Fisheries
5.1 Introduction
A comprehensive survey of fish populations in the tidal freshwater
portion of the Cape Fear basin was initiated in January 1997. The objectives of this
survey were to 1) document the incidence of fish disease, 2) characterize fish community
structure, and 3) track the effects of Hurricane Fran in both the Cape Fear and Northeast
Cape Fear rivers. The survey was a cooperative effort between the Cape Fear River Program
(Dr. Mary Moser, gillnets), the North Carolina Division of Marine Fisheries (NCDMF,
Wilmington Office, trawl) and the North Carolina Wildlife Resources Commission (NCWRC,
Districts 2 and 4, electroshocking). Using three gear types, a broad segment of the fish
population was examined and the preliminary results of the survey (January - May 1997)
were reported in Mallin et al. (1997).
A Marine Resources Grant (administered through N.C. Sea Grant) provided
supplemental funding to continue the survey through May 1998. A specific objective of this
project was to assess fish disease incidence and to compare research catches to those made
by fishermen using commercial gear. For many years, recreational anglers and commercial
fishermen alike have complained that resident fish communities are dominated by only a few
species and that fish disease incidence is high. A recent survey of anadromous fishes in
the lower Cape Fear River also noted the proliferation of hardy non-native species:
flathead catfish, hybrid striped bass, blue catfish, and common carp (Moser and Ross
1993). These species may be able to out-compete native fishes, particularly in areas that
experience periods of poor water quality. Using data from this year's fish survey (June
1997 - May 1998) we were able to evaluate the incidence of fish disease at stations
located in the Cape Fear and Northeast Cape Fear River, and to compare these levels of
disease to those reported by commercial fishermen working in the same areas.
A continued benefit of the fisheries survey has been the opportunity to
assess the recovery of fish populations following the periods of anoxia that occurred
after Hurricane Fran made landfall in September 1996. In the Northeast Cape Fear River
drainage the entire water column went anoxic for over two weeks (Mallin et al. 1997). In
the main stem of the Cape Fear River, hypoxic conditions were recorded during this period,
but anoxia was confined to the lower part of the drainage. A number of fish kills,
primarily in the Northeast Cape Fear and its tributaries were reported, but the effect of
acute anoxia on the entire fish population was not known. In addition to direct mortality
resulting from low dissolved oxygen levels, fish populations were potentially indirectly
affected by reductions in food resources and restricted habitat. The past year of survey
sampling has provided the unique opportunity to examine long-term effects of such a
dramatic water quality event.
5.2 Methods
Study Sites
Fish monitoring was conducted at nine study sites (Figure 5.1). Five sites were selected in the Cape Fear River: approximately 1.5 km above the NC11 bridge (NC11), the lower limb of the oxbow downstream from Syke's Landing near Acme (AC), the mouth of the Black River below Lyon Thoroughfare (BBT), the mouth of Indian Creek (IC), and in Horseshoe Bend (HB). One site was selected in the Brunswick River between the Belville boat ramp and the Highway 74/76 bridge (BRR). The remaining locations were in the Northeast Cape Fear River: approximately 2 km downstream of the NC117 bridge at Castle Hayne (NCF117), opposite the Hoechst Celanese dock (NCF6), and at the mouth of Smith Creek (Smith). The locations were all near water quality monitoring stations (see Section 2) and were limited to areas where all three gear types (gillnets, trawls, and boat electroshocker) could be employed. A 183 m reach at each site was marked off and sampling was conducted in the same reach each month.
Gillnets
Gillnet sampling was designed to sample very large resident and anadromous fishes that are less susceptible to electroshock and trawl collection. Monofilament nets with 11 cm stretched mesh were weighted to sink, thereby sampling the lower half of the water column at each station. Nets were 50 m in length and were deployed from the shoreline, perpendicular to the current. At the Horseshoe Bend and Smith Creek sites, 30 m nets were used because the channel at these locations was too narrow to set longer nets. In each sampling month, the nets were set over a three day period. This resulted in two, 24 h soak times at each station, and allowed sampling both during the day and at night. After the first 24 hr soak, the nets were checked and re-deployed to reduce fish mortality. All fish captured were identified, measured (nearest mm total length, TL), and examined for external evidence of disease (e.g., ulcers, lesions, fin rot, structural deformities, etc.). Anadromous species and catfish were tagged to allow later identification of individuals. All fish were released at the sampling site. The number of species collected, catch per unit effort (CPUE = number of fish / net-day), % diseased fish (number of diseased fish divided by total catch), and % non-native fish (number of non-native fish divided by the total catch) were determined for each site and sampling month.
Boat Electroshocker
District 2 and 4 biologists with the N.C. Wildlife Resources Commission (NCWRC) conducted boat electroshocking surveys monthly at eight of the nine sites (the conductivity at the Brunswick River site was too high to allow reliable sampling with this gear). This technique targets shoreline oriented species that are difficult to capture with either trawls or gillnets. The District 2 biologists used a 7500 watt electrofishing system with an 18 dropper anode array operated from an aluminum boat. A less powerful system (5000 watt) was used by District 4. The stations they sampled had lower conductivity and therefore required less output then stations sampled by District 2. At each site, a 183 m reach was sampled by making a pass with the current down each shoreline and one pass down the middle of the river. All stunned fish were dipnetted and placed in an aerated holding tank until the entire reach had been sampled. Fish were then identified, measured (nearest mm total length, TL), and examined for external evidence of disease (e.g., ulcers, lesions, fin rot, structural deformities, etc.). All fish were released at the sampling site. The number of species collected, catch per unit effort (CPUE = number of fish / 183 m reach), % diseased fish (number of diseased fish divided by total catch), and % non-native fish (number of non-native fish divided by the total catch) were determined for each site and sampling month.
Trawl
North Carolina Division of Marine Fisheries (NCDMF) personnel conducted monthly trawl sampling at each station following primary nursery trawl sampling protocol. This sampling method targets small, bottom-oriented fishes that are generally not collected with either of the other sampling methods. For each sample. a 3.2 m flat otter trawl with 0.64 cm mesh in the body, a 0.32 mesh bag, and a tickler chain was towed with the current for one minute. All fish were identified, measured (nearest mm total length, TL), examined for external evidence of disease, and released at the study site. The number of species collected, catch per unit effort (CPUE = number of fish / tow), % diseased fish (number of diseased fish divided by total catch), and % non-native fish (number of non-native fish divided by the total catch) were determined for each site and sampling month.
Commercial Fisheries Survey
Commercial fishermen that operate gillnets in the areas where survey sampling is conducted were contacted and those willing to participate in the study were paid to collect data on their catches. The fishermen were asked to record the exact location, soak time period (date and time nets were deployed and retrieved) and type of gear used (mesh size, floating vs sinking, set vs drifting, etc.). They identified and measured (nearest 12.7 mm) each fish captured and recorded descriptions of any external evidence of disease or abnormality. The number of species collected, catch per unit effort (CPUE = number of fish / 30 m net-day), % diseased fish (number of diseased fish divided by total catch), and % non-native fish (number of non-native fish divided by the total catch) were determined for each site and sampling month.
5.3 Results and Discussion
High flow conditions occurring in the period from January to April of 1998 precluded sampling using some gear types in some months. Gillnet samples were obtained at all stations and in all months except at Highway 11 (NC11) and Syke's landing (AC) in February 1998. The NCDMF trawl samples were made at all stations and months except for the Highway 11 (NC11) site in February and March. Due to high water, the NCWRC District 4 biologists were unable to obtain electroshocking samples in February and March at Highway 11 (NC11), Syke's Landing (AC), Indian Creek (IC), or the Black River (BBT), and in January at the Black River (BBT) and Indian Creek (IC) sites. The NCWRC District 2 biologists were unable to sample in April at Castle Hayne (NCF117), GE (NCF6), Smith Creek (Smith), and Horseshoe Bend (HB). The Horseshoe Bend site was also not sampled with the electroshocker in February, August, or December due to problems accessing the site.
Species Richness
A total of 77 species were collected in this survey (Tables 5.1 -
5.36). Electroshocking collections had the highest number of species (n = 62) and the
greatest range of fish sizes (Tables 5.25 - 5.36). Gillnetting collections were the least
speciose (n = 20 species) and trawl catches were intermediate in species richness (n = 42
species). The gillnets generally collected very large fishes (> 400 mm TL) and trawling
collected primarily the smallest size classes (< 100 mm TL). At the upper Cape Fear
River stations (NC11 and AC), we collected eastern silvery minnows (Hybognathus regius)
with the electroshocker. This species was not previously known to occur in the lower Cape
Fear drainage. Also of note was the collection of one highfin carpsucker (Carpiodes
velifer) at the NC11 site using gillnets. This fish has only recently been found in
the lower Cape Fear drainage (Moser et al. 1998) and has never before been found so far
downstream.
While fish diversity varied with sampling, method, time of year and
location (Figures 5.2 - 5.10), some trends in species diversity emerge when the data are
pooled by either month (Figure 5.11) or station (Figure 5.12). All three methods indicated
that diversity peaked in late summer - early fall (September-October) and then slowly
dropped off (Figure 5.11). The fewest species were collected in February. This trend
probably reflects the emigration of estuarine transient species in fall and the reduced
activity of fish in the winter months. As in early 1997, we found that electroshocking
collections in both the Cape Fear and Northeast Cape Fear sub-basins were most diverse at
upriver stations (NCF117, NCF6, NC11). At these stations, 10 or more species were
regularly captured (Figures 5.2, 5.9, 5.10). Gillnet catches from the Cape Fear River were
similar in species richness; but, gillnet sampling in the Northeast Cape Fear River
generally produced fewer species (Figure 5.12). Trawl samples had similar species richness
at all stations.
When species diversity is pooled for all stations in the Cape Fear
River (Figure 5.13), there is a clear seasonal pattern. At these stations, species
diversity was high in spring and early summer, and fluctuated at high but constant levels
through the summer and into fall (Figure 5.13). In contrast, pooled species diversity for
the Northeast Cape Fear River stations steadily increased from February to September,
outstripping the number of species collected in the Cape Fear River by August (Figure
5.14). This increasing trend was statistically significant (Spearman's Rank Correlation,
p<0.05) for both electroshocking and trawl collections and can be attributed primarily
to increases in freshwater and estuarine resident species recruitment. There was no
statistically significant trend in species diversity in the Cape Fear River stations
(p> 0.05). We hypothesize that the pattern observed in the Northeast Cape Fear River
reflects the recovery of resident fish populations following extreme anoxia after
Hurricane Fran (Mallin et al. 1997). The less extreme hypoxia that occurred in the Cape
Fear River apparently did not impact resident fish populations enough to produce a
recognizable recovery period. Future sampling in the two drainages will help to establish
whether the high species diversity we obtained in the Northeast Cape Fear River (over 30
species) is "normal". It is likely that species normally found in higher
salinity water, such as jack crevalle (Caranx hippos) and gray snapper (Lutjanus
griseus) were opportunistically invading the relatively de-pauperate stations in the
Northeast Cape Fear River, and that when freshwater residents become established, these
early colonists will no longer be captured in this area.
Fish Abundance
In spite of the high variability in fish abundance among sampling
methods, months, and locations, some trends in the data are notable (Figures 5.2 - 5.10).
Electroshocking and gillnetting collections were largest in late spring and summer months;
but one large trawl sample also occurred in winter (Figure 5.11). The large trawl
collections were primarily of small-bodied, benthic species: hogchokers (Trinectes
maculatus), juvenile blue catfish (Ictalurus furcatus), and freshwater gobies (Gobionellus
shufeldti). These fish are not susceptible to either electroshocking or gillnetting.
The only other very large trawl sample was in Smith Creek in May (Figure 5.8) when a large
school of Atlantic menhaden (Brevoortia tyrannus) was captured.
Trawling and electroshocking produced more individuals than gillnetting
at most sites and in most months (Figures 5.2-5.10). All three methods had higher CPUE in
the upper Cape Fear River stations than in more estuarine ones (Figure 5.12). In contrast,
the upper Northeast Cape Fear River stations generally yielded lower CPUE than those
farther downstream (Figure 5.12). In early 1997, we observed that fish abundance was lower
in the Northeast Cape Fear River than in the Cape Fear River (Mallin et al. 1997). In
contrast, the sampling from this year indicated that fish abundance was the same or higher
in the Northeast Cape Fear than in the Cape Fear in most months (5.13 and 5.14). Again,
this data provides evidence for the recovery of fish populations in the Northeast Cape
Fear River following Hurricane Fran.
Disease Incidence
Of the 8724 fish we captured, 128 exhibited external evidence of
disease (1.5%). Disease incidence ranged from 0 - 100% of the fish collected in an
individual sample (Figures 5.2 - 5.10), and was primarily in the form of fin rot or
ulcerated red sores. No disease was noted in any of the trawl collections. This was
probably due to the fact that the trawl collected small juveniles and transient species
that would not have experienced prolonged exposure to degraded water quality. Disease
incidence was highest in fish collected using gillnets (4.0%). Most of the diseased fish
collected using gillnets had fin rot (54%), with 26% having lesions or ulcerated sores. In
contrast to Mallin et al. (1997), the incidence of disease was lower in electroshocking
surveys (3.3% of fish collected) than in the gillnet survey. Most of the disease incidence
in electroshocking collections was noted in resident freshwater species (bowfin, Amia
calva; largemouth bass, Micropterus salmoides; and sunfish Lepomis sp.),
with nearly all bowfin exhibiting reddened sores or lateral lesions. Because these species
reside in the same area for extended periods, they are more likely to be exposed to
extreme or prolonged doses of low quality water than transient species. Similarly, species
with the highest incidence of disease in gillnet collections were all catfish. These fish
maintain close contact with the substrate and remain in the same area for extended
periods, potentially increasing their exposure to sediment or water borne contaminants.
Overall disease incidence in this survey (1.5%) was slightly lower than
in early 1997 (2.1%, Mallin et al. 1997). This is surprising in light of the fact that the
initial survey (Mallin et al. 1997) did not extend into the summer months. We found that
the overall incidence of fish disease was highest (1.6 %) in periods of warm water
temperature (April - October), and was reduced (0.6 %) in winter (Figure 5.11). Similarly,
if trawl collections are not included in this analysis, disease incidence is still higher
in summer (3.8%) than in winter (2.1%). Warm water temperatures tend to be more stressful
for fish, and lead to reduced dissolved oxygen. In addition, low flow periods in summer
may result in higher concentrations of contaminants due to reduced dilution (lower
discharge). The combined impacts of these summer water quality conditions apparently
result in greater external manifestation of disease in resident fish species.
Some diseased fish were captured in all stations. As in early 1997
(Mallin et al. 1997), we found that disease incidence was lowest in estuarine stations
(Figure 5.12; BRR, Smith, HB). In Mallin et al. (1997) we reported generally higher
incidence of diseased fish in the Cape Fear River gillnet stations than in the Northeast
Cape Fear gillnet stations, and we report a similar result in this year of sampling
(Figures 5.13 and 5.14).
Non-native Species
Non-native fishes are species that have been either accidentally or
purposefully introduced into a system by man and are not naturally indigenous to the area.
These species prey on and compete with native species, and can have profound effects on
the entire ecosystem. In addition, they are often hardy, opportunistic fish that can
thrive in stressed systems. In 1997-98, non-native fishes were most commonly caught with
the gillnets and were less frequently captured with the electroshocker and trawl (Figures
5.2 - 5.10). The most abundant non-native species were: common carp (Cyprinus carpio),
blue catfish, hybrid striped bass (Morone saxatilis x M. chrysops), and flathead
catfish (Pylodictis olivaris). We also collected two grass carp (Ctenopharyngodon
idella). Both of these fish were large adults (>700 mm total length) and were
captured using the electroshocker. One of the grass carp was collected at Syke's Landing
(AC) in August 1997 and the other in Indian Creek in November. In early 1997, we also
collected large grass carp in the upper Cape Fear River (Black River and Indian Creek,
Mallin et al. 1997). Capture of grass carp is significant because they have only recently
been introduced into the drainage and they are potentially very destructive herbivores.
For this reason, the state of North Carolina requires that all grass carp used for aquatic
weed control be certified as sterile individuals before they are introduced into any North
Carolina water body. The fact that only adults have been captured indicates that these
fish are escapees from local reservoirs or farm ponds and there is no evidence that this
species is reproducing in the Cape Fear River basin.
The percentage of non-native fish was highest in the upriver gillnet
stations of the Cape Fear River (NC11 and AC, Figure 5.12) due to large numbers of blue
catfish captured in this area. While non-native fishes were taken at all stations, they
were least abundant in the Brunswick River (BRR). This station was the most saline of the
stations we sampled and consequently there were probably fewer non-indigenous freshwater
species in this area. Non-native fishes were as abundant or more abundant in the Northeast
Cape Fear River collections than in the Cape Fear River (Figures 5-13 and 5-14). In
contrast, in Mallin et al. (1997) we noted that non-natives represented a higher
percentage of the catch in the Northeast Cape Fear River gillnet sites in early 1997. The
apparent decrease in relative abundance of non-indigenous fish may reflect recovery of
native species following Hurricane Fran. Because non-native species can tolerate more
extreme water quality conditions, they may have been least impacted by the low oxygen
conditions that occurred in the Northeast Cape Fear River in fall of 1996. Consequently,
they were apparently some of the first fishes to colonize the Northeast Cape Fear River in
spring of 1997. Although the percentage of non-native species in the Northeast Cape Fear
River has gone down, these fishes are still a significant component of the fish community.
Throughout the Cape Fear River drainage, non-native species dominated the largest size
classes of fish we captured, and therefore have the greatest potential for impacting the
drainage by preying on, or out-producing, native fishes. Because the apparent success of
these species in the Cape Fear River drainage may signal both degraded habitat and the
decline of native fish, it is particularly important that future sampling efforts monitor
the relative abundance of these species.
Commercial Fishermen Surveys
Commercial gillnet fishermen in the Cape Fear and Northeast Cape
Fear River participated in this study during the spring anadromous fish migrations. The
gillnet season closed on April 15 and we did not obtain any samples after that time. The
fishermen sampled a variety of locations in the drainage (n = 21 sites), but generally
used two gear types: drift and set nets. The set nets were moored to an anchored buoy or
the shoreline and were fished perpendicular to the current. In some cases these were sink
nets that fished the lower half of the water column and in some cases the fishermen
floated the nets so that they fished the upper half of the water column. The drift nets
were allowed to drift freely with the current and were manned at all times. These nets
generally are very deep and fish the entire water column. All the nets we surveyed had
mesh sizes between 13.3 and 14 cm stretched mesh. However, there was a large variety of
net lengths. For this reason, the fish catches were standardized to a 30 m net day. For
example, a 100 m net set for 36 h would represent 5.0 net-days of effort.
Four of the fishermen that participated in this survey had nets set in
the same location that research set nets were used. In some cases, they also drifted nets
in the same areas. We used this subset of the commercial fishermen's data to compare to
catches from research nets (Figure 5.15). These data represent a total of 68.6 net days of
commercial fishing effort.
At the Highway 11 (NC11) site, the commercial set and drift nets caught
fewer species than the research nets set in exactly the same locations (Figures 5.2 and
5.15). While research set nets caught similar numbers of fish as commercial set nets in
this location, the commercial drift nets were much more effective than either research or
commercial set nets (Figure 5.15). In addition, these drift nets fished very selectively
for American shad, and no bycatch was reported for drift nets in this area. The CPUE of
American shad collected using drift nets at this site was higher than in any other
commercial sampling location. Higher shad abundance in this area probably resulted because
Dam #1 (less than 1 km upstream) blocks upstream migration of these fish and tends to
concentrate them in the area immediately below the dam. Both commercial and research nets
indicated equally low disease incidence; but, the research nets caught a much higher
percentage of non-native species (blue and flathead catfish) at the Highway 11 site..
The commercial nets set at the mouth of the Black River (BBT), Indian
Creek (IC) and Castle Hayne (NCF117) had higher species diversity, abundance, fish disease
and non-native fish incidence than our research sets (Figures 5.4, 5.5, 5.10 and 5.15).
However, the commercial nets were set in slightly different locations (within 0.5 km)
which may have produced these differences. In contrast to upriver comparisons, the drift
nets in the Brunswick River (BRR) were less productive than the commercial set nets, but
both commercial gear types captured more fish, greater fish diversity and higher numbers
of non-native and diseased fish than were recorded from research gear set near-by.
In the period from February to April, the incidence of fish diseases
noted by commercial set net fishermen was 0.9 % and consisted primarily of lesions,
ulcers, and fin rot. No fish disease was noted for fish collected in research nets at the
same locations and in the same months. The percentage of non-native species in commercial
set nets was 23.5 % and the percentage of these species at the same research sites was 8.7
%. Commercial fishermen regularly complain of high incidence of fish disease and unwanted
fish species (both non-natives, such as the common carp, and natives, such as longnose
gar). Our data support their claims that commercial catches may have higher incidence of
both diseased fish and non-indigenous species than research catches. However, disease
incidence was still quite low in commercial catches, in spite of the fact that sampling
effort using commercial gear was much higher than with research gear. Unfortunately,
commercial fishermen only participated in this survey during months when water
temperatures were low. It is likely that the incidence of diseased fish in commercial gear
is much higher in summer because we found that disease incidence doubled in research
collections when water temperatures were high (April - October).
5.4 Summary
Fisheries sampling continued to document the long-term effects of anoxic conditions that occurred in September 1996, following the passage of Hurricane Fran. Seasonal trends in fish diversity and abundance indicated that Northeast Cape Fear River fish populations continued to recover through the summer of 1997, while fish populations in the Cape Fear River (which experienced less extreme hypoxia) did not change as dramatically. Overall disease incidence was 1.5%, and was approximately twice as high in periods of warm water temperature (April-October). As in early 1996, the data from this year indicated that disease incidence is highest among large resident fishes collected using gillnets and the boat electroshocker. As in previous gillnet surveys, non-native species dominated gillnet collections in the upper Cape Fear River stations. However, lower relative abundance of non-native fishes in the Northeast Cape Fear River this year may reflect recovery of native populations after the passage of Hurricane Fran. Commercial drift nets were both more effective and selective then commercial set nets. Commercial fishermen reported higher incidence of fish disease in gillnets set at the same stations as research nets in the period from February - April 1998; however, disease incidence in these gillnets was low (0.9%) relative to overall disease incidence in fish collected using research gear.
|
Back to Table of Contents for 1997-1998 Annual Report |
| Back to Lower Cape Fear River Program Homepage |