PATTERNS OF NEKTON HABITAT USE ALONG A MARSH COENOCLINE*
Rodney Rountree and Francis Juanes, Dept. Natural Resources Conservation, UMASS-Amherst
A contributed paper presented at the New England Estuarine Research Society Spring Meeting, Portland, Maine, May 18-20, 2000.
A growing body of evidence suggest that communities of marsh nekton are stratified along a depth gradient partly in response to the influence of tidal and diel changes in physical conditions. In addition, ontogenetic shifts in habitat use along this gradient are key to our understanding of marsh function and the process of trophic relay, yet have rarely been specifically addressed. The goal of this study was to develop sampling methodologies and collect preliminary data to test the hypothesis that nekton faunal assemblages, species densities, and ontogenetic stages of Atlantic silversides and other species, are stratified along a marsh creek-to-bay coenocline (i.e., the marsh gradient). Preliminary results include: 1) successful development and use of a new seine sampling technique that provides density estimates of nekton in marsh creek and shallow bay environments, 2) we obtained sufficient preliminary data (69 seine samples) to compare the effectiveness of this new technique against a more conventional standardized seining method, 3) we obtained preliminary data on the diel and tidal changes in temperature along the marsh gradient, and 4) we obtained preliminary data illustrating strong diel changes in nekton density distributions and faunal assemblages along the marsh gradient. Amazingly, low tide fish densities reached values well over 500 fish sq. m.
*Title and abstract of actual talk modified from that submitted prior to the meeting.
[Click on thumbnails to view full size figures]
Slide 2. Marsh gradient - Introduce concept of a marsh gradient following a tidal elevation gradient. Environmental gradients result from interactions among atmospheric, terrestrial, and aquatic influences mediated by the depth gradient. Tidal and diel cycles interact in complex ways with these gradients resulting in gradient reversals, etc. Faunal assemblages associated with any habitat located along the gradient will be strongly influenced by their physiological response to the marsh gradient. Diel, tidal and ontogenetic movements are likely to be directly influenced by tidal and diel cycles in the marsh gradient.
Slide 3. A random stratified sampling design was used. A subtidal marsh creek was divided into three strata, Upper creek > 246 m above the creek mouth (located at the main fork in the creek channel), Lower creek (0-246 m), and Bay (0 to -125 m from the creek mouth).
Slide 4. Paired 5-m wide sections of the creek were to be simultaneously blocked off at low tide at random locations within each creek strata (this technique could not be used in the bay strata). Three "clotheslines" were set out at 5-m intervals along the creek at high tide prior to sampling. A block net was attached to each clothesline and stacked up on one creek side. At low tide the nets would be set by simultaneously pulling the nets with "pull" lines from the opposite back in a manner similar to pulling a shower curtain. Once the block nets were in place at low tide, a 5-m seine was used to make repeated seines inside each of the paired sections. Similar sampling was conducted just outside of the block nets for comparison (see next slides).
Slide 5. Top block net attached to a clothesline and ready for deployment. Bottom - setting the block nets at low tide by pulling them from the opposite bank. Technicians approached the area keeping well away from the bank to avoid disturbance. Once in position, they waited 2 minutes for any disturbance to die down before simultaneously pulling the block nets across. In generally took about 15 seconds to set the nets in this manner.
Slide 6. Top- three block-nets in place at low tide forming two adjacent 5-m sample sections. Each block net was lined with 1/4" gauge galvanized chain. Bottom - three repeated seines with a 5-m tunnel seine with attached live car were made within each 5-m creek section.
Slide 7. From 1 to 3 repeated seines were also made 5 meters above the block nets to allow a comparison of standardized seine methods with the block net methods. A 5-m restriction line was used on the seine to insure a 5-m wide sweep area.
Slide 8. A somewhat different approach was necessary in the bay strata. Six temporary poles were set in the bay at high tide to form a 5-m by 15-m rectangular sampling area. Two random locations were chosen to obtain replication. At low tide a single large block net was walked around the perimeter of the stakes to enclose the 75-m2 area. Top - Simultaneously setting two enclosures, Center - conducting standardized seines outside of the enclosures for comparison, Bottom - conducting 3 repeated seines within each enclosure.
Slide 9. See Figure 6 in the interim report.
Changes in water temperature over three days at different locations along the gradient. Note strong diel and tidal interactions in temperature cycles. Strong damping of the cycles moving down the gradient. Water temperatures exhibited diel cycles of 10 C in the upper creek, but only a couple degrees in the bay.
Slide 10. See Figure 7 in the interim report. Same as Slide 10.
Slide 12. A comparison of densities between experimental (block nets) and control seines. Most species, with the notable exception of Cyprinodon variegatus, were more abundant in the experimental nets.
Slide 13. Depletion of catch (mean densities) among repeated seines. Densities of most species were rapidly depleted by the third seine, with the exception of the Gobiidae, which showed little change.
Slide 14. Difference between depletion rates between experimental and control seines. Regression analysis indicates that catches would approach zero by the 4th seine in block nets, but not until the 7th or 8th seine in the control seines. Catches of Atlantic menhaden, Brevoortia tyrannus, are shown as an example in the bottom graph.
Slide 15. Comparison of catches between day and night samples. Total catch were twice as high, and Atlantic silversides were 10 times as high at night, while catches of many species traditionally considered marsh residents were typically greater during the day.
Slide 18. Canonical discrimination analysis of strata group means showing 95% C.I. of strata canonical means (ellipses). Igenvectors of species with original variates (densities) strongly correlated with canonical axis are shown (Species shown in red are significantly correlated with CAN2, blue with Can1, and green with both axes).
Slide 20. Adult weakfish were observed feeding in the creeks at the earliest stages of the flood tide, apparently taking advantage of hugh concentrations of prey with little danger of stranding or exposure to low dissolved oxygen (since they moved in with the flood tide). Weakfish were observed as far up as 400 m above the creek mouth.
Return to :| TOP |
This page was last modified on July 20, 2001
Copyright (c) 1999 by Rodney Rountree. All rights reserved