986). Feral cats and stoats, which
may limit ship rat numbers on the New Zealand
mainland (Daniel 1972, 1978; data from the
Orongorongo Valley, North Island; B. M. Fitzgerald
& B. J. Karl, unpubl.) are absent from Rosa I.
Ship rats
Demography
The demography of ship rats in podocarp forest at
Pureora is consistent with that reported from similar
non-commensal habitat elsewhere on mainland New
Zealand, as summarised by Daniel (1978) and Innes
(1990).
Table 10 Fleas carried by rodents at Pureora in winter 1987.
Host
No.
of hosts
examinedFlea species
No.
and sex
of fleas
Rattus rattus
R.
norvegicus
Mus musculus
15
2
8
Nosopsyllus fasciatus
Pygiopsylla hoplia
Leptopsyllus segnis
Nosopsyllus fasciatus
Pygiopsylla hoplia
Nosopsyllus fasciatus
Leptopsyllus segnis
13M-
2M +
3F
1 F
4M +
1F+
1
3M +
f20F
2F
2F+ 1?
M
5F+1?
72New Zealand Journal of Zoology, 2001, Vol. 28
In New Zealand, although male ship rats are
capable of producing sperm all year round, pregnant
or lactating females have been trapped mainly
between September and April. Winter breeding is
recorded occasionally, apparently following heavy
fruit fall. Otherwise, breeding normally stops in
winter, even in Northland, where temperature and
food quality do not seem to be limiting (Smith 1986).
At Pureora, pregnant rats were, as expected, trapped
most often in summer and autumn. Most males had
scrotal testes and visible tubules in the cauda
epididymes in spring, summer and autumn, and
young rats were significantly more abundant (from
a third to almost a half of the total catch; Table 3) in
autumn and winter after spring and summer breed-
ing. We collected significantly fewer ship rats in
Fenn traps in summer (King et al. 1996c), which is
consistent with sparse winter breeding.
Innes (1990) suggested that this seasonal
breeding, plus an annual pulse in recruitment of
young, would result in a corresponding seasonal
fluctuation in abundance, from a low in spring and
early summer to a high in autumn and winter. Fenn
trap success at Pureora increased between January
and July in 13 of
15
line x year comparisons, and in
eight comparisons it had increased again by October
(Fig. 2). Long-term studies of ship rat abundance
(Daniel 1978; King et al. 1996a) do indeed show a
weak annual cycle of abundance, more so in the
Orongorongo Valley (fig.
1
of Daniel 1978) than at
Pureora, perhaps because there was more winter
breeding at Pureora.
However, seasonal changes were small compared
with annual changes. At Pureora, Fenn traps detected
significant differences between years; abundance
was highest on all Fenn lines during July to October
1985.
Lines in unlogged native forest and older
exotics also caught many rats in July 1987, although
this second peak was not detected by line FL in
logged native forest. A similar clear peak in abun-
dance was recorded in the Orongorongo Valley in
late 1971, following the heaviest fruit fall of an 8-
year live-trapping study (Daniel 1978). Daniel
suggested (p. 146) that "...both the length of the
breeding season and the over-winter survival of ship
rats are directly controlled by the size of the autumn
seed and fruit crops".
Comparable increases at Pureora were observed
simultaneously in both the native and older exotic
forests. This suggests either that the plant species
responsible were the common shrub hardwoods (e.g.
mahoe Melicytus ramiflorus, five-finger Pseudo-
panax arboreus, wineberry Aristotelia serrata, pate
Schefflera digitata, Coprosma robusta) abundant
under both canopies, or that some other mechanism
is involved. We did not measure fruit abundance or
rat mortality in this study, and so cannot advance
Daniel's (1972) hypothesis that "...the combined
effects of continual predation by feral cats and
possibly mustelids and moreporks, and the
occasionally abundant food supply in winter, are the
two most important factors affecting the density of
ship rats.". Neither did we find any evidence at
Pureora to support Daniel's (1978) suggestion, that
many Orongorongo ship rats starve to death in
winters following light seed years. At Pureora the
rats were not lighter in winter than other seasons
(Table 4).
The increase in density indices for rats in all
habitats in 1985 followed a substantial reduction in
density indices for stoats and cats over the first two
years of the study (King et al. 1996c). In the
Orongorongo Valley, an increase in density of ship
rats following the removal of cats was interpreted by
Fitzgerald & Karl (1979) as evidence that predation
by cats controls the numbers of ship rats there.
However, Blackwell et al. (1998) experimentally
removed mustelids and feral cats from the Puketuku-
tuku Peninsula at Lake Waikaremoana, and rodents
(ship rats and mice) did not increase there relative
to a non-treatment block. They suggested that rodent
populations were more likely to be limited by food
supply than by predation.
We cannot explain why the proportions of female
ship rats pregnant, and the mean testis volume of
males,
were lower in 1985 and 1987 than in other
years,
as in mice collected from the same habitats
(King et al. 1996b). The declining pregnancy of
females in 1985 can be linked to a reduction in
recruitment of young rats and the consequent
population decline of October 1985-July 1996
(Table 5).
Indices of unmanaged ship rat populations
sometimes decrease steadily as a breeding season
progresses. We observed this effect from October
1985 to July 1986 on the Fenn line FU (native
unlogged) at Pureora (Fig. 2B), and in other studies
elsewhere (Innes et al. 1995, 1999). Such changes
in abundance remain inexplicable from these data.
Recruitment of young ship rats seemed to be more
effective in the forest interior than along the road
edges;
in mice the difference ran in the opposite
direction and was much more pronounced (King et
al.
1996b). Nevertheless, ship rats were indifferent
to the nett effects of the removal of
large
podocarps
and the increase in ground cover after logging, both
Innes et al.—Population biology of ship and Norway rats73
in their numbers (King et al. 1996c) and in their
breeding biology.
Smith (1986), Dowding & Murphy (1994) and
Alterio et
al.
(1999) caution that variation in kill-trap
success may be caused by factors other than abun-
dance, such as trappability, food supply and home
range size. For these reasons, kill trap studies are
easiest to interpret when they are compared with
non-treatment blocks, for examining a treatment
perturbation such as a control operation (Innes et al.
1995).
Female age and productivity
Laboratory studies show that ship rats wean at 21
28 days (Cowan 1981), and may reach sexual
maturity at 2-A months (Watts & Aslin
1981;
Brooks
& Rowe 1987). The actual chronological ages of
wild-caught ship rats in relation to their tooth-wear
indices are unknown, but it would be valuable and
interesting to know. Without this information, it is
impossible to check how far the lab data can be
applied to field conditions. Most age class 5-7
females trapped in spring had bred in the previous
season, suggesting they were at least eight months
old. having matured (minimum age two months) by
at latest the previous April. By contrast, most age
class 4 females trapped between September and
November had not yet bred, suggesting that they did
not mature in the previous breeding season and were
therefore only 6-8 months old by the following
spring. These ages are similar to those derived by
Karnoukhova (1971) from caged rats, and repeated
by Miller & Miller (1995) in their Rangitoto Island
research.
Overall mean litter size was 5.33 (±S.D. 1.2), and
the distribution of uterine scars (Fig. 6) indicated that
no female had had more than three litters when
trapped. These data would be consistent with a peak
lifetime productivity of 16 young per female. This
is less than the calculated 29 young per year of
Daniel (1972), who recorded a maximum longevity
of
17
months for
a
female ship rat. Daniel's mortality
data suggest that females in the Orongorongo Valley
may breed in one or two, but not three, seasons.
Ship rats in exotic forests
Exotic forests are apparently poor habitat for ship
rats,
especially the younger plantations with open
canopy. Only two ship rats were collected from line
RE (set in thick grass under young pines planted in
1978),
where mice were common. Ship rats were
always present but not abundant in older exotic
plantations established in or before 1966 (line FE2).
The rats that did live at low density in the older exotic
forests were also smaller and lighter than rats from
other habitats. In a 15-year old second-crop stand
at
Tokoroa,
a
brief study recorded a somewhat higher
density index for ship rats
(13.5/1 OOCTN)
although
with a different trapping regime (Clout 1980).
Comparisons with overseas studies
In Australia, northern Africa, southern Europe, North
America and South America, the ship rat
is
primarily
coastal, rarely found more than a few hundred
kilometres inland, and is an important pest damaging
citrus fruits, macadamia nuts, cocoa, coconut,
sugarcane, date palms, carob and avocado fruits
(Brooks & Rowe 1987). It greatly extended its range
in the Pacific during the Second World War, and
non-commensal ship rats have had their greatest
impacts on endemic fauna on these and other islands
(Atkinson 1985). Ship rats are remarkably adaptable,
occupying for example tussock grassland on sub-
antarctic (54'S) Macquarie Island (Pye et al. 1999),
sugarcane fields and rainforest from sea level to
2500m in Hawaii (Tomich 1986; Lindsey et
al.
1999)
and five different non-agricultural habitats in the
Galapagos Islands (Clark 1980). Clark attributed its
extraordinary world-wide success as an invasive
species to its "dispersibility, competitive superiority
over similar species in disturbed or secondary
habitats, and the ability to reproduce successfully in
a wide variety of environments".
The few accounts of non-commensal ship rat
populations overseas (Tamarin & Malecha 1971;
Clark
1980;
Watts & Aslin
1981;
Stroud
1982;
Tobin
et al. 1994; Downes et al. 1997; McDonald et al.
1997;
Lindsey et al. 1999; Pye et al.1999) confirm
that, as in New Zealand, ship rats are omnivorous,
nocturnal and frequently or occasionally arboreal.
They usually breed for 10-12 months per year in the
tropics (Tamarin & Malecha 1972; Taylor et al.
1990;
Tobin et al. 1994), but for 7-9 months in
temperate zones such as New Zealand. However,
even in Hawaii, with year-round breeding, there was
still a seasonal breeding peak in summer and autumn
(June to November; Lindsey et al. 1999) and higher
populations from late autumn to winter (Tomich &
Bridges 1981).
The somewhat irregular seasonal cycle of
repro-
duction and abundance which we have documented
at Pureora has been described before for non-
commensal
R.
rattus by Tamarin & Malecha
(1971,
1972) in Hawaii, and by Clark (1980) in the
Galapagos Islands. In Hawaii, fortnightly survival
rates of marked rats decreased with increasing
74New Zealand Journal of Zoology, 2001, Vol. 28
density, and, as in New Zealand, the relative impor-
tance of environmental and social factors in
controlling the timing of the cycle was unclear. In
the Galapagos, rat density was positively correlated
with vegetation density, suggesting an important role
for food supply, but Clark (1980) could not explain
the seasonality of breeding there (Galapagos ship rats
do not breed for five months or more of each year).
Mortality estimates calculated by Daniel (1972)
of
93%
per annum for non-commensal Orongorongo
Valley females and 99% for males were similar to
another estimate for ship rats in Puerto Rico
(97%:Weinbren et al. 1970). If this short longevity
is general for individuals in non-commensal popu-
lations, then the seasonal cycle of overall abundance
is due primarily to seasonal breeding and recruitment
followed by the rapid disappearance of each annual
cohort as its members die without replacement until
the following breeding season. The cycle will be
blurred by fluctuations both in extended breeding
and in increased longevity. Our five year study of
kill-trapped ship rat populations at Pureora, and the
eight-year live-trapping study of Daniel (1978),
provide data series longer by far than any other
published accounts in the world. Clearly the even
longer unpublished studies undertaken by the former
Department of Scientific and Industrial Research in
the Orongorongo Valley, Wellington and on Mt
Misery, Nelson, are uniquely long term projects
whose results should be analysed and published.
Effect of trap type on captures
Fenn and rat traps showed the same trends of
abundance for ship rats (Figs 2A,C), but rat traps
caught significantly more young ship rats in age
classes 1-2 (Table
2).
In Fiordland, trapping success
for kiore {Rattus exulans, range of body weights
from 29 to 69 g, n = 9) was 47 times higher in rat
traps than in Fenns (King & Moller 1997). Perhaps
the older and heavier ship rats learned to avoid rat
traps,
or were less likely to be detained by them, or
preferred the meat bait in Fenns. More importantly,
the heavier trigger mechanism on Fenns probably
selected against the lighter, younger animals.
Where rats are not abundant, large samples are
needed in order to detect population changes. The
total number of catches recorded, both at Pureora and
in Fiordland, was much higher in Fenn traps than in
the standard Supreme rat traps, since Fenns were set
for longer. In Fiordland, highly significant post-
seedfall increases in numbers of ship rats were
detected by Fenn traps in both the Eglinton and
Hollyford Valleys, which the rat traps, set in the
same areas on standard lines, missed altogether.
Conversely, Fenn traps failed to detect the near-
continuous presence of kiore in the Hollyford Valley
since well before the seedfall (King & Moller 1997).
Therefore, these authors concluded, it is important
to allow for these differences when presenting
detailed comparisons of population samples of rats
collected by different methods.
At Pureora, Fenn traps again detected significant
differences in ship rat abundance between seasons
and years which were missed by rat traps, and only
the Fenn traps detected the presence of Norway rats.
However, we cannot say whether the heavier Nor-
way rats escaped from rat traps, or were confined to
streamside trapsites that were sampled by Fenns but
not by rat traps.
Norway rats
Remarkably little is known or published about
Rattus
norvegicus in New Zealand, other than on islands.
We are aware of no mainland study which has
targeted them since that of Wodzicki (1950), and
trapping at that time yielded only three Norway rats!
Wodzicki's statements that "cornricks, cornfields,
haystacks, barns and sheds, orchard stores, and
homesteads" contain both R. norvegicus and R.
rattus, while "River and stream banks and shores of
ponds and lakes seem to be occupied almost entirely
by
R.
norvegicus "are based on a questionnaire sent
to "some 50" observers around the country. Our
research at Pureora has therefore provided the only
demographic account of a mainland R. norvegicus
population, whether commensal or non-commensal.
The trapping of most Norway rats in streamside
traps at Pureora is consistent with the general con-
clusions of Wodzicki (1950) and Watson (1961), that
mainland non-commensal populations are relictual
near water. By contrast, Norway rats range widely
on islands such as Stewart (Hickson et al. 1986),
Kapiti (Bramley 1999), and others in the eastern Bay
of
Islands
(Moller & Tilley 1986) which are free of
mustelids, especially stoats. In the 1840s-1870s,
huge populations of Norway rats caused extensive
damage on the mainland (Gillies 1878; Reischek
1930:251;
King 1984:68-70). Taylor (1978, 1984)
contended that, since then, Norway rats have been
confined to their present limited distribution by the
introduction of stoats in the 1880s. But the cor-
relation between the historical decline of Norway
rats and spread of stoats is confounded by the
contemporaneous spread of ship rats, especially in
the South Island (King 1984). Stoats do not seem to
limit the broad-scale distribution of Norway rats in
Innes et al.—Population biology of
ship
and Norway rats75
UK (Corbet & Harris
1991;
Tapper 1992) or of ship
rats in New Zealand. On the local scale, Norway rats
in commensal habitats such as waste landfills, or
small streamside colonies with well-established deep
burrows, might be able to avoid or defend them-
selves against
stoats.
Whether effective stoat control
near these habitats would result in the expansion of
Norway rats depends on whether Norway rats are
also limited by competition with ship rats. Evidence
from distribution patterns on islands and from size-
based dominance relations suggests that Norway rats
may limit ship rats, but not the reverse (Yom-Tov
et al. 1999). These questions could be practicably
tested by field experiments.
Interactions between ship rats and Norway rats are
of interest because in historical time, ship rats have
largely replaced Norway rats in non-commensal
habitats in New Zealand (Atkinson 1973), whereas
Norway rats have replaced ship rats in all habitats
in U.K. (Watson, 1961;Yalden 1999). In U.K., ship
rats now survive in any numbers only on a few
offshore islands (McDonald et al. 1997), while in
New Zealand non-commensal Norway rats seem to
survive in patches on the mainland, but not in great
abundance as on islands. Clearly, more long-term
and detailed study of the Norway rat in New Zealand
ecosystems is required, as noted by Moller & Tilley
(1986).
Implications for conservation
At Pureora (altitude 550-700 m above sea level), the
winter climate is severe; the average annual temp-
erature for 1947-70 was 10.3 C, with ground frosts
on an average of 87 days a year. If ship rats living
in podocarp forests can breed year-round in such
conditions, they are probably capable of rapid re-
covery after control operations conducted at any
season. However there is a clear seasonal peak in
recruitment in summer and autumn, so it is likely that
post-control recovery will be faster then. Most con-
trol operations against ship rats in New Zealand
forests are conducted in spring, but even when rat
densities were reduced by 90%, they recovered
within 2-5 months (Innes et al. 1995).
If winter rat mortality is high and there has been
no winter breeding, ship rat numbers may be low
when North Island kokako start nesting in November
(Innes & Hay 1995). However, results from projects
comparing areas in which ship rats were reduced to
varying degrees suggest that both the percentage of
kokako pairs fledging young, and the percentage of
nesting attempts which succeeded, declined rapidly
and substantially when ship rat tracking rates
(indexed using particular techniques) exceeded 5%
(Innes et al. 1999). In the absence of control, rat
tracking rates in North Island broadleaved-podocarp
forest are typically 40-80%, rarely as low as 15%
(Innes et al. 1995, 1999). Our data have confirmed
that ship rats are abundant and widespread at
Pureora, and that, although rat abundance can vary
greatly between years, even in 'low' years there are
still too many ship rats for most kokako to nest
successfully.
Our research has verified
a
pattern of demography
of ship rats at Pureora which is also typical of non-
commensal ship rats in other countries. There is a
seasonal peak in breeding, high annual mortality
(Daniel 1972), and year-to-year variation in total
numbers controlled by some combination of food,
predation and social limitation of
breeding.
Under-
standing these patterns better should assist in the
development of rational and efficient strategies for
ship rat control.
The conservation impact of Norway rats on main-
land fauna is unknown, but at Pureora it can hardly
be regarded as serious when they are apparently
confined to small local colonies.
Future research
We suggest that the highest priority should be given
to further (Blackwell et al. 1998) manipulative exp-
lorations of the key hypotheses already suggested by
Daniel (1972, 1978) regarding the relative import-
ance of predation and food supply as factors limiting
ship rat abundance. More research which examines
the actual impact of ship rats on New Zealand's
native biodiversity, and the effectiveness of popu-
lation control to reduce that impact, is also badly
needed to improve the targeting and sustainability
of pest control on the New Zealand mainland.
Finally, much remains to be learned about the
distribution, demography, ecology and conservation
importance of Norway rats on the New Zealand
mainland. Taylor's (1978, 1984) hypothesis that
stoat predation limits Norway rat distribution and
abundance could be tested directly by reversible
removal experiments like those of Tapper et al.
(1996),
and documenting the consequences for the
numbers and productivity of Norway rats.
ACKNOWLEDGMENTS
For supporting the original proposal we thank A. E.
Beveridge and the former Auckland Conservancy of NZ
Forest Service (especially A. H. Leigh). The Auckland
76New Zealand Journal of Zoology, 2001, Vol. 28
Conservancy, and the Department of Conservation which
inherited their commitment after April 1987, granted
permission to work on land under their control, and
provided all the encouragement and support we needed
through the long succession of crises which complicated
the project. For funding we are especially grateful to the
Lottery Board, for contributing both to the five years of
field and laboratory work, and to two of the many
following years of analysis and writeup. Other financial
help was given by the Auckland Conservancy NZFS, the
J. S. Watson Trust, the Department of Conservation and
the Foundation for Research, Science and Technology
(contract CO 9x004). Logistical support and manpower
to operate the Fenn lines was cheerfully supplied by a
succession of field staff based at Pureora, especially J.
Mason, P. Novis, A. Carpenter, R. Hepi, G. Harris and
M. Thackroy, guided by J. Gaukrodger. We thank the
then Forest Research Institute, Rotorua (now New
Zealand Forest Research Institute Ltd. and Manaaki
Whenua - Landcare Research New Zealand Ltd.) for
computer support; B.M.Fitzgerald for identifying the
fleas;
F.
Bailey for help with the figures, and two
anonymous referees for their improvements to the
manuscript.
REFERENCES
Alterio, N.; Moiler, H.; Brown, K. 1999: Trappability and
densities of stoats (Mustela erminea) and ship
rats (Rattus rattus) in a South Island Nothofagus
forest, New Zealand. New Zealand Journal of
Ecology
23:
95-100.
Atkinson, I. A. E. 1973: Spread of the ship rat (Rattus r.
rattus L.) in New Zealand. Journal of the Royal
Society of New Zealand
3:
457-472.
Atkinson, I. A. E. 1985: The spread of commensal spe-
cies of Rattus to oceanic islands and their effects
on island avifaunas. In: Moors, P. J. ed. Conser-
vation of island birds. Cambridge, 1CBP Techni-
cal Publication 3. Pp.
35-81.
Atkinson, I. A. E.; Moller, H. 1990: Kiore. //;: King, C.
M. ed. The handbook of New Zealand mammals.
Auckland, Oxford University Press. Pp. 175-192.
Blackwell, G. L.; Potter, M. A.; McLennan, J. A.; Minot,
E. O. 1998. The dynamics and distribution of the
small mammal/predator assemblage in mixed for-
est in the North Island, New Zealand.
Proceed-
ings of
the
Australian
Vertebrate
Pest Conference
17:235-239.
Bram ley, G.N.I
999:
Habitat use and responses to odours
by rodents in New Zealand. Unpublished PhD
thesis,
University of Waikato, Hamilton, New
Zealand.
Brockie, R. E. 1992: A living New Zealand forest. Auck-
land, Bateman. 172 p.
Brooks, J. E.; Rowe, F. P. 1987: Commensal rat control.
World Health Organisation Training and Infor-
mation Guide. Geneva, World Health Organisa-
tion. 107 p.
Clark, D. B. 1980: Population ecology of Rattus rattus
across a desert-montane forest gradient in the
Galapagos Islands. Ecology
61:
1422-1433.
Clout, M. N. 1980: Ship rats (Rattus rattus L.) in a Pinus
radiata plantation. New Zealand Journal of Ecol-
ogy
3:
141-145.
Corbet, G. B.; Harris, S. ed 1991: The handbook of
British mammals. Third ed. Oxford, Blackwell
Scientific Publications.
Cowan, P. E. 1981: Early growth and development of
roof rats, Rallus rattus L. Mammalia 45: 239—
250.
Cunningham, D. M.; Moors, P. J. 1983: A guide to the
identification and collection of New Zealand ro-
dents.
New Zealand Wildlife Service Occasional
Publication
4:
1-18.
Daniel, M. J. 1972: Bionomics of the ship rat (Rattus r.
rattus) in a New Zealand indigenous forest. New
Zealand Journal of Science
15:
313-341.
Daniel, M. J. 1978: Population ecology of ship and Nor-
way rats in New Zealand. In: Dingwall, P. R.;
Atkinson I. A. E.; Hay, C. ed. The ecology and
control of rodents in New Zealand nature re-
serves. Wellington, Department of Lands and Sur-
vey. Pp. 145-152.
Dowding, J. E.; Murphy, E. C. 1994: Ecology of ship rats
(Rattus rattus) in a Kauri (Agathis australis) for-
est in Northland, New Zealand. New Zealand
Journal of Ecology
18:
19-28.
Downes, S. J.; Handasyde, K. A.; Elgar, M. A. 1997:
Variation in the use of corridors by introduced
and native rodents in south-eastern Australia. Bio-
logical Conservation
82:
379-383.
Fitzgerald, B. M.; Karl, B. J. 1979: Foods of feral house
cats (Felis catus L.) in forest of
the
Orongorongo
Valley, Wellington. New Zealand Journal of
Zo-
ology 6:107-126.
Gillies, R. 1878: Notes on some changes in the fauna of
Otago.
Transactions
of the New Zealand Institute
10:
306-322.
Hickson, R. E.; Moller, H.; Garrick, A. S. 1986: Poison-
ing rats on Stewart Island. New Zealand Journal
of Ecology
9:
111-121.
Holdaway, R. N. 1999: A spatio-temporal model for the
invasion of the New Zealand archipelago by the
Pacific rat Rattus exulans. Journal of the Royal
Society of New Zealand 29:91-105.
Innes,
J. G. 1990: Ship Rat. In: King, C. M. ed. The
handbook of New Zealand mammals. Auckland,
Oxford University Press. Pp. 206-225.
Innes et al.—Population biology of
ship
and Norway rats77
Innes,
J. C; Hay, R. 1995: The nesting of the North
Island kokako (Catlaeas cinerea wilsoni) - a re-
view of accounts from 1880 to 1989. Notornis 42:
79-93.
Innes,
J. G.; Hay, R.; Flux, I.; Bradfield, P.; Speed, H.;
Jansen, P. 1999: Successful recovery of North
Island kokako Callaeas cinerea wilsoni popu-
lations, by adaptive management. Biological Con-
servation 87: 201-214.
Innes,
J. G.; Warburton, B.; Williams, D.; Speed, H.;
Bradfield, P. 1995: Large-scale poisoning of ship
rats (Rattus rattus) in indigenous forests of the
North Island, New Zealand. New Zealand Jour-
nal of Ecology
19:
5-17.
Karnoukhova, N. G. 1971: Age determination of brown
and black rats. Ekologia
2:
71-76.
King, C. M. 1983: The relationships between beech
(Nothofagus sp.) seedfall and populations of mice
(Mus musculus), and the demographic and di-
etary responses of stoats (Mustela erminea) in
three New Zealand forests. Journal of Animal
Ecology
52:
141-166.
King, C. M. 1984: Immigrant killers: Introduced preda-
tors and the conservation of birds in New Zea-
land. Auckland, Oxford University Press. 224 p.
King, C. M. 1990: The handbook of New Zealand mam-
mals.
Auckland, Oxford University Press. 600 p.
King, C. M.; Edgar, R. L. 1977: Techniques for trapping
and tracking stoats (Mustela erminea): a review
and a new system. New Zealand Journal of Zool-
ogy
4:
193-212.
King, C. M; Flux, M.; Innes, J. G.; Fitzgerald, B. M.
1996a: Population biology of small mammals in
Pureora Forest Park: 1. Carnivores (Mustela
erminea, M.furo, M nivalis and Felis catus). New
Zealand Journal of Ecology
20:
241-251.
King, C. M.; Innes, J. G.; Flux, M.; Kimberley, M. O.
1996b: Population biology of small mammals in
Pureora Forest Park: 2. The feral house mouse
(Mus musculus ). New Zealand Journal of Ecol-
ogy
20:
253-269.
King, C. M.; Innes, J. G.; Flux, M.; Kimberley, M. O.;
Leathwick, J. R.; Williams, D. S. 1996c: Distri-
bution and abundance of mammals in relation to
habitat in Pureora Forest Park. New Zealand Jour-
nal of Ecology
20:
215-240.
King, C. M.; Moller, H. 1997: Distribution and response
of rats Rattus rattus, R. exulans to seedfall in
New Zealand beech forests. Pacific Conservation
Biology
3:
143-155.
Lindsey, G. D.; Mosher, S. M.; Fancy, S. G.; Smucker, T.
D.
1999: Population structure and movements of
introduced rats in an Hawaiian rainforest. Pacific
Conservation Biology
5:
94-102.
McDonald, R. A.; Hutchings, M. R.; Keeling, J. G. M.
1997:
The status of ship rats Raltus rattus on the
Shiant Islands, Outer Hebrides, Scotland. Bio-
logical Conservation
82:
113-117.
Miller, C. J.; Miller, T. K. 1995: Population dynamics
and diet of rodents on Rangitoto Island, New
Zealand, including the effect of a 1080 poison
operation. New Zealand Journal of Ecology 19:
19-27.
Moller, H.; Tilley, J. A. V. 1986: Rodents and their
predators in the eastern Bay of Islands. New Zea-
land Journal of Zoology
13:
563-572.
Moors, P. J. 1990: Norway Rat. In: King, C. M. ed. The
handbook of New Zealand mammals. Auckland,
Oxford University Press. Pp. 192-206.
Nelson, L, Jr.; Clark, F. W. 1973: Correction for sprung
traps in catch/effort calculations of trapping re-
sults.
Journal of Mammalogy
54:
295-298.
Pye,
T.; Swain, R.; Seppelt, R. D. 1999: Distribution and
habitat use of the feral black rat (Rattus rattus) on
subantarctic Macquarie Island. Journal of Zool-
ogy, London 247: 429-438.
Reischek, A. 1930: Yesterdays in Maoriland. London,
Jonathan Cape.
Smith, K. M. 1986: The diet of the ship rat Rattus rattus
rattus L. in a kauri forest in Northland, New
Zealand. Unpublished MSc thesis, University of
Canterbury, Christchurch.
Stroud, D. 1982: Population dynamics of Rattus rattus
and
R.
norvegicus in a riparian habitat. Journal of
Mammalogy
63:
151-154.
Tamarin, R. H.; Malecha, S. R. 1971: The population
biology of Hawaiian rodents: demographic pa-
rameters. Ecology
52:
383-394.
Tamarin, R. H.; Malecha, S. R. 1972: Reproductive pa-
rameters in Rattus rattus and Rattus exulans of
Hawaii, 1968-1970. Journal of Mammalogy 53:
513-528.
Tapper, S. C; Potts, G. R.; Brockless, M. H. 1996: The
effects of
an
experimental reduction in predation
pressure on the breeding success and population
density of grey partridges Perdix perdix. Journal
of Applied Ecology
33:
965-978.
Tapper, S. C. 1992: Game heritage: an ecological review
from shooting and gamekeeping records.
Fordingbridge (UK), Game Conservancy.
Taylor, J. M.; Calaby, J. H.; Smith, S. C. 1990: Repro-
duction in New Guinean Rattus and comparison
with Australian Rattus. Australian Journal of
Zo-
ology
38:
587-602.
78New Zealand Journal of
Zoology,
2001,
Vol. 28
Taylor, R. H. 1978: Distribution and interactions of ro-
dent species in New Zealand.
In:
Dingwall, P. R.;
Atkinson I. A. E.; Hay, C. ed. The ecology and
control of rodents in New Zealand nature re-
serves. Wellington, Department of Lands and Sur-
vey. Pp.
135-141.
Taylor, R. H. 1984: Distribution and interactions of intro-
duced rodents and carnivores in New Zealand.
Ada Zoologica Fennica
172:
103-105.
Tobin, M. E.; Koehler, A. E.; Sugihara, R. T. 1994:
Seasonal patterns of fecundity and diet of roof
rats in a Hawaiian macadamia orchard. Wildlife
Research 21: 519-526.
Tomich, P. Q. 1986: Mammals in Hawai'i, 2nd edition.
Honolulu, Bishop Museum.
Tomich, P. Q.; Bridges, K. W.
1981:
Rodents.
In:
Mueller-
Dombois, D.; Bridges K. W.; Carson, H. L. ed.
Island ecosystems - biological organisation in
selected Hawaiian communities. Massachusetts
(USA),
Hitchinson Ross Publishing Company,
and US/1BP Synthesis Series 15.
Watson, J. S. 1961: Rats in New Zealand: A problem of
interspecific competition.
Proceedings
of
the Ninth
Pacific
Science Congress
1957
(Zoology):
15-17.
Watts,
C. H. S.; Aslin, H. J. 1981: The Rodents of Aus-
tralia. Sydney, Angus & Robertson. 321 p.
Weinbren, M. P.; Weinbrcn, B. M.; Jackson, W. B.;
Villella J. B. 1970: Studies on the roof rat (Rattus
rattus) in the El Verde forest. In: Odum, H.;
Pigeon, R. F. ed. A tropical rainforest: a study of
irradiation and ecology at El Verde, Puerto Rico.
Washington DC, US Atomic Energy Commis-
sion. Pp.
169-181.
Wilson, P. R.; Karl, B. J.; Taylor, R. H. 1998: The role of
introduced predators and competitors in the de-
cline of kaka (Nestor meridionalis) populations
in New Zealand.
Biological Conservation
83:
175-
185.
Wod/icki, K. A. 1950: Introduced mammals of New
Zealand: an ecological and economic survey.
Wellington, DSIR Bulletin 98. 255 p.
Yalden, D. 1999: The history of British mammals. Lon-
don, T& AD Poyser.
Yom-Tov, Y.; Yom-Tov S.; Moller, H. 1999: Competi-
tion, coexistence, and adaptation amongst rodent
invaders to Pacific and New Zealand
islands.
Jour-
nal of Biogeography
26:
947-958.