Key words: Temporary pools, vernal pools, crustaceans, wetlands, diversity ... pools occur worldwide, provide habitat for organisms with a variety of life history ...
8S
Hvdrobiologia 32S : 85-11 6, 1996.
© 1996 Kluw er Academic Publishers. Primed in Belgium.
Species richness, endemism and ecology of crustacean assemblages in northern California vernal pools Jamie L. King ', M arie A. Sirnovich ' & Richard C. Brusca.' of Ev olution & Ecology, University of Californi a, Davis, California 95616, USA of Biology, Unive rsity of San Diego, Alcala Park, Sa n Diego, Californi a 92110, USA lGrice Marin e Biological Laboratory, University of Charl eston, 205 Ft. johnson, Charleston, South Carolina 29412, USA I Sec tion
2 Departm ent
Received 20 June 1995; in revised form 23 November 1995; accepted 29 November 1995
Key words: Temporary pools, vernal pools, crustaceans, wetlands, diversity
Abstract Ephemer al pool s occur worldwide , provide habita t for organisms with a variety of life history strategie s, and may have served as evolutionary refugi a for some taxa since Meso zoic times. Yet, our understanding of the ec ology and evolutionary history of ephemeral pool communities is hampered by a pauci ty of such basic data as .the species co mposition of pool assemblages. We surveyed S8 vernal (ephemeral spring-time) pools from 14 sites in northern California for crustaceans, and found diverse assemblages composed largely of endemic and rare species . Sixty-seve n species of crustaceans were found, and as many as 30 of these may be new, undescribed species. Differen ces in species composition among pools correspond with physical and chemical aspects of the habitat (depth, solutes concentration, elevation, biogeographic region), and with existing geologi clfJoristic-based habitat descripti ons. Spe cies richness is positively correlated with both depth and surface area. Thi s relationship can be explained in terms of hydroperiod (accommodation of species wi th slower developmental rates in long-Jived pool s, greater time for temp oral resource partitioni ng) and size (spatial habitat heterogeneity), High species richness and numerous co-occ urrences of congeneric species in temporary pools may be due to super-abundant resources, low levels of predat ion, and annual truncation of the community which prevents ecological interaction s from go ing to completion. The results of this survey underscore the need for conservation of the vernal pool habitat and endem ic vernal pool spec ies in California. The best preservation strategy will include many pools at each site, multipl e site s of each habitat type, and all identified habitat types.
Introduction
Ephemeral pool s are an ecological feature of virtually every continent on the globe (Thorn e, J 984; Williams, 1985; Zedler, 1987 and references therein). These ephemeral wetland s, which are also referred to In the U.S.A. as temp orary pools, vernal or autumnal pools, hogwallows, pans, and playas, occ ur wherever season al precipit ation and impervious substrates combine 10 create surf ace pools during the wet season (Holland & Jain, 1981 ). Ephemeral pool communities include both opportuni stic (vagrant) species (e.g . flying insects, birds, amph ibians) which arc present during the aquat ic phase and kay e as the pool dries, and less-vagile
(resident) species (e.g. plants, crustaceans, gastropods, many beetle species) , which survive the drought phase as dormant adults . juveniles, or propagu les such as seeds or cysts (see Wiggins et al., I no for a more exhaustive synthesis of Iife history strategi es in tempo rary pools). The ecology of these unique communities thus reflects the interplay of a temp orally variable habi tat and the activities of many organi sms with different Iile history strategi es. Evolutionarily, eph emeral pools may be important both as refuges for certain taxa and as sites of allopatric speciation. At least one group , the large branchi opod crustac eans, became restricted to temporary pools concurrent with the rise of predatory, suction-feeding bony fishes in Mes ozoic seas and lakes
86 (Hutchinson, J.9 67; Kerfoot & Lynch, 1987 ; Wagele, 1992; Wiggins et al., 1980). Rate s of spec ia tion and local ex tinc tion for so me epheme ral pool taxa may be relatively high, reflecting the ge ograp hically fragment ed natur e of the habit at and the isol ation of sm all pop ulations (Fugate, 1992 ; Holl and &Jain, 1988; Ho over, 1937). Despite their worldwid e d istribution, ecologic al interest, and evol utionary importan ce ,ep hemeral pools are usually overlooked in lim nolog ica1 and ecolog ical textbooks (Williams, 198 5). Also , in contrast with the large bod y of knowledge dealin g with zoo pla nkton communities of permanent waters such as lakes, the aquatic liter atur e contains very few surveys of eph emeral pool fauna, Co nsequently, the litera ture of temporary waters is largel y scattered am ong species- and taxon-specific papers, an d rar ely addr ess es com munity-level pa tterns and pr ocesses (with the exc eptio n of reviews dealin g with life history strate gies of ephemeral pool inhabitants, e .g., Wiggins et a!., 1981 ; Williams, 198 5). The often uncertain future of eph emeral pool habi tats and thei r endemic species pre sents an additional imperativ e for basic research and con servati on of these sys tems . Losses of vern al pool habitats in Californ ia's Ce ntral Valley, whi ch began in the la te 1800's with the agr icultural settleme nt of the st ate and continue today du e to urbani zation, are es tim ated at be tween 60 and 85% of pre-settlement acreage (Federal Regi ster, 1994b). In highl y-urbani zed co astal areas habitat loss es have been even greater: approximately 97% of the vernal pools in San Diego County have been destroyed (Weir & Bauder, 1990). This massi ve loss of habit at po rtends ex tinction fo r many species which are uniqu e ly adapted and end emi c to Califo rn ia vernal pool com munities. Fifteen species of vernal pool plants (Fed era l Register, 1991 ; 1992 ; 1993a ; 199 3b) and seven species of vern al pool invertebrates (Fe deral Regi ster, 1980 ; 1993a; 1994a ; 1994b ) are now federa lly listed or proposed for listin g as threaten ed or enda ngered. Inhabitan ts of ephemeral pools in ot her parts of the world are also disapp earing because of habit at losses. For example, one species of ephemeral pool anostracan fro m Fran ce is beli eved to be locally ex tinct, and anoth er is con sidered endange red (Belk, 1994); of the three eubranchiop od spec ies known fro m G reat Brit ain , one is locally extinct, and the oth er two are greatly redu ced in ra nge and are now protected by the Briti sh Wildl ife and Cou ntr yside Ac t 1981 (Gra inger, 1994). This pa per rep orts the resu lts of a survey of the crus tacean fauna inhabiting vernal pools in north ern Cal
iforni a. Our study focuses on crustacea ns be cause of their num erical dominance and ecologica l importance in the poo l co mmunity. We pre sent data on the physical habitat, species assemblages, p atterns of species rich ness, and co-occurre nce of co ngeneric species, and we evalu ate several ecologi cal hyp otheses for thei r poten tial to expl ain the obse rved patt erns. Spec ifically, we report the discovery of 30 potentially new crustacean species , asse ss the correspo nde nce of ph ysical habitat charac teristics with spec ies assemblage sim ilarity, an d examine the extent to which ecological principles su ch as spec ies-area relation ship s, co mpetitio n, pr ed ati on, productivity, and the effects of temp oral var iability may de term ine patterns and lev els of species ric hness in these vern al po ols. This su rvey, alth ough lim ited to crus taceans , is the first of its ki nd for north ern C alifornia vern al pools and represents an important step toward understand ing these largely unk now n comm unities. Thi s first step is cru cial , because without a basic und er stand ing of patt ern s and sca les of var iatio n in a sys tem it is exceedingly difficult to design informa tive e cologi cal field experimen ts. Thu s, whi le the d ata we present are necess arily de scriptive and exploratory, our main goal is to stim ulate more ecolog ica l and evo lutio nary study of Ca liforn ia 's vernal po ols in particul ar, and eph emeral freshwater pool s in ge neral. Th e dat a from our survey pr ovid e a significan t foundation upon which hyp ot hesis -o riented field experiments can be built.
Description of sites studied We surveyed 58 vern al pools from 14 study sites situ ated along a 30 0 Ian north-south tran sect throu gh the Sacramento Valley , from Fall Ri ver in Sh asta Cou nty to Jep son Pr airi e in Solan o Co unty durin g the spr ing of 1992 (Figure I). Th e tran sect route followed a Pacif ic Gas & El ect ric natur al gas pip elin e corridor that extends approximately 1360 kil ometers from Ki ngs gate, Briti sh Colum bia, Ca nada to Pa noc he, Ca lifor nia, U.S. A . The su bterranea n pip eline was ori gin all y construc ted in 1960. We co nsider the pipeline route to be a rand om (w ith respect to the distributi on and typ es of vern al pool habitats encountered ) tran sect thr ough the Ce nt ra l Valley. Our survey included und isturbed pool s adj acen t to the pipelin e corridor, as well as pools on the pipeli ne right-of- way which have re-establi shed or wh ich lie o utside the trench zone and appear to have been only moderately disturbed by the 1960 con struc tion . All pools were visited three times duri ng the 1992
37
Alturas
.... Fall Riv er Burney
Reddi ng
+ Manton Plain
Th omes Creek
Susanville
• Dales Plain • Tuscan Buttes \ + Red Bluff ;+ Coyote Creek Truckee Creek
+ Hall Creek of Stony Creek + Wilso n Creek +
Corning Chico
Truckee Marysville
• Jepson Prairie ) Stoc kto n
~
erkeley
Key to Habitat Types
San Jose
.... Northern Transmontane • Volcanic Mudflow + Hardpan • Claypan
o I o
10
20
30 Kilometers
i ' i i
10
20 Miles
Figure J. Map of study sues ill northern California. Transect extends approximately 3UOkrn from FaJI River in the north to Jepson Prairie in the south.
88 wet seas on betwe en February and May , although some po ols were dry be fore the third sam pling . The transect study sites fa ll into four ve rna l pool habit at typ es according to the geological and floris tic cla ssi fication es tablished by Cheatham & Haller (1975, as ci ted in Holland, 1986) , desc rib ed by Holl an d ( l 986), and modified by Taylor et al . (1992). F aJl R iver (F R) is the single representative of northern transrnon tan e verna l pool s and contains very sm all pools (refer to Appe ndices A and B for pool sizes) within a m ou n tain mead ow at 1100 m ele vation. All the other sit es are be low l OO m ele vation and ar e situated in the Sacr a me nto River Valley. Three sites , Manton Pla in (MP), Da les Plain (DP), and Tuscan B uttes (T B), con tai n pri mar ily north ern Sacramento Valley volcani c mudflow vernal pools . Vernal poo ls at these volcanic boul der s tr ew n si tes are few in number but med iu m to very large in size, despite ea rl ier descriptions of such pool s as typ ically sma ll (viz. Holland 19 86). The Jepson Pr a irie (JP) site is cha racte rized by bot h s mall mim a mound (hummock and swa le topo graphy, se e Cox & Gakahu 1983) pools and very large, turbid, wind s we pt play as of the Sacrame nto Valley clay pa n vernal poo l ty pe . Only the large playas are inc luded in this su rvey because the sm all pools at this site were not sufficie ntly inundat ed during the sampli ng per iods. The remai ning ni ne si tes, Red Bl uff (RE ), Coyote Creek (CC) , Truckee Creek (TR), Thomes Creek (T H), Corning (CR ), H all C ree k (HC), Ston y Creek (SC), Wilson Creek (WC) , and Alle nda le (AL), co ntain Sacramento Valley hard pan vernal poo ls. These sites inclu de numerou s very sm all to me d ium sized pool s scattered amo ng mi rn a mo unds in valley grasslands .
Methods Sampling procedures. All pools were sampled three time s, except for mos t of the hardpan pools whi ch were sampled only twice because they were dry atthe time of the third samp ling. The three sampli ng tri ps (Fe bruary 16-24 , March 23- A pril 3, and Ap ril 22 May 1) spanned roughly the ea rly, middle, and late por tion s of poo l hydrocycles We sampled aquatic organ isms from poo ls by volume -standard ized plankton tows using a sq uare han d net frame ( 17 em x 25 em) fitted with 400 lim mes h plank ton net. Whe rever possi ble , we loo k samples separately from the sur face, bo ttorn, and ed ge areas of the pool. For sur face tows , the net was dropped vertically into the water to a de pth of 8 ern, then dra wn horizontally through the water over
a dis tance of one meter. For bott om tow s, the net was dropp ed vert ically into the water unt il the lower edge of the frame rested on the pool bottom . Th e net was drawn across the botto m for one meter, then brou gh t verti cal ly to the surface. Tills me th od sampled app roxima tely the top 1 cm of bottom sediment and the overly ing 16 ern of water. In poo ls tha t we re not deep eno ugh fo r separate surfac e and bot tom plankton tows , we com bined these as a sing le tow. We immedia tely preserved pla nkto n samples in 70 % ethy l alco hol. Water chemistry and physical p roperties. We col lected water chemis try data in the field, includin g co n ductivi ty (li D - I), total d issolved solids (T DS in mg 1- 1), alkalinity (mg 1-1 ), salinity (%0) , p H, dissolved oxyge n (mg I-I ), and tem perat ure (D C). We es timated surface area and maxim um and average pool dep th at the lime of sampling, as we ll as long-term ma ximum surfac e area and depth bas ed on observed poo l ch ar act er istics such as filling , veg etation , an d top ogra phy . We quant itatively defined descriptive terminology fo r water chemi stry, pool size , and de pth pr ior to dat a analysis (Appendix A). Sample processing and curatio n. In the la bor ato ry, pla nkton samp les were sc ree ned throu gh a 500 lim sieve becau se mo st crustaceans sma lle r than 500 li m are primarily lar vae or j uve niles and canno t be iden tified, although this pro ced ure may have red uce d the number of harpac ticoi d copepods and very small species of cypridopsine and lim nocyth erid ostracods in our samples. T he 500 li m fract ion of ea ch sa mp le was sort ed into the follo wing categories : Anos traca, Notostraca, Conc hos trac a (L aev icaudata, Spi nica uda ta), Cladocera, Copepoda, Ostracoda , In sect a, and 'o ther' (e.g. gastro pods, annelids, arach nids, platy helminths, algae, plant frag ments, etc.). For a few samples , where thousands of individua ls were prese nt, aliquots were so rted . All sa mp les were preserved in 70% ethyl alcoh ol, exc ept so m e representati ve os tra cod specimens whic h wer e dried and permanently cura ted in glass via ls or on pa leo ntology slide s . All sam ples are perma nently arc hived at the Lo s A nge les Cou nty Mus eum of Natural History, and are accompa nied by I :150 000 scale map s show ing the geog rap hic locations of collection si tes . We en courage inves tiga tors inte rested in acc essing any speci mens or the ' other' taxa to contact the appropriate L ACM curator. Species identifications. A nostracan, no tostr acan, and co nchostracan species were iden tified by the aut hors, with some samp les ve rified by Jo el M arlin (Los Ange les Cou nty Museum of Nat ura l His tory ), Denton Belk (Our Lady of the Lake Un iversi ty, Sa n
89 A ntonio), a nd Mi chael F ugate (Universi ty of Cal ifor nia , River sid e ). Cladocerans were ide ntified by Bren da Hann (U niversity of Man itob a). Co pepods wer e ide ntified by Jan et Reid (U.S . Natio nal Museum of Na tura l H istor y, Smithso nian Insti tutio n). Ostracod s were identified by both Claire Carter (U .S . Geo log i ca l Sur vey, Menlo Pa rk) and Richard For ester (U. S. Geo logi cal Survey, Denver). Some s pecime ns co uld not be iden tified as a know n spec ies, and these are listed as ' sp.' Other spec ime ns clo sely resembled a described species, but differed su f ficie ntly from the c urrent species description to be co n s idered a ' possible' or 'probable ' new spe cie s. In these cas es the pu tative species nam e is given, but the unce r ta in tax onomi c sta ndi ng of the specim e n is note d . Two os tracod ta xa co uld not be ass igned to a know n ge nus and are listed as ' U nknown sp .'. Taxonomic descrip tions (an d redescrip tio ns) of cl adoceran and co pepod species fr om this ma ter ial are in progress and will be repo rted elsew here by the specialists noted above. Data analysis. Sali nity da ta, alth oug h po te ntial ly biologically imp ortant, wer e not includ ed in water chem istry analyses beca use they we re always below I %0, and our eq uipmen t was not sensi tive to val ues in the parts per m il1 ion ra nge. Dissolved oxyge n, altho ugh vari able, was always abo ve 5 rng ml- I and , there fore , not bio logica lly limitin g. Howev e r, we collected these dat a typicall y between 0900 and 1700 ho urs ; late nig ht or ea rly-m orn ing me asureme nts may have shown periods of oxygen de plet ion (Sch oln ic k, 1994). Tem per a ture ran ges are prov ided in Ap pe nd ix B, but are not inc luded in further da ta analyses. We performed a pri ncipa l com pone nts ana lysis on the remain ing water chemistry data a nd the pool size and elevation data usi ng the S tat view SE + Gra phics soft wa re package for the Maci nt osh (Ab acu s Concepts , Inc . 1988). We add ressed patterns of commu nity co mpos itio n w ith a c lusteri ng analysi s based upon species pres e nce /absence amo ng poo ls. For each pool we sc ored all species th at wer e presen t as pos itive ch ar acter states and all spec ies fo und in the survey but ab se nt from the pool as negat ive charac ter states. We excl uded fro m this an alysis 24 of the 67 tota l species because the y were present in few er than three pools; rare sp ecies (i.e. those occ ur ring in less tha n 5% of the poo ls) can be unin form ative and often prob lematic in mu l tivariate comm unity sim ilarity analys es (G auc h, 1982 , H iII & Ga uch, 1980) . We used the hierarchical clus te rin g f uncti on of the SYSTAT 5.0 s oftware pac kage for the Macin tosh (SY STAT, Inc . 1989) and specified
the ave rage linkage (U PGMA) m e th od of de ndrogram con struc tion to generate a ph enogram of pools . We examined the relat ion ship betwee n spec ies asse m blage and geograph ic prox im ity by 'reg ressing Jaccard 's Coefficie nt of co m m u nity simi la ri ty aga inst linear dist ance betwe en poo ls fo r al l pairw ise co m pari so ns among poo ls of the hardpa n type. Ja ccar d 's Coefficient is defi ned as
Sj = a/ (a + l> + c), where a = the num ber of spe cies in both pools, b = the nu m ber of species in pool B but not in p ool A , a nd c = the num ber of species in pool A but no t in pool B. Lin ear distances we re measu red amon g po ols wi thin sites using 1:1200 sca le maps, a nd amo ng pool s at di ffer en t sites us ing 1:150 0 00 scale m aps. A ltho ug h these meas ureme nts vary in acc ur acy across th e tw o differen t scales, we use the m as rough es tim a tes.
Results Water chemistry and physica l pro pe rties. (see Appe n d ix B for raw da ta) The six water chemist ry an d phys ical var iables ca n be redu ced to a representati ve se t of four (solutes, depth, pH, and elevation) . O ve r all sa mp les fro m all poo ls, stro ng co rrela tions exis t fo r TDS versus condu ctivity (,1 = 0 .92 1; F ig ure 2a), TDS ver su s alkalinity (?- = 0.67 5; Fig ure 2b), co nd uc tiv ity vers us alkali nity (,1 =0 .680; Fi gure 2c ), and m aximu m depth vers us log surface area (?- = 0 .394 ; Fig ure 2d ). Prin cipal com pone nts anal ysis revealed three fac tor s whi ch acco unt for 78% of the varia nce in phy s ica l and wa ter chemistry da ta (Table I ). A c utoff at three fac tors is s uggested by the c um ula tive varian ce exp lained, the fact or eigenvalue mag nitudes , a nd a scree (e ig enva lue ve rs us factor) plot (as recommended by Cattell, 1966 ; E ver itt & Dunn, 1992 ; and Joliffe 1972). Th e co rrelat ions and the com m o n fac tor-load ing of alkalinity, co nd uc tivity, and TDS s upport the coll apse o f these varia bles to a si ng le sol u tes variab le, her ea fte r re prese nted by TDS . Simi larly, d ept h and surface ar ea , both of which determine oth er biologicall y importan t properties such as pool volume a nd hyd rope riod (dura tion of pool inundationj.can be rep rese nted usi ng de pth as a single surrogate variable . A lth oug h eleva tio n does not covary with any other phys ical or c he m ical vari ables in these an alyses, it indi cates a n important d iffe r ence between Fa ll Rive r ( loca ted in th e Sie rra Nevad a M oun tains) a nd all the oth er sites (located in the va lley prope r), beca use it corresponds to major d ifferences
90
a.
b.
400
400
r 2=0.921
r2::O.675
300 r-
r-
-E 200
,
'"";'
'--"
C/J
0 f-<
•
300
•
100
••
•
OIl
E .......,
• •
t3
O-t=---.,.---..---~-~
o
200
400
.
200
(/)
600
•
100
•
•
•
• •
0 0
800
Conductivity (~- 1 )
50
100
150
Alkalinity (mg 1-1 )
d.
c .
800
r2::O.680
'"";'
c
E,- ti ::l
• • • ..:•
••
S
200 01 0
~ . 50
E x
•
.
• 0.75
•
'--"
G
;>-
c: 0 U
E ..c: c.. o
•
.g 400 "0
•
600
~
I
100
150
Alkalinity (mg 1-1 )
II
•
I.• •
•• •• •• • • •• •
0.5
..
0.25
..... •
... ......._•
o • .0 1 . 1
10 100 1000 10000
Maximum Surface Area (m2 ), log scale
Figure 2. Wate r chemistry and pool size correlation s, a, TDS vs condu ctivity (y:= O.50 3x + 1.872 ), b. TDS vs alkalinity (y :=J. 839 x- 36 ,988), c. Conductivity vs alkalinity (y :=3.7 19x -7 6,032), d. Maximum depth vs max imum surface area , log scale (y;co 0, 10910g x - 0 ,023 ).
in climatic regimes and biogeographic provinces (see Durrenberger & Johnson , 1976) , Hab itat charac teristics. The redu ced array of po ol physic al and water chem istry variables (solutes, depth, pH, and elevation) spe ci fies four group s which co rre spo nd to the four ge ologic/fl oristic habi tat types previ ously identified . Cla yp an pools have significantl y high er solute lev el s, clay pan and volcanic mud flow poo ls are signifi cantly de eper, and transmontane pools are significantly higher in elev ation than all others , None
of the po ol types show any clear pattern for pH . Based on these analyses and field ob servations the following generalization s em erge for the are as included in this study (F igure 3): Northern transmoru ane vernal po ols are low in solutes, shallow, fairl y turbid, and high in elevatio n; northern Sacramento Valley volcanic mud flow vernal po ols are low in so lutes, de ep , long-lived (4-6 mo .), very clear, and low in elevation; Sa cramento Valle y hardpan vern al pool s are low in solutes, usually s hallow, often very short-lived ( 1-3 m o.), clear, and
91 Ta hl« I. Principal compo nents factor analysis matrix (uurotated) Alkalinity, TDS, and conduc tivity load heavily on factor I, depth and surface area load heavily on factor 2; pH and elevation both load heavily on factor 3, bur the latter has u negative loading. Cumulative percent variance explained by these three factors is approximately 78% .
Factor Depth Surface area pH Alkalinity TDS Elevation Conductivity % variance
-
J
Factor 2
Facto r 3
0 .61 I
0.6 14
-
0.061
0.493
0.606
-
0. 119
0 .273
-
0.844
- 0.33 2
0 92 (,
-
0 303
- 0. 188
-
0.45 3
0. 92 6
-
0.0270
45 .070
0 194
18.03
0 .740 -
0 .02 1 0.093
-
0.67 1 0. 111 14 .820
= 77.1)2
low in ele vatio n; Sa cramento Vall ey clay pan verna l pools are high in so lu tes, ofte n ve ry deep, lo ng-li ved, highly tur bid, and low in elevat ion . Similarity of spe cies assemb lages. Spec ies assem blage s of verna l pool crus tace a ns ap pear to be strongly related to habi tat ty pe. Majo r nod es of the species assemblage phenogr am coincide with ph ysical and che mica l diffe re nces amo ng po ol s, an d with the geo logiclftoristic habita t cl ass ifica tions p resented earlier (Figure 4) . Wi th o nly a si ngle ex ce ption, poo ls from the same habitat typ e clu ster togeth er on the basis of species assem blages. Th e Jepso n Prairie (claypan) pools differ m ost fr om all others, a nd the separation of this cluster from all o ther poo ls at node I correspo nds to the large difference in sol utes between Jeps on Prai rie a nd all other si tes . Th e M ant on P lai n, D ales Plain , and Tuscan Buttes (v olc a nic mud fiow ) pools cl uster togeth er and separa te at node 2 from a ll the hardp an and transruonlane po ols, with the e xcep tio n of one Manton Plai n pool (MP2) wh ich fall s into the latt er clus ter. Thi s second node coincid es with the maxim um po ol depths at these sites, becau se the for me r grou p co ntains medi um to deep po ols and the latt er con tain s primarily very sha llow to sha llow poo ls . Th e Fa ll R iver (transmon lane) pools break aw ay from the hard pan cluster at node 3, co rrespond ing to their locat ion at a high er e le vation and in a different biogeographi c pro vince . Th e topo logy ofthis pho nog ra m is high ly ro bust ;c ha nges in both the sim ilarity coefficien t al gorith m and the lin k age method of de ndrogr am co ns truc tio n do not alte r poo1 me m ber shi p in the maj or cl usters . Two lines of ev idc nce ind ica te tha t, for pools of the same habitat typ e , geograph ic pro ximi ty is not correlated with species ass emblage similarity. The 9
ha rdpan sites, wh ich inc lude 44 poo ls and span the g rea tes t dis tance alo ng the transect o f all habi tal types, arc the most exh austi ve po rtion of our d atase t. Th e phonogram fo r th is s ubse t of d at a s hows th at po ols at the sa me site do not tend to cluster tog ethe r (F ig ure 4). A lso , a plot of al/ pai rw ise com pariso ns among hard pan pools shows no clear rel ation ship between com m unity simi lari ty (Jaccard 's Coefficient) a nd d istan ce (? = 0.00007 ; Fi g ure 5) . Crustacean species, endemism, and rarity. Six ty se ven species of Crustacea occurred in the po o ls in this s urvey, includ ing 4 specie s of Anostraca, 2 Not ostra ca, ] Spinica uda ta, I Lae vica uda ta, 18 C ladoce ra, 17 Copepoda (7 Cyclopoida, 6 Calanoida, 4 Harp act icoi da), an d 24 Os traco da (Tab le 2) , Taxo nom ic sp ecial ists iden tified 17 of the 67 species (25%) as proba bly unde scribed an d a noth er 13 (19%) as po ss ibly undescribed, based on dev ia tions of these specime ns fro m exi s ting taxo nomi c descript ions an d exam inat ions of compara tive mater ial. The se numbers indi cat e that up to 44% of the crustacean species inh abitin g nor thern Ca lifor n ia vern al poo ls cou ld be new spec ies . Thirty-five of the 67 species (52 %) fo und in th is sur vey had not pre viously bee n re ported from Ca lifornia . Some of the spe cies e nco untered, such as
Branchin ecta conse rvatio, Branchinecta lync hi, Lin der iella occldentalis, a nd Lepidurus packardi , are restric ted to Cal iforn ia vern al poo ls (Eng e t al ., 199 0), while man y othe r species are widespread an d also oc cur in permane nt freshwater ha bitats su ch as la kes, ponds, or streams . T he 30 poten tia lly new species fo und in th is s urvey are no t kn own to oc cur elsewhere a nd may be endem ic to vernal pools an d to th e region. Thus, the level of e ndem ism for crus tac ea ns in n orth ern Cali fo rnia ve rna l poo ls may be as high as 51 % (potentially 34 en demics out of 67 tot al sp eci es). Also, a large proportion of the spec ies in this su rvey are re l atively rar e. O ver 25% (17) of the 67 to tal specie s had only a si ngle pool occ urre nce (F ig ure 6a), and 40 % (27 ) ha d only a sin gle site oc currence (Figure 6b ) . Species richness and pa tterns ofdiversity. The total species ri ch ness fo und in this survey is pa rt ition ed hier archi call y amo ng poo ls, sites, an d habit at types. For e xa m ple, spec ies assembl ages vari ed, eve n am ong poo ls with in the same site, so tha t all s pec ies OCC UlTing at a si te were not fo und within a sin gle pool. C o nse q ue ntly, the s pecies richn ess of any pa rt icular si te was al way s gr ea ter tha n that of the mo st diverse p ool at tha t s ite. T he num be r of species pe r poo l ran ged f rom I (TR 17, TR24 , TROFF) to 27 (Dales Poth oles ) and averaged 9.6 spec ies per pool (s.d =: 6 . I, N = 58) . T he
92 100
o
+
E ~ se
•
75
o
A Northern Transm ontane
•
ll)
Q
o
c: 50
•
+
o
-t
+
25
+
+ +HH-++ +
+1i1't r" +
o Volcanic Mu dflow
+
Hardpan
•
Claypan
+ +
+f.+ 0+--------,----,--------;---......------, o 50 10 0 150 200 250 IDS (mg 1-1 ) Figure 3, Bivariate plot of water chemistry and pool size: maximum depth vs TDS,
total number of spec ies at a si te range d fr om 4 (Hall Creek ) to 39 (Dales Plain), with an avera ge of 18.6 speci es per site (s.d . = 8.7, N = 14). Spec ies richness by hab itat type ranged from 14 (nort hern transm ontane) to 46 (volc a nic mudflow) with an average of 31 .8 species pe r habit at type (s.d. =13.9, N =4). Thu s, on average, a given pool, site, or habitat type holds roughly 14%, 28 %, and 47%, respec tively, of the total dive rsity in the sys tem. The numb er of crustace an spec ies occ urri ng in a pool was positi veJy correlated with maximu m pool depth (,2 0.5 84 ; Figure 7a) and with the Iogari thmj, of observed ma ximum poo l sur face area (,2 = 0.300; Figure 7b). However, some species, such as B ran chinecta d issimilis, Bran chin ecta lynchi, Lep uluru s sp. A, Acaruhacyclops vern alis, Cantho camp tus sp. A, Hesperodiaptomus hirsutus, Skistodiaptomus pallidus, Hcterocyprisaff. H. rotu ndatus , and unkn ow n ostra cod sp. G, were on ly fo und in sm all, sha llow poo ls whi ch had short hydroperiods (i.e . 1-3 mo nths). Th ere fore , altho ugh deeper, larger pool s (temporary 'l akes' ) with longer hydroperiods (i.e. 4--6 mo nths) gen eral ly co ntained more spec ies, the to tal di versity encou n tered in this survey is distribut ed between both highl y eph emeral pools and pools of longer du ration. Co-occu rrence of congeneric species. Con gen eric species co-occurred in 55% (32 out of 58) of the pools.
=
Three cladoceran genera (Alona, Daphnia, and Simo ceph alus), two copepod genera (Diacyclops and Hes perodiaptomu s), an d fo ur ostracod genera tCandona, Eucypris, Heterocypris , and Limnocyth ere) incl ude co occ urri ng co nge ners (Table 3) . Some small, sho rt lived pools in this study had co-occurring co nge ner ic spec ies, but sites with the largest, most lon g-lived pools (Jepso n Prairie and Dal es Plain ) were pa rtic u larly rich in sympa tric conge ners. For example, the Jep son Pra irie site co ntai ned a total of five con gen eri c associati ons, and a sing le pool at this site (JP5/0Icott Lak e) held four congeneric species pa irs. Ni ne sets of congener ic spec ies we re fou nd at the Dal es Plain site. T hree of these com binations occ urred in a si ngle pool (Da les Lake); five others oc cu rred in another sin gle pool (Potholes) and included a co nge neric ostr acod trio. In mos t ca ses, pop ula tio ns of co-occurring co n geners overlapped both temporall y (wi thin the sa me sampl ing period ) and spa tially (wi thin the same sub sampl e).
Discussion Our survey of north ern Ca liforni a verna l pools reveal s a highly diverse cru stacea n fau na, a large propor tion of whic h co nsists of ende mic, relativel y rare ,
93
TYPE P OOL# DEPTH
n V
v V V
V V V
If H H V
H H H H
H H H
j H~ H H H H If
H H H H H H H If
H If rI H If H H H H H H H H H H H H
H
[~
JP3 l PA2 l PS T UITI2 IU S
D D D
MPI
M
M
MPIMS DPJDIO DPPotholes DPDaies Lak TH42a TH3 TH20 MP2 CR4 CRf2 CR2
D
;vJ
D
0
D
D
M
M
ALI CRr , TH,\
S
vs
\=' R 1 (\~J
oS
"Ne!!
:::
SC2
;:
.~C~
S
CRli HC IO TROFF
T Rl? TR24 HC9 T Rl 6 CR?4 TR? TR IS CC2 THl 4 THl] TH33 RB16 TH40 THtl? RBD RB I3 CC4 T H41 CRl 4A WCI S WC18
S
S
ALA RBI S RB3 RB20 RB4 CCS FR I FR7 FR6
_ _ _1
S
VS
S
S
VS
S
SCS
1
2
>-
~s l J
S S
S
S _ _ _ _ _--I~
S
S _ _ _ _ _ _--'l - M M
S S S
S S S
S
S
M
S
S
S
S
---_--I~-
3
M
S
S
S
D
S
I
vs
S
I 0.0
0.5
1.0
Figure 4. Similarity dendrogram of pool species assemblages. Euclidean distance, average linkage method. C =claypan, Y =: volcanic mudflow, If =hardpan, NT =: northern transmontune, D =deep. M =: medium, S =shallow. VS =very shallow. The first ( W O letters of each pool code represent the sue name. Nodes J. 2. and 3 correspo nd to solutes, depth, and elevation/biogeographic province, respe ctively. • MP2 appears to be all artifact of pipeline construction rather than a naturally occurring pool. Because of the more recent origin of th is pool relative to the other volcanic mudftow pools, MP2 may no: yet be fully colonized. Alternatively. this artificial pool may not provide the same habitat quality found in the natural pools or the region.
94
11
r 2= .00007
?::'
'i: .iuicoulho' at l
o
o
o
o
o
o
o
o
o
o
o
o
o
Cyz icUJ cufijomicuJ
o
Number uf 50inicaudatans ner slte
I
,=-a~i£~'!.d'la tl )
LYllce''''_0!! chY/lrils Number orlaevlcaud atans ner site
o
1-_ _ 1_ _ 1_ _ 1
I
~~ I--j _ _
o I
I +
+
+
o
1
I
I
I
o
o
I
I
o
+
o I
o
I
o
._ _
o
C la d uce r a ( 1M)
I ~ -
I I 5 3
-- -
2
I ~ -11---.--"
I
-~-I - -I
\
:
\
+ +
-
- 1-
--+---'--1
- -' --'-- ~ -- i--- , - -,--, --
+
+ -+ 1- +
+
+
+
+
+
+
7
+
+
+ +
+
+
+
+
+ '2
- ·-----R I
I
12
x
+
,.
I
+
+
-~
+
X
x
+ +
17
:=h
_ _ _,__+ _I.~_L...!.__, 4
2
+
_
H=i + 4
+ 3
Prob. Undcs .s-prob ably undescribed
Poss Undes.=possibl y undescribed ISl C'A Recordvflrst recorded occurrence: oflhis species from the stale of Califomi a NT:::::. northern rransn-ontane vernal pools VM ~ northern Sacramen to Valley volcanic mud now vernal pools l l = Sac ramento Valley hardpan vernal pools C"" Sacramento v alley claypan vernal pools FR=Fnll River M!'=M. nlon Plain Dl'
......L'.L
+ +
+
+ +
~
5
14
14
1 2 28
7
13
7
•
+
+
+
-
X
+
+
..
X
X
+
+
+
'i
~
+
•
+
X
X X
I
5
+
X X
I 7 I 10 J 2 5 3
H en
+
I
-
Til
.;
X
7
-
Tn
-
X
X
4
ueterocyp ris of[ u. , orundarus
CC
X X
(Hererocy? ,!., sP.cL. Heteroc voris aff.ll. ca rohnen sis
RU
+
X
X
(,
5 2 3 I I II 5 II
TU
+
(24
Brad ieyuro ndena rericltlata ? [Ca ndona .' p.) Candono caudota
Dr
+
I 7 2 9 4 I 5 I 9 I 1
Diacycfap:' navlI.' ICv! Dsacyclops sp. A (Cy) Eucyc/ops .regans t9:!.
Mlcrocyclop'
X
12 I I 5 J I
fAcanthncy cJop.t carolin ian/Is (Cyl
MP
11
g
19
Ig
+ 5 9
0
7
2
21
17
5
1 11
.
\0 -..I
98 30
a.
r2::0.584
•
• •
25 til
Pt.ln n. IesL.k, 111 -10 r othoies
+
+
+
+
+
+
_._-
....
.....
+
......
.. - . . ..... ..... ............... - - _ .. ..
T Oscln Dnlln IT l2 TUS
+
Rn lJ
+
Rtd muff Rnt6 ROts
+
o
R020
+
+
+
+
+
+
Bryocamptuswashtngtonensu (110) Co nthocamptu. ,o~rtcoA "i (lfa) Canrhocalnpfu, 'p. A (Ha) Olneyc/ol" era ss/cawl. vor. broc hycercus rCy) Oiacyc/ops novu. rCy) Dtocyctops sp. A (Cy) c/tcyclop., elegans (Cy) F/"porod/aplomu. codacus (C o) H.sp.rodlaplomus e ,... ,1 (Ca) II.spcrodlaplomus birsutn (Cn)
+
+ +
+
+
-I
+
+
+
+
+ +
+
+
+
+
+
+
+
+
+
+
+ +
+ +
+
l~plodlap'unrJ>p. II (Cy) E;lIcyc/op> ,ltgalU (Cy) H.~iapIOmas cadacv: (Co) Iluprrodlaplomus etsen! (Ca) Hupc-odJaptom •• ltlrsulus (Ca) ,lLprodlaplom u. Iyn'rlll (Co) Mlcrccyc/ops ru bellus (Cy) 1S1i.,oJ laJ]lo", u. pallldu. (Co) N... ber or ton.nod, ner 0001 Oslruodl (14\ BrodIrystrrmdrs lo ret tcukn a? (CoM otta m} Co..J01tO cavdato Candono if. C. stognalts CoMona n. sp. C Candona n. sp. D CoM ono sp. II Cyprldop.l. vidva CJ7JrU S"hK/ohroa f;ncyprls sp. II E;. Wls s./. sf'. II E:uqprls s.l. p . B CucyprU vlrens m,dla (H'/rtYx ypr l. p.) /ltlrrrx:ypr1s off. If. caroltnensts I HtI,rocwrls off. II. ronmdatu s 1f",rocyprlJ IlIo>rtgnJrlU (LimnocyI~'"
d-
.....-
--.
d ...fc R.d BlutT
-
_....
11..
- ~.
s
- listed alnhabeticall
~-._-
Coyot. IcJUMt.J
h - - uth . ---
b- . .
- - ~-
Ane ndal. J '!>to. Pnlrl e A-I A-A J-3 J-A1 J ·S 294S2 687 31 259 804 118 21382 237S83 0.13-1J.25 0.08-0.20 0.11·0.22 0.12-1J.l5 0.10-1J.20 0.40-1).60 0.50-1).80 0.40·0.55 65-16Q 110-340 30-40 2G-30 3G-SO 30 4G-70 8G-250 44-48 40-44 44-60 46-98 48-100 6Q·140 56-64 36-58 50-70 60-80 60-100 60-70 95-120 180-520 130·340 25~80 6.4.7.2 8.G-8,4 6.2-7.8 6.0-7.4 7.2·7.6 6.2-8.0 7.6-7.8 8.G-8.9 20-24 13.5-)2 20·23 20.5·24 13.2-21 17-21 18-21 11-33
+
+
+
0
I
I
I
0
0
+
+
0
0
0
0
0
0
0
I
I
0
0
0
0
0
0
0
+ I
+
0
I
+ I
0
0
0
0
0
0
0
0
0
0
0
+
+ +
+
+
+ +
+
+
+ + 4
3
+
+
+
+
I
I
I
I
0
0
0
0
0
0
0
sp. A
/..eplcJuMt.J pacMrdl N.",""r ornolorlratans eer nool Spinlcauda" (I) Cyx tcus ca1!fornl= N... ""r or mlnlcaudll.., ee r r-oot
Lanlt1lud... (I) LY"
