Open Access
How to translate text using browser tools
1 January 2014 The Cast Net: an Overlooked Sampling Gear
Stein William, Patrick W. Smith, Galen Smith
Author Affiliations +

Sampling fishes and decapods in shallow, estuarine, marsh habitats is challenging because of depth variation, soft substrates that make standing difficult, and the presence of submersed aquatic vegetation. Though often overlooked, a cast net is an efficient gear type for sampling nekton in this environment. Complaints about the cast net are that it is difficult to deploy successfully and the results are not repeatable. We present a method of standardization of cast-net deployment that reveals that most of the variation in area covered is among individual operators rather than within one individual. The 1.8-m cast net collected more species, more biomass, and faster-swimming nekton than the throw trap. There was no statistical difference in the total number of animals collected or the total number of species between the gear types. We conclude that the cast net is a useful gear type for sampling nekton communities in estuarine habitats especially when there are differences in the spatial and temporal occurrences and relative abundances of many different organisms in an area with complex and varying habitats.

Estuaries and coastal wetlands are among the most productive ecosystems and are subject to constant change, both natural and anthropogenic (Day 1981; Stevens et al. 2006c; Day et al. 2008; Able and Fahay 2010). Sampling of fish and decapod communities in estuarine marshes is important for assessing the health of these areas and for monitoring anthropogenic impacts (Beck et al. 2001; Solomon et al. 2006; Rotherham et al. 2007; Sheaves and Johnston 2008). Information about the habitat requirements of all life stages of the fishes and decapods that inhabit these areas is critical in making decisions regarding preservation and restoration (Sargent and Carlson 1987; Beck et al. 2001). Sampling at a high spatial resolution may be necessary to understand habitat requirements for many species. The proper design of sampling regimens and the choice of gear is important to the success of any monitoring program (Rotherham et al. 2007). In addition to the intent of the study the choice of sampling gear depends on many factors that include, but are not limited to, the habitat being sampled and the species targeted (Sargent and Carlson 1987; Rojas and Minello 1997). Sampling in estuaries may be difficult because of the variation in depth, the variety of habitats, the nature and quality of the substrate and vegetation, and the variety of species (Rojas and Minello 1997).

Assemblage studies in estuarine marshes have typically used otter or beam trawls, drop samplers, seines, throw traps, cast nets, or a combination of multiple gear types (Winemiller and Leslie 1992; Martino and Able 2003; Solomon et al. 2006; Stevens 2006b; O'Connell et al. 2007, in press). The physical conditions in the study area may make sampling difficult and the use of some gear difficult or impossible. Direct observation and electrofishing may not be practical for estimating nekton densities in vegetated habitats (Jordan et al. 1997). Trawls and seines are often impractical because of depth, substrate, or vegetation. Shallow depths and limited accessibility may limit the use of drop samplers. Deep or fast-moving water may render throw traps ineffective.

Enclosure traps that quickly and efficiently surround a welldefined area in a habitat are very effective for sampling fish assemblages (Rozas and Reed 1993; Jordan et al. 1997; Rojas and Minello 1997). One type of enclosure trap, the throw trap, has been used extensively in Louisiana tidal marshes (Rojas and Minello 1997; Minello et al. 2012). It is easy to deploy and has high catch efficiency and high gear clearance efficiency. It is an excellent gear for estimating species density. The throw trap has become the “gold standard” for sampling shallow marshes and marsh edge particularly where the determination of nekton densities is the goal. However, it becomes increasingly difficult to use as depth increases. Therefore, alternative gear types may be useful in some marsh environs.

The cast net has been used by other investigators to collect fishes in shallow habitats and to supplement impoundment collections (Meador and Kelso 1990; Stevens 2006a). It has also been utilized in the sampling of shallow wetland pools (Sheaves and Johnston 2008). The cast net covers a large area per deployment and large areas can be sampled quickly. Gear avoidance by large or fast-moving nekton can be decreased with proper use of this gear. Some complaints about the cast net are that it is difficult to deploy and that the area covered by the net is not consistent (Leber 1995; Emmanuel et al. 2008). Catch efficiency is low compared with the throw trap as small animals can escape through the cast-net mesh.

In June 2010, we initiated a study of the fish and decapod community structure of the Lake Pontchartrain-Lake Borgne Land Bridge, a highly productive estuarine marsh in the Pontchartrain Estuary of southeastern Louisiana. The purpose of this study was a large scale initial assessment of seasonal species, relative abundance, composition, and richness during a 2-year period. We were concerned about the limitations of the cast net and how it might bias our results. To better understand these potential limitations, cast-net samples were compared with throw-trap samples. Specifically, this study examined the cast net as a sampling gear and compared it with the throw trap in shallow estuarine environments. We addressed the issue of inconsistent deployment by presenting a method of standardizing the area covered by the net for different users by standardizing the mean and variation of coverage for a series of deployments. We then compared the effectiveness of the throw trap to that of the cast net for estimating relative abundance in a nekton community in southeastern Louisiana.


Study area.—The Lake Pontchartrain-Lake Borgne Land Bridge, located in the Pontchartrain Estuary in southeastern Louisiana, comprises more than 100 km2 of predominantly Spartina spp. salt marsh with tidal creeks and irregular shallow ponds (Penland et al. 2002). This area is frequently affected by both winter and tropical storms and is subject to compaction and subsidence, saltwater intrusion, and erosion. For purposes of this study, the Land Bridge was divided into five distinct areas of approximately equal size, and each area was subdivided into 250 × 250-m squares. Each month for 24 months, four randomly selected squares in each of the five areas were sampled at five random sites each. Depth typically ranged from 20 cm to over 3 m. The substrate is composed primarily of poorly compacted decaying organic material, clays, and silts. It is not possible to stand or walk on the substrate or the sparse land of the marsh platform. Flow velocity in the creeks may exceed 1 m/s especially with the combined effect of wind stress and tides. Submersed aquatic vegetation (SAV) and algae are dense in many of the ponds in the summer and fall.

Gear.—The throw trap used in this study was a 1-m2 aluminum box, open at both ends and 78 cm deep. It was emptied by means of a 1-m2 aluminum frame, which fit just inside the throw trap, covered with 1-mm nylon mesh. The monofilament cast net used for our study was 1.8 m in radius with 2 kg of lead per radius meter. We chose to use the cast net with 6-mm mesh for our study as this mesh size was commonly used in bag seines and trawls in similar studies and because this is the smallest mesh size that is commonly commercially available.

Cast-net deployment standardization.—To assess the variation of cast-net deployment, three different operators of different skill levels and experience each threw the net 10 times onto a flat lawn of 5-cm-high St. Augustine grass. A work platform at the end of a 21-m articulated boom mounted on a truck bed (bucket truck) was maneuvered directly over the net after it was deployed. A plumb-weight suspended from the platform was maneuvered until it was in the center of the net. A 50 × 50-cm white square was then placed directly under the weight and a digital photograph was taken from a height of 8 m directly over the weight to minimize any angular foreshortening.

Using Image J software, the area of the net and the area of the white square were measured in each digital image and the area of the net throw was computed from the ratio (Rasband 1997— 2011). The mean area covered by the net and the variance and SD were computed for each of the operators. We used repeatability (r) as a measure to quantitatively describe the variation that occurs among rather than within operators (Lessells and Boag 1987). Mean values of the areas of each operator's throws and the variance were obtained and compared using an ANOVA and post hoc analyses were performed with Tukey's honestly significantly different (HSD) function in R Statistical Software version 3.0.1.

Repeatability can be used to examine the consistency of the operators. Repeatability (Lessells and Boag 1987) was computed as follows:

where is between group variance and is within group variance.

Field study: throw trap and cast net.—Sampling in the field was conducted at 37 different sites on the Land Bridge during June, September, and November 2011, and January and July 2012. All sampling was performed using a shallow-draft, motorized, aluminum, fiat-bottomed boat outfitted with a 2-m anchor pole and 5 m of rope. The boat was allowed to drift into the sampling area and the pole was used to anchor the boat and the end of the 5-m tether. For the first 20 samples, either the net or the throw trap was arbitrarily deployed first. The order was randomly determined by a coin flip for the remaining 17 samples. After the first gear was deployed and emptied, the boat position was shifted on the mooring rope such that the second gear was deployed at a different undisturbed location. In this manner, the same community was sampled in each area by both the cast net and the throw trap. Initially, the times required for sampling with each gear type were recorded for the first six samples with each gear type.

The cast net was deployed by the most-experienced operator from the deployment standardization method. It was randomly deployed <3 m from the boat, allowed to sink, and then slowly retrieved and emptied into a large tub. All SAV was carefully removed and inspected. All fishes and decapods were placed into a sealable plastic bag and euthanized in an ice-water bath, as per University of New Orleans, Institutional Animal Care and Use Committee protocol 09–016.

The throw trap was deployed into undisturbed water no deeper than 78 cm. If the water level came over the side of the throw trap, the sample was terminated and the boat was moved to a new location at least 50 m distant. If the net sampling was done first, it was repeated in the same order at the new location. The throw trap was then swept with the net frame until three successive sweeps failed to produce any new animals. All SAV was removed from the throw trap and inspected for presence of animals, which were placed in bags and euthanized as described above. All animals were identified to species level and weighed. Standard length of fishes, carapace width of crabs, and total length of penaeid shrimp were recorded. The number of each species, the total number of species, and total biomass were also recorded for each sample. If a sample contained more than 25 animals of a single species, only the largest and smallest were weighed and measured along with 23 additional specimens randomly selected.

The cast-net and throw-trap data were compared across all of the samples rather than as a pairwise comparison of the samples, because the same community was sampled with each pair of gear deployments. A multivariate ANOVA (MANOVA) was performed using R Statistical Software (a = 0.05), with weight, number of animals, and number of species per sample as dependent variables and gear type, i.e., cast net or throw trap, as predictor variables (Ihaka and Gentleman 1996). The weight and number of animals in each collection were log transformed to satisfy the assumptions of normality required for the test. If significant differences were found between gear types, pairwise t-tests for each dependent variable (weight, number of animals, and number of species per sample) were performed.

A resemblance matrix of the samples (numbers of each species) was created using Primer 6 software in order to assess the relative distances apart of the samples in the same rank order as the dissimilarities (Clarke and Gorley 2006). A one-way analysis of similarity (ANOSIM) using cast net and throw trap as predictor variables was performed on the samples as number of animals per species per sample. Similarity percentages analysis (SIMPER) was performed to determine which species were driving the dissimilarity between gear types.


Cast-Net Standardization

An ANOVA of “areas” as a function of the operators demonstrated a significant difference among the operators relative to the difference within the operators (F = 32.28, P < 0.001). However, since the within-operator error is small compared with the between-operator difference, it suggests that the operators were more consistent than different operators were alike (Vanhooydonch et al. 2005).

Repeatability (r) was 0.758, which means that approximately 76% of the variation was between operator pairs rather than within operators. This showed that although there is a difference among operators, individual operators were consistent in their performance. There was a difference in skill level and more skillful throwers were able to cover more area, but each thrower appeared to be consistent in their throws and cover a similar area each time (Figure 1). This did not imply that an operator's throws would cover the same area each throw, but rather there would be little variation over a large number of throws. This method can be repeated periodically to determine the area a single operator's throws cover as they gain experience or proficiency and to allow for the comparison of different operators. This experiment was conducted with a relatively light-weight net (<6 kg). Fatigue did not appear to influence the results. It is possible that a heavier net or a greater number of repetitions may negatively influence consistency. These results imply that an individual operator's performance can be standardized.

Field Study: Throw Trap and Cast Net

A difference between times required for sampling with the two gear types was immediately noted and this part of the experiment was discontinued. It took less than 5 min for a single sample with the cast net, but at least 15 min for a single sample with the throw trap depending upon the number of sweeps required with the clearing net, as some samples required more than 30 min to complete. Furthermore, if the depth was more than 0.78 m or if the trap sank in the soft substrate, sampling with both gear types had to be repeated.

Abiotic data were obtained at each sampling site. Water depth was measured directly and temperature and salinity measurements were made with a Yellow Springs Instruments Professional Plus meter. All sampling was performed at depths ranging from 0.23 to 0.78 m and was limited by the depth of the throw trap. Water temperatures ranged from 13°C to 32.4°C. Salinity varied from a minimum of 0.71‰ to 10.35‰. Secchi disk depth varied from 0.2 to 0.6 m.

A MANOVA was performed which showed that the gear types were significantly different (F = 7.35, P = 0.008; Figure 2). Pairwise t-tests indicated that the weight of the samples was significantly different between gear types (t = 5.02, P < 0.001; Table 1), but the number of animals (t = 1.51, P = 0.135) and species (t = 0.282, P = 0.779) collected per sample were not (Table 1).

Collections included 22 different species of fishes and decapods (Table 2). Sixteen different species were collected in the throw trap and 21 in the cast net. All of the cast-net samples contained fishes or decapods or both. Five of the 37 throw-trap samples did not contain any animals. The two gear types differed in the total numbers of animals collected and in the weight of the samples. The cast net collected 616 animals, the total weight of which was 2,966 g and the mean weight per sample was 80.18 g. The weight of the 946 animals collected in the throw trap was 187 g and the mean weight per sample was 5.06 g. The mean number of animals collected per cast-net throw was 17 (range, 1–139). The mean number of species per throw was 3.08 (range, 1–6; SD = 1.31). For the throw trap, the mean number of animals collected was 26 (range, 0–247) and the mean number of species was 2.54 (range, 0–6; SD = 1.7).


Box plot depicting the mean and quartiles of area covered by a cast net thrown 10 times by three different operators. Operators 1 and 2 were experienced while operator 3 was inexperienced. Efforts by operator 3 exhibited a greater range, variance, and SD.


Box plots depicting the median and quartiles for centroid values, log weight, species, and log individuals for each gear type: cast net and throw trap. Centroid values are standardized combined values that included all three response variables analyzed. An asterisk (*) indicates significant difference.

Red Drum and Sailfin Mollies were both collected in the throw trap and not in the cast net. The throw-trap species did not include Striped Mullet, Gulf Menhaden, Sand Seatrout, Pinfish, Silver Perch, or Leatherjack, all of which were collected by the cast net. Of the 946 animals collected by the throw trap, 765 were Rainwater Killifish and the next most common was the blue crab. Gulf Menhaden was the most common species in the cast net collection (278) followed by Rainwater Killifish (109; Table 2). The cast net covered four times the area of the throw trap and small Rainwater Killifish likely escaped through the larger mesh.

Although the cast net collected almost all of the species collected with the throw trap, the converse was not true. The throw trap tended to collect more benthic and SAV-associated animals while the cast net collected not only those, but also more pelagic species. The one exception to this was the single throwtrap deployment that accounted for the only eight small juvenile Red Drum collected in the sampling. All of the Striped Mullet, Gulf Menhaden, Leatherjack, and Sand Seatrout were collected in the cast net. Most of the Rainwater Killifish, Gulf Pipefish, and the gobies were collected in the throw trap although all of these species were represented in the cast-net samples. An ANOSIM showed a significant difference among the samples (Global R = 0.063, significance level: P = 0.003), which suggests, but not strongly, that the two gear types sampled different communities. However, the very low Global R-value implies that the variation within samples is high. This may partly be responsible for the significant difference observed between gear types.


Results of pairwise t-tests for cast-net and throw-trap samples. The significance level was adjusted for the t-tests by dividing the original a level (0.05) by the number of variables (three). Significant results are shown in bold italic text. Only the weights of the samples were significantly different. The number of species and the number animals collected were not statistically different for the different gear types. The significance level had to be adjusted for the t-tests. This was done by taking the original level (0.05) and dividing it by the number of variables (three).


Species composition for 37 collections with each gear type including total numbers of each species collected. The cast net collected essentially all of the species collected in the throw trap but surpassed the throw trap in capturing pelagic species. The throw trap collected more benthic animals and those associated with SAY.

Distinguishing taxa in the cast-net samples relative to the throw-trap samples (percent contribution to average dissimilarity) included more numerous Gulf Menhaden (8.5%), Gulf Killifish (5.75%), Sheepshead Minnow (4.44%), Spot (3.76%), and Striped Mullet (3.05%) and fewer Rainwater Killifish (35.57%), Bay Anchovy (7.70%), Gulf Pipefish (3.81%), and Naked Goby (5.93%; Table 3). The differences in species collected by the two gear types are reflected in the differences in standard lengths of fish collected between the two gears. Rainwater Killifish less than 12 mm SL were collected by both gear types although the smallest fish (10 mm SL) was collected by the cast net. Blue crabs less than 10 mm carapace length were collected by both gear types. The largest animals were collected in the cast net (Figure 3) and included a Striped Mullet (220 mm SL). The difference in organism lengths between the two gear types indicated a difference in the animals collected in the communities sampled; the cast net collected more motile nekton with longer SLs. The same community was sampled but the two gear types separated out different components.


We have addressed concerns that the cast net is difficult to deploy and standardize when sampling estuarine habitats. A certain amount of skill is required to master the use of any gear type and the cast net is no exception. Just as there are techniques for employing seines, trawls, and throw traps, there are techniques for properly deploying a cast net. Assuming the operator has mastered the basic technique, we have shown that it is possible to standardize the net throw for each operator, in terms of coverage area, regardless of the amount of training or experience that the operator may possess. It is reasonable to assume that each operator will be consistent in the area that is covered by a series of throws. One method previously used to estimate the area covered by a cast net was to throw it from a boat 14 times onto grass and measure the maximum and minimum radius for each throw (Stevens 2006b). These measurements were then used to calculate the approximate area covered with the assumption that the spread net approximated an ellipse. Using a net with radius of 1.15 m, the area covered was estimated to be 2.8 ± 0.2 m2 (mean ± SD) (Stevens 2006b). The calculated maximum coverage area using the reported net radius (without knowing the length of the lead line) is 4.15 m2. The approximate coverage area was 67% of the maximum. Our method produced a more accurate estimate of the covered area by measuring the actual area rather than approximating it. Both experiments confirmed that the area actually covered by the thrown net cannot be calculated from the length of the brails but must be measured. Furthermore, our data revealed that although there is variance among operators, individual operators are consistent from cast to cast. As operators gained experience, we found that at a certain point their coverage area remained relatively constant. Experience may decrease the amount of individual variation. For this reason, it may be necessary to conduct an initial assessment of each operator and to perform a periodic reassessment. Our method will allow a semiquantitative approach for using this gear because a statistically meaningful evaluation can be made across a number of samples, although a comparison cannot be made from cast to cast.


SIMPER Analysis of cast net versus throw trap collections (Primer 6). The two gear types differ in that the cast net collected more of the motile species that may escape the throw trap and the throw trap collected more of the smaller more benthic species and SAV associated nekton that may escape through the mesh of the cast net.

Sampling may be quantitative or qualitative in nature. For some sampling regimes, the goal of the research may be to determine the number of animals present within a habitat (density) and their response to various treatments (Rojas and Minello 1997). In others, the goal may be to determine the number of species within a specific area or habitat, define the community, or determine diversity or productivity (O'Connell et al. 2004). In each case, the choice of gear type may be different. The accuracy and precision of the gear employed is determined not only by the ability of the gear to entrap all of the animals in the area sampled (collection efficiency), but also on the ability of the operator to remove all of the animals collected from the gear (clearance efficiency) (Jordan et al. 1997; Steele et al. 2006). Once collected in the gear, all of the animals must be removed and counted (gear efficiency). The suitability of various gear types to accomplish these goals has been the subject of much debate and investigation (Sargent and Carlson 1987; Rojas and Minello 1997; Able et al. 2005; Rotherham et al. 2007; Able and Fahay 2010).


A density plot for all measured fish from each gear type is depicted. Only up to 25 fish were measured per species per sample (see Materials and Methods for more details). The majority of fish sampled using a throw trap were from a smaller SL range than those sampled using a cast net (˜ 10–40 mm and ˜10–80 mm, respectively).

We analyzed our data by comparing the results of each gear deployment rather than comparing the results in terms of collections per square meters sampled. The area covered by the cast net was approximately four times greater than the area covered by the throw trap. Had we analyzed the data by area rather than by deployment, the results may have been different. We did not test the catch efficiency of gear in our study (Kjelson and Colby 1977). We relied on previous work, which showed that cast nets range from 10% to 28% catch efficiency, combining capture and recovery efficiency, with a mean of approximately 20% (Jordan et al. 1997; Rojas and Minello 1997; Stevens 2006b, 2006c). The capture efficiency of the throw trap has been estimated to be 60–75% and the recovery efficiency greater than 85% (Rozas and Reed 1994). If the capture efficiency of the throw trap was four times greater than that of the cast net and the cast net covered four times the area of the throw trap, the results for the two gear types should have been similar, but they were not. Capture efficiency of the throw trap is marginally greater for larger animals than for smaller ones (Rozas and Reed 1994; Rojas and Minello 1997). Our throw trap was constructed of aluminum and was heavy and difficult to deploy. Throw traps constructed of lightweight plastic pipe and mesh netting undoubtedly would be much easier to use. A throw trap made of lightweight material could be deeper than our 78-cm throw trap, which would allow sampling in deeper water.

In our study, the cast net collected larger and more motile animals than did the throw trap. Gear avoidance was decreased because the net was deployed farther from the boat than was the throw trap. The cast net was made of nylon mesh and was less visible than the aluminum throw trap. The cast net covered an area four times as large as the throw trap making it difficult for larger animals to swim out from under it. The cast net collected more species and more biomass than the throw trap and because it successfully collected larger and more motile fishes, such as Striped Mullet. Stevens (2006b) also found that the cast net collected more Striped Mullet, Ladyfish Elops saurus, and White Mullet Mugil curema than did the throw trap.

At the other end of the size spectrum, the throw trap was more efficient at collecting smaller nekton. Small animals were able to escape through the mesh of the cast net but could not escape the throw trap. It is noteworthy in our study that the cast net also collected very small animals as well, despite the relatively large minimum mesh size of 6 mm, but animals were generally of greater length than those collected by the throw trap (Figure 3). Our cast-net collections did not contain as many SAV-oriented animals such as Rainwater Killifish or as many of the benthic gobies as did the throw trap because small nekton were able to escape through the mesh. The cast net did not bring up all of the SAV it encompassed, so undoubtedly, additional fishes that may have been present were not collected. Fishes such as gobies are capable of burrowing into the substrate and are capable of avoiding the cast net but not the throw trap. As evidenced by the large number of gobies and Rainwater Killifish in the collections, the throw trap is biased towards benthic and smaller animals as these animals are less able to avoid the trap than are larger more motile animals. The clearing net used with the throw trap had 1 mm mesh whereas the cast net had 6 mm mesh, which may account for some of the differences in our study in that fewer small animals were likely to escape the throw trap than the cast net during clearing.

The absolute numbers of animals collected was not markedly different between the two gear types nor were the number of species collected. The purpose of our study was to characterize the nekton community structure on the Land Bridge. We were not interested in absolute densities of animals, but rather their relative abundance. We were sampling a very large area and time and resources were finite. The cast net allowed us to gather the most data per unit effort to accomplish this goal. Had we been interested in the absolute densities of nekton, the throw trap may have been a better choice.

As with other gear types, the choice of the best cast net for the study is important. Sufficient net radius size must be chosen to adequately cover the anticipated area to be sampled without being too large for the operator to handle successfully. Nor should it be too large for the area being sampled. The size of the animals collected will be determined to some extent by the mesh size of the net. We collected fishes <12 mm SL with a 6mm-mesh net. However, the smaller the mesh size is, the smaller the animals that are collected. Smaller cast nets, though, create more air and water resistance. In deep water, drag produced by small mesh size may cause the net to collapse prematurely thus decreasing the effective coverage area. The weight of the lead line is equally important as this will determine how fast the net sinks, how well it counteracts the drag of the mesh, and how well it will enclose animals particularly in the presence of SAV. It is better to use too much weight than too little. Our experience has shown that the cast net cannot be used in the presence of very dense SAV or emergent vegetation, but is efficient in the presence of moderate, less-dense SAV. We have compared sampling of fish and decapod communities with the throw trap and have characterized a particular type of habitat where the cast net is a useful sampling gear. The present study differs from previous studies in which the cast net was employed for specific purposes in specific habitats: canals, impoundment ditches, and marsh pools (Meador and Kelso 1990; Stevens 2006a; Sheaves and Johnston 2008). The use of the cast net has enabled us to obtain a large number of samples over a large geographical area and across a wide range of habitats in an efficient amount of time. We have demonstrated that in our study, the cast net was more versatile and faster to use than the throw trap and was not limited by depth and boat size. In our habitats sampled it was less destructive of vegetation than were beach seines and trawls and was also easier to use, given the shallow depths and soft substrate. The cast net was less labor intensive than the throw trap. As more coastal marshes become shallow open water as a result of a rise in sea level and saltwater intrusion, the number of areas where the cast net will be a useful gear type will increase. However, in choosing the gear type that is best suited to each application, the researcher must first consider the limitations imposed by the habitats and species to be sampled, then consider using the cast net.


We thank Chris Canedo for technical assistance with bucket truck and methodology and Jon McKenzie, Ph.D., University of New Orleans, for help with the cast-net standardization. We thank F. Jordan for the use of his throw trap. A special thanks to Martin T. O'Connell, Ph.D., University of New Orleans, for research support and reviewing previous drafts. This manuscript represents publication no. 13 for the Nekton Research Laboratory, Pontchartrain Institute for Environmental Sciences.



K. W. Able , and M. P. Fahay . 2010. Ecology of estuarine fishes: temperate waters of the western North Atlantic. John Hopkins University Press, Baltimore.  Google Scholar


K. W. Able , K. J. Smith , and S. M. Hagan . 2005. Fish composition and abundance in New Jersey salt marsh pools: sampling technique effects. Northeast Naturalist 12:485–502. Google Scholar


M. W. Beck , K. L. Heck Jr ., K. W. Able , D. L. Childers, D. B. Eggleston , B. M. Gillanders , B. Halpern , C. G. Hays , K. Hoshino , T. J. Minello , R. J. Orth , P. F. Sheridan , and M. P. Weinstein . 2001. The identification, conservation, and management of estuarine and marine nurseries for fish and invertebrates. Bioscience 51:633–641. Google Scholar


K. R. Clarke , and R. N. Gorley . 2006. Primer v6: user manual/tutorial. Primer-E, Plymouth, UK. Google Scholar


J. H. Day , editor . 1981. Estuarine ecology with particular reference to southern Africa. A. A. Balkema, Rotterdam, The Netherlands. Google Scholar


J. W. Day , R. R. Christian , D. M. Boesch , A. Yáñez-Arancibia , J. Morris , R. R. Twilley , L. Naylor , L. Schaffner , and C. Stevenson . 2008. Consequesnces of climate change on the ecogeomorphology of coastal wetlands. Estuaries and Coasts 31:447–191. Google Scholar


B. E. Emmanuel , L. O. Chukwu , and L. O. Azeez . 2008. Cast net design characteristics, catch composition and selectivity in tropical open lagoon. African Journal of Biotechnolology 7:2081–2089. Google Scholar


R. Ihaka , and R. Gentleman . 1996. R: A language for data analysis and graphics. Journal of Computational and Gaphical Statistics 5:299–314. Google Scholar


F. Jordan , S. Coyne , and J. C. Trexler . 1997. Sampling fishes in vegetated habitats: effects of habitat structure on sampling characteristics of the 1-m2 throw trap. Transactions of the American Fisheries Society 126:1012–1020. Google Scholar


M. A. Kjelson , and D. R. Colby , editors. 1977. The evaluation and use of gear efficiencies in the estimation of estuarine fish abundance, volume II. Circulation, sediments, and transfer of material in the estuary. Academic Press, New York. Google Scholar


K. M. Leber 1995. Significance of fish size-at-release on enhancement of Striped Mullet fisheries in Hawaii. Journal of the World Aquaculture Society 26:143–153. Google Scholar


C. M. Lessells , and P. T. Boag . 1987. Unrepeatable repeatabilities: a common mistake. Auk 104:116–121. Google Scholar


E. J. Martino , and K. W. Able . 2003. Fish assemblages across the marine to low salinity transition zone of a temperate estuary. Estuarine Coastal and Shelf Science 56:969–987. Google Scholar


M. R. Meador , and W. E. Kelso . 1990. Physiological responses of Largemouth Bass, Micropterus salmoides, exposed to salinity. Canadian Journal of Fisheries and Aquatic Sciences 47:2358–2363. Google Scholar


T. J. Minello , L. P. Rozas , and R. Baker . 2012. Geographic variability in salt marsh flooding patters may affect nursery value for fishery species. Estuaries and Coasts 35:501–514. Google Scholar


M. T. O'Connell , R. C. Cashner , and C. S. Schieble . 2004. Fish assemblage stability over fifty years in the Lake Pontchartrain estuary; comparisons among habitats using canonical correspondance analysis. Estuaries 27:807–817. Google Scholar


M. T. O'Connell , A. M. U. O'Connell , and C. S. Schieble . In press. Response of Lake Pontchartrain fish assemblages to Huricanes Katrina and Rita. Esutuaries and Coasts. DOI: 10.1007/sl2237-013-9675-3. Google Scholar


M. T. O'Connell , T. D. Sheppherd , A. M. U. O'Connell , and R. A. Myers . 2007. Long term declines in two apex predators, Bull Sharks (Carcharhinus leucas) and Alligator Gar (Atractosteus spatula), in Lake Pontchartrain, an oligohaline estuary in southeastern Louisiana. Estuaries 27:807–817. Google Scholar


S. Penland , P. McCarty , A. Beall , and D. Maygarden . 2002. Environmental overview. In S. Penland , A. Beall , and J. Kindinger , editors. Environmental atlas of the Lake Pontchartrain basin (CD-ROM). U.S. Geological Survey, Open File Report 02–206, Reston, Virginia. Available: (February 2014). Google Scholar


W. S. Rasband 1997–2011. ImageJ. U.S. National Institutes of Health, Bethesda, Maryland. Google Scholar


L. P. Rojas , and T. J. Minello . 1997. Estimating densities of small fishes and decapod crustaceans in shallow estuarine habitats: a review of sampling design with focus on gear selection. Estuaries 20:199–213. Google Scholar


D. Rotherham , W. G. Macbeth , S. J. Kennelly , and C. A. Gray . 2007. Reducing uncertainty in the assessment and management of fish resources following an environmental impact. ICES Journal of Marine Science 64:1512–1516. Google Scholar


L. P. Rozas , and D. J. Reed . 1993. Nekton use of marsh-surface habitats in a Louisiana (U.S.A.) deltaic salt marshes undergoing submergence. Marine Ecology Progress Series 96:147–157. Google Scholar


L. P. Rozas , and D. J. Reed . 1994. Comparing nekton assemblages of subtidal habitats in pipeline canals traversing brackish and saline marshes in coastal Louisiana. Wetlands 14:262–275. Google Scholar


W. B. Sargent , and P. R. Carlson . 1987. The utility of Breder traps for sampling mangrove and high marsh fish assemblages. Pages 194–205 in F. J. Webb , editor. Proceedings of the 14th annual conference on wetlands restoration and creation. Hillsborough Community College, Tampa, Florida. Google Scholar


M. Sheaves , and R. Johnston . 2008. Influence of marine and freshwater connectivity on the dynamics of subtropical estuarine wetland fish metapopulations. Marine Ecology Progress Series 357:225–243. Google Scholar


J. J. Solomon , R. B. Brodie , and G. S. Ehlinger . 2006. Distribution and abundance of fish assemblabes and select macroinvertebrate s from the lower St. Mary's River basin in northwest Florida. Florida Scientist 69:1–18. Google Scholar


M. A. Steele , S. C. Schroeter , and H. M. Page . 2006. Sampling characteristics and biases of enclosure traps for sampling fishes in estuaries. Estuaries and Coasts 29:630–638. Google Scholar


P. W. Stevens 2006a. Patterns of fish use and piscivore abundance within a reconstructed saltmarsh impoundment in the northern Indian River Lagoon, Florida. Wetlands Ecology and Management 14:147–166. Google Scholar


P. W. Stevens 2006b. Sampling fish communites in saltmarsh impoundments in the northern Indian River Lagoon, Florida: cast net and culvert trap gear testing. Florida Scientist 69:135–147. Google Scholar


P. W. Stevens , C. L. Montague , and K. J. Sulak . 2006c. Fate of fish prodution in a seasonally flooded saltmarsh. Marine Ecology Progress Series 327:267–277. Google Scholar


B. Vanhooydonch , A. Y. Herrel , R. Van Damme , and D. J. Irschick . 2005. Does dewlap size predict male bite performance in Jamaican Anolis lizards? Functional Ecology 19:38–42. Google Scholar


K. O. Winemiller , and M. A. Leslie . 1992. Fish assemblages across a comples, tropical freshwater/marine ecotone. Environmental Biology of Fishes 34:29– 50. Google Scholar
© American Fisheries Society 2014
Stein William, Patrick W. Smith, and Galen Smith "The Cast Net: an Overlooked Sampling Gear," Marine and Coastal Fisheries: Dynamics, Management, and Ecosystem Science 6(6), 12-19, (1 January 2014).
Received: 9 May 2013; Accepted: 6 November 2013; Published: 1 January 2014
Back to Top