ArticlePDF Available

Whitebait conservation and protected areas at non-tidal rivermouths: integrating biogeography and environmental controls on īnanga (Galaxias maculatus) spawning grounds


Abstract and Figures

Galaxias maculatus is a declining amphidromous fish that supports New Zealand’s culturally-important whitebait fisheries targeting the migratory juvenile stage. Spawning ground protection and rehabilitation is required to reverse historical degradation and improve fisheries prospects alongside conservation goals. Although spawning habitat has been characterised in tidal rivers, there has been no previous study of spawning in non-tidal rivermouths that are open to the sea. We assessed seven non-tidal rivers over four months using census surveys to quantify spawning activity, identify environmental cues, and characterise fundamental aspects of the biogeography of spawning grounds. Results include the identification of compact spawning reaches near the rivermouths. Spawning events were triggered by periods of elevated water levels that were often of very short duration suggesting that potential lunar cues were less important, and that rapid fish movements had likely occurred within the catchment prior to spawning events. Spawning grounds exhibited consistent vertical structuring above typical low-flow levels, with associated horizontal translation away from the river channel leading to increased exposure to anthropogenic stressors and associated management implications for protecting the areas concerned. These consistent patterns provide a sound basis for advancing protective management at non-tidal rivermouths. Attention to flood management, vegetation control, and bankside recreational activities is needed and may be assisted by elucidating the biogeography of spawning grounds. The identification of rapid responses to environmental cues deserves further research to assess floodplain connectivity aspects that enable fish movements in emphemeral flowpaths, and as a confounding factor in commonly-used fish survey techniques.
Content may be subject to copyright.
Whitebait conservation and protected areas at non-tidal
rivermouths: integrating biogeography and environmental
controls on ¯
ınanga (Galaxias maculatus) spawning grounds
Shane Orchard
and David R. Schiel
Marine Ecology Research Group, University of Canterbury, Christchurch, New Zealand.
Waterways Centre for Freshwater Management, University of Canterbury and Lincoln University,
Christchurch, New Zealand.
Corresponding author. Email:
Abstract. Galaxias maculatus is a declining amphidromous fish that supports New Zealand’s culturally important
whitebait fisheries targeting the migratory juvenile stage. Spawning ground protection and rehabilitation is required to
reverse historical degradation and improve fisheries prospects alongside conservation goals. Although spawning habitat
has been characterised in tidal rivers, there has been no previous study of spawning in non-tidal rivermouths that are open
to the sea. We assessed seven non-tidal rivers over 4 months using census surveys to quantify spawning activity, identify
environmental cues, and characterise fundamental aspects of the biogeography of spawning grounds. Results include the
identification of compact spawning reaches near the rivermouths. Spawning events were triggered by periods of elevated
water levels that were often of very short duration, suggesting that potential lunar cues were less important, and that rapid
fish movements had likely occurred within the catchment prior to spawning events. Spawning grounds exhibited
consistent vertical structuring above typical low-flow levels, with associated horizontal translation away from the river
channel leading to increased exposure to anthropogenic stressors and associated management implications for protecting
the areas concerned. These consistent patterns provide a sound basis for advancing protective management at non-tidal
rivermouths. Attention to flood management, vegetation control, and bankside recreational activities is needed and may be
assisted by elucidating the biogeography of spawning grounds. The identification of rapid responses to environmental cues
deserves further research to assess floodplain connectivity aspects that enable fish movements in ephemeral flowpaths, and
as a confounding factor in commonly used fish survey techniques.
Keywords: coastal lagoons, fisheries conservation, floodplain connectivity, migratory species, riparian zones, spatial
Received 1 February 2021, accepted 29 March 2021, published online 11 May 2021
Galaxias maculatus is a small-bodied amphidromous fish that is
widespread across the Pacific where it is known by various
names including ¯
ınanga, common jollytail, puye and puyen
(McDowall 1991;Zattara and Premoli 2005;Barile et al. 2015).
Traditional ‘whitebait’ fisheries targeting juvenile fish have
been established in several countries including New Zealand,
Australia, and Chile, all of which have suffered historical
declines that impact the viability of continued harvest and raise
conservation concerns (Campos 1973;McDowall 1984;Fulton
2000;Mardones et al. 2008;Vega et al. 2013). These fisheries
are situated near the coastal interface and exploit the postlarval
stage of amphidromous populations as they migrate into fresh-
water systems following an oceanic development period of
,6 months (McDowall and Eldon 1980). At this stage, juvenile
G. maculatus are largely colourless and typically in the 40–
55 mm size range before undergoing rapid developmental
changes soon after entering fresh water (McDowall 1968;
McDowall et al. 1994). New Zealand’s fishery has considerable
recreational, cultural and economic value that contributes to the
wellbeing of local communities nationwide, and can be partic-
ularly important in small towns (McDowall 1984). In this paper,
the management context of whitebait conservation recognises
that G. maculatus is the most abundant species in the whitebait
catch in the majority of New Zealand rivers (McDowall 1965;
Yungnickel et al. 2020), while also being an ‘at risk – declining’
species (Dunn et al. 2018) that requires attention to all
life stages.
The lifecycle of migratory galaxiids has important implica-
tions for their conservation and the sustainability of whitebait
fisheries at a variety of scales (McDowall 2008). At the
subglobal scale, the general pattern of amphidromy involves
an oceanic development period in which larvae may be trans-
ported long distances and potentially colonise distant shores
(McDowall 2002,2007). This life history may be evolutionarily
favourable in small island states otherwise depauperate in
Pacific Conservation Biology
Journal compilation CSIRO 2021 Open Access CC BY-NC
freshwater fish fauna by providing a mechanism for recruitment
to new habitats (Berra et al. 1996;McDowall 2010a). At more
local scales, the replenishment of larval sources remains critical
to the maintenance of existing populations. In many cases this
relies on the continued health of adult fish populations for egg
production coupled with the bioregional partitioning of larval
pools leading to recruitment pulses, or absences, as the case may
be (McDowall 2010b).
In addressing these dynamics, there is considerable debate
about the relative importance of oceanic dispersion pathways
and larval retention effects, alongside the potential for natal
homing, all of which may affect the biogeography of amphi-
dromous larval pools (Waters et al. 2000;McDowall 2010a;
Augspurger et al. 2017). Hickford and Schiel (2016) investi-
gated natal homing in the South Island of New Zealand, and
found no evidence of juvenile G. maculatus returning to the
same rivers in which they hatched. However, some recruits were
identified returning to rivers close to their natal origin on the
same coast, and 15 juveniles originating from the west coast
were identified in east coast whitebait samples (Hickford and
Schiel 2016). These results are consistent with oceanic currents
having an important role in the structuring of regional larval
pools (Chiswell and Rickard 2011).
Several additional lines of evidence point to the importance of
local recruitment sources for the maintenance of local popula-
tions at certain times. At one extreme, the capacity for a residen-
tial and largely non-migratory lifecycle has been identified
through studies on land-locked water bodies in Australia
(Pollard 1971;Chapman et al. 2006)andSouthAmerica
(Cussac et al. 2004;Barriga et al. 2002,2007;Carrea et al.
2013). Inherent consequences include larval retainment such that
all phases of the life cycle are completed within a significantly
smaller spatial domain in comparison to a diadromous lifecycle
(McDowall 2010c;Rojo et al. 2020). There is also evidence for a
wide spectrum of life histories somewhere between these
extremes, such as where larvae drift relatively slowly down-
stream and complete variable proportions of their growth and
development in retentive freshwater bodies (Augspurger et al.
2017), or nearshore coastal environments such as river plumes
(Sorensen and Hobson 2005;Shiao et al. 2015).
For G. maculatus, the potential effects of larval retention and
bioregional partitioning are heightened in comparison to other
whitebait species because of its shorter, generally annual, life
cycle (McDowall 2010b). An additional piece of this puzzle is
presented by variance in the biogeography of the spawning
grounds themselves, with local egg production sources being
disproportionately more important to the availability of post-
larval recruits where partitioning occurs. This illustrates that
effective management demands an understanding of spawning
grounds in all regions and river types.
Although the biogeography of G. maculatus spawning
grounds has been relatively well characterised in tidal rivers
where they are strongly influenced by spring high tides (Benzie
1968;Taylor 2002;Orchard et al. 2018a,2018b), many non-
tidal rivers also support fish populations. In New Zealand, these
rivers are particularly common on the east coast of the South
Island and south-eastern North Island where they are often
associated with mixed sand–gravel beaches at the coast (Kirk
1980,1991). Many such rivers form perched lagoons, or ha
under the influence of high wave energies and associated barrier
formation that promotes the establishment of non-estuarine
conditions at the rivermouth (Kirk 1991;Hart 2009). Other
examples are found in regions with significant alluvial transport
from glaciated and paraglacial areas (e.g. Iceland, Russia,
Alaska and Canada), and also on fine-grain shorelines in Israel
and South Africa (Carter et al. 1989;Cooper 1994;Lichter and
Klein 2011;Green et al. 2013;Smith et al. 2014;McSweeney
et al. 2017). To date, however, there has been no comprehensive
assessment of the location, timing and environmental controls
on amphidromous G. maculatus spawning activity at non-tidal
The present study addresses this information gap in a region
characterised by a high density of non-tidal rivermouths which
facilitated both concurrent investigations at several study sites,
and the generation of a wider landscape view. Specific hypoth-
eses included that spawning grounds were likely to be found in
rivermouth lagoon environments due to similarities with tidal
lagoons, including their ability to support periodically inundated
riparian vegetation, and that lunar cycles associated with spring
high tide periods were unlikely to act as a trigger for spawning
activity, as is the case in tidal rivers, due to the apparent lack of
appreciable hydrodynamic effects. Instead, we considered that
water level variations would be more likely to trigger spawning
events, as has been reported in land-locked rivers (Pollard 1971),
and tidal lakes (Richardson and Taylor 2002). Despite these
expectations, the research was largely exploratory and involved
the characterisation of previously unsurveyed rivers and their
associated environments. In the remainder of the paper we report
conclusive results that include strong spatiotemporal patterns
likely to be transferable to other localities. These provide a
sound basis for advancing the management of non-tidal river-
mouths to support their role in whitebait conservation.
Study area
The Kaiko
¯ura coast is characterised by a predominance of non-
tidal rivers, as is the case elsewhere in the Canterbury region of
New Zealand (Supplementary Material, Fig. S1). Seven non-
tidal waterways of varying character were chosen within a
,35 km section of coastline between Oaro and Mangamaunu
Bay (Fig. 1). These study catchments included two relatively
large rivers (Oaro and Kahutara) and five smaller streams.
Within this area there were no previous records of G. maculatus
spawning grounds in the National ¯
Inanga Spawning Database
(NISD) although Taylor and Marshall (2014) observed spawn-
ing shoals in Middle Creek. Records in the New Zealand
Freshwater Fish Database (NZFFD) include G. maculatus
observations in five of the seven study catchments, and further
populations were expected to be present in other waterways
(Supplementary Material, Fig. S2). Whitebait fishing occurs in
at least three of these rivers (Oaro, Kahutara and Lyell Creek).
The study catchments range from 440 ha to over 23 000 ha in
size, with the smallest (Blue Duck) representing a 2nd order
stream originating in coastal hill-country, and the largest
(Kahutara) being a 5th order river originating in the Seaward
¯ura Range (Table 1). Predominant catchment geologies
partly reflect the inland extent of each catchment in relation to
BPacific Conservation Biology S. Orchard and D. R. Schiel
hill-country and coastal outwash plains, with the latter being
extensive either side of Kaiko
¯ura Peninsula where they form
mixed sand–gravel beaches at the coast (Fig. 1). As is
characteristic of many waterways in the region, five of the study
sites featured non-estuarine lagoons at the rivermouth. At Blue
Duck and Harnetts creeks no such lagoon was present during the
Table 1. Characteristics of study catchments as recorded in field observations and selected data from the New Zealand River Environment
Classification (Snelder and Biggs 2002)
Study catchment Order
Catchment area (ha) Climate Geology Land cover Salinity (ppt)
Source of flow
Oaro River 4 4725 Cool wet Hard sedimentary Scrub 0.01 Hill country
Kahutara River 5 23 095 Cool wet Hard sedimentary Scrub 0.01 Hill country
Lyell Creek 3 1633 Warm dry Alluvium Pastoral 0.01 Low elevation
Middle Creek 4 2588 Cool dry Alluvium Pastoral 0.01 Low elevation
Swan Creek 3 1541 Cool wet Alluvium Pastoral 0.01 Low elevation
Harnetts Creek 3 1201 Cool wet Alluvium Pastoral 0.01 Low elevation
Blue Duck Creek 2 440 Cool wet Soft sedimentary Scrub 0.02 Hill country
Stream order of the lowest reach.
Maximum bottom salinity recorded from monthly hand-held meter sampling in the study reach.
Bathymetry (m)
0 - 2
2 - 5
5 - 10
10 - 30
30 - 200
> 200 Study sites
Residential areas
15 km
0200 km
Fig. 1. Location of study sites on the Kaiko
¯ura Coast, South Island, New Zealand.
G. maculatus spawning at non-tidal rivermouths Pacific Conservation Biology C
study period, but they have been recorded in the past. All of the
study sites, with the exception of Oaro, were affected by tectonic
uplift of between 0.6 and 0.9 m during the 2016 Kaiko
earthquake (Clark et al. 2017;Schiel et al. 2019), which has
generally increased the elevation of these lagoons and lower
river reaches in relation to sea level. In the case of Lyell Creek
and Middle Creek high spring tides would occasionally inundate
the lower reaches pre-earthquake, but this no longer occurs since
the uplift event (P. Adams, pers. comm.). All of the study sites
were open to the sea during the study period. This partly reflects
the effects of a flood event in February that cleared the river-
mouths and also illustrates that these rivers are not land-locked.
However, observations during low flow periods in other years
have shown that at least four of these waterways (Blue Duck,
Harnetts, Middle and Swan) may develop temporary closures
associated with the development of semipermeable barriers at
the coast (S. Orchard, pers. obs.).
Spawning ground surveys
Sampling design
In the first month of the study (February 2019) intensive eggs
searches and environmental measurements were completed in
all catchments by a team of three researchers in all catchments
beginning at the rivermouth and working upstream, with the
distribution of spawning sites (if any) being unknown. Although
March and April were expected to be the months of peak
spawning activity based on previous work in tidal waterways
in the Canterbury region (Orchard et al. 2018a,2018b), signifi-
cant spawning activity was detected in February that was
apparently triggered by a rain event. This finding helped to
establish the general location of spawning in the study catch-
ments and define areas for repeat surveys in the following
months. This resulted in the delineation of survey reach lengths
of between 400 and 700 m from the rivermouth, with the
upstream limit marked by a prominent riffle or general increase
of gradient in the riverbed. None of the spawning sites detected
in the following months were located near the upstream limit of
these survey areas, improving confidence that they were suffi-
cient to detect the spawning activity that occurred. Data from
permanent stage-height loggers operated by the Canterbury
Regional Council were available for the Lyell and Middle Creek
catchments (Environment Canterbury 2020). In both cases,
these loggers were located several hundred metres upstream of
the identified spawning reach but nonetheless provided a useful
indication of water level fluctuations in relation to the timing of
spawning events. Following discovery of the February spawn-
ing sites, additional water level loggers (Odyssey ODYWL,
Dataflow Systems Ltd) were installed within the observed or
potential/estimated spawning reach in the five remaining catch-
ments to provide further information on water level changes in
relation to subsequent spawning events. For all spawning sites
detected, real-time kinematic (RTK) GPS surveys were con-
ducted using a Trimble R8 GNSS receiver to measure the upper
and lower elevation of the egg band. Geodetic benchmarks were
included within all RTK-GPS surveys to achieve an estimated
vertical accuracy of 3.5 mm þ0.4 ppm RMS. Heights were
referenced to the New Zealand Vertical Datum (NZVD) (Land
Information New Zealand 2016).
Detection and measurement of spawning sites
Field surveys
Spawning surveys were completed over four consecutive
months (February–May). Following the standard approach in
tidal waterways in which spawning events have an apparent
relationship with the lunar cycle (Taylor 2002;Orchard and
Hickford 2018), each round of surveys commenced a few days
after the full moon of the month, which was associated with the
largest monthly spring tides within the study period (Table 2).
The field survey technique involved systematic searches for eggs
laid in riparian vegetation by a team of three people. For every
5 m length of river bank three searches were made at random
locations. Each search involved inspection of the stems and root
mats along a transect line oriented perpendicular to the river
bank. Typically, a 0.5 m wide swathe of vegetationwas inspected
on each transect, with these being of variable length depending
on the bank slope and vegetation extent. In this case, the transect
lengths were adjusted to include all vegetation within areas
inundated by the February flood event as judged by the position
of strand lines and silt deposition. As this was the highest water
level recorded during the study period, the spatial extent estab-
lished in February was applied to all subsequent surveys.
Area of occupancy
All egg detections were associated with a given location that was
identified as a spawning site (Orchard and Hickford 2018).
Confirmed identification of G. maculatus eggs was made on the
basis of their characteristic size (,1 mm diameter), position on
the riverbank, developmental attributes (i.e. development of
‘eyes’) and observations of adult fish in the vicinity of spawning
sites. Individual sites were defined as continuous or semi-
continuous patches of eggs with dimensions defined by the
pattern of occupancy. In each case, the upstream and down-
stream extents of the patch were established in the field.
Coordinates were recorded using a hand-held GPS and corrected
Table 2. Survey dates and lunar cycle (day/month)
Survey dates Oaro Kahutara Lyell Middle Swan Harnetts Blue Duck Full moon date
February 25/2 25/2 27/2 26/2 26/2 26/2 27/2 20/2
March 26/3 26/3 27/3 27/3 26/3 27/3 27/3 21/3
April 24/4 25/4 25/4 23/4 24/4 24/4 25/4 19/4
May 24/5 24/5 23/5 25/5 25/5 25/5 23/5 19/5
The full moon period was associated with the highest spring tides of the month following the lunar cycle for all months of the study period.
DPacific Conservation Biology S. Orchard and D. R. Schiel
in QGIS ver. 3.12 (QGIS Development Team 2020) with the
assistance of site photographs and landmarks. For each spawn-
ing site, the length along the riverbank was measured along with
the width of the egg band at the position of each of the original
search transect, and additional transects where needed to pro-
vide a minimum of three measurements at all sites. Zero counts
were recorded when they occurred within a spawning site as is
common where the egg distribution is not a continuous band.
The area of occupancy (AOO) was calculated as length mean
Spawning site productivity
Egg production was assessed by direct egg counts using a sub-
sampling method (Orchard and Hickford 2016,2018). At each
transect, as above, a 10 10 cm quadrat was placed in the centre
of the egg band and all eggs within the quadrat counted. Egg
numbers in quadrats with high egg densities (.200 per quadrat),
were estimated by further subsampling using five randomly
located 2 2 cm quadrats and the average egg density of these
subunits used to calculate an egg density for the larger
10 10 cm quadrat. The mean egg density was calculated from
all 10 10 cm quadrats sampled within the site, inclusive of
zero counts. The productivity of each site was calculated as
mean egg density AOO.
Statistical analyses
Differences in AOO and egg production trends were tested using
two-way ANOVAs with fixed factors of catchment and month,
followed by post hoc Tukey HSD tests for significant main
effects. Model assumptions were checked via Shapiro–Wilk and
Levene tests using the car package in R (Fox and Weisberg
2011). Variation in AOO and egg production was also examined
using robust one-way ANOVAs (Field and Wilcox 2017)to
assess the effect of catchment and month separately using the
robust package in R (Wang et al. 2020). These analyses con-
sidered catchment effects in relation to the pooled data from all
months, and also for the two main spawning events treated
separately (February and April). Similar tests were applied for
the dependent variables of distance upstream, and vertical ele-
vation based on the centre point of each spawning site in the
horizontal and vertical dimensions, respectively. A regression
analysis was used to explore the relationship between egg pro-
duction and AOO using pooled data from all sites. All analyses
were completed in R ver. 3.5.3 (R Core Team 2019).
Occurrence, size and productivity of spawning grounds
Spawning sites were detected in all seven catchments, although
at different times over the study period (Fig. 2). Two natural
spawning events were responsible for the majority of spawning
activity, the first occurring in February at the beginning of the
study period, and the second in April. A total of 34 individual
spawning sites were measured in these two events (14 and 20
sites, respectively). A single site was detected in March in one
of the study catchments (Lyell Creek). However, this spawning
event was triggered by the artificial raising of water levels as
part of an ecological engineering experiment conducted with
the local council, and was thus atypical of the wider regional
trend. No spawning was detected in any of the study catch-
ments in May.
Most of the spawning sites were relatively small though often
dense patches of eggs. The total AOO recorded over all months
(23.4 m
3.3, s.e.) was significantly different between rivers
(d.f. ¼6, F¼13.71, P,0.001), and similar results were
obtained when these analyses were restricted to results from
February and April (Table 3). The mean egg density across all
sites and months was 3.8 eggs cm
. A regression on the
combined data showed a significant relationship between the
AOO and egg production of individual spawning sites
¼0.86, F¼189.1, P,0.001). Decomposition to the indi-
vidual months showed that the February relationship was
stronger (R
¼0.85, F¼69.38, P,0.001) than in April
¼0.34, F¼9.28, P¼0.007), reflecting the influence of
the large and highly productive Kahutara River sites in Febru-
ary, but also the observation of relatively few sparse sites with
low egg densities in relation to AOO.
The total egg production across catchments and months was
,1.8 million eggs (Table 3). Egg production differed signifi-
cantly between catchments over the study period (d.f. ¼6,
F¼20.42, P,0.001) reflecting the influence of relatively high
0 0
Egg production
Area of occupancy (m2)
10 000
100 000
1 000 000
Harnetts Blue
Fig. 2. Summary of ¯
ınanga (Galaxias maculatus) spawning activity in seven non-tidal rivers on the Kaiko
¯ura Coast of New Zealand in 2019. Note log
scale on y-axis for egg production (columns) and linear scale for area of occupancy on right. Error bars are standard error of the mean derived from the
subsampling approach used to calculate each metric for each spawning site. The x-axis shows individual spawning sites numbered by river for each of
the three months in which spawning occurred. A fourth month (May) was also surveyed but no spawning was found.
G. maculatus spawning at non-tidal rivermouths Pacific Conservation Biology E
(b) Comparison between the two major spawning events (February and April). Values shown are the total AOO and egg production, and mean distance inland and elevation of spawning sites recorded within
each month. Statistical analyses are for robust one-way ANOVAs (Wang et al. 2020), with fixed variables of catchment and month
Metrics Oaro Kahutara Lyell Middle Swan Harnetts Blue Duck Contrasts d.f. Robust FPSignificance
AOO (m
) Feb 3.0 9.6 2.2 1.8 0.7 0.5 0.0 Catchment 6 15.34 5.13E-07 ***
Apr 0.2 0.0 2.1 0.6 0.0 2.3 0.3 Month 1 4.55 2.98E-02 *
Egg production Feb 27 4 644 1 198 965 74 850 19 092 99 167 24 251 0 Catchment 6 14.24 9.82E-07 ***
Apr 8660 0 39 582 0 2763 28 384 1134 Month 1 8.89 0.00545 **
Distance inland (m) Feb 91 212 238 150 180 185 na Catchment 6 53.26 1.03E-13 ***
Apr 83 na 200 na 254 229 280 Month 1 2.28 0.124 ns
Elevation NZVD (m)
Feb 1.29 1.73 0.96 1.95 2.53 3.52 na Catchment 6 273.3 2.20E-16 ***
Apr 1.20 na 0.86 na 2.52 4.08 2.45 Month 1 0.18 0.667 ns
No. of sites Feb 5 3 3 1 1 1 0
Apr 2 0 9 0 3 5 1
Significance codes: ***, P,0.001; **, P,0.001; *, P,0.01; ns, not significant; a, 0.05.
For comparison, elevation of Mean High Water Springs is approximately 0.5 m NZVD.
Table 3. Summary statistics for ¯
ınanga (Galaxias maculatus) spawning grounds in seven non-tidal rivers on the Kaiko
¯ura Coast of New Zealand in 2019
AOO, area of occupancy; NZVD, New Zealand Vertical Datum 2016
(a) Summary of biogeographical metrics over three months (February–April). The mean and standard error (s.e.) reported for AOO and egg productionmetrics derive from subsampling measures used for their
estimation. A further month (May) was also surveyed but no spawning was detected. Statistical analyses are for between-river comparisons using robust one-way ANOVAs (Wang et al. 2020), across all
months. See text for two-way ANOVA results
Metrics Oaro Kahutara Lyell Middle Swan Harnetts Blue Duck All sites d.f. Robust FPSignificance
AOO (m
) mean 3.15 9.59 4.57 2.31 0.74 2.76 0.28 23.4 6 13.71 0.000162 ***
s.e. 0.60 1.15 0.75 0.29 0.06 0.40 0.04 3.3
Egg production mean 28 3 304 1 198 965 11 6 321 19 092 10 1 930 52 634 1134 1 773 380 6 20.42 4.13E-06 ***
s.e. 18 1 943 69 8 446 86 113 11 150 73 138 32 977 715 74 5 600
Distance inland (m) mean 89 212 223 150 236 222 280 202 6 45.02 8.08E-12 ***
s.e. 6 22 19 0 28 16 0 13
Elevation NZVD (m)
mean 1.26 1.73 0.89 1.95 2.52 3.98 2.45 2.11 6 281.9 2.20E-16 ***
s.e. 0.04 0.05 0.03 0.05 0.06 0.11 0.03 0.05
No. of sites 7 3 13 1 4 6 1 35
FPacific Conservation Biology S. Orchard and D. R. Schiel
egg production at Kahutara in comparison to the other catch-
ments, even though spawning was recorded there only in
February (Table 3b). The between-month comparisons also
showed significant differences for both AOO (d.f. ¼1,
F¼4.55, P¼0.03) and egg production (d.f. ¼6, F¼8.89,
P¼0.005) that is indicative of much greater spawning activity
in February across the region as a whole. Two-way ANOVAs
also showed significant differences between catchments
(d.f. ¼6, F¼15.34, P,0.001) and between months
(d.f. ¼1, F¼7.80, P¼0.01), with no significant interaction
between the two (d.f. ¼4, F¼1.42, P¼0.26).
Catchment position
All of the spawning sites were located surprisingly close to the
coast, despite the presence of apparently suitable habitat for
spawning further upstream (Fig. 3). This pattern held in all
catchments with the mean upstream limit of spawning being
248 m from the rivermouth. The maximum distance upstream of
the rivermouth for any spawning site was 388 m in Lyell Creek,
and the minimum distance ranged from 65 m to 275 m. This
resulted in the identification of well-defined and relatively
compact spawning reaches in all catchments by the end of the
study period (Fig. 3). These spawning reaches ranged from 5 m
in length (associated with a single spawning site) to 251 m at
Lyell Creek, which also had the highest number of individual
spawning sites (Table 3).
The downstream limit of spawning exhibited a close rela-
tionship with the downstream limit of riparian vegetation that
was subject to at least periodic inundation during the study
period (Fig. 3). This relationship was consistent between spawn-
ing events in the four study catchments where spawning was
recorded more than once, as well as being common to all seven
rivers despite marked differences in their morphology and size.
The apparent high fidelity to a favoured spawning reach can be
seen in the statistical comparisons that show a significant
difference between catchments (d.f. ¼6, F¼45.0,
P,0.001) for the mean position of spawning sites expressed
as distance from the rivermouth (Table 3a). This fidelity is
partially obscured in the seasonal totals visualised in Fig. 3, but
can clearly be seen in the between-month comparisons for the
two main spawning events (Table 3b). Differences in the
distance inland were statistically significant between catch-
ments (d.f. ¼6, F¼53.3, P,0.001), but not between months
(d.f. ¼1, F¼2.28, P¼0.124).
Vertical position
Significant differences were found between rivers in the vertical
dimensions of spawning grounds. The highest elevation sites
were in Harnetts Stream up to 4.5 m NZVD (New Zealand
Vertical Datum 2016), and the lowest elevation sites (0.7 m
NZVD) were in Lyell Creek (Fig. 4). In the other rivers
spawning grounds fell within the range of 1.0–2.5 m NZVD,
Blue Duck
100 200 300 400
Distance from river mouth (m)
500 600
Median position
Survey limit
Coastal interface
Riparian vegetation
Vegetation limit
Fig. 3. Horizontal position of ¯
ınanga spawning ground in seven non-tidal rivers on the Kaiko
¯ura Coast
expressed as the range and median distance of spawning site locations upstream fromthe rivermouth. The
survey area extended from the coastline to the upstream limit indicated by red crosses. The green line
indicates the downstream limit of riparian vegetation within the river channel as observed during the
survey period. Seaward of this position all rivers are characterised by active mixed-sand gravel and
boulder rivermouth systems that are subject to regular reworking in storm and flood events.
G. maculatus spawning at non-tidal rivermouths Pacific Conservation Biology G
with the mean height across all individual spawning sites being
2.1 m 0.05 (s.e.). For comparison, the Mean High Water
Springs (MHWS) elevation on the Kaiko
¯ura coast was ,0.4 m
NZVD within this period (LINZ 2018a,2018b). One-way
ANOVAs showed that the mean elevation of spawning sites
was significantly different between sites across all months
(d.f. ¼6, F¼273.3, P,0.001), and similarly when only
February and April were considered (Table 3). However, mean
elevation was not significantly different between February and
April (d.f. ¼1, F¼0.18, P¼0.667). This reflects the occur-
rence of spawning sites in similar three-dimensional locations
between months at Oaro, Lyell and Swan and Harnetts despite
the addition of higher elevation sites at the latter in comparison
to February (Fig. 4). These impressions are supported by two-
way ANOVA results showing significant differences between
catchments (d.f. ¼6, F¼458.2, P,0.001), but not months
(d.f. ¼1, F¼0.15, P¼0.669), and a significant interaction
between the two (d.f. ¼3, F¼0.19, P¼0.001).
Role of environmental cues
There was a strong relationship between water level fluctuations
and the timing of spawning events. The first of these occurred
unexpectedly and involved an unseasonal heavy rain event in
late February (Fig. 5a). Subsequent spawning surveys showed
that this event triggered widespread spawning activity through
the region that contributed 76% of the AOO and 95% of the egg
production measured in the study period as a whole. Stage height
data showed that the rain event on 23 February caused a rapid
spike in water heights of 70 cm in Lyell Creek and 35 cm in
Middle Creek (Fig. 5a), with this difference partly reflecting
channel morphology and associated constriction effects at the
logger positions as well as catchment-specific discharge rates.
The apparent triggering effect of water level fluctuation had a
considerable influence on the three-dimensional geography of
the spawning grounds, with the lowest elevation eggs being
located ,30 cm above typical low-flow levels in the spawning
reach at all sites.
By late March, the continuing lack of rain suggested that the
February eggs were unlikely to be reinundated naturally and an
ecological engineering experiment was designed in collabora-
tion with the regional council involving a temporary (4 h)
closure of the Lyell Creek mouth using a gravel bund sufficient
to raise water levels upstream by ,40 cm. In addition to
facilitating the hatching of February eggs, the March spawning
surveys showed that this artificial inundation event triggered
spawning in the affected reach despite its short duration. In
contrast, no March spawning was recorded in any of the other
study catchments (Fig. 2). However, a series of moderate rain
events occurred in April sufficient to raise water levels region-
wide, and spawning was detected in five of the study catch-
ments. Concurrent water level and spawning site elevation data
from the April events provide conclusive evidence of the
triggering effect of water level fluctuations (Fig. 5b). Despite
marked differences in the hydrographs, all spawning sites were
located in an elevated position relative to the water level during
the previous period of low flows.
These observations also provide insights into the likely
duration of spawning events as revealed by the time at which
water levels dropped to below the minimum elevation of
spawning grounds. For example, the developmental stage of
eggs at Blue Duck Creek suggested that spawning occurred on
the rain event on 7 April and the vertical position of eggs
indicates that the spawning site was inundated for ,54 h
(Fig. 5b). In contrast, the mean spawning site elevations at Lyell
Blue Duck
All months
Elevation NZVD (m)
Fig. 4. Vertical position of ¯
ınanga spawning grounds in seven non-tidal rivers on the Kaiko
¯ura Coast.
Coloured boxes represent the mean elevation of the egg band calculated from all sites recorded in the
respective months. Error bars are one standard deviation of the mean position for top and bottom
elevation of the egg band. The grey shading (all months) shows the maximum and minimum height of the
egg band in each river over all survey months. Note that the single spawning site recorded in March
(in Lyell Creek) lies within the elevation band of other months.
HPacific Conservation Biology S. Orchard and D. R. Schiel
Stage height (m)
Middle Creek
Lyell Creek
Elevation NZVD (m)
Blue Duck
Elevation of April spawning sites
Fig. 5. Hydrographs from the study catchments. (a) Stage height records
from permanent loggers operated by the Canterbury Regional Council in
Lyell Creek and Middle Creek for the summer of 2019. Each plot is split into
two panels to facilitate comparison with the other catchments below. (b)
Hydrographs from water level loggers deployed within the spawning reach
of five catchments after the discovery G. maculatus spawning sites near the
rivermouths in February. Note that these plots are referenced to New
Zealand Vertical Datum (NZVD) rather than stage heights. Yellow boxes
show the NZVD elevation range of spawning sites discovered in April in four
of these catchments. The length of the box represents the maximum window
in which spawning could have occurred. Results from Oaro, Swan and Blue
Duck clearly show the coincidence of spawning activity with periods of
elevated water levels. At Harnetts this relationship is obscured by the
presence of lower elevation spawning sites located downstream of the logger
position combined with considerable gradient between the two. The pro-
longed periods of elevated water levels at Oaro and Kahutara result from
interactions between river flows and barrier formation at the rivermouth.
Similar effects can be seen at Swan in consideration of water level before and
after the April rain events.
G. maculatus spawning at non-tidal rivermouths Pacific Conservation Biology I
Creek were 10 cm lower than the February sites but still a
minimum of 20 cm above typical low flow levels. Together with
the obviously fresh (non-eyed) appearance of the eggs, the
elevation range indicates that spawning occurred during the
relatively sharp spike in water levels on 22 April, for which
the period of sufficient inundation lasted only a few hours. Harnetts
Creek showed a similar pattern, with spawning occurring close to
the peak of the 22 April rain event within a short period. These
temporal patterns suggest that the fish are both highly attuned
to detecting water level fluctuations and likely to be highly mobile
within the river at these times. The surprisingly short lag times
between the onset of water level fluctuations and spawning events
indicate very rapid responses to environmental cues.
This study adds to previous evidence of plasticity in G. macu-
latus life histories (Augspurger et al. 2017), and is consistent
with the prediction of Barbee et al. (2011) that the substantial
differences in environmental characteristics found across the
G. maculatus range could manifest as variability in spawning and
migration cues. Moreover, the recognition of diverse life histories
in several amphidromous species has also led to calls for reori-
entation of the term towards relationships between benthic and
pelagic life stages rather than necessarily involving a marine–
freshwater distinction (Closs et al. 2013;Closs and Warburton
2016). This highlights the need for further research on the char-
acteristics of recruitment sources, and in turn emphasises the need
for a solid understanding of the biogeography of spawning
grounds, which are the ultimate larval source (McDowall 2010b).
In New Zealand, there is a noticeable information gap on the
characteristics of non-tidal rivers despite a considerable body of
work on G. maculatus spawning in estuarine locations
(McDowall 1968,1991;Hickford and Schiel 2011;Orchard
et al. 2018a), and contrasts with numerous studies on landlocked
G. maculatus populations in Australian and South American
rivers and lakes (Pollard 1971;Cussac et al. 2004;Chapman
et al. 2006;Barriga et al. 2007). Recent advances in New Zealand
have included the identification of non-diadromous recruitment
in amphidromous galaxiid species, including G. maculatus,
where suitable freshwater pelagic habitat was present in rivers
open to the sea (Hicks et al. 2017;David et al. 2019). However,
despite confirming the potential for spawning in non-tidal
environments, these studies provide only limited information
on the specific location and relative importance of the associated
spawning sites. This is, in part, due to the choice of sampling
methodologies that take advantage of otolith microchemistry
markers as an indicator of previous life histories such as exposure
to marine environments or other chemically distinct water bodies
such as lakes (Ruttenberg et al. 2005;Warner et al. 2005).
Consequently, these approaches are well suited to identifying
broad-scale phenomena such as the presence of non-diadromous
individuals, but may lack the resolution needed to ascertain finer-
grain aspects such the contribution of recruitment from different
sources, and the associated biogeography of spawning grounds
(Carson et al. 2013).
This study extends and complements the previous New
Zealand work by applying a census survey approach with a
focus on the characterisation of spawning grounds (Orchard and
Hickford 2018). Although such surveys are resource-intensive
in comparison to sparse sampling designs, they have the poten-
tial to elucidate fundamental spatial ecology aspects and gener-
ate information that is directly useful for site-based management
needs. Findings from this study provide the first comprehensive
account of spawning ecology in amphidromous G. maculatus
populations at non-tidal rivermouths and include strong spatio-
temporal patterns that were consistent across study catchments
and months. The following sections provide a brief discussion of
these aspects with a focus on transferable learning for conserva-
tion in other non-tidal rivers, and comparison with the existing
knowledge base.
Position in catchment
The observed biogeographical pattern provides strong support
for the prediction that spawning grounds would be found in non-
estuarine lagoon environments due to similarities with known
spawning areas in the lower reaches of tidal waterways. Five of
the seven catchments exhibited a dynamic lagoon feature near
the rivermouth, and in each case spawning was found within the
lagoon or a short distance upstream. In the remaining two
catchments (Blue Duck and Harnetts), no rivermouth lagoon
feature was present during the study period but the spawning
grounds were located in the first slow-flowing reach upstream.
At the catchment scale, all spawning grounds occupy a relatively
compact reach within the lower river and in all cases high-
quality unused spawning habitat was present further upstream.
Although this pattern supports the notion of downstream fish
movement to a preferred low-elevation and low-gradient reach
as occurs in tidal situations (McDowall 1968), there was also
some variation in the observed upstream limit of spawning
between events. For example, in Lyell Creek in March, and
Harnetts Creek in April, spawning was found further upstream
of the February sites, which translated to a larger confirmed
spawning reach than if these sites had not been observed.
Consequently, different river conditions and potentially also
fish-centric factors, such as the size of the spawning population,
could introduce further variation and result in a larger reach
being important at other times. Despite this, the downstream
limit of spawning was relatively consistent and strongly related
to the most downstream position of suitable vegetation in the
periodically inundated riparian zone (Fig. 3). This suggests that
fish are actively searching the lower reaches of the river during
spawning events and are selecting appropriate spawning habitat
in that vicinity in preference to similar habitat further upstream.
In comparison to other studies, these results show similarities
with previous reports of G. maculatus spawning in tidal lowland
lakes such as Te Waihora/Lake Ellesmere in Canterbury, and
lakes Waihola and Waipori in Otago, where spawning sites have
been reported near the confluence of inflowing streams
(McDowall 1968;Richardson and Taylor 2002). In some cases,
spawning has been recorded out-of-phase with the tidal influ-
ence in response to other sources of water level fluctuations
that include wind-driven variations in water level height
(Richardson and Taylor 2002). In addition, studies from lake
environments in other countries have reported a variety of
spawning behaviours including upstream migrations to spawn-
ing grounds in tributary streams (Pollard 1971), and lacustrine
JPacific Conservation Biology S. Orchard and D. R. Schiel
spawning in landlocked lakes (Barriga et al. 2002;Chapman
et al. 2006). These variations demonstrate the need for further
characterisation of spawning ecology across a range of environ-
ments and also the important role of adult fish migrations that
are inherent in the selection of spawning sites.
Vertical dimension
Strong structuring in the vertical dimension was a common
feature of all spawning sites (Fig. 4). At the reach scale, this
resulted in the spawning grounds being 0.4–1.0 m above normal
low-flow levels and translated to a wide variety of positional
nuances in relation to human activities in, and adjacent to, the
river bed. At Blue Duck, Harnetts, Swan and Middle creeks the
spawning grounds are located on relatively steep banks close
the river channel where they are relatively protected from
nearby recreational activities and stock grazing in adjacent land.
However, at Lyell Creek, which features well-defined banks in a
more urban context, the spawning grounds overlap with areas
subject to herbicide spraying and landscaping works. At
Kahutara and Oaro rivers, which feature wider river beds and
more extensive rivermouth lagoons, spawning grounds are
exposed to river diversion earthworks for flood control, vege-
tation clearance and off-road vehicle use.
These results have important consequences for river man-
agement since many human activities are found in the same
elevation zone as spawning grounds, as is the case in tidal
waterways (Richardson and Taylor 2002;Orchard and Hickford
2018;Orchard et al. 2018a). In the non-tidal context, the major
differences include the lack of a regular (i.e. tidal) inundation
cycle. This is generally useful from a management perspective
by providing a convenient indicator of the most likely elevation
range in which spawning will be found, leading to planning
approaches based on tidal height as a proxy for the confirmed
location of spawning sites (Greer et al. 2015). Despite this,
multimonth surveys in tidal river systems have also shown that
high-discharge events can have the effect of raising spawning
grounds above normal tidal levels, and this may be associated
with appreciable horizontal translation away from the river
channel into vulnerable areas such as recreation reserves
(Orchard 2019). Consequently, attention to the role of discharge
rates in the horizontal and vertical structuring of spawning
grounds is important. This provides a sound basis for the
integration of flood management, vegetation control, and bank-
side recreational activities to achieve effective conservation at
all rivermouths where G. maculatus populations are found.
Timing of spawning, fish movement and environmental cues
The timing of spawning in all seven catchments provided
additional support for the hypothesis of spawning events being
triggered by rain-induced water level fluctuations. Multiple
lines of evidence, including high-precision elevation data and
concurrent water level data for independent spawning events,
confirm the importance of this environmental cue with no
exceptions across all 35 individual spawning sites. This func-
tional relationship is consistent with G. maculatus ecology in
tidal situations (McDowall 1968), and in landlocked riverine
populations in which spawning was found in riparian vegetation
inundated by freshes following rain events (Pollard 1971).
Alongside these potentially widespread trends, it is evident that
atypical variations are also possible such as in dry seasons where
spawning in pools has been reported, and lacustrine spawning
examples for which the spawning cues are poorly known
(Chapman et al. 2006).
In this case, insights on the timing of spawning events
obtained from fine scale spatiotemporal data suggest an impor-
tant role for poorly understood aspects involving fish movement
and migration triggers that precede the spawning event. The
finding of a very short time lag between manifestation of the
environmental cue and the onset of spawning deserves particular
attention since this may be indicative of rapid fish movements at
these times. Results from this study show general support for the
hypothesis that lunar cycles associated with spring high tide
periods would be unlikely to act as a trigger for non-tidal
spawning migrations, as has been reported to be a potential
cue in tidal waterways (McDowall 1968;Taylor 2002). How-
ever, the full moon period coincided with the heavy rain event of
late February and could therefore have had a bearing on the
position of fish within the wider catchment at that time (Table 2).
In comparison, some of the April spawning events show diver-
gence from the lunar cycle, and maintained an apparent syn-
chrony with water level changes. Nonetheless, it remains
difficult to disentangle the major factors influencing fish move-
ment towards rivermouths prior to the timing of spawning
events, especially on a seasonal basis. For example, lunar cues
might still be functioning as an influence on fish movements at
the catchment scale in connection with progression of the
G. maculatus lifecycle within these waterways.
Questions surrounding the functional trigger for fish move-
ments in these non-tidal rivers add intrigue to the already
mysterious process by which tidal river fish initiate their migra-
tions to estuarine spawning grounds to coincide with spring high
tides (McDowall 1968,1991). Further research on these aspects
would be insightful for improving the understanding of migration
triggers, as these have a fundamental influence on spawning
behaviour and the associated biogeography of spawning grounds.
Understanding these movements is also important for the design
of sampling approaches that assume comparable study areas for
size-class measurements (Ravn et al. 2018), or, in the case of
mark–recapture techniques, require accounting for within-
population detectability differences over time (MacKenzie et al.
2003;Lindberg and Lindberg 2012). For G. maculatus,the
potential for variable dynamics between size or age classes
combined with strong directional migration effects presents con-
founding factors that are difficult to control for, yet could directly
influence size-frequency and catch-per-unit-effort measures over
relatively short periods, including successive sampling days.
Protected area implications and concluding remarks
Spawning grounds provide an example of a geographically
small but disproportionately important critical habitat that is
vital for successful completion of the whitebait life cycle
(Mitchell 1994). Protecting spawning grounds is therefore an
important focus for management, and is indeed the subject of
New Zealand legislation under the Conservation Act (1987) and
the Resource Management Act (1991). Spatially explicit infor-
mation on actual or likely spawning locations can assist their
conservation by identifying areas for protection and enabling
G. maculatus spawning at non-tidal rivermouths Pacific Conservation Biology K
robust impact assessments to assess potential threats and
determine management needs. For G. maculatus, spatial plan-
ning at relatively fine scales has the potential to help address the
common occurrence of riparian and river management activities
that are incompatible with habitat protection goals. Applications
include improving the efficiency of area-based management
tools through avoiding unnecessary trade-offs between com-
peting demands for space (Faith and Walker 2002;Orchard and
Hickford 2020). This study supports the quest for integrated
management that embraces these aspects by providing the first
comprehensive account of G. maculatus spawning ecology at
non-tidal rivermouths and identifying biogeographical patterns
that were consistent between rivers and over time. It is hoped
that this may assist the development of both of regulatory
measures and voluntary approaches to protect sensitive areas at
the necessary times.
Conflicts of interest
The authors declare no conflicts of interest.
We thank the Ministry of Business, Innovation, and Employment (MBIE)
and the New Zealand Ministry for Primary Industries (MPI) for funding
support. Particular thanks to staff at the Canterbury Regional Council for
helpful discussions and access to water level records, and to staff and stu-
dents at the University of Canterbury who supported this work.
Augspurger, J. M., Warburton, M., and Closs, G. P. (2017). Life-history
plasticity in amphidromous and catadromous fishes: a continuum of
strategies. Reviews in Fish Biology and Fisheries 27, 177–192.
Barbee, N. C., Hale, R., Morrongiello, J., Hicks, A., Semmens, D., Downes,
B. J., and Swearer, S. E. (2011). Large-scale variation in life history traits
of the widespread diadromous fish, Galaxias maculatus, reflects geo-
graphic differences in local environmental conditions. Marine and
Freshwater Research 62, 790–800. doi:10.1071/MF10284
Barile, J., Escudero, M., Carren
˜o, E., and Martı´n, D. S. (2015). Efecto de la
salinidad en la supervivencia embrionaria de puye Galaxias maculatus
(Jenyns, 1842)/Effect of salinity on survival of embryos of jollytail
Galaxias maculatus (Jenyns, 1842). Latin American Journal of Aquatic
Research 43, 282. doi:10.3856/VOL43-ISSUE2-FULLTEXT-2
Barriga, J. P., Battini, M. A., Macchi, P. J., Milano, D., and Cussac, V. E.
(2002). Spatial and temporal distribution of landlocked Galaxias macu-
latus and Galaxias platei (Pisces: Galaxiidae) in a lake in the South
American Andes. New Zealand Journal of Marine and Freshwater
Research 36, 345–359. doi:10.1080/00288330.2002.9517092
Barriga, J. P., Battini, M. A., and Cussac, V. E. (2007). Annual dynamics
variation of a landlocked Galaxias maculatus (Jenyns 1842) population
in a northern Patagonian river: occurrence of juvenile upstream migra-
tion. Journal of Applied Ichthyology 23, 128–135. doi:10.1111/J.1439-
Benzie, V. (1968). Some ecological aspects of the spawning behaviour and
early development of the common whitebait Galaxias maculatus attenuatus
(Jenyns). Proceedings of the New Zealand Ecological Society 15, 31–39.
Berra, T. M., Crowley, L., Ivantsoff, W., and Fuerst, P. A. (1996). Galaxias
maculatus: an explanation of its biogeography. Marine and Freshwater
Research 47, 845–849. doi:10.1071/MF9960845
Campos, H. (1973). Migration of Galaxias maculatus (Jenyns) (Galaxiidae,
Pisces) in Valdivia Estuary, Chile. Hydrobiologia 43, 301–312.
Carrea, C., Cussac, V. E., and Ruzzante, D. E. (2013). Genetic and
phenotypic variation among Galaxias maculatus populations reflects
contrasting landscape effects between northern and southern Patagonia.
Freshwater Biology 58, 36–49. doi:10.1111/FWB.12036
Carson, H. S., Lo´ pez-Duarte, P. C., Cook, G. S., Fodrie, F. J., Becker, B. J.,
DiBacco, C., and Levin, L. A. (2013). Temporal, spatial, and interspe-
cific variation in geochemical signatures within fish otoliths, bivalve
larval shells, and crustacean larvae. Marine Ecology Progress Series
473, 133–148. doi:10.3354/MEPS10078
Taylor, R. B. (1989). Barrier and lagoon coast evolution under
differing relative sea-level regimes: examples from Ireland and
Nova Scotia. Marine Geology 88, 221–242. doi:10.1016/0025-
Chapman, A., Morgan, D. L., Beatty, S. J., and Gill, H. S. (2006). Variation
in life history of land-locked lacustrine and riverine populations of
Galaxias maculatus (Jenyns 1842) in Western Australia. Environmental
Biology of Fishes 77, 21–37. doi:10.1007/S10641-006-9051-2
Chiswell, S. M., and Rickard, G. J. (2011). Larval connectivity of harbours
via ocean currents: a New Zealand study. Continental Shelf Research 31,
1057–1074. doi:10.1016/J.CSR.2011.03.012
Clark, K. J., Nissen, E. K., Howarth, J. D., Hamling, I. J., Mountjoy, J. J.,
Ries, W. F., et al. (2017). Highly variable coastal deformation in the
2016 MW7.8 Kaiko
¯ura earthquake reflects rupture complexity along a
transpressional plate boundary. Earth and Planetary Science Letters
474, 334–344. doi:10.1016/J.EPSL.2017.06.048
Closs, G. P., and Warburton, M. (2016). Life histories of amphidromous
fishes. In ‘An Introduction to Fish Migration’. (Eds P. Morais, and
F. Daverat.) pp. 102–122. (CRC Press: Boca Raton.)
Closs, G. P., Hicks, A. S., and Jellyman, P. G. (2013). Life histories of
closely related amphidromous and non-migratory fish species: a trade-
off between egg size and fecundity. Freshwater Biology 58, 1162–1177.
Cooper, J. A. G. (1994). Sedimentary processes in the river-dominated
Mvoti estuary, South Africa. Geomorphology 9, 271–300. doi:10.1016/
Cussac, V., Ortubay, S., Iglesias, G., Milano, D., Lattuca, M. E., Barriga,
J. P., et al. (2004). The distribution of South American galaxiid fishes:
the role of biological traits and post-glacial history. Journal of Biogeog-
raphy 31, 103–121. doi:10.1046/J.0305-0270.2003.01000.X
David, B. O., Jarvis, M., O
¨zkundakci, D., Collier, K. J., Hicks, A. S., and
Reid, M. (2019). To sea or not to sea? Multiple lines of evidence reveal
the contribution of non-diadromous recruitment for supporting endemic
fish populations within New Zealand’s longest river. Aquatic Conserva-
tion: Marine and Freshwater Ecosystems 29, 1409–1423. doi:10.1002/
Dunn, N. R., Allibone, R. M., Closs, G. P., Crow, S. K., David, B. O.,
Goodman, J. M., et al. (2018). Conservation status of New Zealand
freshwater fishes, 2017. New Zealand Threat Classification Series 24.
Department of Conservation, Wellington.
Environment Canterbury (2020). River flow data. Canterbury Regional
Council. Available at
Faith, D. P., and Walker, P. A. (2002). The role of trade-offs in biodiversity
conservation planning: linking local management, regional planning and
global conservation efforts. Journal of Biosciences 27, 393–407.
Field, A. P., and Wilcox, R. R. (2017). Robust statistical methods: a primer
for clinical psychology and experimental psychopathology researchers.
Behaviour Research and Therapy 98, 19–38. doi:10.1016/J.BRAT.
Fox, J., and Weisberg, S. (2011). ‘An R Companion to Applied Regression.’
2nd edn. (SAGE Publications: Thousand Oaks, CA.)
Fulton, W. (2000). Tasmanian Whitebait – a multispecies fishery targeting
migrating fishes. In ‘Fish Movement and Migration. Australian Society
LPacific Conservation Biology S. Orchard and D. R. Schiel
for Fish Biology Workshop Proceedings, Bendigo, 28–29 September
1999’. (Eds D. A. Hancock, D. C. Smith, and J. D. Koehn.) pp. 256–260.
(Australian Society for Fish Biology: Sydney.)
Green, A., Andrew, J., Cooper, G., and LeVieux, A. (2013). Unusual barrier/
inlet behaviour associated with active coastal progradation and river-
dominated estuaries. Journal of Coastal Research 69, 35–45.
Greer, M., Gray, D., Duff, K., and Sykes, J. (2015). Predicting inanga/
whitebait spawning habitat in Canterbury. Report No. R15/100. Envi-
ronment Canterbury, Christchurch.
Hart, D. E. (2009). Morphodynamics of non-estuarine rivermouth lagoons
on high-energy coasts. Journal of Coastal Research 1355–1359.
Hickford, M. J. H., and Schiel, D. R. (2011). Population sinks resulting from
degraded habitats of an obligate life-history pathway. Oecologia 166,
131–140. doi:10.1007/S00442-010-1834-7
Hickford, M. J. H., and Schiel, D. R. (2016). Otolith microchemistry of the
amphidromous Galaxias maculatus shows recruitment to coastal rivers
from unstructured larval pools. Marine Ecology Progress Series 548,
197–207. doi:10.3354/MEPS11701
Hicks, A. S., Jarvis, M. G., David, B. O., Waters, J. M., Norman, M. D., and
Closs, G. P. (2017). Lake and species specific patterns of non-
diadromous recruitment in amphidromous fish: the importance of local
recruitment and habitat requirements. Marine and Freshwater Research
68, 2315–2323. doi:10.1071/MF16387
Kirk, R. M. (1980). Mixed sand and gravel beaches: morphology, processes
and sediments. Progress in Physical Geography 4, 189–210.
Kirk, R. M. (1991). River–beach interaction on mixed sand and gravel
coasts: a geomorphic model for water resource planning. Applied
Geography 11, 267–287. doi:10.1016/0143-6228(91)90018-5
Land Information New Zealand. (2016). LINZS25009 Standard for New
Zealand Vertical Datum 2016. Land Information New Zealand, Wellington.
Lichter, M., and Klein, M. (2011). The effect of river floods on the
morphology of small river mouths in the southeastern Mediterranean.
Zeitschrift fu
¨r Geomorphologie 55, 317–340. doi:10.1127/0372-8854/
Lindberg, M. S., and Lindberg, M. S. (2012). A review of designs for
capture–mark–recapture studies in discrete time. Journal of Ornithology
152, 355–370. doi:10.1007/S10336-010-0533-9
LINZ. (2018a). Local mean sea level datums. Updated 22 March 2018.
Available at
jections-and-heights/ [accessed 4 September 2020].
LINZ. (2018b). New Zealand Nautical Almanac, 2018/2019 edition, NZ
204. Land Information New Zealand, Wellington.
MacKenzie, D. I., Nichols, J. D., Hines, J. E., Knutson, M. G., and Franklin,
A. B. (2003). Estimating site occupancy, colonization, and local extinc-
tion when a species is detected imperfectly. Ecology 84, 2200–2207.
Mardones, A., Vega, R., and Encina, F. (2008). Cultivation of whitebait
(Galaxias maculatus) in Chile. Aquaculture Research 39, 731–737.
McDowall, R. M. (1965). The composition of the New Zealand whitebait
catch, 1964. New Zealand Journal of Science 8, 285–300.
McDowall, R. M. (1968). Galaxias maculatus (Jenyns), the New Zealand
whitebait. New Zealand Marine Department, Fisheries Research Bulle-
tin 2, 1–83.
McDowall, R. M. (1984). ‘The New Zealand Whitebait Book.’ (Reed:
McDowall, R. M. (1991). Conservation and management of the whitebait
fishery. Department of Conservation Science and Research Series 38,
McDowall, R. M. (2002). Accumulating evidence for a dispersal biogeog-
raphy of southern cool temperate freshwater fishes. Journal of Biogeog-
raphy 29, 207–219. doi:10.1046/J.1365-2699.2002.00670.X
McDowall, R. M. (2007). On amphidromy, a distinct form of diadromy in
aquatic organisms. Fish and Fisheries 8, 1–13. doi:10.1111/J.1467-
McDowall, R. M. (2008). Diadromy, history and ecology: a question of
scale. Hydrobiologia 602, 5–14. doi:10.1007/S10750-008-9290-7
McDowall, R. M. (2010a). Why be amphidromous: expatrial dispersal and
the place of source and sink population dynamics? Reviews in Fish
Biology and Fisheries 20, 87–100. doi:10.1007/S11160-009-9125-2
McDowall, R. M. (2010b). ‘New Zealand Freshwater Fishes: an Historical
and Ecological Biogeography.’ (Vol. 32.) (Springer: New York.)
McDowall, R. M. (2010c). Historical and ecological context, pattern and
process, in the derivation of New Zealand’s freshwater fish fauna. New
Zealand Journal of Ecology 34, 185–194.
McDowall, R. M., and Eldon, G. A. (1980). The ecology of whitebait
migrations (Galaxiidae: Galaxias spp.). New Zealand Ministry of Agri-
culture and Fisheries, Fisheries Research Bulletin 20, 1–172.
McDowall, R. M., Mitchell, C. P., and Brothers, E. B. (1994). Age at
migration from the sea of juvenile Galaxias in New Zealand (Pisces,
Galaxiidae). Bulletin of Marine Science 54, 385–402.
McSweeney, S. L., Kennedy, D. M., Rutherfurd, I. D., and Stout, J. C.
(2017). Intermittently closed/open lakes and lagoons: their global
distribution and boundary conditions. Geomorphology 292, 142–152.
Mitchell, C. P. (1994). Whitebait spawning ground management. New
Zealand Department of Conservation, Science and Research Series 69,
Orchard, S. (2019). River restoration opportunities in Amelia Rogers
Reserve. Report prepared for Christchurch City Council, April 2019.
Orchard, S., and Hickford, M. J. H. (2016). Spatial effects of the Canterbury
earthquakes on ¯
ınanga spawning habitat and implications for waterways
management. Report prepared for IPENZ Rivers Group and Nga
¯i Tahu
Research Centre. Waterways Centre for Freshwater Management and
Marine Ecology Research Group. University of Canterbury,
Orchard, S., and Hickford, M. J. H. (2018). Census survey approach to
quantifying ¯
ınanga spawning habitat for conservation and management.
New Zealand Journal of Marine and Freshwater Research 52, 284–294.
Orchard, S., and Hickford, M. J. H. (2020). Protected area effectiveness for
fish spawning habitat in relation to earthquake-induced landscape
change. In ‘Sustainable Bioresource Management: Climate Change
Mitigation and Natural Resource Conservation’. (Eds R. Maiti, H. D.
Rodrı´guez, C. A. Kumari, D. Mandal, and N. C. Sarkar.) Chapter 22.
(Apple Academic Press.)
Orchard, S., Hickford, M. J. H., and Schiel, D. R. (2018a). Earthquake-
induced habitat migration in a riparian spawning fish has implications
for conservation management. Aquatic Conservation: Marine and
Freshwater Ecosystems 28, 702–712. doi:10.1002/AQC.2898
Orchard, S., Hickford, M. J. H., and Schiel, D. R. (2018b). Use of artificial
habitats to detect spawning sites for the conservation of Galaxias
maculatus, a riparian-spawning fish. Ecological Indicators 91,
617–625. doi:10.1016/J.ECOLIND.2018.03.061
Pollard, D. A. (1971). The biology of a landlocked form of the normally
catadromous salmoniform fish Galaxias maculatus (Jenyns). I. Life
cycle and origin. Australian Journal of Marine and Freshwater Research
22, 91–123. doi:10.1071/MF9710091
QGIS Development Team (2020). QGIS Geographic Information System.
Open Source Geospatial Foundation Project. Available at
Ravn, H. D., Lauridsen, T. L., Jepsen, N., Jeppesen, E., Hansen, P. G.,
Hansen, J. G., and Berg, S. (2018). A comparative study of three different
methods for assessing fish communities in a small eutrophic lake.
Ecology of Freshwater Fish 28, 341–352. doi:10.1111/EFF.12457
R Core Team. (2019). ‘R: A Language and Environment for Statistical
Computing.’ (R Foundation for Statistical Computing: Vienna, Austria.)
G. maculatus spawning at non-tidal rivermouths Pacific Conservation Biology M
Richardson, J., and Taylor, M. J. (2002). A guide to restoring inanga habitat.
NIWA Science and Technology Series 50, 1–29.
Rojo, J. H., Ferna´ ndez, D. A., Figueroa, D. E., and Boy, C. C. (2020).
Phenotypic and genetic differentiation between diadromous and land-
locked puyen Galaxias maculatus.Journal of Fish Biology 96, 956–967.
Ruttenberg, B. I., Hamilton, S. L., Hickford, M. J. H., Paradis, G. L., Sheehy,
M. S., Standish, J. D., et al. (2005). Elevated levels of trace elements in
cores of otoliths and their potential for use as natural tags. Marine
Ecology Progress Series 297, 273–281. doi:10.3354/MEPS297273
Schiel, D. R., Alestra, T., Gerrity, S., Orchard, S., Dunmore, R., Pirker, J.,
et al. (2019). The Kaiko
¯ura earthquake in southern New Zealand: loss of
connectivity of marine communities and the necessity of a cross-
ecosystem perspective. Aquatic Conservation: Marine and Freshwater
Ecosystems 29, 1520–1534. doi:10.1002/AQC.3122
Shiao, J.-C., Tzeng, C.-S., Li, P.-C., and Bell, K. N. I. (2015). Upstream
migration and marine early life history of amphidromous gobies inferred
from otolith increments and microchemistry. Environmental Biology of
Fishes 98, 933–950. doi:10.1007/S10641-014-0329-5
Smith, A. M., Guastella, L. A., and Goble, B. J. (2014). Forecasting lagoon
outlet erosion: KwaZulu-Natal, southeast Africa. Journal of Coastal
Research 70, 151–155. doi:10.2112/SI70-026.1
Snelder, T. H., and Biggs, B. J. F. (2002). Multi-scale river environment
classification for water resources management. Journal of the American
Water Resources Association 38, 1225–1240. doi:10.1111/J.1752-1688.
Sorensen, P. W., and Hobson, K. A. (2005). Stable isotope analysis of
amphidromous Hawaiian gobies suggests their larvae spend a substantial
period of time in freshwater river plumes. Environmental Biology of
Fishes 74, 31–42. doi:10.1007/S10641-005-3212-6
Taylor, M. J. (2002). The national inanga spawning database: trends and
implications for spawning site management. Science for Conservation
188. Department of Conservation, Wellington.
Taylor, M., and Marshall, W. (2014). Inanga spawning survey of the
Canterbury Region. Report prepared for Environment Canterbury. 59pp.
Vega, R., Dantagnan, P., Mardones, A., Valdebenito, I., Zamorano, J., and
Encina, F. (2013). Bases biolo´ gicas para el cultivo del puye Galaxias
maculatus (Jenyns, 1842): una revisio´ n/Biological bases for whitebait
culture Galaxias maculatus (Jenyns, 1842): a review. Latin American
Journal of Aquatic Research 41, 369. doi:10.3856/VOL41-ISSUE5-
Wang, J., Zamar, R., Marazzi, A., Yohai, V., Salibian-Barrera, M., Maronna,
R., et al. (2020). robust: Port of the Sþ‘Robust Library’’. R package ver.
0.5-0.0. Available at¼robust
Warner, R. R., Swearer, S. E., Caselle, J. E., Sheehy, M., and Paradis, G.
(2005). Natal trace-elemental signatures in the otoliths of an open-coast
fish. Limnology and Oceanography 50, 1529–1542. doi:10.4319/LO.
Waters, J. M., Dijkstra, L. H., and Wallis, G. P. (2000). Biogeography of a
southern hemisphere freshwater fish: how important is marine dispersal?
Molecular Ecology 9, 1815–1821. doi:10.1046/J.1365-294X.2000.
Yungnickel, M. R., Hickford, M. J. H., and Schiel, D. R. (2020). Spatio-
temporal variation in species composition of New Zealand’s whitebait
fishery. New Zealand Journal of Marine and Freshwater Research 54,
679–694. doi:10.1080/00288330.2020.1745854
Zattara, E. E., and Premoli, A. C. (2005). Genetic structuring in Andean
landlocked populations of Galaxias maculatus: effects of biogeographic
history. Journal of Biogeography 32, 5–14. doi:10.1111/J.1365-2699.
Handling editor: Alan Lymbery
NPacific Conservation Biology S. Orchard and D. R. Schiel
... In this study, the timing of the surveys during the summer months coincided with the īnanga spawning season in which mature adult fish are expected to have migrated downstream (Benzie 1968;Orchard & Schiel 2021). This seasonal effect likely explains the low numbers of adult īnanga caught at lower catchment sites for which there are no obvious connectivity issues (sites 2, 4 and 5) in comparison to site 1 where īnanga were more abundant than elsewhere. ...
Technical Report
Full-text available
The Environment Canterbury (ECan) Regional Fish Habitat initiative provides a coordinated region-wide approach to identifying, prioritising and remediating fish habitat. Within this context, there is a particular need to assess barriers to fish passage which is important consideration for the management of migratory fish. This contributes to the conservation of species that rely on waterway connectivity for completion of their life cycle. This report summarises a series of investigations designed to support management decisions on the status and potential remediation of barriers to fish passage in the Middle Creek catchment on the Kaikōura plains. The primary objectives of the project were the development of a connectivity assessment to evaluate environmental risks and benefits associated with potential fish passage interventions at identified barriers. Because migratory species are already known in the catchment, the major unknowns related to the composition of fish populations upstream of the barriers, and the need for improved information on the diversity of species in the catchment as a whole. Secondary objectives included further assessment of potential connectivity barriers, particularly those at Mt Fyffe Road on Middle Creek and Schoolhouse Road on Luke Creek, and recommendations for remediation following the assessment of fish population trends.
... They could include a focus on improving adult fish or spawning habitat, or both. The availability of īnanga spawning habitat is of particular note since the īnanga population must spawn in this reach and the spawning grounds likely conform to the patterns observed in other non-tidal river mouth systems on the Kaikōura coast (Orchard & Schiel 2021). ...
Technical Report
Full-text available
This project investigated and trialled a spotlight-based approach for fish surveys with the objective of obtaining relative abundance data for a target species at a relatively large scale. The initial intention was to trial this general approach to explore the feasibility of completing catchment-wide surveys for shortjaw kōkopu (Galaxias postvectis) in hill-country catchments on the Kaikōura coast. The initial intention for the 2021 summer was to attempt a whole-of-catchment survey for shortjaw kōkopu in Blue Duck Stream using a combination of the rapid spotlighting method described here and eDNA sampling at the bottom of the catchment. However, complexities in contacting and securing approval from the many landowners in the Blue Duck catchment were encountered, leading to a back-up plan being developed for the spotlight surveys in the Rakautara River catchment following contact with the landowners there. This report provides a brief summary of the above project. The primary objectives involved the refinement and trialling of existing Department of Conservation (DOC) guidelines for fixed reach spotlight surveys. The scope was limited by the time and resources available to the team but greatly assisted by the generous support of several volunteers. In total, these resources allowed for the survey of 2.8 km of river comprising seven survey reaches (each 400 m). Because of the larger size of the Rakautara catchment in relation to the original trial site in the Blue Duck, two key decisions were taken to maximise the value of the results obtained that are reflected in the location of the survey areas. First, the lower 2 km of the catchment below a major (c. 10 m) waterfall was surveyed in its entirety to assess the fish population in this area which was suspected to be potentially enriched due to a natural barrier effect. Second, the remaining survey effort was deployed in the upper catchment above the waterfall. Although the latter consisted of only two survey reaches (800 m total) it provides an indication of the barrier effect.
Technical Report
Full-text available
River restoration opportunities in Amelia Rogers Reserve. Prepared for Christchurch City Council, April 2019.
Full-text available
We studied phenotypic and genetic differences between individuals of puyen Galaxias maculatus from two sites in the same river basin in Tierra del Fuego National Park, southern South America. Individuals from the two sampling sites presented morphometric and genetic differences. The morphometric differences indicated that individuals from Laguna Negra (LN) were shorter, more robust, and with larger eyes, whereas those from Arroyo Negro (AN) were thinner, elongated, and with smaller eyes. Genetic differences showed that AN individuals had a greater genetic structuration and an older demographic history than LN individuals. From our results, we could affirm that the individuals from the two sampling sites belong to different populations with a high degree of isolation. The demographic history could indicate that the individuals of G. maculatus which migrated to northern areas during the last glaciation settled in the Beagle Channel after its formation. The LN population could have originated after the retreat of the glaciers, migrating from AN. This article is protected by copyright. All rights reserved.
Technical Report
Full-text available
The Canterbury earthquakes resulted in numerous changes to the waterways of Ōtautahi Christchurch. These included bank destabilisation, liquefaction effects, changes in bed levels, and associated effects on flow regimes and inundation levels. This study set out to determine if these effects had altered the location and pattern of sites utilised by inanga (Galaxias maculatus) for spawning, which are typically restricted to very specific locations in upper estuarine areas. Extensive surveys were carried out in the Heathcote/Ōpāwaho and Avon/Ōtākaro catchments over the four peak months of the 2015 spawning season. New spawning sites were found in both rivers and analysis against pre-earthquake records identified that other significant changes have occurred. Major changes include the finding of many new spawning sites in the Heathcote/Ōpāwaho catchment. Sites now occur up to 1.5km further downstream than the previously reported limit and include the first records of spawning below the Woolston Cut. Spawning sites in the Avon/Ōtākaro catchment also occur in new locations. In the mainstem, sites now occur both upstream and downstream of all previously reported locations. A concentrated area of spawning was identified in Lake Kate Sheppard at a distinctly different location versus pre-quake records, and no spawning was found on the western shores. Spawning was also recorded for the first time in Anzac Creek, a nearby waterway connected to Lake Kate Sheppard via a series of culverts. Overall the results indicate that spawning is taking place in different locations from the pre-quake pattern. Although egg survival was not measured in this study, sites in new locations may be vulnerable to current or future land-use activities that are incompatible with spawning success. Consequently, there are considerable management implications associated with this spatial shift, primarily relating to riparian management. In particular, there is a need to control threats to spawning sites and achieve protection for the areas involved. This is required under the New Zealand Coastal Policy Statement 2010 and is a prominent objective in a range of other policies and plans.
Whitebait comprise a culturally, commercially and recreationally important fishery in New Zealand, where post-larvae are netted while returning from their marine phase. In this study, we expanded an historical (1964) sampling programme to gain a contemporary understanding of the species composition of the whitebait fishery; 87 rivers were sampled over six months in 2015. Over the entire country, >12 species were found in samples and 84.6% of these were īnanga (Galaxias maculatus). Kōaro (G. brevipinnis) and banded kōkopu (G. fasciatus) were abundant in some rivers and regions at particular times of the year. Buller was the most variable region, spatially and temporally, for species composition; Canterbury was the least variable. Banded kōkopu whitebait migrated one month earlier north of Cook Strait than in the south. There was a positive association between the abundance of kōaro and banded kōkopu in samples and the level of indigenous forest cover in catchments. Compared to samples from 50 years ago, there was a greater proportion of kōaro and banded kōkopu whitebait throughout the country. This spatio-temporal variability requires fishery regulations to be more tailored and flexible if they are to conserve the diversity of life-histories present in the catch and sustain the whitebait fishery.
• The Mw 7.8 earthquake that struck the north‐east coast of the South Island of New Zealand in November 2016 caused extensive upheaval, of up to 6 m, over 110 km of coastline. Intertidal habitats were greatly affected with extensive die‐off of algal communities, high mortalities of benthic invertebrates, and greatly reduced ecosystem functioning, such as primary productivity. Only isolated pockets of key species remained in these areas, many of which were within protected areas around Kaikōura. • The loss of key species of algae and invertebrates fragmented marine populations and compromised connectivity and recovery processes because of the large dispersal distances needed to replenish populations. Severe sedimentation from terrestrial slips and erosion of newly exposed sedimentary rock compromised settlement and recruitment processes of marine species at many sites, even if distant propagules should arrive. • The combination of habitat disruption, loss of species and their functioning, and impacts on commercial fisheries, especially of abalone (Haliotis iris), requires multiple perspectives on recovery dynamics. • This paper describes these effects and discusses implications for the recovery of coastal ecosystems that include the essential involvement of mana whenua (indigenous Māori people), fishers, and the wider community, which suffered concomitant economic, recreational, and cultural impacts. These community perspectives will underpin the protection of surviving remnants of intertidal marine populations, the potential use of restoration techniques, and ultimately a successful socio‐ecological recovery.
• Otolith microchemistry was used to identify marine‐ versus freshwater‐derived recruitment of three native freshwater fish species belonging to the southern hemisphere family Galaxiidae, in New Zealand's longest river system, the Waikato River. • Water chemistry data for trace elements and ⁸⁷Sr/⁸⁶Sr isotope ratios were collected from five lentic and 10 lotic water bodies throughout the lower river floodplain. Potential spawning sites for galaxiids were compared with values obtained by laser ablation inductively coupled mass spectrometry (LA‐ICPMS) depth profiling of young‐of‐the‐year otoliths sampled from fish in nine lower river catchment sites. • Otolith chemical signatures from the larval rearing period indicated that catchment‐scale recruitment for two species, Galaxias argenteus (Gmelin, 1789) and Galaxias fasciatus Gray, 1842, was driven predominantly by non‐diadromous recruitment from one lake (Lake Waahi). In contrast, diadromous recruitment appeared to be more common for Galaxias maculatus (Jenyns, 1842); however, non‐diadromous specimens were also identified for the first time from a New Zealand river. • Reversing lake outlet flows linked to river stage appears be important in facilitating the dispersal of rheotactic larvae out of lakes, suggesting that lake outflow management at key times could be used to sustain this ecologically important function. • This study highlights that some water bodies can supply a disproportionately large number of recruits to support fish populations within the wider riverscape. Identifying these water bodies and managing them to sustain recruitment is key to the conservation of non‐diadromous Galaxiidae in this modified lowland environment
Sound decisions on the management of fish stocks depend on knowledge about the species composition, number, biomass and size structure of existing populations. Accordingly, the ability to make solid population estimates is essential. In this study, a 2.15 ha lake was completely drained and the total number of fish was recorded and amounted to 180,915 individuals divided into seven species having a total weight of 1,395 kg. Before the draining, three commonly used methods in fish surveys were applied: multi‐mesh gillnets, point abundance sampling by electrofishing (PASE) and mark–recapture. Following the determination of the actual number and size distribution of each species, we evaluated the efficiency of the methods and found that gillnets caught a relatively high number of species (five out of seven) and thus proved to be the best tool for mapping species richness. However, gillnets were size selective towards larger individuals of perch (Perca fluviatilis) and did not catch roach (Rutilus rutilus) <5 cm. In contrast to gillnets, PASE was very effective at catching YOY fish in the shore zone but selected for larger‐sized roach. In sum, gillnetting proved to be the most accurate method for estimating species composition, PASE also being useful. Overall, mark–recapture provided relatively good estimates of population size but small‐sized (<11 cm) roach proved not to be well suited for mark–recapture surveys. We conclude that the best method(s) surveying fish stocks depends on various factors such as target species, size distribution and the purpose of the survey.
Galaxias maculatus is a diadromous riparian-spawning fish that supports an important fishery. Eggs develop terrestrially as with several other teleost fishes. Spawning habitat occurs in specific locations near rivermouths and its protection is a conservation priority. However, quantifying the areas involved is hampered by high egg mortality rates on degraded waterway margins. We hypothesised that temporary artificial habitat would detect spawning in these situations producing a useful indicator for riparian management. We installed arrays of straw bales as artificial habitat in two independent experiments over consecutive years and assessed their impact using pairwise Before-After-Control-Impact (BACI) experimental designs. We tested degraded gaps within the distribution of known spawning sites and also areas further upstream and downstream. Nine spawning occurrences were recorded on artificial habitats in 2015, 22 in 2016, and two on paired controls. Both experiments produced a significant effect for artificial habitats deployed in degraded gaps within the known spawning site distribution (p=0.0001) providing evidence that these locations should be regarded as actual or potential spawning sites. In 2016 the technique also produced a significant effect downstream of known sites in one of the study catchments (p=0.0375). We believe the use of artificial habitats as a detection tool could be useful in a variety of management contexts. These include identifying areas for protection, as confirmation of site suitability prior to making restoration investments, and in investigations to support the migration of habitats to new locations under climate change, since these may currently be degraded.
• Galaxias maculatus is a riparian spawning fish that supports an important recreational fishery in New Zealand, with spawning habitat requirements strongly structured by salinity gradients at river mouths. This study reports changes to the spawning habitat following a series of large earthquakes that resulted in the widespread deformation of ground surfaces in the vicinity of waterways. • Assessments of habitat recovery focused on two river systems, the Avon and Heathcote, with pre‐disturbance data available over a 20‐year period. Recovery dynamics were assessed by field survey and mapping of spawning habitat prior to and on seven occasions after the disturbance event. Riparian land‐use and management patterns were mapped and analysed using overlay methods in a geographical information system (GIS). • Habitat migration of up to 2 km occurred in comparison with all previous records, and several anthropogenic land uses have become threats because of changed patterns of co‐occurrence. Incompatible activities now affect more than half of the spawning habitat in both rivers, particularly in areas managed for flood control purposes and recreational use. • The results are an example of landscape‐scale responses to salinity and water‐level changes driven by tectonic dynamics. These dynamics are not the source of the stress per se; rather, they have increased the exposure of the species to pre‐existing stressors. • The case illustrates important principles for managing subtle, yet widespread, change. Adaptive conservation methods and investments in information are priorities for avoiding management failure following environmental change.