TRANSLITTORALTALITRIDAE (CRUSTACEA: AMPHIPODA) AND THE NEED TO RESERVE TRANSITIONALHABITATS: EXAMPLES FROM TASMANIAN
SALTMARSHESAND OTHERCOASTALSITES
AlastairM.M. Richardson, Roy SwainandVeraWong1
Departmentof Zoology, University of Tasmania,GPOBox 252C, Hobart,Tas. 7001, Australia.
'Present address: 8 CedexCt, Glen Waverley,Vic. 3150, Australia.
Abstract
Richardson,A.M.M.,Swain, R.andWong, V., 1997. Translittoral Talitridae (Crustacea:
Amphipoda)andtheneedtoreserve transitional habitats:examplesfrom Tasmaniansalt-
marshes andother coastal sites. MemoirsoftheMuseumofVictoria 56(2): 521-529.
Transitionalhabitats,e.g.,those thatliebetweenseaandland, are often ignoredbecause of thedifferentbackgroundsofmarine andterrestrial biologists,but they areimportantforthe seriesofspecieswhichspantheboundary.The amphipodfamily Talitridae hasperhapsthe widesttranslittoralrangeandsoprovides evidence fromwhichthe evolutionary processof landcolonisationcan be deduced.Thetalitridsfound ontwosea-land transitionsin Tas- mania,asandy-beachto rainforestandasaltmarshtodry woodland,are described,andthe significance oftheirdistributions isdiscussed.The need to reserve thefull rangeof such translittoral seriesis noted. Thiswill notnecessarily be achieved byestablishing reserve boundarieson the basisofvegetation associations.
Introduction
Transitionalhabitats, thatisthoselyingbetween major habitats (land and sea, freshwater and land) tend to fall through the net ofbiological investigations (Little, 1990). Marine biologists andterrestrialbiologistsdevelopfromsuchdif- ferent backgrounds that they have little to do witheach otherand so the zones ofinteraction between their spheres of operation tend to be ignored. While marine biologists oftenwork in the intertidalzone,their studiestendtobe con- centratedwell belowthe high watermark,while fewterrestrialbiologistsapproachthehigh water markfromthe otherdirection.Thustheamount ofstudyon the intermediate zoneissmall, and
thisisparticularlyobviouswhenthe transitional zone is extensive, as it is in the most sheltered sea-land interfaces, in saltmarshes and man-
groves.
Andyet the interface between thetwo major
habitats is in many waysmore important than the sea or land alone. The colonisation of land was one of the greatest evolutionary events in the history of those seven phyla that have achievedterrestriality.Although mostofthecol- onisation eventshappened a long timeago, the habitats in which they occurred are still there, and in some cases species can be found filling many ofthe intermediate stages. In a fewtaxa, which might be called translittoral (Chester,
1992),seriesofspeciescan befound withdistri- butions spanning the entire range from fully
marinetofullyterrestrial.Thebestexamplesare theamphipod crustacean family Talitridaeand thepulmonatemollusc familyEllobiidae.Trans-
littoralgroups ofthistype provide very valuable opportunities to reconstruct the evolutionary events leading to life on land (Little, 1989,
1990),andtheybecome morevaluablestillwhen
the species involved are relatively closely- related, as in the two families mentioned above.
In thiscontribution wewill presentevidence of translittoral distributions in the Talitridae anddemonstratethatsuchdistributions require a broader approach to reserve design than has been adopted hitherto.
Talitridae
TheTalitridae include theonly fully terrestrial
members ofthe Amphipoda, one of only two crustacean groups (the other is the Isopoda) which havemembersthatarefully terrestrial,ie.
independent of free water for all oftheir life stages. The Talitridae have been looselyclassi-
fied by Bousfield (1984) into four groups, the palustral talitrids (marsh-hoppers), the beach-
fleas,thesandhoppers andthelandhoppers.The marsh-hoppersinclude the mostplesiomorphic forms and they arefoundin thelower levelsof saltmarshesandmangroves,wheretheyhavean amphibious lifestyle, entering and swimming
efficiently inseawater even atlowtide. In some
places (Tasmania:Chester, 1992,South Africa:
521
https://doi.org/10.24199/j.mmv.1997.56.46
522 A.M.M. RICHARDSON
Griffiths, 1976) their distributions extend into fullyfreshwaterstreams. InsoutheastAustralia the genus Eorchestia is widespread in salt-
marshes and the lower reaches of associated streams (Richardson, 1993).Thebeachfleasand sandhoppersarethe typicaltalitridsoftheinter- tidal zone, distinguished by the adaptations of sandhoppersfortheirparticularhabitat,ieloose sand. Thus the sandhoppers are large, robust and spiny talitrids, sometimes with unusual morphological adaptationsforlifeinsand.They
are efficient burrowers. The beachfleas, in con-
trast,arcnotwell-developedforburrowing,pre- ferring to seek refuge under cast seaweed or stones (non substrate modifiers: Macintyre, 1963,ornestlers: Griffiths, 1976); they are gen- erally more lightly built and weakly spined.
Beachfleas are usuallyfoundonrocky ormuddy
shores, but there is at least one undescribed species inTasmania which hasbecome conver- gent with thesandhoppers and forages on mid-
tidalwetsand duringlowtide,taking refugeina shallow burrow during high tide.
The landhoppersare found ina range ofter- restrial habitats in the Southern Hemisphere, Japan, the Phillipines, Central America, the Caribbean and many oceanic islands. In Tas- maniathey arcfound throughouttheisland,but wheretheyhavecolonised continents(Australia, southernAfrica)theyareconfinedtothehigher rainfall zones around the continental margins.
Bousficld (1984) recognises two subgroups, the plesiomorphic simplidactylate genera, which seemtobedescendedfromamarsh-hopperlike ancestor, and the cuspidactylate genera which apparently arose from the beachfleas. The sim- plidactylate formsareconfinedtoAustraliaand southern Africa.
Thusthe Talitridae showa truly translittoral range from the semi-aquatic marsh-hoppers to thelandhoppers.Thesisterfamilyto theTalitri- dae is the Hyalidae, and it is widely- accepted that talitridsdescendedfroma hyalid-like ances- tor (Bousficld, 1984). Aquatic hyalids may be found adjacent to intertidal talitrids in fully
marine and brackish waters, providing a valid extension to the translittoral series.
Itseemslikelythat thetalitridamphipodscol- onised land relatively recently compared to otherterrestrialarthropods. Little (1990) noted that the fossil record ofthe amphipods is only from theQuaternary, but Bousfield (1984)and Bousficld and Poinar (1995) suggests a mid- Cretaceous origin for the simplidactylate land- hoppers of the Southern Hemisphere conti-
nents,contemporarywithanextensive radiation ofthe angiosperm flora in Gondwana.
Tasmaniahasaremarkablyrich talitridfauna (Richardson, 1993, 1996) which includes rep- resentatives of all Bousfield's groupings. The taxonomy of the landhoppers is well-known (Friend, 1987)andthe diverseintertidalfaunais currently being described by the senior author.
Friend (1987) divided theTasmanian landhop- perfaunaintothree ecological groupings,oneof which he called the coastal group. Five or six species inthreegenera make upthisgroupand they have distributions which are strongly restricted to a narrow zone, less than 100 m
wide, immediately behind the hightide mark.
Tasmaniaalso retainssomeplaces,especially in the west, where transitions from land to sea remain undisturbed, and clothed with native vegetation.Theseprovideimportant opportuni-
tiesto studythe natural distributionsoftalitrid
amphipods, isopodsandothergroupsacross the transition. Wewill considertwo cases, a sandy beach to wet sclerophyll forest at South Cape Bay and asaltmarshtodry woodlandtransition at Lutregala Marshon Bruny Island. Details of these studies are described elsewhere (South Cape Bay: Richardson et al., 1991; Lutregala Marsh: Richardson and Mulcahy, in press).
Studyareasand methods South Cape Bay
SouthCapeBayisintheextremesouth of Tas- mania(Fig 1). It facesalmost duesouth, but is shelteredfromthe prevailingsouthwesterlyand westerly winds by promontories to the west.
Because ofstrongwave action, the sand above thehighwatermarkiscolonisedbyfew foredune plants, so that low coastal scrub and forest appears within a few metres of high water.
Closesttotheshoretheshrubs Correa backhou- siana, Cyathodesabietina, Westringiabrevifolia
and Olearia phlogopappa form a low dense cover, fiveto ten metresdeep, whichgives way
to lowforest with emergent treesof Eucalyptus nitida, sassafras,Athewspermamoschatum,cel- ery top pine, Phyllocladus aspleniijblius, and southernbeech,Nothofagus cunninghamii. This coastal forest has an understorey of native laurel,Anopterusglandulosus, andoften cutting grass, Gahnia grandis. The ground layer is
mostly absent, apart from a few mosses and herbs.The vegetation becomeslowerand more wind-pruned fromthewesterntotheeasternend ofthe bay.
Kilometers
LutregalaMarsh
S.CapeBay
Figure 1.Tasmania, showingthe locationsofthetwo studysites mentioned in thetext.
Thetopography immediatelyinlandfromthe beach consists of a series oftwo or three low
ridges,apparentlystabiliseddunes, separatedby damper swales. The most landward ridge
becomeshigherfromwesttoeast,butattheeast- ern end of the bay erosion has cut into these structures, eliminatingthelowerseawardridges and formingasandcliffapproximately8metres high.The soilsaresandy, butatthe eastern end ofthebaywaterloggedclayeysoilsappear within
30m inland ofthe top ofthesand cliff.
Two transect lines were set up in February
1989, at right angles to the high water mark, from below the extreme high water mark of
spring tides (EHWS) to between 40 and 70 m
inland.At eachmetermarkalongthese transects a pitfall trap was set. The traps consisted ofa plastic container of 5 cm diameter and 10 cm
depth, let into thesoilsothat thelipofthe trap wasflushwiththe surfaceofthe sand,soilorleaf litter. Saturated picric acid solution (5-10 ml) wasplacedinthe traps as a non-volatilepreserv- ativeand Petridishlidsweresupportedontwigs overthe traps to prevent rainwaterfrom enter- ing them. Traps were set for one series often nightsinFebruary andanother of14 daysinJuly 1989.
LutregalaMarsh
LutregalaMarsh issituated onthe west coast ofBruny Island, which isjust offthe southeast coast of Tasmania (Fig. 1). The marsh has a
diverse talitrid fauna, and, importantly,a rela- tivelyundisturbedtransitionfromsaltbushsalt-
marshthroughsalinetussockgrasslandtonative terrestrial vegetation, despite some grazing by cattleon the higher levels. Themarsh occupies approximately 10ha.
Two sampling methods were employed to determine the distribution of amphipods.
Firstly,twotransectsofpitfalltraps,moreorless at right angles to the high water mark, were placed across the marsh, andextendedintoter- restrial vegetation up to 50 m beyond the extreme high water mark. Pitfalls were con- structed inthesame way asat SouthCapeBay.
Transect A was set and retrieved three times between March and October 1990 duringover- nightlow springtides, but Transect B wasonly used once because ofpersistent flooding.
In the second approach to sampling, collec- tions were made extensively over the marsh.
Fromarecent, colouraerialphotograph,collec- tionsiteswere chosenthatcouldeasilybeident- ified onthe ground, and which representedthe range ofvegetation types present. Ateach sam- plingsite, at least 20 amphipods werecollected with anaspirator, and the sameenvironmental variables thatwere collected at each pitfall site
were measured. These collections were also
made between March and October 1990. This period covers both the non-breeding (winter) and early breeding season of the species present.
Results South Cape Bay
Twelvespecies, threeofwhich weresupralit- toral sandhoppers, are listed in Table 1. Two
undescribed speciesof Tasmanorchestia Friend were identified, and none ofthe sandhoppers has beenplacedinaspecifictaxon because ofthe poor taxonomic knowledge ofthisgroup.
The catchesfromthe pitfall traps (Figure 2a- d)showdistinctzonationsinthe distributionof species.Amongthesandhoppers,there isa clear series,with'Talorchestia' species1 foundclosest tothe sea, followed by 'Talorchestia' species2,
which in turn is replaced by 'Talorchestia' species 3. The replacement of sandhoppers ('Talorchestia' species 3) by landhoppers(Aus- trotroidesmaritimusFriend)is verysharp. The overlap was never more than 2 traps (= 2m), and onseveral nights therewasnooverlapatall.
524 A.M.M. RICHARDSON
Table 1. Talitrid amphipodscollected at South Cape Bay and Lutregala Marsh, with their phylogenetic affinities, ecologicalgroupings and abundance. Bousfield's (1984) groupings are as follows. I:i: palustral talitrids. II:iii: cuspidactylate beachfleas havinga 4-dentate lacinia mobilis on the left mandible. Ilia: sandhoppers havinga 4-dentate lacinia mobilison the left mandible;
III:ii: sandhoppers havinga 5-dentate lacinia mobilison the left mandible. IV:i: simplidactylate landhoppers. IV:ii: cuspidactylatelandhoppers.
Taxon Bousfield Landhopper Abundance: Abundance
Group Ecological at South at (Bousfield Group (Frienc1 Cape Bay Lutregala
1984) 1987)
Eorchestiapalustris Richardson I:i Common
Beachflea ll:iii
—
CommonTalorchestiasp. 3 III:i Common
—
Talorchestiasp.1 Ill:ii Common
—
Talorchestiasp. 2 IIF.ii Common
—
Austwtwides maritimns Friend IV:i Coastal Common Common
A. longicomis Friend IV:i Westernforest Rare
—
Keratroides rex Friend IV:i Coastal Scarce Scarce
Mysticotalitrus Iasmaniae(Ruftb) IV:i Eastern forest Common Common
M. ayptusFriend IV:i Eastern forest Scarce Scarce
K. vulgaris(Friend) IV:i Eastern forest Common Scarce
K. angulosus(Friend) IV:i Eastern forest Common
—
NeorchestiaplicibranchaFriend IV:i Western forest Common
—
Tasmanorchestia Friend sp.1 IV:ii Coastal Rare
Tasmanorchestia Friend sp.2 IV:ii Coastal Rare Rare
OntransectAtherewasacompletereplacement ofsandhoppersbyA. maritimusfrom trap9to trap 10, with no sandhoppers intrap 10andno
A. maritimus in trap 9 after 14 nightsofcatch- ing. Once beyond the landward range of the sandhoppers, the catch at S. Cape was domi- nated by A. maritimus for the next 20-30 m
inland. Tasmanorchestiasp.andKeratroidesrex also appeared, but in very low numbers, and only on transect B in thesummer series.
Samples collected by hand, during the day, (Table 2) from the four swaleson S. Cape Bay transect Bdidnotcontainthesameproportions ofspecies as thepitfall trapsat all sites. Atthe seawardend, handsamplesfailedtocatchsand- hoppersbut caughtamuchhighernumberofA.
maritimus and K. vulgaris. Further landward, mostspeciesweremuchlesslikelytobetrapped than caught by hand,andonlyA.maritimusand M. tasmaniae were over-represented in the traps.
Furtherinland,beyondthetraplines,theland- hoppercommunity in forestlitterchanges,with notraceofthe coastalgroupspeciesor Mystico- talitrustasmaniae.K.angulosusandNeorchestia plicibrancha become the dominants, with Austwtwides longicomis Friend appearingata
drier site about 1 km from the beach. Further inlandstill,N.plicibranchawastheonlyspecies collected in rainforest several kilometers from the coast.
LutregalaMarsh
Eight species of talitrid amphipods were
identified at Lutregala Marsh. These included onespeciesfromeach ofthe palustralandbeach-
fleagroups, three coastal speciesandthreeforest species(Table3). Oneofthe coastal specieswas an apparently undescribedspeciesofTasmanor- chestia Friend.
Profiles ofthe transects, the distributions of amphipodsandthe levelsoftheenvironmental parameters are shown in Figs 3a and 3b. Both transectscross aslight rise attheirseawardends, butTransectB isclearly muchlower than Tran- sectAandliabletotidalfloodingovermostofits length. At their landward ends both transects riseslightly andpass into terrestrialvegetation.
The most striking contrast in the amphipod fauna is the absence of Eorchestiapalustris on the drierTransect A,whereit isreplaced inthe lowersites bythe beachflea. OnTransect Bthe beachflea ismostly confinedto theupper levels ofthetransect,beyondthe range ofE.palustris.
t
TRANSLITTORALTALITRIDAEAMPHIPODA INTASMANIA 525
Talorchestia sp1
fe
Talorchestiasp2
3.0 10 20 30
Talorchestiasp3
10 20 30
A Mysticotalitrustasmaniae
10 20 30
3Neorchestiaplicibrancha
A:S CapeA:February
\t
Talorchestia sp1
3.0
10 20 30
Talorchestia sp2
it
10 20 30
Talorchestiasp3
3.0 -\
Austrotroides maritimus
»•
3.0 10 20 30
j Mysticotalitrustasmaniae
3.0"I 10 20 30
Keratroidesvulgaris—
•-
Keratroides angulosus
3.0 10 20 30
Neorchestiaplicibrancha
C:S.CapeA: July
3.0
Talorchestiasp1
h:
20 30 40
Talorchestiasp2
3.0
10 20 30 40 50 80
3.0 Austrotroides maritimus
o i
.
20 30
Talorchestiasp3
30 40 50 60
Austrotroidesmaritimus
30 40 50
Mysticotalitrustasmaniae
3.0 1
o
10 20 30
jKeratroides angulosus—
ooo » O
40 50 60
Keratroides rex
Keratroidesvulgaris 3.0
o
60 metres
I Tasmanorchestiasp.
^»
Neorchestiaplicibrancha
10 20 30
B: S.CapeB: February
J^^^l Talorchestiasp1
60 metres
60 metres
1
3.0 "
3.0 "
w
10 20 30^^^^ Talorchestiasp2
40 50 60 70
w*
10 20 30
^^^^ Talorchestiasp3
40 50 60 70
w*
3.0
;
10 20 30
Austrotroidesmaritimus
40 50 60 70
:
3.0 -
10 20 30
Mysticotalitrustasmaniae
40 50 60 70
3.0 '
10 20 30
Keratroidesvulgaris-4
40 50 60 70
Keratroides angulosus <m
3.0"
10 20 30
Neorchestiaplicibrancha
40 50 60 70
•'•* T
10 20 30 40 50 60 70
D: S.CapeB: July
Figure2a-d Catches oftalitridamphipodsinpitfalltrapsatSouthCapeBay,Tasmania.Thehorizontalax, sism
metresfrom Trap1 which wasset attheseaward edge ofthe strandl.ne vegetation; theverticalaxisis log,,,(X + M whereXistheaveragedaily catchof eachspeciesineachpitfall.Thetransectsextended40m(A) or45m(B) inland duringthe February series,but were extendedto 70m inthe July series.
526 A.M.M. RICHARDSON
Table 2. Catches oftalitrid amphipodsbyhand (first figure) orpitfall (second figure) at or near South Cape Bay (nopitfallswere set inland ofthe 4th swale). Species abbreviations are as follows. Tal 2: Talorchestia sp. 2; Tal 3: Talorchestiasp. 3; Austr mar:Austrotroidesmaritimus;
Kera vulg: Keratroides vulgaris; Keraangu: Keratroidesangulosus; Mysttas: Mysticotalitrus tasmaniae; Myst cryp:Mysticotalitrus crytpus; Neorplic: Neorchestia plicibrancha;Austrlong:
Austrotroides longicornis.
Tal 2
Tal 3
Austr mar
Kera vulg
Kera angu
Myst
tas
Myst cryp
Neor
plic
Austr long
Swale Swale Swale Swale 70 m
1 km
5 km
0/1 3/96 125/1 20/0
— —
0/3 125/4 78/0 6/0 30/0
—
2/4 128/1 11/0 9/8— —
40/0 75/2 4/4— —
1 14—
14
3/0 8/0 35/6
6 14
On both transects thereisconsiderableoverlap between the distributions of the beachflea, coastal landhoppers, andforest landhoppers.
Fig4 shows the distribution ofspeciesfrom the four ecological groups ofamphipods that were present (marsh-hoppers, beachfleas, coastaland forestlandhoppers) overthe marsh and the major vegetation types. Eorchestia palustrisisrestrictedtothewetmarshvegetation whereitisregularlyinundated, whilethebeach-
flea extends more deeplyinto the upper marsh and terrestrial vegetation. The forest landhop- pers arerestricted totheterrestrialscrub,but the coastallandhoppersarefound boththereandin the uppermarsh.
Discussion
At South Cape Bay the interface between the beach and the adjacent forest is quite sharp.
There was complete replacement of sandhop- persby landhopperswithin4metres,atthe point wherewoodyvegetationformeda closedcanopy and substantial amounts of organic matter began to appear in the sandy soil. The coastal landhoppers extend inland for another 50 m, beyond which the landhopper community becomes much less diverse than in the coastal zone.Furtherworkat SouthCapeBay andelse-
where in the west of Tasmania (Richardson, 1993, in prep.)shows that the distance that the coastal species penetrate inland is related to exposuretowind, perhaps becausethisincreases theamount ofsalt spray deposited behind the shore.
At Lutregala Marsh, in contrast, there is no obviouspoint where intertidal talitrids (in this
case beachfleas) are replaced by landhoppers, since there are substantial overlaps both between the distributions ofthe beachflea and the coastallandhoppersand betweenthebeach-
fleaandthemarsh-hopperswhich occupyazone closerto the sea.
The numbersoftalitridspecies presentatboth
sitesis high, 11 speciesat South Cape Bay and eight at Lutregala Marsh, and phylogenetically diverse. Although the phylogeny ofthe Talitri- dae is not clearly understood, it is generally accepted (Bousfield, 1984, Moore et al., 1993) that sandhoppers form a distinct, specialised group that is unlikely to have produced any forms ancestraltothe landhoppers. This, com- bined with the absence of beachfleas and the abrupt transition from beach fauna to forest fauna, suggeststhatsandy beacheswereunlikely to have been the sitewhere the colonisation of landbytalitridsbegan.Onthe saltmarshes,both ofthelikelyancestralgroupsare present(marsh- hoppersandbeachfleas),the transitiontolandis
gradual, distributions overlapsubstantiallyand the saltmarsh habitat foreshadows land in the structure and nutritional nature of its
vegetation.
Inboth these examplesthere arereasonswhy
the phenomena ofinterest covera widerrange than "shore" or"land", howeverthey might be defined. Certainly vegetation patterns do not giveanyindicationoftheinlandlimitsofcoastal landhoppers in either location. There hasbeen some interest in the reservation of saltmarsh vegetation in Tasmania (Kirkpatrick &Glasby 1981),butbydefinition thathasreferred tosalt-
marshplants.Therearemanyrecords of beach-
Mysticotalitrustasmaniae
Mysticotalitruscryptus
Keratroidesvulgaris
10 animals per trapnight
Austrotroides maritimus
Keratroides rex Tasmanorchestia sp.
A2 A4 A6 A8 A10 A12 A14 A16 A18
10 animals per trapnight
Eorchestia palustris
Mysticotalitrus cryptus <
Keratroides rex
Austrotroides maritimus —
Tasmanorchestiasp.
Figure 3a, b. Profile oftwo transects on Lutregala Marsh showingthelevelsofsoilmoisture,soilorganic contentandsoilsalinity ateach trappingposition,and therelative abundanceofamphipodstrappedin pit-
falls. MHW: meanhigh waterlevel.
•:Mysticotalitrustasmaniae O :Austrotroidesmaritimus
:M.cryptus :Keratroides rex
A:Keratroides rex A : Tasmanorchestia sp.
250m
O:Eorchestiapalustris
•:Beachflea
Figure 4a, b. Maps ofLutregala Marsh, showingthe distributions ofthe four ecologicalgroups oftalitrid amphipods. A: marsh-hoppers and beachfleas; B:
coastal and forestlandhoppers. Thedashedline rep- resents the approximate boundary of woodland.
Sampleswerecollectedby hand,and from thepitfall transects(A: startoftransect A:B: startoftransectB).
Filledsymbolsrefertospeciesoftheforestlandhopper group, or the beachflea, open symbolsto the coastal landhopper grouportothepalustralgroup(Eorchestia palustris).