Occurrence and biological impacts of fishing gear and other marine debris in the Florida Keys

Full text

Occurrence and biological impacts of fishing gear and other
marine debris in the Florida Keys
M. Chiappone
*
, A. White, D.W. Swanson, S.L. Miller
Center for Marine Science and NOAA’s National Undersea Research Center, University of North Carolina at Wilmington,
515 Caribbean Drive, Key Largo, FL 33037, USA
Fishing constitutes one of the most significant threats
to marine biodiversity and ecosystem function, docu-
mented by a growing body of information on the nu-
merous impacts to populations, community structure,
and habitats (Dayton et al., 1995; Roberts, 1995; Jen-
nings and Polunin, 1996). Besides the more obvious ef-
fects on species population structure, fishing activities
may also reduce the structural complexity of habitats or
cause corresponding changes in ecological processes
such as competition and predation (Russ, 1991; Jones
and Syms, 1998; Auster and Langton, 1999). These
patterns are most obvious in areas where explosives,
poisons, or other destructive fishing methods are used
(Hatcher et al., 1989). However, ecological effects can be
expected in any area where traps, mobile fishing gear
such as trawls, and, potentially, even large numbers of
recreational fishers operate (Russ, 1991; Jennings and
Lock, 1996).
The Florida Keys (Monroe County, Florida) have a
long history of commercial and recreational fisheries
that target a great diversity of fish and invertebrate
species using a multitude of gears (Tilmant, 1989;
Bohnsack et al., 1994). In terms of volume of seafood
landed, the Florida Keys is the most important area in
the state in landings, dockside value, and numbers of
commercial fishing vessels, especially for highly valued
invertebrate fisheries (Adams, 1992). There are also
significant, but largely undocumented effects of tens of
thousands of recreational fishers (Davis, 1977), who
target hundreds of species using mostly hook-and-line
and spear guns (Bohnsack et al., 1994). Baseline data on
fishing gear and other marine debris were collected as
part of a larger assessment of benthic community
structure in the Florida Keys National Marine Sanctu-
ary, a large (9500 km2) marine protected area bordering
three national parks in southern Florida (Fig. 1). These
data are particularly timely because this coastal ecosys-
tem continues to experience a growing number of rec-
reational fishers, and both commercial and recreational
fishers exploit hundreds of invertebrates and fish species
(Bohnsack et al., 1994; Ault et al., 1998). This study
addressed several issues on marine debris occurrence
in shallow-water coral reef and hard-bottom habitats.
First, what is the spatial extent and frequency of remnant
fishing gear at multiple spatial scales in the Florida
Keys? Secondly, what factors, such as habitat type
(depth) or management regime (closed or open to fish-
ing) affect the spatial variability of marine debris oc-
currence? Thirdly, what are the biological impacts of
marine debris, especially from remnant commercial and
recreational fishing gear, on reef biota such as hard
corals and sponges?
Forty-five sites were surveyed southwest of Key West
to Big Pine Shoal in the lower Keys region of the
Sanctuary, spanning 60 km from southwest to northeast
and 12 km from nearshore to offshore (Fig. 1). Sites
were visited between July and August 2000 and were
selected using a two-stage, stratified random sampling
design (Cochran, 1977; Ault et al., 1999). Five of the 23
no-fishing zones (designated as sanctuary preservation
areas (SPAs), research only areas (RO), and ecological
reserves (ER)) in the Sanctuary were surveyed (indicated
in Fig. 1). Based on the spatial distribution of coral reef
habitat types (FDEP, 1998) and the depth limits of the
zones, the following habitat strata were sampled: near-
shore hard-bottom, mid-channel patch reef, offshore
isolated patch reef, offshore aggregate patch reef, back
reef and rubble/hard-bottom matrix, shallow fore reef
(4–7 m depth), and deeper fore reef (8–12 m) (Table 1).
Two random sites were sampled in each no-fishing zone,
within a particular habitat stratum that consisted of pre-
designated 200 200 m2areas randomly selected from a
grid constructed using a geographic information system.
*
Corresponding author.
0025-326X/02/$ - see front matter Ó2002 Published by Elsevier Science Ltd.
PII: S0 0 2 5 - 3 2 6 X ( 0 1 ) 0 0 2 9 0 - 9
www.elsevier.com/locate/marpolbul
Marine Pollution Bulletin 44 (2002) 597–604

Reference sites were randomly assigned by habitat type
(according to FDEP, 1998 data). Preliminary data on
debris density were not available, thus sample allocation
was based on the spatial distribution of hard-bottom
habitats and the amount of funded field days. At each
site, four random sampling points using differential GPS
Table 1
Sampling effort for fishing gear and other debris in the lower Florida Keys
Habitat type Management type No. sites No. transects Area sampled (m2)
Nearshore hard-bottom Western Sambo ER 2 8 400
Reference areas 2 8 400
Mid-channel patch reef Western Sambo ER 2 8 160
Reference areas 2 8 160
Offshore patch reef Western Sambo ER 2 8 160
Reference areas 2 8 160
Offshore aggregate patch Looe Key RO 2 8 400
Reference areas 2 8 400
Back reef rubble Western Sambo ER 1 4 200
Sand Key SPA 1 4 200
Reference areas 7 28 1400
Shallow fore reef (4–7 m) Looe Key SPA 2 8 400
Reference areas 3 12 600
Deeper fore reef (8–12 m) Sand Key SPA 2 8 400
Western Sambo ER 2 8 400
Eastern Sambo RO 2 8 400
Reference areas 9 36 1800
Total 45 180 8040
No-fishing zones are ER, RO, and SPAs.
Fig. 1. The lower Keys region of the Florida Keys National Marine Sanctuary, southeastern Florida, with marine debris sampling locations and no-
fishing zones.
598 M. Chiappone et al. / Marine Pollution Bulletin 44 (2002) 597–604

were located. At each GPS point, one 25 m transect was
deployed, typically from inshore to offshore (total of 4
transects per 200 200 m2site). Marine debris were
surveyed by searching an area 1 m out from each tran-
sect side. The transect dimensions were selected to max-
imize the area sampled given the number of personnel
available and the number of other variables mea-
sured during the study. The type of gear, dimensions
(length, width, height), and numbers of sessile inverte-
brates impacted (touched and/or scarred) were noted.
Surveys employed plastic slates, transect reels, and a
meter stick.
Comparisons of the mean frequency of occurrence
(number of incidences per 100 m2) of debris categories
and numbers of organisms impacted per 100 m2were
made among habitat types and between no-fishing zones
and reference areas by habitat type. Statistical compar-
isons of means at the two spatial scales were conducted
by calculating confidence intervals (CI) based on the
equation CI ¼mean t½a;dfstandard error, with stan-
dard errors estimated by the two-stage, stratified ran-
dom sampling design (Cochran, 1977). CI were adjusted
for multiple comparisons using the Bonferroni proce-
dure (Miller, 1981). Goodness of fit procedures using
chi-square analyses were employed to test whether the
frequency of debris was independent of habitat type and
management regime (Zar, 1996).
Nine major categories representing 18 individual types
of debris items were recorded from 45 sites encompass-
ing 8040 m2(Table 2). Of the 110 debris occurrences,
monofilament line (38%), wood from lobster pots (20%),
combined fishing weights, leaders, and hooks (16%), and
rope from lobster traps (13%) were the most frequently
encountered, representing nearly 90% of all debris. De-
scriptive statistics for rope from lobster pots, monofila-
ment, and wire leaders were generated from estimates of
length in the field. The mean length of 14 occurrences of
rope was 6:81:83 m ðmean 1SEÞ, with a range 0.7–
20 m. Of the 42 occurrences of monofilament line, the
mean length was 1:00:22 m ðmean 1SEÞ, with a
range 0.09–7.1 m. Wire leaders were found 17 times
within the transects, had an average length of 0:8
0:15 m ðmean 1SEÞand ranged from 0.13 to 2 m in
length. Total lengths of rope, monofilament line, and
wire leaders recorded from all sites were 95.2, 43.2,
and 13.6 m, respectively.
Total debris density was significantly greater (com-
parison-wise a¼0:002) on aggregate offshore patch reefs
compared to nearshore hard-bottom, and the density of
debris from lobster traps was significantly greater on
aggregate offshore patch reefs compared to nearshore
hard-bottom and deeper fore reef strata (Fig. 2). Al-
though the average density of hook-and-line gear was
3.4 per 100 m2for the shallow fore reef, there was
Table 2
Frequency of fishing gear and other marine debris in the lower Florida Keys
Debris type Frequency % of total debris No. habitats recorded No. sites recorded
Remnant lobster traps
Wood 22 20.0 3 7
Rope 14 12.7 4 10
Cement 1 0.9 1 1
Plastic opening to pot 1 0.9 1 1
Hook-and-line gear
Monofilament linea42 38.2 3 11
Fishing weights, leaders, hooks 18 16.4 4 12
Glass bottles 1 0.9 1 1
Fabric (cloth windsock) 1 0.9 1 1
Plastic
Band 1 0.9 1 1
Mesh bag 1 0.9 1 1
Aluminum
Anchor 1 0.9 1 1
Cans 1 1.8 1 1
Can flip-top 1 0.9 1 1
Lead and other metals
Diving weights 1 0.9 1 1
Bar 1 0.9 1 1
Mesh grating 1 0.9 1 1
Rod 1 0.9 1 1
Boat motor 1 0.9 1 1
Total 110 100.0 5 28
a
Includes monofilament line with or without attached wire leaders, sinkers, and/or hooks.
M. Chiappone et al. / Marine Pollution Bulletin 44 (2002) 597–604 599

substantial inter-site variability. A disproportionate
amount of hook-and-line gear (55.7%) was found on the
shallow fore reef (X2¼136:7, df ¼6, P<0:001), espe-
cially in reference areas, despite only representing 11.1%
of the total sampling effort (Fig. 3). Similarly, a dis-
proportionate amount of lobster trap debris (57.9%
relative to 8.9% of effort allocation), primarily wood
slats and rope, was recorded from aggregate offshore
patch reefs (X2¼127:3, df ¼6, P<0:001), and to a
lesser extent on the shallow fore reef (23.7% vs. 11.1%
of sample allocation).
Although we expected no-fishing zones to yield lower
debris densities, especially for fishing gear, no significant
density differences (comparison-wise a¼0:05) for com-
bined debris types, hook-and-line gear, and lobster traps
were detected between zones and reference sites by
habitat type (Fig. 2). Chi-square analysis indicated that,
for all habitat types combined, the frequencies of hook-
and-line gear (X2¼0:06, df ¼1, P>0:75) and remnant
lobster traps (X2¼0:19, df ¼1, P>0:50) were not
significantly different between no-fishing zones and ref-
erence areas. While the frequencies of hook-and-line
gear (24 incidences in no-fishing zones, 37 in reference
sites) and debris from lobster traps (16 incidences in no-
fishing zones, 22 in reference sites) were slightly greater
in reference areas, appreciable amounts of debris were
Fig. 2. Mean frequency (95% CI) of combined debris types, combined hook-and-line gear, and combined lobster trap debris per 100 m2in the
Florida Keys between no-fishing zones (NTZ, ) and reference areas (REF, ) by habitat type. Habitat types are arranged from inshore to offshore
along the x-axis, with the number of sites sampled in each stratum in parentheses.
600 M. Chiappone et al. / Marine Pollution Bulletin 44 (2002) 597–604

recorded from the zones. For example, hook-and-line
gear was found 13 times from two sites within Sand Key
SPA at 8–12 m depth, or approximately three incidences
of gear per 100 m2. Similarly, hook-and-line gear was
recorded from the shallow fore reef of Looe Key SPA
seven times from two sites (two incidences per 100 m2),
with one site yielding six occurrences.
Fifty-four of the 110 occurrences (49%) of marine
debris caused tissue abrasion, other damage, and/or
mortality to 161 individuals or colonies of sessile inver-
tebrates, represented by sponges, branching gorgonians,
fire coral (Millepora alcicornis), scleractinian corals, and
the colonial zoanthid Palythoa mammilosa. Gorgonians
(37%) and sponges (28%) were the most commonly af-
fected, followed by fire coral (19%), scleractinian corals
(9%) and colonial zoanthids (8%). No significant differ-
ences (comparison-wise error a¼0:002) were detected in
the mean number of impacted organisms per 100 m2for
any of these taxa among habitat strata. Between no-
fishing zones and reference sites by habitat type, signifi-
cant differences (comparison-wise a¼0:05) in mean
impact frequency were detected for sponges (mid-chan-
nel patch reef zones >reference sites), scleractinian
corals (aggregate offshore patch reef zones >reference
sites), and colonial zoanthids (mid-channel patch reef
zones >reference sites), but not for fire coral or
branching gorgonians (Fig. 4).
Debris types causing the greatest degree of damage
were hook-and-line gear (68%), especially monofilament
line (58%), followed by debris from lobster traps (26%),
especially rope (21%). Hook-and-line gear accounted for
the majority of damage to branching gorgonians (69%),
fire coral (83%), sponges (64%), and colonial zoanthids
(77%) (Fig. 5). Remnant lobster traps were also impor-
tant, accounting for 64% of the stony corals impacted,
22% of gorgonians, and 29% of sponges. The frequency
of impacts for all debris types (X2¼49:3, df ¼4,
P<0:001) and hook-and-line gear (X2¼38:7, df ¼4,
P<0:001) were not equivalent among the invertebrate
taxa. In particular, branching gorgonians and sponges
were disproportionately more affected than other biota,
while scleractinian corals and colonial zoanthids were
less so. While the frequency of lobster trap impacts
differed among the taxa, they were more equally im-
pacted relative to hook-and-line gear (X2¼10:9, df ¼
4;0:05 <P<0:025), which caused disproportionately
more damage to gorgonians and sponges.
Considering the intensive commercial fishing effort
and the significant increases in registered recreational
boats and angler days in the Florida Keys (Bohnsack
et al., 1994), patterns in the distribution and frequency
of marine debris recorded during this study, especially
remnant fishing gear, are not surprising. However, this is
the first study we are aware of in the Florida Keys that
also attempted to document potential biological impacts
to subtidal organisms from marine debris, with com-
parisons made among multiple coral reef habitat types
and between no-fishing zones and reference areas.
Although 18 different debris items were recorded, hook-
and-line gear, especially monofilament line, and rem-
nant lobster traps, especially buoy lines, were the
predominant debris items and were differentially ap-
portioned to where fishing effort is concentrated for reef
fishes and spiny lobster, respectively. That is, hook-and-
line gear was most common in both shallow and deeper
fore reef areas further offshore, while lobster trap debris
was most abundant on offshore and mid-channel patch
reefs located between the shoreline and the offshore fore
reef. Not unexpectedly, debris density, particularly
hook-and-line gear, was very low or absent from near-
shore hard-bottom, mid-channel patch reef, and back
reef habitats. We initially assumed that, independent of
habitat type, the mean density of debris, especially
fishing gear, would be lower in no-fishing zones ‘‘pro-
tected’’ since 1997 compared to reference areas. How-
ever, no significant density differences for combined
debris, hook-and-line gear, and lobster traps were de-
tected. There are several possible explanations for these
patterns. First, non-compliance may have occurred in
some of the no-fishing zones sampled, especially on the
Fig. 3. Proportion of recreational hook-and-line gear and commercial lobster trap debris by benthic habitat type and management zone relative to
sampling allocation (number of sites).
M. Chiappone et al. / Marine Pollution Bulletin 44 (2002) 597–604 601

Fig. 4. Mean frequency (95% CI) of biological impacts per 100 m2for combined invertebrate taxa, sponges, and gorgonians in the Florida Keys by
benthic habitat type and inside (NTZ) and outside (REF) of no-fishing zones. Numbers of sites sampled in each stratum are in parentheses.
Fig. 5. Frequency of biological impacts to invertebrates by fishing gear and other debris in the Florida Keys.
602 M. Chiappone et al. / Marine Pollution Bulletin 44 (2002) 597–604

fore reef where hook-and-line gear was relatively com-
mon. Second, and specific to remnant lobster trap gear,
it is possible that storms or other factors distributed
ropes and wooden slats from the traps. Since lobster
traps are most commonly deployed in seagrass beds,
with some deployed on the periphery of patch reefs,
wave action from storms could have transported trap
components elsewhere. It is quite likely, for example,
that Hurricane Georges during 1998 resulted in the de-
struction and transport of numerous lobster traps.
Third, and plausible for both hook-and-line gear and
lobster traps, the gear encountered in 2000 could have
been remnant debris prior to 1997. Qualitative obser-
vations concerning whether debris is recent or heavily
fouled with organisms should be incorporated into fu-
ture surveys.
To put the study results into perspective, it is worth
noting that there has been a concerted effort on the part
of the FKNMS, non-governmental institutions (The
Nature Conservancy and The Bacardi Foundation), and
dive operators to organize and conduct an annual vol-
unteer reef cleanup effort (Adopt-a-Reef Program) in
the Florida Keys since 1994 (M. Enstrom, The Nature
Conservancy, personal communication). The program
targets recreational divers and provides data placards
for recording debris collected at designated locations,
usually popular dive sites, coordinated through local
dive shops. The types of data collected include date
and location of the cleanup, diver name, bottom time,
number of trash bags filled, a rank-order of the five most
important debris items, and an estimated biomass of
debris collected. From 1994 to 2000, 866 divers collected
nearly 7500 kg of debris, with hook-and-line gear, alu-
minum cans, plastic, cardboard, wood, and rope from
lobster pots constituting the most common items. These
are similar to results from the present study, the ex-
ception being the predominance of remnant lobster
traps in our surveys due to the inclusion of offshore
patch reefs. Despite 16 volunteer dives at Looe Key SPA
comprising 13 h of bottom time during 1999, hook-and-
line debris was not recorded, even though this was the
same general area as our 2000 surveys. In contrast, we
recorded six incidences of hook-and-line gear (mean of
three incidences per 100 m2) from the central fore reef
area of this site. This pattern may reflect differences in
specific locations surveyed, intensity of searches per unit
area for debris (i.e., roving diver versus small strip
transects), or non-compliance with regulations. Similar
explanations are plausible for Sand Key SPA. Despite
its designation as a no-fishing zone, hook-and-line gear
was found an average of 3.3 times per 100 m2on the
deeper fore reef. We are not aware of volunteer cleanup
efforts at this site, so this result may reflect remnant gear
prior to zone creation or to non-compliance.
Methods of fishing that cause habitat modification
or damage to benthic organisms represent potentially
serious consequences of fishing (Russ, 1991; Benaka,
1999). Although there is increasing recognition of the
consequences to benthic habitats from the use of mobile
fishing gear (Watling and Norse, 1998; Auster and
Langton, 1999) and other destructive fishing practices
(Saila et al., 1993; Jennings and Polunin, 1996), we are
not aware of any studies in the Florida Keys that have
evaluated biological impacts from marine debris and
specifically fishing gear. Interpretation of the biological
impact data is complicated by several factors. Both the
debris density and the distribution of sessile inverte-
brates sampled in this study are related to habitat type,
and secondarily by management type. Future efforts
need to consider the scaling of debris occurrence with
impacts relative to these two factors. For example, it
is probable that a coral-dominated reef with a given
amount of hook-and-line gear will not be affected in the
same way as a gorgonian–sponge dominated reef with
the same density of gear. Estimates of the proportion
of different taxa impacted by debris relative to total
abundance estimates are also useful for placing the de-
bris impact assessment into context. Also, the long-term
impacts to biota and the degree of recovery are un-
known. For example, we observed several instances
where hook-and-line gear, especially monofilament, was
overgrown by sponges, and it seems plausible that some
debris will be incorporated into the habitat matrix. We
also recognize that the future biological assessments
would be more useful if data on the severity of each
impact (e.g., amount of tissue damage) relative to the
size of the organism were collected.
We suggest that future debris surveys in the Florida
Keys should compare debris densities between no-fish-
ing zones and reference areas, as well as the impacts to
sessile biota and whether fishing gear is relatively recent
or biologically fouled. Despite these considerations, re-
sults from this study suggest that overall estimates of
biological impact from marine debris may be consider-
able, and such impacts are among a suite of factors that
affect the structure and condition of Florida Keys reefs.
As visitation and fishing pressure increase in this area,
it can be expected that the extent of marine debris and
the impacts to organisms will also increase.
Acknowledgements
Grants from Emerson Associates International, the
Florida Keys National Marine Sanctuary, and field
support from NOAA’s National Undersea Research
Center, University of North Carolina at Wilmington
supported this research. The authors would like to
particularly thank A. Creedon, M. Enstrom-Warner, E.
Franklin, B. Haskell, M. Tagliareni and The Nature
Conservancy. Research in the Florida Keys was granted
under NMS Permit FKNMS-074-98.
M. Chiappone et al. / Marine Pollution Bulletin 44 (2002) 597–604 603

References
Adams, C., 1992. Economic activities associated with the commercial
fishing industry in Monroe County, Florida. Staff Paper SP92-27,
Food and Resource Economics Department, Institute of Food and
Agricultural Sciences, University of Florida, Gainesville, FL, USA.
Ault, J.S., Bohnsack, J.A., Meester, G.A., 1998. A retrospective (1979–
1996) multispecies assessment of coral reef fish stocks in the Florida
Keys. Fishery Bulletin 96, 395–414.
Ault, J.S., Diaz, G.A., Smith, S.G., Luo, J., Serafy, J.E., 1999. An
efficient sampling survey design to estimate pink shrimp population
abundance in Biscayne, Bay Florida. North American Journal of
Fisheries Management 19, 696–712.
Auster, P.J., Langton, R.W., 1999. The effects of fishing on fish habitat.
In: Benaka, L. (Ed.), Fish Habitat: Essential Fish Habitat and
Rehabilitation. AFS Symposium 22, Bethesda, MD, pp. 150–187.
Benaka, L.R., 1999. Fish Habitat: Essential Fish Habitat and
Rehabilitation, AFS Symposium 22, Bethesda, MD, USA.
Bohnsack, J.A., Harper, D.E., McClellan, D.B., 1994. Fisheries trends
from Monroe County Florida. Bulletin of Marine Science 54, 982–
1018.
Cochran, W.G., 1977. Sampling Techniques, third ed. Wiley, New
York, USA.
Davis, G.E., 1977. Effects of recreational harvest on a spiny lobster,
Panulirus argus, population. Bulletin of Marine Science 27, 223–236.
Dayton, P.K., Thrush, S.F., Agardy, M.T., Hofman, R.J., 1995.
Environmental effects of marine fishing. Aquatic Conservation:
Marine and Freshwater Ecosystems 5, 205–232.
FDEP (Florida Department of Environmental Protection), 1998.
Benthic habitats of the Florida Keys. FMRI Technical Report TR-
4, St. Petersburg, FL, USA.
Hatcher, R.G., Johannes, R.E., Robertson, A.I., 1989. Review of
research relevant to the conservation of shallow water tropical
marine ecosystems. Oceanography and Marine Biology: An
Annual Review 27, 337–414.
Jennings, S., Lock, J.M., 1996. Population and ecosystem effects of reef
fishing. In: Polunin, N.V.C., Roberts, C.M. (Eds.), Reef Fisheries.
Chapman and Hall, New York, USA, pp. 193–218.
Jennings, S., Polunin, N.V.C., 1996. Impacts of fishing on tropical reef
ecosystems. Ambio 25, 44–49.
Jones, G.P., Syms, C., 1998. Disturbance, habitat structure and the
ecology of fishes on coral reefs. Australian Journal of Ecology 23,
287–297.
Miller, R.G., 1981. Simultaneous Statistical Inference. Springer, New
York, USA.
Roberts, C.M., 1995. Effects of fishing on the ecosystem structure of
coral reefs. Conservation Biology 9, 988–995.
Russ, G.R., 1991. Coral reef fisheries: effects and yields. In: Sale, P.F.
(Ed.), The Ecology of Fishes on Coral Reefs. Academic Press, New
York, USA, pp. 601–636.
Saila, S.B., Kocic, V.L.J., McManus, J.W., 1993. Modeling the effects
of destructive fishing practices on tropical coral reefs. Marine
Ecology Progress Series 94, 51–60.
Tilmant, J.T., 1989. A history and an overview of recent trends in
the fisheries of Florida Bay. Bulletin of Marine Science 44, 3–
22.
Watling, L., Norse, E.A., 1998. Disturbance of the seabed by mobile
fishing gear: A comparison to forest clearcutting. Conservation
Biology 12, 1180–1197.
Zar, J.H., 1996. Biostatistical Analysis, third ed. Prentice Hall, Upper
Saddle River, NJ, USA.
604 M. Chiappone et al. / Marine Pollution Bulletin 44 (2002) 597–604