Estuarine and Coastal Habitat Use of Gulf Sturgeon (Acipenser oxyrinchus desotoi) in the North-Central Gulf of Mexico
- First Online:
- Cite this article as:
- Ross, S.T., Todd Slack, W., Heise, R.J. et al. Estuaries and Coasts (2009) 32: 360. doi:10.1007/s12237-008-9122-z
- 167 Views
Gulf sturgeon are anadromous, spawning in freshwater and returning to the marine environment to feed. Herein, we document the marine distribution and timing of movement in and out of the marine environment of Gulf sturgeon natal to the Pascagoula and Pearl rivers (MS and LA). From 1999 to 2004, we attached sonic transmitters to 194 fish averaging 151 (MS) to 160 (LA) cm in fork length. We located 56 different Gulf sturgeon in the estuarine or marine environments, some multiple times. Fish were distributed nonrandomly, being found primarily in shallow water (mean = 3.9 m) in barrier island passes. Benthic samples taken at Gulf sturgeon telemetry location sites were dominated by Florida lancelets, sand dollars, annelids, haustoriid amphipods, and mollusks—all documented prey of Gulf sturgeon. Movement into salt water consistently occurred in October and November; movement back into rivers or low salinity estuaries was complete by the end of March.
KeywordsGulf sturgeonMarine habitat useAcipenser oxyrinchus desotoiAcipenseridaeConservationAnadromous fishesGulf of Mexico
Anadromous sturgeon, once abundant and the basis for various commercial fisheries, today exist as diminished and fragmented populations and worldwide are considered one of the most threatened groups of fishes (Rochard et al. 1990; Gross et al. 2002; Van Winckle et al. 2002; Munro et al. 2007). In North America, there are two species and three recognized taxa of oceanic anadromous sturgeon (Munro et al. 2007). Green sturgeon, Acipenser medirostris, occur along the Pacific coast, and Atlantic sturgeon, Acipenser oxyrinchus, comprising two subspecies, occur along the Atlantic and Gulf coasts. The nominal subspecies of Atlantic sturgeon, A. oxyrinchus oxyrinchus, ranges from Labrador, throughout the Gulf of St. Lawrence drainage, and southward at least to the St. Johns River in Florida, and Gulf sturgeon, A. oxyrinchus desotoi, occurs in Gulf of Mexico drainages from Tampa Bay westward to the Mississippi River (Gruchy and Parker 1980).
Populations of Gulf sturgeon occurring in the Pascagoula and Pearl rivers of Mississippi and Louisiana are the focus of this study. Gulf sturgeon have never been reported from the Escatawpa River, a major lower tributary of the Pascagoula River that flows primarily through Alabama (Ross 2001; Boschung and Mayden 2003). The Gulf sturgeon is listed as threatened under the Endangered Species Act (USFWS 1991) and as endangered in Mississippi (MMNS 2001) and threatened in Louisiana (LDWF 2008), with the principal reasons for population declines being habitat loss due to dams, commercial fishing, and general water quality deterioration (USFWS and Gulf States Marine Fisheries Commission 1995).
Herein, we address marine habitat use of Gulf sturgeon. Until recently, there has been much greater attention given to the freshwater phase of both the Gulf sturgeon (Carr et al. 1996; Parauka et al. 2001; Fox et al. 2002; Edwards et al. 2003) and Atlantic sturgeon (Bain 1997; Smith and Clugston 1997; Stein et al. 2004). In fact, Fox et al. (2002) concluded that the marine foraging locations for Gulf sturgeon were still not known. Stein et al. (2004) used fisheries data to study the marine distribution of Atlantic sturgeon off the northern US coast. Even though fish were primarily found over gravel and sand substrata, they proposed that Atlantic sturgeon should show little if any marine habitat selection as long as prey were available.
In all cases of anadromous sturgeon, movement out of freshwater into estuarine and coastal environments is associated with feeding (Bemis and Kynard 1997). For Gulf sturgeon, as is generally true for other anadromous sturgeon species, feeding by subadults and adults is primarily limited to estuarine or marine waters, with the fish fasting and losing weight during the period of freshwater residency (Wooley and Crateau 1985; Mason and Clugston 1993; Carr et al. 1996; Sulak and Randall 2002; Heise 2003). This has been further documented by analysis of carbon isotope ratios that show subadult and adult Gulf sturgeon to be almost totally dependent upon the marine food web for growth (Gu et al. 2001).
Some information on marine habitat use by Gulf sturgeon exists for the northeastern Gulf of Mexico, especially for estuaries. For instance, Fox et al. (2002) found adult Gulf sturgeon in Choctawhatchee Bay in depths of 2–4 m with high sand content; three fish that moved into the Gulf of Mexico were in shallow water (mean = 8 m), again with a high sand content. Dominant potential prey items at Gulf sturgeon relocation sites in Choctawhatchee Bay were annelids and crustaceans. Sulak and Clugston (1999) released four Gulf sturgeon near the mouth of the Suwannee River, Florida, and tracked an additional two fish that emigrated from the Suwannee River. Fish were found at depths of 5–10 m. Harris et al. (2005) did an extensive analysis of potential prey distributions and Gulf sturgeon locations in the Suwannee River estuary. They found that fish were associated primarily with sandy areas having high potential prey abundance (especially brachiopods, amphipods, and brittle stars). Rogillio et al. (2007) mentioned marine habitat use of Gulf sturgeon natal to the Pearl River drainage, but did not include specific information on locations or habitats. Two studies have thus far concentrated on marine habitat use of Gulf sturgeon in the easternmost portion of its range. Edwards et al. (2003) tracked Gulf sturgeon offshore of the Suwannee River estuary in the northeastern Gulf of Mexico using sonic telemetry. They located seven fish in 1996 and 18 fish in 1998–1999 in the Suwannee Sound region at generally shallow depths (0.5–4.5 m). Edwards et al. (2007), studying marine habitats use of Gulf sturgeon native to various Florida rivers, also documented their use of shallow depths (2–6 m).
There are no published studies of marine habitat use of Gulf sturgeon that focus primarily on the north-central Gulf of Mexico—an area important to fish natal to streams in the Mobile Basin and to the Pearl and Pascagoula rivers. A detailed understanding of marine habitat use is important for evaluating potential environmental impacts as part of the critical habitat designation of Gulf sturgeon (USFWS and National Marine Fisheries Service 2003). In fact, data from our preliminary reports (e.g., Heise et al. 1999; Ross et al. 2001a, b) contributed extensively to the original critical habitat designations off the Alabama, Mississippi, and Louisiana coasts.
Because of the lack of information on marine habitats used by Gulf sturgeon natal to the Pearl and Pascagoula rivers, our first objective was to determine offshore locations of these fish. Objectives 2–4 were to test the hypothesis that Gulf sturgeon were randomly distributed within the marine environment, to describe the marine habitats where Gulf sturgeon were located in terms of physical and biotic variables and then test the hypothesis that these parameters remained the same among winters, and finally to characterize timing of movement into and out of salt water. We present data from two independent concurrent studies of Gulf sturgeon natal to the Pascagoula River of Mississippi and the Pearl River of Louisiana and Mississippi, conducted by personnel at the University of Southern Mississippi and the Mississippi Museum of Natural Science (MS) and the Louisiana Department of Wildlife and Fisheries (LWF). For the MS study, we address all four objectives for Gulf sturgeon tagged in the Pascagoula and Pearl rivers and relocated by MS personnel; for the LWF study, we address objectives 1 and 4 for Gulf sturgeon tagged in the Pearl River and relocated in Mississippi Sound and the associated barrier islands by LWF personnel.
Methods and Materials
Tagging—MS, Pascagoula River
After capture, Gulf sturgeon were measured to the nearest cm in fork length (FL), total length (TL), and modified standard length (SL; measured to the posterior edge of the last lateral scute; Bailey and Cross 1954), placed in a mesh sling and weighed to the nearest kg. Fish were fitted with sonic and radio tags (sonic 90 × 18 mm, pulse coded, 48 month, 42 g air, 20 g water, 69–76 KHz, Model CT-82-3AA, Sonotronics, Inc., Tuscon, AZ, USA; radio 110 × 18 mm, 40 month, 55 g air, 26 g water, 48–49 MHz, 60 cm trailing antenna, Model 2-5955, Advanced Telemetry Systems Inc., Isanti, MN, USA). Radio tags were used in a companion study of riverine habitat (Heise et al. 2004, 2005).
Tags were attached by drilling two holes in two adjacent dorsal scutes (i.e., fourth–fifth or fifth–sixth). Plastic-coated stainless steel attachment cables that were built into the radio transmitter were then threaded through holes in the scutes and into the holes on the sonic transmitter and secured by aluminum cable clamps. In addition, each sturgeon was given an AVID passive integrated transponder (PIT) tag (Norco, CA, USA), inserted into the muscle near the base of the dorsal fin, and an external dart tag that was placed near the base of the dorsal fin (1999) or pectoral fin (2000–2003). A 204-cm FL Gulf sturgeon captured on 17 September 2003 was also fitted with a satellite pop-up archival transmitting (PAT) tag (75 g in air, 175 × 21 mm plus antenna, Wildlife Computers, Inc., Redmond, WA, USA) attached by a heavy monofilament lanyard inserted through a hole drilled in a middorsal scute. The PAT tag is designed to detach from the fish at an assigned date and transmit data to an ARGOS satellite.
Fish were handled following guidelines developed by the United States Fish and Wildlife Service (USFWS 1992) and, other than weighing, all processing was done while the fish were tethered by a line around the caudal peduncle in the river, at a depth sufficient for normal gill ventilation. The entire procedure required about 10–15 min per fish; to minimize stress, all areas where the skin was penetrated were treated with the antibiotic Stress Coat ® (Aquarium Pharmaceuticals, Chalfont, PA, USA) and fish were released near the point of capture as soon as possible after processing. We did not observe any immediate mortality due to capture, tagging, and handling, over the course of this study; however, five postcapture mortalities (1–2 days later) occurred during the first 2 years of the project (1999—3; 2000—2). By stopping all netting once water temperatures reached 28°C, we eliminated all postcapture mortality in subsequent field seasons. Over a 5-year period, we attached a total of 109 sonic transmitters (not counting tags which were known to have failed or to have been shed) to juvenile (n = 1), subadult (n = 3), and adult Gulf sturgeon from the Pascagoula River (1999—14, 2000—22, 2001—31, 2002—20, 2003—22). Fish averaged 151-cm FL (range = 72–204) and 37.1 kg (range = 3.5–87.1).
Tagging—LWF, Pearl River
Gulf sturgeon were tagged, with one exception (Tickfaw River), in the Bogue Chitto, a major tributary of the lower Pearl River, near Pages Lake (30°36′54″ N, 89°50′53″ W; Fig. 1). Fish were captured using a multifilament nylon gill net (2.4 m deep × 45.7 m long, with 203 mm stretch mesh) and fitted with a coded sonic tag (67 × 18 mm, 10 g in water, Sonotronics CT-82-3, frequency 70–75 KHz) and a radio tag (ATS series 2090, frequency 49–50 MHz, 78 × 17 mm, 32 g air). These external tags were placed just below the dorsal fin where two holes were made using a sharp needle or cordless drill. Tags were attached with nylon covered stainless steel or Monel wire. AVID PIT tags were inserted into the muscle near the posterior base of the dorsal fin; Floy FT-4 lock-on tags were also attached near the posterior dorsal fin base. Other aspects of handling were similar to those described for MS. The date, location captured, latitude, longitude, tag numbers, TL (for 2002–2003), FL, and weight were recorded for each fish. Over a 4-year period, a total of 85 Gulf sturgeon, not counting replacement tags, were fitted with sonic tags (2000—25, 2001—24, 2002—18, 2003—18). Fish averaged 160-cm FL (range = 104–213) and 43.2 kg (range = 9.1–147.4).
Location, Movement, and Habitat Characterization—MS Personnel
We attempted to search for sonic-tagged Gulf sturgeon on an approximate weekly basis (contingent on weather), beginning with the movement of fish into coastal waters (usually October; Heise et al. 2005) and extending past the time that Gulf sturgeon return to freshwater (usually April; Heise et al. 2004, 2005). Weather and sea conditions often precluded listening for sonic tags at offshore sites south of the barrier islands. For instance, in 2004, all offshore trips were either cancelled or aborted at sea due to unsafe boating conditions.
Tracking dates (86 unique days; 231 region-days) by MS personnel for sonic-tagged Gulf Sturgeon in the north-central Gulf of Mexico, 1999 to 2004
1/12/00, 2/3/00, 2/4/00, 2/15/02, 12/8/03
10/28/99,11/10/99,11/18/99,11/23/99, 12/8/99,1/21/00, 3/23/00, 4/5/00, 6/1/00, 3/29/00, 7/5/00, 7/6/00, 7/7/00, 2/21/01, 3/7/01, 3/13/02, 3/14/02, 3/15/02, 10/18/2002, 11/25/03, 12/08/03, 12/15/03, 1/28/04, 3/15/04, 3/17/04
9/15/99,10/28/99,11/10/99, 12/8/99, 1/21/00, 2/9/00, 3/22/00, 3/29/00, 4/5/00, 11/9/00, 3/7/01, 5/4/01, 1/29/02, 3/15/02, 3/29/02, 5/29/02, 10/18/02, 11/8/02, 3/11/03, 4/11/03, 10/16/03, 10/31/03, 11/14/03, 11/21/03, 11/25/03, 12/12/03, 3/16/04, 3/17/04, 3/19/04, 4/16/04
1/12/00, 2/3/00, 2/4/00, 1/12/01, 1/10/02, 2/15/02, 12/8/03
9/15/99,10/28/99,11/3/99,11/10/99,11/18/99,11/23/99,12/8/99,1/12/00,1/21/00, 2/24/00,3/23/00,6/1/00,7/5/00,7/6/00, 7/7/00, 1/12/01, 2/21/01, 5/4/01,12/18/01, 1/10/02, 3/13/02, 3/15/02, 7/18/02, 10/18/02,11/8/02, 11/25/03, 12/08/03, 12/12/03, 12/15/03, 1/12/04, 3/15/04, 4/02/04
9/15/99,10/28/99,11/3/99,11/10/99, 12/8/99,1/21/00,2/9/00,3/22/00, 11/9/00,2/7/01,3/7/01, 6/21/01, 12/5/01, 1/29/02,3/15/02, 3/29/02, 5/29/02, 7/18/02, 10/18/02, 2/28/03, 3/11/03, 3/21/03, 4/11/03, 11/21/03, 11/25/03, 1/12/04, 3/16/04, 3/19/04
1/12/00, 2/3/00, 7/7/00, 12/6/00, 6/20/01, 12/18/01, 1/10/02, 2/15/02, 11/20/2002, 1/8/2003, 1/15/2003, 12/08/03, 12/15/03, 3/13/04, 3/18/04, 4/02/04
11/3/99,11/18/99,1/12/00, 2/3/00, 2/24/00, 3/23/00, 6/1/00, 7/5/00, 7/6/00, 7/700, 11/29/00, 12/6/00, 1/26/01, 2/21/01, 5/4/01, 6/20/01, 8/10/01, 11/18/01, 11/23/01, 12/18/01, 1/10/02, 1/29/02, 2/15/02, 2/21/02, 3/13/02, 3/14/02, 7/18/02, 10/18/02, 1/6/03, 1/7/03, 1/8/03, 1/15/03, 2/12/03, 3/21/03, 11/25/03, 12/08/03, 12/12/03, 12/15/03, 1/12/04, 2/9/04, 3/13/04, 3/17/04, 3/18/04, 4/02/04, 4/28/04
10/28/99,11/10/99, 2/8/00, 2/24/00, 3/22/00, 7/6/00, 11/29/00, 12/20/00, 1/26/01, 2/7/01, 3/7/01, 5/4/01, 6/21/01, 12/5/01, 1/29/02, 3/14/02, 3/15/02, 3/29/02, 5/29/02, 7/18/02, 11/8/02, 12/11/02, 1/6/03, 1/8/03, 2/19/03, 2/28/03, 3/21/03, 11/25/03, 12/12/03, 1/12/04, 1/21/04, 2/9/04, 3/16/04, 3/19/04, 4/28/04
06/20/01, 8/10/01, 11/18/01, 2/21/02, 1/31/2003, 2/12/03
Using a directional hydrophone, proximity to a sonic-tagged fish can be determined to approximately 5–10 m using changes in tonal characteristics of the signal (Heise et al. 1999). Gulf sturgeon locations were noted by GPS position and water depth (determined with boat mounted and hand-held fathometers), and we characterized bottom water conditions at each site by temperature (°C), dissolved oxygen (ppm), and salinity (ppt) using a YSI meter. We measured turbidity (cm) with a Secchi disk and took five bottom samples with a petite Ponar dredge at each Gulf sturgeon location. From the five samples, we visually scored the dominant and subdominant substrata as 1—clay, mud; 2—fine sand; 3—medium to coarse sand; and 4—shell fragments. We used one-way analysis of variance, or a Kruskal–Wallis test if variances were heteroscedastic, to test for differences in physical habitat characteristics among winters (α = 0.05) using BIOMstat 3.3 (Sokal and Rohlf 1995). Because of the generally low number of macrobenthic organisms in individual grab samples, the five benthic samples were pooled for analysis of the prey base. In the laboratory, all samples were sorted for macrobenthic organisms, and all organisms were identified to the lowest taxonomic level possible and enumerated. We analyzed macrobenthic data using PRIMER software (version 5.2.9; Clarke and Gorley 2001). Changes in macrobenthos among winter seasons were analyzed using both hierarchical agglomerative cluster analysis and nonmetric multidimensional scaling (MDS), because the two-dimensional MDS plot showed only moderately good representation of structure (i.e., stress [equivalent to goodness of fit] between 0.1 and 0.2; Clarke and Warwick 2001). Both cluster analysis and MDS were based on a Bray–Curtis dissimilarity measure with a square-root transformation to reduce the influence of highly abundant species. Because the volume of material obtained by the grab samples varied depending upon bottom conditions, we also standardized data by sample totals. To determine the significance of seasonal differences, we used one-way analysis of similarity (ANOSIM) with 999 iterations (Clarke and Warwick 2001). The analysis was limited to macrobenthos comprising ≥1% relative abundance within any of the five winter seasons. All mapping was done in ArcView GIS (ESRI version 3.3).
We evaluated the spatial distribution of Gulf sturgeon in the marine environment using the program FishTel 1.4 (Rogers and White 2007). To test the hypothesis that Gulf sturgeon were randomly distributed within the search area, we first constructed a polygon around all search points, with the northern boundary being the coastline, then used the Random function of FishTel to compare the actual fish locations to an equal number of randomly generated locations. The resultant variance, based on 10,000 iterations, was then compared to the variance generated by another 10,000 iterations of randomly distributed points (n = number of observed fish locations) to determine the P value (Rogers and White 2007). Because some fish were located more than once, we repeated the analysis after randomly removing multiple locations based on the same fish.
Movement and Habitat Use—LWF Personnel
Louisiana Department of Wildlife and Fisheries tracking effort by year and region for sonic-tagged Gulf sturgeon in the north-central Gulf of Mexico, 2000 to 2003 (90 unique days, 94 region-days)
Mississippi Sound and Barrier Islands
East Pearl River/Bay St. Louis
Mississippi River Gulf Outlet/Breton Sound
Relative to the four north–south telemetry regions (cf., Fig. 1) and based only on locations and search stations recorded by MS personnel (63 Gulf sturgeon locations; 1,375 telemetry stations), the percent of telemetry locations of Gulf sturgeon in the inshore zone was equivalent to sampling effort in the region; no fish were located in the midsound nor in the offshore area, and more fish occurred in the barrier island region relative to sampling effort (inshore tracking 14%, locations 13%; midsound tracking 14%, locations 0%; barrier island tracking 60%, locations 87%; offshore tracking 12%, locations 0%). Of the fish located in the barrier island region, 93% were found in the passes between the islands, including the two small passes cut through Ship Island by hurricanes. Although most locations occurred in the passes, we did locate one fish on three successive days as it moved along the south side of Horn Island. Over the 3-day observation period (approximately 48 h), this fish moved at least 15.8 km to the east (straight-line distance), with daily movements of 6.6, 7.8, and 1.4 km.
For fish tagged in the Pascagoula River, the number of relocations each year was relatively constant, ranging from three to seven, even though the number of sonic-tagged fish varied from 14 to 31. As a consequence, there was a negative correlation between the number of fish tagged each year and the number located at least once in the same year (n = 5, r = −0.88, P < 0.05). Thus, even though the fish we located were grouped along the barrier islands, we have no information on the distribution of fish that we did not locate. An exception to this was fish 70 2.9.3 that was tagged on June 14, 2000 in the Pascagoula River, located off Horn Island on November 29, 2000 and next captured in the Yellow River, Florida (30°34′24″ N, 86°54′41″ W) on June 12, 2001. Clearly, extensive movements, likely via the nearshore marine environment, are possible.
Habitat characteristics in the vicinity of Gulf sturgeon locations
Mean (95% CI)
Bottom temperature (°C)
Dissolved oxygen (ppm)
Secchi depth (cm)a
Dominant substratum (coded)
Subdominant substratum (coded)
Percent relative abundance and percent occurrence of macrobenthic organisms taken from Ponar grabs at Gulf sturgeon locations during winters 2000–2001, 2001–2002, 2002–2003, and 2003–2004 (n = 27). Only taxa comprising ≥1% relative abundance are shown
Average percent relative abundance
Cumulative relative abundance
Acanthohaustorius cf. intermedius
Unidentified bivalve A
Nephtys cf. bucera
Tellina sp. B
Combined MS and LWF
Based on the shortest, overwater distance from the river mouth to the point of the marine locations, Gulf sturgeon leaving the mouth of the Pascagoula River traveled an average of 22 km to the winter feeding areas in Mississippi Sound and along the barrier islands (n = 17 unique fish; range = 15–36 km; multiple locations of individual fish were combined as an average). In contrast, fish coming from the Pearl River traveled an average of 80 km (n = 22 unique fish; range = 46–122 km).
The life cycle of the Gulf sturgeon in the north-central Gulf of Mexico integrates riverine, estuarine, and coastal habitats—spawning in specific freshwater sites (Heise et al. 2004), moving to summer holding areas still within freshwater (Heise et al. 2005), moving offshore, primarily to barrier island passes, to feed (this study; Rogillio et al. 2007), and with juvenile fish likely overwintering and feeding in estuaries prior to initiating the migratory cycle (Sulak and Clugston 1999). Gulf sturgeon occupy coastal waters of Mississippi beginning in October or November. This is corroborated by the timing of downriver migration in both the Pascagoula and Pearl rivers, which generally occurs from late September to November (Heise et al. 2005; Rogillio et al. 2007). In the Pascagoula River, seaward migration is associated with shorter day length, falling water temperature, and heightened river flow (Heise et al. 2005). Gulf sturgeon leave the marine environment in February and March, and we had no telemetry locations in coastal waters from April through September. The marine data complement our studies within the Pascagoula River which indicate that Gulf sturgeon move into freshwater from late February and March through April, generally reaching spawning habitats 250 river km upriver by mid-April (Ross et al. 2003; Heise et al. 2004). In Choctawhatchee Bay, Florida, Gulf sturgeon typically enter the marine environment during September and October and return to freshwater in April and May (Fox et al. 2000; Parauka et al. 2001). Ripe fish tend to enter freshwater somewhat earlier than nonbreeding fish, usually from March to mid-April (Fox et al. 2000). Gulf sturgeon natal to the Suwannee River show a similar timing for movement into salt water (Edwards et al. 2003).
Gulf sturgeon tagged in the Pascagoula and Pearl rivers occupy the same marine feeding habitats. The habitat association of Pascagoula and Pearl river Gulf sturgeon can also continue into freshwater. Based on an analysis of short tandem repeat length variation (i.e., nuclear microsatellite DNA), Gulf sturgeon occupying the Pearl River during the freshwater phase of their life cycle are almost exclusively natal to the Pearl River. In contrast, Gulf sturgeon captured from the Pascagoula River not only are predominately natal to the Pascagoula River but also include a substantial number of nonripe fish that are natal to the Pearl River (Dugo et al. 2004). This unequal sharing of riverine habitat may be due to the closer proximity of the Pascagoula River to the winter habitat compared to the Pearl River.
Once they left freshwater, the predominant habitats we identified for Gulf sturgeon were the relatively shallow waters of the barrier island passes. These areas, characterized by strong tidal currents and clean sand substrata, were consistently used over each of the five winters of our study. Gulf sturgeon were never located in the middle regions of Mississippi Sound, suggesting that they moved quickly from the river mouths to the barrier islands. Working in Choctawhatchee Bay, Florida, Parauka et al. (2001) and Fox et al. (2002) also documented the association of adult and subadult Gulf sturgeon with shallow water and clean sand substrata. Fox et al. (2002) suggested that fish moved quickly over deeper, less favorable foraging areas to the preferred shallow sandy sites. They found Gulf sturgeon most commonly at 2–4 m depths with substrata having 80% or more of sand. In both late fall and early spring, Gulf sturgeon in the Suwannee River estuary were likewise associated with substrata having high percentages of sand (Harris et al. 2005). In a study most similar to ours, Edwards et al. (2007) studied marine habitat use of Gulf sturgeon natal to northwestern Florida rivers. Using both sonic and satellite pop-up archival transmitting tags, they found that Gulf sturgeon occurred in shallow water (2–6 m) at distances of 2–12 km offshore, even though coastal movement in both directions was common and ranged from 30 to 180 km.
Edwards et al. (2003) found that telemetered Gulf sturgeon in the marine environment off the Suwannee River moved relatively slowly, with speeds of 1 km h−1 or less, and Parkyn et al. (2007) recorded even slower speeds of 0.8 to 2.2 km day−1 for adult fish in the Suwannee River estuary. These rates are similar to our observation of a speed of approximately 0.2 km h−1 for a single fish as it moved eastward off Horn Island. Our estimate is likely low, given that we were only recording straight-line distances taken over a 3-day period.
Although our telemetry stations in the Gulf of Mexico south of the Alabama, Mississippi, and Louisiana barrier islands were limited because of inclement weather, we never located Gulf sturgeon in this offshore region. Subsequent work over three winters by LWF in the vicinity of the Chandeleur Islands also failed to locate a single Gulf sturgeon (Rogillio and Ruth 2003, 2005). The recovered data from the single PAT tag also indicated a location within a barrier island pass and not further offshore. Sulak and Clugston (1999) proposed two depth options for Gulf sturgeon movement in the marine environment—either they move along the nearshore area at depths of 10 m or less or, alternatively, move far offshore to depths of 40–100 m. Thus far, the results of our study and work by Rogillio and Ruth (2003, 2005), Sulak and Clugston (1999), Fox et al. (2000, 2002), and Edwards et al. (2007) support the first option. However, because of the relatively low percent of fish that we located after tagging, we cannot rule out movement of Gulf sturgeon onto deeper continental shelf areas. A point also emphasized by Edwards et al. (2007). As suggested also by Fox et al. (2002), female Gulf sturgeon, which may spawn only every 2–5 years (Smith 1985; Fox et al. 2000) may make more extensive offshore movements.
Atlantic sturgeon show a similar movement pattern, emigrating to salt water from September through November and generally returning to freshwater from February through July, with later arrival dates corresponding to higher latitudes (Smith 1985). The winter distribution of Atlantic sturgeon also includes estuarine and offshore habitats. Collins and Smith (1997), in a survey of long-term capture records for South Carolina waters, documented winter occurrences of Atlantic sturgeon in generally shallow (<40 m) areas of the Atlantic Ocean from November into April. Similar nearshore shallow water marine habitat use was indicated by Laney et al. (2007) for primarily juvenile fish captured off North Carolina. Savoy (2007) captured subadult and adult (mean = 105-cm FL) Atlantic sturgeon in Long Island Sound during October and (Stein et al. 2004), using fisheries by-catch data, documented movement onto the shallow areas of the continental shelf off the northeastern USA, especially at depths of 10–50 m. Fish were usually associated with gravel or sand substrata.
Green sturgeon, the third taxon of North American, oceanic anadromous sturgeon, typically spend more time at sea than the Atlantic or Gulf sturgeon, with adults moving into freshwater in the spring for spawning and back into salt water from October through December (Moyle 2002; Erickson et al. 2002; Erickson and Hightower 2007; Erickson and Webb 2007). Marine locations tend to occur in a narrow nearshore bathymetry zone, but with broad latitudinal movements. At sea, green sturgeon typically occupy depths of 40–110 m, which is relatively shallow, albeit much deeper than Atlantic and Gulf sturgeon, given the nearshore bathymetry of the Pacific coast (Erickson and Hightower 2007). Based on incidental fishery captures, the European sturgeon (Acipenser sturio), also an oceanic anadromous species (Munro et al. 2007), is similar to the green sturgeon in spending more time at sea during the winter. Reported depths are generally <100 m with fish reported over a range of substrata (Rochard et al. 1997).
As suggested by Stein et al. (2004) and others, the marine distribution of Atlantic sturgeon is likely driven by prey availability, an observation that appears to apply as well to Gulf sturgeon (Edwards et al. 2003). However, Stein et al. (2004) proposed that Atlantic sturgeon would show little if any habitat selection instead being found wherever their prey might occur. In contrast, our results for Gulf sturgeon indicate a highly structured spatial distribution, although it too is likely caused by the distribution of preferred prey taxa.
Benthic samples taken at sites where we located Gulf sturgeon were dominated in terms of percent occurrence by Florida lancelets, sand dollars, annelids, haustoriid amphipods, and mollusks, all of which are documented prey of large subadult and adult Gulf sturgeon (Huff 1975; Mason and Clugston 1993). Densities of Florida lancelets in clean sand substrata can be quite high reaching 4,000–5,000 m−2 at depths of 4–6 m at Perdido Key, Florida (Rakocinski et al. 1993), densities over 1,000 m−2 at depths of 5 and 15 m along Santa Rosa Island, Florida (Rakocinksi et al. 1998), and densities of 1,200 individuals m−2 in Tampa Bay, Florida (Stokes 1996). In March 2002, a 62-kg Gulf sturgeon was found dead on Dauphin Island (death was perhaps due to injuries suffered during capture in a trawl) that had hundreds of large lancelets in its stomach (Richard Heard, personal communication). In addition, the overall benthic fauna is characteristic of the bottom type that support large populations of ghost shrimp (Lepedophthalmus, family Callianassidae), which are known to be an important prey of Gulf sturgeon (Carr et al. 1996; Fox et al. 2000), but which were not susceptible to our sampling gear (Fox et al. 2002; Richard Heard, personal observation). In Choctawhatchee Bay, Florida, annelids and crustaceans were the numerically dominant taxa in benthic samples taken at Gulf sturgeon telemetry sites (Fox et al. 2000). Atlantic sturgeon captured off the New Jersey coast also consumed primarily annelids as well as isopods and amphipods (Johnson et al. 1997). Similar prey were also documented for Atlantic sturgeon in Long Island Sound, except that in addition to polychaetes and amphipods, fish consumed numerous pea crabs (Pinnixia; Savoy 2007).
The inshore regions of Mississippi Sound, although not used extensively by the large adult Gulf sturgeon that are the subject of our research, are likely important nursery areas for younger fish. This is supported by our capture in February of four juvenile Gulf sturgeon (mean = 48 cm, range 38–63-cm FL; presumed ages, based on Huff’s (1975) age–length relationship, of 1.1–2.7 years) in the Pascagoula River estuary (Ross et al. 2003). Sulak and Clugston (1998) also found that juvenile Gulf sturgeon remained in nearshore waters throughout the winter in the Suwannee River estuary, Florida. Additional study of habitat use of juvenile Gulf sturgeon in Mississippi Sound and other nearshore environments is clearly needed.
Our work, the first multiyear study of habitat use of adult Gulf sturgeon in the nearshore marine environment of the north-central Gulf of Mexico, documents the use of barrier islands offshore of Mississippi Sound as important foraging grounds for this threatened species. Gulf sturgeon tagged in the Pearl River drainage of Mississippi and Louisiana and the Pascagoula River drainage of Mississippi and Alabama occupy the same winter foraging habitats. Recovery of the northern Gulf populations of Gulf sturgeon requires continued protection not only of the riverine habitats with access to historical spawning sites but also protection of coastal waters and, especially, barrier island passes.
Overall, the manuscript benefited greatly from the comments of two anonymous reviewers and from the editorial help of Y. Ross. For the MS portion of the study, we thank Mollie Cashner, Tanya Darden, Richard Darden, John Ewing, Billie Joe Johnson, Yvonne Ross, and Brian Kreiser for help with fieldwork and logistical support and Randy E. Edwards and Wendy Gierhart for providing the location data for the pop-up tag. We are grateful to Frank Parauka, USFWS, for sharing his knowledge of Gulf sturgeon, for providing tag information, and for supplying the PAT tag. We thank Jerry McLelland for the help with invertebrate identifications and Captain Kyle Jarreau aboard the vessel Jus Ad Water for providing access to offshore telemetry sites. Gulf sturgeon in the Pascagoula River were captured under annual administrative scientific collecting permits issued to STR and WTS by the Mississippi Museum of Natural Science, Mississippi Department of Wildlife, Fisheries and Parks; capture of Gulf sturgeon in the Pascagoula River estuary was done under annual scientific collecting permits issued to STR and WTS by the Mississippi Department of Marine Resources. This study was supported by the National Sea Grant College Program of the US Department of Commerce’s National Oceanic and Atmospheric Administration under NOAA Grant # NA16RG2258, the Mississippi–Alabama Sea Grant Consortium, the US Fish and Wildlife Service, the Mississippi Department of Wildlife, Fisheries, and the National Fish and Wildlife Foundation Shell Marine Habitat Program. Additional support was provided to S. T. Ross by the T. W. Bennett Jr., Distinguished Professorship in the Sciences at the University of Southern Mississippi. The views expressed herein do not necessarily reflect the views of any of these organizations. For the LWF portion of the study, we thank Elizabeth Rabalais and Tim Ruth as assistant Project leaders who coordinated sampling, data records, and scheduling of telemetry day trips. Fieldwork and logistical support by Jeff Thompson and Brian McNamara was commendable. Funding for portions of the projects was through the National Fish and Wildlife Foundation Shell Marine Habitat Program in 2000 and Section 6 funding though the Louisiana Wildlife and Fisheries Department Heritage Section from 2001 to 2003. Collecting of Gulf sturgeon was authorized by LWF.