Distribution, population assessment and conservation of the endemic Bermuda killifishes Fundulus bermudae and Fundulus relictus

Please download to get full document.

View again

of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Document Description
Distribution, population assessment and conservation of the endemic Bermuda killifishes Fundulus bermudae and Fundulus relictus
Document Share
Document Tags
Document Transcript
  ENDANGERED SPECIES RESEARCHEndang Species ResVol. 3: 181–189, 2007Printed October 2007 P  ublished online August 2, 2007  INTRODUCTION Bermuda (32°18’N, 64°46’W) consists of over 120islands, with a total land area of 55 km 2 , 960 km fromthe USA coast. It lies on top of a volcanic seamount thatrises 4000 m from the seafloor, capped by limestoneand coral reefs. All islands are concentrated along thesouthern edge of the seamount, the larger ones form-ing a narrow chain linked by causeways and bridges.Bermuda has no permanent surface freshwaterstreams or lakes and fewer than 20 anchialine pondsscattered across the islands (Thomas & Logan 1992).These ponds (3 of which are man-made) are isolated,saline, land-locked bodies of water with permanentsubterranean connections to the ocean. Natural fresh-water ponds are few, and mostly man-made. Wide-spread drainage of freshwater marshes was employedas part of mosquito control methods in the first half ofthe 20th century, as the government attempted to pre-vent the spread of mosquito-borne diseases. In the1930s, wetlands were used for the disposal of garbage,and after the 1950s this activity became the main fea-ture of marsh reclamation. During this period, largetracts of marshland and numerous ponds were com-pletely filled, leading to major losses in biodiversity. Incontrast, the total acreage of saline ponds hasremained unchanged since the end of the 19th century(Sterrer & Wingate 1981) and may function as refugiafor endemic species. Thomas et al. (1991, 1992)reported detailed descriptions of the physical andbiotic characteristics of the 6 largest anchialine pondsin Bermuda. Most of the ponds in Bermuda (irrespec-tive of salinity) are small and shallow with low habitatcomplexity.To date, 433 species of fishes have been recorded inBermuda, of which 8 are currently recognized asendemic (Smith-Vaniz et al. 1999). Two of these 8endemics belong to the killifish genus Fundulus  : F.bermudae  Günther 1874 and F. relictus Able & Felley1988 . These are believed to be descendants of the © Inter-Research 2007 · www.int-res.com*Email: mouterbridge@gov.bm Distribution, population assessment and conservation of the endemic Bermuda killifishes Fundulus bermudae  and Fundulus relictus  Mark E. Outerbridge 1, *, John Davenport 2 , Anne F. Glasspool 1 1 Bermuda Biodiversity Project, Bermuda Zoological Society, PO Box FL 145, Flatts, FL BX, Bermuda 2 Department of Zoology, Ecology &Plant Science, University College Cork, Distillery Fields, North Mall, Cork, Ireland ABSTRACT: Fundulus bermudae and Fundulus relictus  are endemic to Bermuda and are protectedunder the Bermuda Protected Species Act 2003. These killifishes were described as abundant andwidespread in the wetland communities of Bermuda during the late 19th and early 20th centuries.Surveys were undertaken during 2004–2005 to determine the current distribution, as well as to esti-mate the size and structure of each Fundulus  population. Killifishes are now found in only 9 isolatedponds. For 6 ponds, populations appear to be large enough to be self-sustaining for the foreseeablefuture; for 1 pond, the population is low enough to be regarded as vulnerable. Estimates were notfeasible in the case of the remaining 2 ponds. KEY WORDS: Anchialine ponds ⋅ Bermuda ⋅ Fundulus bermudae   ⋅ Fundulus relictus   ⋅ Killifishes ⋅ Mark and recapture ⋅ Visible implant elastomer Resale or republication not permitted without written consent of the publisher   Endang Species Res 3: 181–189, 2007 F.heteroclitus–F. grandis species group, and to havesrcinated from populations on the east coast of theUSA (Able & Felley 1988) at least 5000 yr beforepresent (Smith-Vaniz et al. 1999, Grady et al. 2001).Uni- and multivariate analyses of adult morphologyand isozyme analysis of natural populations indicatethat both of the Bermuda forms are endemic anddistinct from F. heteroclitus (Able & Felley 1988).Theoverall shape of both species is typical for the F.heteroclitus–F. grandis  group. Both sexes have adeep vertical body profile, a rather short and robustbody with posteriorly placed dorsal and anal fins ofapproximately equal size, a deep caudal peduncle, anda somewhat rounded caudal fin (Fig. 1). Bermuda’sfundulids are sexually dimorphic: females are uniformolive to brown in colour, darker above, lighter below,without definite markings; males are darker in colour,usually dark green or olive with a yellowish underside,and when spawning have a dark spot towards the rearof the dorsal fin. F. bermudae  and F. relictus  are, how-ever, indistinguishable in the field, though it appearsthat no pond contains mixed populations (Table 1). Formore detailed physical descriptions see Able & Felley(1988). Fundulus bermudae  and F. relictus  are bentho-pelagic in the pond habitat, appear to form looseschools made up of equally sized individuals, and areomnivorous opportunistic feeders. Stomach contentanalysis has revealed that F. bermudae  inhabitingMangrove Lake, the largest pondin Bermuda, eats filamentousgreen algae and plant material,molluscs, crustaceans and insects(Rand 1981).The annual spawning cycle for Fundulus bermudae  is near-syn-chronous for both sexes. Thespawning season for males andfemales begins in February,reaches a peak in May and June,respectively, and falls abruptlyafter June until September; the endof the spawning season (Outer-bridge et al. 2007). Laboratoryobservations of spawning site pref-erences for F. bermudae  and F.relictus  suggest that both speciesmay deposit eggs in algal mats orsubmerged vegetation (Able &Hata 1984). Such solid surfacesmay include the roots and pneu-matophores of the red and blackmangrove tree Rhizophora mangle  and Avicennia nitida , as well aswidgeon grass Ruppia maritima ,allof which are very common in many of the Fundulus  ponds in Bermuda. The lifespan of Bermuda’s killi-fishes is unknown; however, individuals have beenkept in captivity at the Bermuda Aquarium, NaturalHistory Museum and Zoo for 5 yr (J. Gray pers. comm.)In 2003, Fundulus bermudae  and F. relictus  werelisted as protected under the Bermuda Protected Spe-cies Act (www.bermudalaws.bm). Additionally, thereare plans to submit an application for both species tobe included on the IUCN (The World ConservationUnion) Red List of Threatened Species. There havebeen few opportunities for these fishes to increasetheir range, largely due to the restriction in habitatavailability and the fragmentation of the wetlands inBermuda. Additionally, quantitative assessments ofeach pond population have been lacking and are limit-ing conservation efforts. The aims of the present study were (1) to determinethe current distribution of Bermuda’s killifishes, (2) toestimate the size of each extant population, and (3) todescribe the basic structure of these populations. MATERIALS AND METHODS A field list of 44 potential killifish habitats wasdrafted, including salt marshes, peat marshes,drainage canals, inland freshwater, brackish and fullymarine ponds, mangrove communities, and sheltered, 182Fig. 1. Fundulus bermudae  . (a) Mature female from Trott’s Pond and (b) male from Mangrove Lake. Photos: Mark Outerbridge (a) and Jennifer Gray (b)  Outerbridge et al.: Endemic Bermuda killifishes shallow bays dominated by seagrass beds in areasimmediately adjacent to coastal mangroves (Outer-bridge et al. 2006). Surveying used a combination ofdirect observation and baited traps.Minnow traps made of 8 mm diagonal wire meshwere used. Each measured 425 mm long and 228 mmwide with a 165 mm long conical entrance at both endsthat tapered to a 28 mm diameter opening throughwhich fish could enter the trap. Each trap was baitedwith a small amount of canned sardines, securely fas-tened, deployed from a canoe and left to soak for 1 h ata random location within each habitat. Such traps havebeen standard materials in numerous investigations of Fundulus  species in North America (Kneib & Stiven1978, Meredith & Lotrich 1979, Sweeney et al. 1998,Kneib & Craig 2001). Kneib & Craig (2001) evaluatedthe efficacy of these traps for measuring relative abun-dance and density of killifishes and suggested thatreliable estimates can be made if the traps are used inisolated pools where habitat complexity is low and kil-lifish density remains constant during the samplingperiod—conditions satisfied in the present study.Wherever extant populations were found, up to 12traps were placed at random locations throughout eachpond (12 in Mangrove Lake, Trott’s Pond, Bartram’sPond and the West Walsingham Ponds; 7 in WarwickPond; 4 in Lover’s Lake; and 3 in the Blue Hole BirdPond). Population censuses were subsequently per-formed based on the Petersen Index methodology ofmark and recapture (Ricker 1958) using a visibleimplant elastomer (VIE) (Northwest Marine Technol-ogy). VIE effectiveness has been documented byDewey & Zigler (1996), Frederick (1997) and Kneib &Craig (2001). The pale area on the abdomen adjacentto the proximal base of the pelvic fins was chosenasthe tagging site. Application of elastomer to fishwasaccomplished by full anaesthesia (FINQUEL MS-222 Argent Chemical Laboratories) andinjectionimmediately below the dermis so that theneedle tipcould still be clearly seen. Immediately after tagging,fish were transferred to a bucket of continuously aer-ated, untreated pond water and given 2 h to recoverbefore being released at random locations within eachpond.Sex and total length, measured to the nearest 1 mm,of all killifishes captured were determined at the timeof tagging. Whenever possible, sex was recordedbased on presence or absence of a dorsal fin ocellusand an anal-sheath on the anal fin, as well as overallbody colouration. Males were positively identifiedwhen they possessed the dark ocellus on the posteriorhalf of the dorsal fin; females were positively identifiedif they possessed an anal sheath on the anal fin. Bodycolouration was used as a secondary identificationtechnique.Second, third and fourth recapture censuses wereperformed on each extant population based on thesrcinal marking event 1, 2 and 3 wk, respectively,after the initial tagging event. All trapped fish wereexamined for tags and the numbers recorded. Petersenpopulation estimates—calculated as N  = [( n 1 + 1)( n 2 +1) / ( m 2 + 1)] – 1, where N  = the total estimatedpopulation, n 1 = number of fish marked in first sample, n 2 = total number of fish in the second sample, m 2 = 183Pond Fundulus  Date ofSourceEstimated numberStatus inspeciesintroductionpopulationof fish translocated2004Lover’s Lake F. relictus  nananaExtantWest Walsingham Ponds F. bermudae  nananaExtantEast Walsingham Ponds F. bermudae  nananaExtantTrott’s Pond F. bermudae  nananaExtantMangrove Lake F. bermudae  nananaExtantWarwick Pond F. bermudae  nananaExtantEvan’s Pond F. bermudae  nananaExtantNonsuch Island F. bermudae  1976Trott’s PondUnknownExtirpated(saltwater pond) a Bartram’s Pond a F. relictus  1986Lover’s LakeUnknownExtantNonsuch Island F. bermudae  1993Mangrove Lake53 b Presumed(freshwater pond) a ExtirpatedBlue Hole Bird Pond a F. bermudae  1995West Walsingham50 c Extant a Man-made pond b Jon Cotter (1993) Acclimation of brackish killifish F. bermudae  to fresh water for introduction into Nonsuch Island, Bermuda(unpubl. paper in the Bermuda Aquarium, Natural HistoryMuseum and Zoo library) c J. Madeiros (pers. comm.)Table 1. Fundulus bermudae  and F. relictus  . Summary of killifish distribution and transfer history among the ponds of Bermuda. na: not applicable  Endang Species Res 3: 181–189, 2007 number of marked fish in the second sample—wereaveraged from the 3 separate recapture events toobtain a mean population size for each pond. RESULTS The 2004–2005 surveys confirmed the existence ofpopulations in the following 9 locations only: Lover’sLake, Bartram’s Pond, Mangrove Lake, Trott’s Pond,Blue Hole Pond, both East and West WalsinghamPonds, Warwick Pond and Evan’s Pond (Fig. 2). Noother Fundulus  populations were discovered in any ofthe peripheral wetland communities of Bermuda,including those described as having them at some timein the past 150 yr.Petersen estimates (Table 2), mean lengths, sizeranges (Table 3), and sex ratio data (Table 4) werecollected for all ponds where extant Fundulus  popu-lations were found, except for the East WalsinghamPonds and Evan’s Pond; no fish could be trapped formarking in these 2 ponds during the survey period.The population estimates (in order of decreasing size)were Mangrove Lake (11325; F. bermudae  ), Trott’sPond (7926; F. bermudae  ), Lover’s Lake (8508; F.relictus  ), Blue Hole Bird Pond (5394; F. bermudae  ),West Walsingham Ponds (2202; F. bermudae  ), Bar-tram’s Pond (1793; F. relictus  ) and Warwick Pond(436; F. bermudae  ). Non-parametric Mood Mediantests performed on all data sets were used to deter-mine if a significant difference occurred betweenmale and female median sizes (Table 5). Statisticallysignificant differences were found only in Lover’sLake (p = 0.006), theWest Walsingham Ponds (p =0.023), and Warwick Pond (p = 0.009). In all 3 popula-tions females were significantly larger than males.Additionally, females outnumbered males in all pondssurveyed, except Trott’s Pond, where the sexesoccurred in equal numbers. However, chi-squaretests performed on the sex ratio data sets for Man-grove Lake, Lover’s Lake, Trott’s Pond, and WarwickPond did not show statistically significant differencesbetween female and male numbers (p > 0.05 in allcases), whereas significant differences were noted inthe Blue Hole Bird Pond, Bartram’s Pond, and theWest Walsingham Ponds populations (p < 0.001 in allcases) (Table 5). DISCUSSIONDistribution The present Bermudiandistributionof killifishes is very restricted. In the19th century, killifishes were abundantand widely distributed throughout themarshes and ponds of Bermuda (Hurdis1897, Smith-Vaniz et al. 1999). Beebe &Tee-Van (1933) described killifishes(referred to as the ‘mangrove minnow’or ‘mangrove mullet’) as ‘abundant inbrackish pools and ponds.’ Historicalcollections also indicate that killifisheswere once extant in Stocks Point on St.David’s Island, in Pembroke Marsh,and in what was once an area known asPaget East swamp.During the 1990s local conservation-ists were aware of extant populationsonly in Lover’s Lake, Mangrove Lake,Trott’s Pond, both East and West Wals-ingham Ponds, Warwick Pond and 184PondPetersen estimate F. bermudae  Blue Hole Bird Pond5394 ±480Mangrove Lake11325 ±1884West Walsingham Ponds2202 ±178Trott’s Pond7926 ±1576Warwick Pond436 ±13 F. relictus  Bartram’s Pond1793 ±224Lover’s Lake8508 ±1347Table 2. Fundulus bermudae  and F. relictus  . Petersenestimates of population (mean ±SE, ind. pond –1 ) for killifish ineach of the ponds studied.Fig. 2. Fundulus bermudae  and F. relictus  . Distribution of killifishes in pondsacross the islands of Bermuda in 2004 and 2005  Outerbridge et al.: Endemic Bermuda killifishes Evan’s Pond. There were 4 other populations living inman-made ponds (created by dredging), where killi-fishes had been intentionally introduced as a precau-tion against possible extinction events. Fundulus relic-tus  from Lover’s Lake had been introduced in 1986 toBartram’s Pond on the Stoke’s Point nature reserve,and in 1995 F. bermudae  from one of the West Wals-ingham Ponds had been introduced to Blue Hole BirdPond in the Blue Hole Park. In 1976, specimens of F.bermudae  from Trott’s Pond had been introduced to anartificially created saltwater pond on the NonsuchIsland nature reserve, while specimens of the samespecies from Mangrove Lake were later introduced,after adaptation, to an artificial freshwater pond on thesame island in 1993 (Table 1) (D. B. Wingate pers.comm.).By 2001, it was suggested that 2 of the natu-rally occurring Fundulus  populations extant in the1990s might have died out (Grady et al. 2001), and in2003 at least one population on Nonsuch Island haddisappeared as a result of total destruction of a salt-water pond during Hurricane Fabian (J. Madeirospers. comm.)The present survey results reconfirmed the presenceof all the naturally occurring populations of Fundulus  known to be extant in the 1990s, as well as 2 of the 4translocated populations in the man-made ponds. Killi-fishes were not, however, found on the Nonsuch Islandnature reserve, despite a recent (1993) introduction inone of the ponds. These results indicate that Bermuda’s Fundulus  species may have completely disappearedfrom the coastal mangrove and the inland marsh com-munities that may have provided a degree of continu-ity between isolated killifish pond populations. Theextant populations were found living mostly in pondsof the eastern parishes of the islands, with 2 excep-tions. Most of these ponds are small, isolated andanchialine in nature. Human modification of historicalkillifish habitats is clearly the single greatest reasonwhy distribution is currently limited.Site visits, direct observation and trapping proved tobe effective in identifying extant populations; how-ever, it remains possible that small killifish populationswere overlooked in the present study. Other Fundulus  species are well known for utilizing shallow water sed-iments for refuge (Minckley & Klaassen 1969). It is fea-sible that a very small population dispersed around afairly large area, and living almost exclusively in thebottom sediment would create the illusion that a pondwas uninhabited. This may have been the case inWarwick Pond, thereby leading conservationists tobelieve (erroneously) that the srcinal population haddisappeared by 2001. Population and individual sizes Prior to this investigation little was known of thehealth and status of Bermuda’s Fundulus  populations.Knowledge of basic population estimates and theirstructure is necessary for conservationists to make 185PondNMeanMedianRange F. bermudae  Blue Hole Bird Pond70455.054.034–97Mangrove Lake66471.967.052–126West Walsingham51348.348.027–72PondsTrott’s Pond50061.556.036–100Warwick Pond15077.179.541–129 F. relictus  Bartram’s Pond50053.853.038–92Lover’s Lake50063.063.041–97Table 3. Fundulus bermudae  and F. relictus  . Mean, medianand range of total length (mm) of killifish in each of the pondsstudiedPondNF:M F. bermudae  Blue Hole Bird Pond7041.78:1Mangrove Lake6641.17:1West Walsingham Ponds5131.86:1Trott’s Pond5001:1Warwick Pond1501.21:1 F. relictus  Bartram’s Pond5002.06:1Lover’s Lake5001.08:1Table 4. Fundulus bermudae  and F. relictus  . Female to male sex ratios (F:M) of killifish in each of the ponds studiedNF vs. MF:Msizesex ratiop  χ 2 p F. bermudae  Blue Hole Bird Pond7040.28554.2 <0.001 Mangrove Lake6640.0623.46>0.05West Walsingham Ponds513 0.023 47.71 <0.001 Trott’s Pond5000.3250>0.05Warwick Pond150 0.009 1.31>0.05 F. relictus  Bartram’s Pond5000.12760.13 <0.001 Lover’s Lake500 0.006 0.8>0.05Table 5. Fundulus bermudae  and F. relictus  . Summary of thestatistical tests for male (M) and female (F) killifish size andsexratio (M:F) comparisons in each pond. Differences weretested with Mood Median tests and chi-squared tests (with1degree of freedom), respectively. Significant differencesbetween the sexes are in bold (p < 0.05)
Similar documents
View more...
Search Related
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks