Reference: Biol. Bull. 202: 43-52. (February 2002) Morphological Variations Among Larval-Postlarval Intermediates Produced by Eyestalk Ablation in the Snapping Shrimp Alpheus heterochaelis Say PAUL S. GROSS1 * AND ROBERT E. KNOWLTON2 ^Department ofBiochemistry and Molecular Biology, Medical University ofSouth Carolina. Charleston. South Carolina 2V425; and Department ofBiological Sciences, George Washington Universitv, Washington, DC 20052 Abstract. The eyestalk of many crustaceans contains the instars, culminating in a distinct change (metamorphosis) to X-organ, the presumptive site of production and release of a "postlarva" or"megalops" that more closely resemblesthe many protein and peptide hormones into the hemolymph. adult form (Williamson. 1982: Felder et ai, 1985). The Removal of the eyestalk deprives the animal of these hor- metamorphic molt is usually ofsufficient magnitude for the mones and is known to affect many physiological processes postlarva to assume the life habit of the adult, but some in the adult and developing larva. In the snapping shrimp larval features may be retained until the next molt, to a Alpheus heterochaelis Say, eyestalk ablation performed juvenile form that is similar to the adult except for size. early in larval development has profound effects on mor- Morphogenesis and molting history can be affected by phogenesis, causing the appearance ofsupernumerary larval environmental factors such as diet, temperature, and photo- stages, accompanied by retardation and even complete ar- period (reviewed in Knowlton, 1974), and by hormonal rest of morphogenesis. In this study, we examined the cues. An "X-organ-sinus gland complex," located in each effects on morphogenesis of bilateral eyestalk removal at of the paired eyestalks just medial to each optic complex, carefully controlled intervals. We found that the crucial has long beenknown as asourceofmany hormones thatcan point for this operation the point at which the animal affect various physiological processes either directly or attains the ability to metamorphose fully isjust before the indirectly in adult decapods (reviewed in Carlisle and onset ofecdysis to the third instar. Additionally, the pattern Knowles, 1959; Passano. 1960, 1961; Welsh, 1960; Kelly, of development and morphogenesis among body segments 1967; Jenkin, 1970; Kleinholz. 1976; Cooke and Sullivan, follows adiscernible double gradient pattern along the ante- 1982; and Fingerman, 1987). This complex also plays arole rior-posterior axis in which the extremities of the animal in control oflarval development (reviewed in Christiansen, attain the potential for morphogenetic advance prior to the 1988; Charmantier and Charmantier, 1998). Eyestalk abla- central thorax. This pattern of morphogenesis, punctuated tion has been shown to produce supernumerary larval by ecdysis, is a continuous rather than a stepwise or com- stages, larval-postlarval intermediates, or both in various partmentalized phenomenon. crustaceans such as Rhithropanopeus harrisii (Costlow. 1966) and Homanis americanm (Charmantier et ai, 1985; Introduction Charmantier and Aiken, 1987). Larvae ofthe snapping shrimpAlpheusheterochaelis Say In decapod Crustacea, larval morphogenesis generally from coastal North Carolina (described by Knowlton, 1973) consists of small increments of structural change (from the exhibit some unique developmental features. The larval initial hatching morphology) manifested as several zoeal phase is abbreviated, lasting only 4-5 days when reared at 22-25C and consisting of three larval (zoeal) and one Received 12 June 2001; accepted 27 November 2001 postlarval stage. The larvae hatch with a large amount of *To whom correspondence should be addressed. E-mail: grossp@ stored yolk and oil. which they apparently use as the sole musc.edu source of energy throughout the larval period. Exogenous 43 44 P. S. GROSS AND R. E. KNOWLTON feeding does not begin until the postlarval stage when the B. an antifungal agent, plus 50 /xg/ml each ofthe antibiotics mouthpurts develop into functional feeding organs (Gross streptomycin sulfate and ampicillin (ICN Biochemicals. and Knowlton, 1991 ). Costa Mesa, CA). Larvae were maintained in 10 ml of Eyestulkless A. heterochaelis larvae have been observed processed seawater, changed daily, in an incubator at a to form larval-postlarval intermediates consistently as long constant 25' C to provide optimum development and a as eyestalk removal is performed prior to the midpoint of "standard" 4-day larval development pattern for control larval Stage III (Knowlton, 1988, 1994). An array of dif- animals (Knowlton. 1973). When animals reached the post- ferent morphologies was discerned, depending on the time larval andjuvenile condition, they were fed freshly hatched (during the second and third larval instars) that the eyestalk Artemia salina (brine shrimp) nauplii ml lihitinn. General extirpation was performed. Knowlton ( 1994) described four laboratory procedures are given in Gross and Knowlton types of larval intermediates, designated stages IVA (1997, 1999). through IVD, each possessing more developmentally ad- vanced features and occurring between stage III and the Protocol ofoperations postlarva. In his study, operations were performed once a day at the same time of day over the 4- to 5-day larval Immediately after all larvae in a particular brood had period. In the present study, this protocol was refined by molted to larval stage II (about 2-3 h post-hatching), a winecrreeaspienrgfotrhmeednumtobesrevoefnti(im.ee.,poaitnt2s aht. wahtic2h0 ohp,eraantdionast swuabssestubojfecttheedbtroooedy,estdaelskigenxattierpdatti2on(.foAr 2sehcopnodstp-ohpautlcahtiinogn) subsequent 10-h intervals over the duration of larval devel- was ablated 18 h after t2 and designated t-,,,. Eyestalk abla- ompaminetnati)n,inugtitlihzeinagnimmoarles oavneirmaslesverataleaincshtartsim(emoplotisn)t.,Tawnod 7ti0onhs, wweirteh tghreonuppserbfeoirnmgeddesatig1n0a-thedinatcecrovradlisngalftye.rIt-n,,,auppret-o series of experiments, each involving larvae hatched from vious study, it was found that larvae ablated at 10 h post- eggs borne by four different females, were executed. Spe- hatching (t,()) closely resembled those ofthe t-,,, group (data cwihfeicthoebrjemcotirvpehsogoefnethseisseoecxcpuerrsimienntdisscwreertee s(t1e)pstoorasicserctoani-n anboltastihoonwsn;atGtr2oasnsdatn20doKcncouwrlrteodnd,ur1i9n9g1S).taTygpeicIIalalnyd,aelylesottahlekr treisnuulotussinanddev(e2l)optmoeensttaalblairsrhesitf(eexittihrepractoiomnplaettevaorriodueslatyiemde)s. aibmlmaetdiioantsedluyriafntgerlatrhvealmoslttagteoIsItI,agweitIIhI.t,A nocucmubrerrinogfallamrovsate and whether there is a critical threshold time after which in each brood were retained as unoperated controls. The development to the postlarval stage is determined. control group and each group of larvae designated for surgery at aparticulartime point had roughly equal numbers of animals. Materials and Methods Collection and maintenance ofadults Analysis ofintermediate morphologies Ovigerous adult female specimens of Alpheus hetero- In the first series of experiments, a total of 274 larvae chaelis were collected periodically during the summer from fourbroods were used. Afterthe final set ofoperations months at three sites ("Duncan's Green," "Research Cove," at t70, each larva was individually monitored as it molted to and Radio Island) in Beaufort, North Carolina. Each shrimp the intermediate form (stage IV of Knowlton, 1994) or was maintained individually in a 2-1 aquarium containing postlarva (in controls). Data pertaining to size differences 1.5 I ofaerated (by pump and stone) seawaterand an oyster are given in Gross and Knowlton (1999). At this time, the shell "house" for the shrimp to hide under. The animals animal was preserved in 70% ethanol (EtOH). Each larva were maintained on a cycle of 14 h daylight and 10 h was then cleared in 5% KOH. stained using carmine-borax/ darkness and were fed Tetramin large flake fish food ad 35% EtOH. and placed in 100% glycerol by serial transfer, lihittnn. Gravid females were assessed daily for brood ma- in preparation for structural examination and measurement turity. To establish the absolute age of any brood, only (Guyer, 1906; Gross, 1995). Glycerin was chosen as the larvae for which the hatching event was actually observed final medium because it maintains the clarity ofthe stained were used. specimen; it is significantly more viscous than water, giving the specimen a greater internal rigidity; and it causes the Reurinu oflarvae otherwise brittle exoskeleton to become pliable, allowing the appendages to be moved and posed easily during anal- Seawater used in all experimental procedures was pre- ysis oflarval characteristics. Both dissecting and compound pared as described in Gross and Knowlton ( 1997). Hatched microscopes were used to analyze morphological featuresof zoea larvae were maintained in UV-treated seawater, which the fourth and fifth instar intermediate forms with respect to was furtherfiltersterilized through a0.45-/im filterto which time ofablation, and were related to analogous morpholog- me following chemicals were added: 50 ng/ml amphotericin ical characters found in control postlarvae and juveniles. MORPHOLOGICAL VARIATIONS IN ALPHEUS In the second series ofexperiments, a separate set of629 Table 1 larvae was used. The procedure was similar to that already Recognition characters ofexpunJeJ inicriiicilnilc morphoxenetic outlined except that the larvae were not sacrificed but were categories {/c/v?/\j allowed to continue development through subsequent in- stars (molts). For consistency, operations on larvae were Form Additional ke\ leatures performed at the same time points. As they developed, live A, Pleopods asetose (or with very short setae) animals were examined for morphological diagnostic char- A, Pleopods setose (=Knowlton's stage IVA) acteristics (along with instar duration and mortality; see A/B Antennal flagellum with 5 segments Gross and Knowlton, 1997). Each larva was assigned a Antennule tlagella as in stage IVA substage or "form" designation based on the system pro- B Antennal rlagellum with 9-16 segments, protopod with basicerite aplossoedrebcyorKdneodwalfttoerne(a1c9h94)m.olTtiwnog:be(h1a)vithoeraolricehnatraatcitoenrsofwetrhee B/C AEnxtosetpynolndoucslearoirfteipne(nr=eerKionaponowddlsotuot1n^e'-r1sfssletatagogeselelIuVmBslegmented, peduncle with larva dorsal side up (walking posture), consistent with an Exopod ofpereiopod 4 shortened adult benthic life habit, or dorsal side down (swimming C, Telson lacks pairofdorsal spines behavior), exhibited by pelagic larva; and (2) consumption C2 Exopods ofpereiopods 1 + shorterbut setose othfeAnsteommiaachs.alina nauplii. as evidenced by food (opacity) in C/D EEnxdo(=oppoKodndsoswolfotfopnepr'eesrieoisptooapgdoesdsI1V-1C3-)5sewtiotshemaonrdeschloarwt-elniekde dachls Exopod ofpereiopod 4 asetose. nearly vestigial Results D Exopods ofpereiopods 1-2 shorterbut setose Exopods ofpereiopods 34 asetose. nearly vestigial Analysis offourth instar (stage IV intermediate) (=Knowlton's stage IVD) morphology D/PL Exopod ofpereiopod 1 setose, nearly vestigial Exopods ofpereiopods 2^J asetose. vestigial The range ofintermediates between stage III (third instar) E Exopods ofpereiopods I J asetose. vestigial (as in postlarva) and the postlarva (normally fourth instar) produced by bi- Telson with two pairs ofspines on dorsal surface (as injuvenile) lateral eyestalk extirpation in these experiments is consis- F Exopods ofpereiopods 1-3 vestigial (absent on 4) tent with the four broad categories (i.e.. IVA-IVD) estab- G TEexlospoondswiotfhptewroeipoapiordssof1-s2pivneesstiognialdo(rasbaslensturofnace3 (aansdi4n)juvenile) lished by Knowlton ( 1994), but upon closer examination of Telson with two pairs ofspines on dorsal surface (as injuvenile) individual specimen morphologies, some modifications of these broad categories are indicated. With more frequent ablation and monitoring. 10 forms were discerned, whose but development in the thoracic region is notyet seen. Much major structural features are summarized in Table 1. variation in form was noted among IVB larvae, especially Stage IVA. as described by Knowlton (1994). differs with regard to the development ofhead appendages. A few from stage III (see Knowlton. 1973) only with respect to larvae had the full (i.e., normal postlarval) complement of4 abdominal development, and two distinct subtypes of the segments on the outer flagellum ofthe antennule, but typi- IVA form were discerned. In general. Stage IVA, is typified cally only 1-2 segments were present. Conversely, the inner by very little morphological change. Stage IVA, animals flagellum of the antennule usually was postlarval in form were "more postlarval" with respect to the tail region (e.g., (possessing 3 segments), but a few larvae had 1-2 segments formation of uropodal endopodite and acquisition of the added. A wide range ofantennal flagellum segments ( 1-22) postlarval number of terminal setae on the telson). How- was displayed by animals in this form, but even when a full ever, the telson of each IVA, larva retained its stage III complement of segments was present (22-26), the overall shape, with the uropodal exopodite being underdeveloped, length of the antennal flagellum was significantly different lacking both the spine located on the outer margin and the (F ---- 77.30. P < 0.05) than in the normal postlarva. distal-mostjoint, and the endopodite being smallerthan that Although the third and final segment of the antennular ofa normal postlarva in most cases (69.6% of those exam- peduncle was usually present, the nasicerite was rarely ined). Stage IVA2 animals often possessed pleopods about added and the statocyst never found among stage IVB normal postlarval size and bearing long plumose setae (as in animals. This formwas mostoften produced wheneyestalks the postlarva); less often (in 31.6% ofthose examined), the wereremoved late in stage II (t,(1) andearly in stage III (t, ), uropodal endopodite was reduced. The IVA, form was but many other intermediate mils were also produced as a encounteredexclusively among larvae ablatedatt2, whereas result of eyestalk ablation during this period (Table 2). the IVA2 form was encountered primarily among t:il larvae Some larvae, primarily in (he t:ogroup IVA/B. showed very (Table 2). little advance over the IVA2 form but did have 1-2 seg- Stage IVB, as described by Knowlton (1994), incorpo- ments added to the antennal flagellum (Table 1 ). The A/B rates the abdominal development seen in the IVA-, form designation indicates that these larvae were not quite equiv- with an array ofmorphogenetic advances in the headregion. alent to either the 1VA, or IVB forms. 46 P. S. GROSS AND R. E. KNOWLTON Table 2 Distribution <>/fourth inshirformsateach operative timepoint (percentoftotalsurvivorsperoperative limepoint) Fourth instar form MORPHOLOGICAL VARIATIONS IN Al.PHEUS 47 26). Some specimens possessed 3-segmented inner and stage VA, since another molt had occurred. However, most outer antennular flagella (rather than the usual 4 segments), (55.5%) IVA, larvae exhibited slight morphological ad- but the peduncle was always 3-segmented, with basicerite vancement (i.e., they were morphologically identical to and statocyst fully developed. These larvae exhibited com- stage IVA2) and were designated stage VA2. A significant plete transformation of pereiopods 3-5 to postlarval form. proportion of animals (39.7', ) advanced to either stage Theendopodofpereiopod 1 was morphologically postlarval VA/B or VB (equivalent to IVA/B and IVB, respectively). in this stage as well, always being composed of the full All but one of the t2 animals died during this instar, with complement of 6 segments, with the elbow-like joint be- only a single stage VB larva surviving to molt unchanged to tween merus and carpus, and completion of the joint be- asixth instar. A similartrendofdevelopmental arrest is seen tween the propodus and dactyl to form a functional chela. among larvae ablated at t20: 75.8% of animals exhibiting Theendopod ofpereiopod 2 still showed some variability in stage IVAi, IVA/B, or IVB morphology at the fourth instar segment number (9-10), but was functionally complete, (Table 2) did not advance morphologically beyond stage with an articulated chela and an elbow-like joint between VB at the fifth instar (Table 3). However, three larvae from the merus and first carpal article. Exopods of pereiopods 3 the t2l) group molted to postlarval form, surviving to molt to and 4 were usually devoid of setation and nub-like (Table a morphologically normal juvenile by the sixth instar. It is e1x),opasodisn wtheerepolsotnlgarevra;thiannstohamteofspaecniomremnasl, phooswtelavrevra,. tEhxeos-e isnptaenrneistnigngthtehagtatphrbeeettw-,e,,elnarfvoaremhDadamnodrpthhoeljougviecnaille.feTathuersees pods ofpereiopods 1 and 2 were not functional but had not three larvae (designated VD/PL/J) possessed setose exopods attained the naked nub-like appearance of the postlarva. on pereiopods 1 and 2 (as in form D) and had a telson Stage IVD animals typically walked ventral side down, like morphology (i.e., two pairs ofdorsal spines) consistent with the postlarva, consistent with the presence of a fully de- that of a juvenile. cvaeplaobpleed oofrgfaenedoinfgba(allasnocelik(estatthoecypsots)t.laTrhvae)y, wasereeviudseunacleldy insMtaarnsyfo(r4m5e.d0%a)woifdethaerlraaryvoaeflaabrlvaatle-dpoastttl,ar,vwalhiicnhtearsmefdoiuartteh by the presence of Anemia sulina nauplii remains in the forms (Table 2), molted to the juvenile form at the next stomach. A few animals (C/D in Table 2) ablated at t30 and t40 tmhoeltf;ifntihneinsatnairma(lTsab(l1e5.3)0.%)Arseisgenmibfilceadntnonrummablerpoostflaarnviamealast possessed setose exopods on pereiopods 1-3. Similarly, a from this group exhibited characteristics, noted under E in significant proportion ofthose larvae ablated at t50-t70 pos- Table 1, that were intermediate between the normal post- Tsehsesseedlsaertvaateioanppoenalryeodntothbeeesximoiploadrtooftpheerpeoisotploadrva1 (inD/oPtLh)e.r larva andjuvenile. Some larvae ablated at t30 were found to respects, except that the second pereiopodal endopod occa- ppoosdsensusbstheonjuvpeenrielieoptoedlsson1-m3o,rpohroolnogy1-w2itohnlvyest(irgeimaalineixnog- sionally had 9 segments rather than 10. In summary, detailed examination of ablated larvae pereiopods being uniramous); these intermediate forms from each time point yielded in essence a continuum of were designated F and G, respectively (Table 3). Of the stage IV morphologies between larval stage III and the broad spectrum of intermediate stage IV types noted among normal postlarva, and larvae ablated at t30 and t40 showed larvae ablated at t40 (Table 2), the majority (70.5%) molted the greatest variability in intermediate form morphology to juvenile at the fifth instar, but most of the others (E in (Table 2). Table 3) retained all pereiopod exopod rudiments onpereio- pods 1 t. All larvae from the t50-t7() groups molted to the juvenile form at the fifth instar. Analysis offifth instar morphology All animals surviving to the sixth instar, which attained Intermediate morphologies encountered at the fifth instar some semblance ofpostlarval morphology at the fifth instar, (i.e., after four molts post-hatching), which is normally the were found to exhibit morphology consistent with classifi- molt to ajuvenile form, are enumerated in Table 3 with the cation as juveniles. Comparing the distribution of larval same nomenclature used to describe fourth instar forms. A intermediates found at the fifth (vs. fourth) instar, larvae stage V designation identifies fifth instar intermediates, ablated priortothe molt tostage II! (t2 and t20) appeared, for which are still morphologically intermediate between larval the most part, to be incapable of completing development stage III and postlarva, while E, F, and G designations are (metamorphosis). Conversely, those ablated after the molt reserved for individuals with characteristics intermediate to stage III (t, through 170) completed metamorphosis by between the postlarva (PL) and juvenile (J). the sixth instar. However, animals ablated at t, (and to At the fifth instar, some larvae continued to exhibit a lag some extent at t40) showed a definite lag in morphogenesis in morphological advancement. Some larvae ablated at t, relative to larvae ablated at latertimes and were more prone (IVA, in Table 2) did not exhibit any structural change after to appear as postlarval-juvenile intermediates during the the molt to the fifth instar (Table 3) and were designated fifth instar. 48 P. S. GROSS AND R. E. KNOWLTON Table 3 Distribution offifth instarforms at each operative timepoint (percentoftotal xnrvivoraperoperative lime point) MORPHOLOGICAL VARIATIONS IN ALPHEVS 49 phological changes among larvae ablated during stage II individuals showing exactly the same changes at each molt; (before attaining the IVA or IVB form), our more detailed or "compartmentalized" (Fraser, 1936), with individuals analysis showed that by the fifth instarmost larvaeexhibited exhibiting morphological variation within a limited range some change in morphology relative to their stage IV coun- (compartment). Indeed, they provide ample evidence that terparts. The majority oft-, animals advanced from form A development in these crustaceans is continuous a gradual , to the A-,. A/B, or B form, and most t-> animals surviving to process in which different body segments attain the poten- a fifth instar also advanced slightly in form. Thus it appears tial for differentiation into their adult morphology at differ- that even though metamorphosis is blocked by ablution ent times. Removal of the eyestalk at intervals makes this before the third instar, morphogenesis is not completely aspect ofgradual morphogenetic change more apparent and arrested. Even among larvae that can complete metamor- shows that molting and morphogenesis are concurrent but phosis to the juvenile form, the effects ofeyestalk ablation independently controlled processes. may only be overcome over multiple instars and elongation of the larval period (Gross and Knowlton, 1999). On the The axis ofmorphogenetic change other hand, even though t3n and t4(, larvae were able to com- plete metamorphosis to a morphologically normaljuvenile by Eyestalk ablation at various times interrupts the normal the sixth instar, there was a definite lag in morphogenesis. process ofmorphogenesis at different points along the mor- Form D animals are substantially different and more phogenetic continuum (described above). Differentiation advanced than those of form C since they possess fully appears to take place at different rates in each segment in a functional (articulating) thoracic endopods (as in the post- "double gradient" pattern. This is well evidenced by the larva) and are able to feed, orient themselves ventral side descriptive data collected from groups of animals at each down, and walk. It is apparent that at about 35-40 h ablation time point. The ability to differentiate at the next post-hatching the third instar larva attains the ability to molt is attained in the tail region first, with uropodal endo- make, after molting to the fourth instar, a more "complete" pods and pleopods gaining the ability to differentiate to morphological conversion from a planktonic larval exis- postlarval form (and size) as early as 2 and 20 h post- tence to a more adult benthic existence. Ablation after about hatching, respectively. Under eyestalk control, ability to 35 h post-hatching more frequently resulted in the stage differentiate is attained in the head region slightly later, IVD form or more advanced forms (up to and including since larvae ablated at t, and t,(l begin to exhibit morpho- normal postlarva Table 2). Most t5(, individuals exhibited genetic advances (changes in form and size ofappendages) morphology consistent with the postlarva (at the fourth in the head region at the fourth instar. The segment bearing instar). But morphogenesis was not "finalized" until late in pereiopod 5 gains the ability to metamorphose (i.e., convert the third instar, since larvae with form D features were seen the styliform dactyl to a blunter form and shorten in overall even among t70 animals. length) at the fourth instar at about the same time as the head appendages (around t,(l), reinforcing the idea that The nature' ofthe morphogenetic continuum determination is advancing forward from the tail. Among eyestalkless animals, determination in the middle region Carlisle and Knowles (1959) postulated that factors lo- (thorax) occurs later than in the head and tail region and is cated in the X-organ and released at the sinus gland control attained in exopods of the third and fourth pereiopods well molting and that morphogenesis is astepwise process which before those of the first and second. coincides with the moltcycle andmay even becontrolledby Differentiation in form (e.g.. adding segments to append- it. This does not appear to be the case for A. hetemchaelis ages) precedes the potential for functional changes (e.g., larvae, which, when their eyestalks are ablated at various formation of movable joints). Pereiopod 5 differentiates to times, display a continuum of intermediate morphologies postlarval length and segment number with the t3(, group, independently ofthe molting cycle. Ifmolting and morpho- but functional joints are not commonly found on this ap- genesis were inextricably tied, then it would be reasonable pendage at the fourth instar in these animals, whereas in the to assume that premature molting would either fail to pro- t40 group the fifth pereiopod is usually fully functional by duce any morphogenetic advance or would yield a stepwise the fourth instar. Among larvae ablated at t3(). the third and array ofdevelopmental intermediates. Knowlton ( 1994) ob- fourth pereiopodal endopo.K were equivalent to controls in served a limited array of intermediates in experiments in length and segment numb i. but functional joints on these which eyestalks were ablated only once per day. However, appendages were generally not seen until later (e.g., among by increasing the number of time points for eyestalk abla- t40 or t50 animals). At ihe fourth instar, endopods ofpereio- tion, amorecontinuous arrayofintermediates was produced pods 1 and 2 both become functional among larvae ablated that encompassed many more intermediate morphological around t40 (ort50). They lag behind pereiopods 3-5 in onset forms. Our results argue against developmental patterns in of morphogenesis (reaching postlarval length and segment higher Crustacea being "fixed" (Gurney. 1939). with all number) and attainment of functionality (articulatingjoints). 50 P. S. GROSS AND R. E. KNOWLTON One can hypothesize that, under normal circumstances, certain segments does depend on a certain MH threshold, the ability to metamorphose (at the molt to the fourth instar) then removal ofthe eyestalk (which deprives the developing is attained at different times by different body regions and larva of any further production of this hormone) should even by specific segments within a body region of the A. produce morphogenetic arrest in those segments presum- heterochaelis larva. Factors in the eyestalk either control ably the more central thoracic ones that were below this process directly, or different body segments become threshold at the time of eyestalk removal. This condition receptive at different times to a factoror factors released by should be permanent for those segments below threshold, the eyestalk. Morphogenetic advance is not manifested until since circulating litersofthe morphogenetic factorwould be ecdysis, but the potential for morphogenesis toward the below threshold level at the time ofablation. Clearly this is postlarval form starts early in the second instar, increasing not the case since almost all larval-postlarval intermediates gradually and continuously over the larval period. This produced by means of bilateral eyestalk ablation during potential for morphogenesis (differentiation at the next stage III are capable of metamorphosis to a normal form, molt) is not the same along the posterior-anterior axis but given enough time. MH instead may be visualized as a "double gradient" extending If specific segments became responsive to at differ- from the posterior and anterior ends toward the thorax. The ent times during the morphogenetic continuum, then in trend is that the capacity to differentiate develops in the order for morphogenesis to continue in those segments that abdomen and abdominal appendages (uropods and pleo- are activated later or at a slowerrate, MH would have to be pods) first, followed shortly thereafter by the head and its available in the circulation well after its source was re- MH appendages, and lastly by the thorax. In the thorax, deter- moved. would thus have to be an extraordinarily long- mination occurs earliest in the segments bearing pereiopod lived factor, which is not typical for circulating hormones. MH 5 and maxilliped I, while the last segment to be determined Alternatively. may be released in small amounts during bears the second pereiopod. Thus, the double gradient si- stage II (enough to instigate abdominal morphogenesis even multaneously extends forward from the fifth pereiopod and after eyestalk ablation), but this release is either not ade- backward from the first maxilliped, converging on the sec- quate to activate other body regions or other body regions ond pereiopod. are unresponsive at this stage. After the molt to stage III, MH may be released in a quantity high enough to instigate Possible mechanisms ofhormonal control morphogenesis in all body regions but in a quantity too low tocontrol the timing ofmorphogenesis (since larvae ablated Several modelscould account forthe production, through at t3() exhibit a wide variety of larval-postlarval intermedi- bilateral eyestalk ablation, of larval-postlarval intermediate ates but can eventually attain the normal juvenile form), forms and for the double gradient pattern of gradual deter- especially in the thoracic region. That is, circulating tilersof mination. One explanation is that eyestalk removal deprives MH may control ihe liming of morphogenesis in each the developing larva of a critical regulatory principle or segment, but aclivalion depends on a certain basal concen- MH hormone that causes a given segment to transform from tration of being released. larval to postlarval morphology, and that this potential is In insects, morphogenesis is not controlled by a hormone realized gradually or at different times in different seg- that instigates the process but by a metamorphosis-inhibit- ments. The functional details of these mechanisms are still ing factor (MIF) orjuvenile hormone (JH) that inhibits its a matter of conjecture. onset until a specific time (pupation). This hormone (JH) is The simplest explanation for the appearance of larval- maintained in high tilers during the larval phase and drops postlarval intermediate forms, advanced by Knowlton precipitously at metamorphosis. If the control of metamor- (1994), posits the existence of a morphogenetic hormone phosis were similar in the Crustacea, it seems more likely (MH) that is secreted by the eyestalk (presumably by the that the eyestalk produces some sort of regulatory factor, X-organ-sinus gland complex). This release would neces- which controls morphogenesis ihrough a second hormone. sarily be gradual and begin early in stage II. Imam (1982) Some inferenlial evidence for a cruslacean juvenile hor- noted that certain tissues having a glandular appearance mone or metamorphosis-inhibiting factor does exist. Free- early in stage III differentiate in the larval eyestalk, only to man and Cosllow ( 1983) found lhal whole-body extracls of degenerate in the postlarval eyestalk, surmising that these stage III Rhithropanopeus harrisii larvae inhibited resorp- structures play a secretory role in the control of morpho- lion of Ihe large dorsal spine (characlerislic of Ihe zoea) in genesis. A morphogenetic hormone would have to activate vitro. Further, Laufer and Borst (1988) have advanced Ihe different segments at different times: eithera segment could idea lhat methyl farnesoale (MF), produced in Ihe mandib- respond when the hormone reached a specific (and variable ular organ (MO), functions as a MIF: they established that from segment to segment) concentration threshold in that control ofits secretion is underthe influenceoftissues in the segment, or it could respond only at a specific time in the eyestalk. It has been shown conclusively that the X-organ- morphogenetic continuum. If morphogenetic activation of sinus gland complex secretes a mandibular organ inhibiting MORPHOLOGICAL VARIATIONS IN ALPHEL'S 51 hormone (MOIH), which controls tilers ofMF by inhibiting timing but one lhal affects hormonal aclivation of the re- MO its production within the (Liu and Laufer, 1996; Wain- modeling process itself. wrighl ct ill.. 1996; Liu et ill.. 1997) The results of the present research cannot conclusively distinguish between Knowllon's "MH hypothesis" and the Literature Cited existence ofa M1F, and it may be that both systems exist in wtaonudledm.beKanobwallatnocne b(1e9t9w4e)enasMsIerFteadndthMatHinthsautcchonatroclassethiet AbdeAufd,fmeicLntIi.,sotnPr.aMtaTicaorknoaobctr,amecGth.hiyuYlmehfearroznskeeesnlob,aetreR^.itihCrlhoaaurygvahoetihh/,e.wA.annJe.dmAiiA/a.iivaSecacuglttio.r.,B1aa9ni9nd8iadit-.s onset and timing ofmorphogenesis. But on the basis ofthis Ki'h 50: 73-SI. and other research (e.g., Freeman and Costlow, 1983), it is Abdu, LI., P. Takac, H. Laufer, and A. Sagi. 1998b. Effect of methyl possible that one or more factors, released from the eye- farnesoateonlate larval developmentand metamorphosisintheprawn stalk, inhibit MIF function or release from a second gland, Mliakcereofbfercat'c1hBiiuoml.rBousleln.be1r9g5i:i.1(1D2e-c1a1p4.oda, Palaemonidae): Ajuvenoid- and that different body segments have unique thresholds to Borst,D. W., H. Laufer, M. Landau, E. S. Chang, W. A. Hertz, F. C. MIF tilers. The hypothesis that the eyestalk produces a Baker, and D. A. Schooley. 1987. Methyl farnesoateand its role in regulatory hormone modulating the release of MIF would crustacean reproduction and development. Insect Biochem. 17: 1123- partially explain the double gradient pattern of morphoge- I 127. Carlisle, D. B., and F. Knowles. 1959. Endocrine Control in Crusta- netic potentiation inA. heterochaelis larvae if, as postulated ceans. Cambridge University Press. London. 120 pp. by Borst ct ul. ( 1987), the mandibularorgan is the source of Charmantier,G.,and I). K. Aiken. 1987. Intermediate larval andpost- MIF and that MIF is in fact MF. Simple proximity to this larval stages ofHomarus americaniis H. Milne Edwards, 1837 (Crus- gland could be related to circulating tilers of MIF. The tacea: Decapoda). J. Crustac. Bio/. 7: 525-535. extreme ends of the animal could be subjected to lower Charmantier, G., and M. Charmantier-Daures. 1998. Endocrine and amounts of MIF naturally, by virtue of being the most vneenuerbore.ndRoecprrionde.reDgeuvl.at3i3o:ns27in3-e2m8b7r.yos and larvae ofcrustaceans. In- distant from the mandibular organ, and thus would be re- Charmantier. G., M. Charmantier-Daures, and I). E. Aiken. 1985. leased from inhibition earlier, while the central body seg- Eyestalk involvement in the control of metamorphosis in Homarus ments would be proportionately more affected by MIF and americaniis. C R. Acad. Sci. I'aris 300: 271-276. released from inhibition later as tilers of MIF dropped off Christiansen, M. E. 1988. Hormones in decapod larvae. Pp. 47-68 in under the influence of eyestalk secretion. Release of MIF ASsocpieecttysooffLDoencdaopno,dAC.rAu.stFacienacnhaBmioalnogdy.P.SSy.mpRaoisnibuomw,ofetdhse.ZColoalroegnidcoanl and circulating tilers may be controlled by a factor released Press, Oxford. from the X-organ-sinus gland complex in the intact animal. Cooke, I. M.,and R. E. Sullivan. 1982. Hormones and neurosecretion. In eyestalkless animals, responsiveness to MIF may drop Pp. 205-290 in The Biologv of Crustacea. Vol. 3. D. E. Bliss, ed. over lime (even though circulating tilers may nol). or cir- Academic Press. New York. culating tilers ofMIF may fall even withoul ihe influence of Costdleovwe,loJp.meDn..tJorf.th1e96m6.udTchraeb.efRfheictthorfopeayneospteaulks ehxatririrspiaitiGoonulodn.laGrevna.l ihe faclor or factors released by ihe eyestalk (though more Comp. Endocrinol. 7: 255-274. gradually than in intact animals). This could produce the Felder, D. L., J. W. Martin, and J. W. Goy. 1985. Patterns in early double gradient effect as each segment becomes progres- postlarval development of decapods. Pp. 163-225 in Lan-al Growth. sively less responsive or liters fall below the unique thresh- Crustacean Issues, Vol. 2. A. M. Wenner, ed. A. A. Balkema, Rotter- dam. oldThoef eiadceha tsheagtmeMnFt siseqtuheentaicatlilvye. JH in crustaceans is slill F'inCgreursmtaacn.,BMio.l.179:871.-24.The endocrine mechanisms of crustaceans. J. very much in doubt, but some compelling evidence does Fraser, F. C., 1936. On the development and distribution of young exist. Exposure to MF has been shown to retard metamor- stages ofthe knll (Eiiphausia superba). Discovery Rep. 16: 40-48. phosis in H. americaniis (Borst et cil.. 1987). to retard early Freeman, J. A., and J. D. Costlow. 1983. Endocrine control of spine development in Macrobrachium rosenbergii (Abdu ct ai. epidermis resorption during metamorphosis in crab larvae. RouxArch. De\: Biol. 192: 362-365. 1998a), and to have a juvenilizing effect on Bali/nits cini- Gross,P.S. 1995. Intrinsiccontrol oflarval morphogenesisandmolting pliitrite larvae (Smith et /., 2000). However, as pointed out in the caridean shrimp Alpheus lictcn', haclis Sa\. Ph.D. dissertation, by Abdu et al. (1998b). these effects could be due to George Washington University. Washington. DC. 137 pp. nonspecific loxicily related lo continuous exposure to MF. Gross, P. S., and R. E. Knowlton. 1991. l-'me structure ofeyestalkless These samMeFaulhors (Abdu et ul.. 1998b) present a strong A31l:ph2e3uAs.heterochaelis zoeal-postlar.,:i intermediate forms. Am. Zoo/. case that is the direct hormonal agent controlling mela- Gross, P. S., and R. E. Knowlton. 1997. Effects of timed eyestalk morphosis; the arguments are particularly convincing when ablation on molting in larvae ol the snapping shrimp.Alpheus hetero- coupled with ihe identification of an eyestalk-relaled chaelis Say. Im-eitebr. Repnxl. L>e\: 32: 119-126. MOIH. In relation lo ihese sludies, the present work clearly Gross, P. S., and R. E. Knowllon. 1999. Variation in larval size after indicates thai MIF, be it MF or some other factor, controls eliysesStaayl.kJa.blCartuisotnaci.nlBairovl.ae1o9f:t8he-1s3n.appingshrimp,Alpheusheterochae- metamorphosis directly and thai the relationship between Gurney, R. 1939. LarvaeofDecapodCrustacea. Ray Society. London. morphogenesis and molting is not one that affects molt 306 pp. 52 P. S. GROSS AND R. E. KNOWLTON Guyer, M. F. 1906. P. 221 inAnnualMicrology: PracticalExercisesin Liu, L.. and H. Lauter. 1996. Isolation and characterization of sinus Z<i,'ilnXHi.ilMicro-icclmiaiie. University ofChicago Press. Chicago. jland neuropeptides with both mandibularorgan inhibiting and hyper- Imam, F. S. 1982. Formation, growth, and spatial relationship ofeye- glycemiceffects from the spidercrabLihinia finarffiiiaia.Arch. Insect stalkneuroendocrinc stuctures inAlphctisheterochaelis. Am. Zee/. 22: Hioilicm. Phvsiol. 32: 375-385. 411A. Liu, L., H. Laufer,V.J. Wang,and T. Hayes. 1997. A neurohormone Jenkin, P. M. 197(1. Control of Growth and Metamorphosis, Purl II. regulatingboth methyl farnesoatesynthesisandglucosemetabolismin AnimalHormones:A ComparativeStudy. PergamonPress.Oxford. 310 acrustacean. Biochem. Biophys. Res. Coinnnm. 237: 694-701. PP- Passano, L. M. 1960. Molting and its control. Pp. 473-535 in The Kelly, G. 1967. Endocrine Coordination in Invertebrates: An Introduc- l'h\sioh>x\ofCrustacea, Vol. 1.T. H.Waterman,ed. AcademicPress. tion To Neuroendoerine Regulation In Animals. Pageant Press. New New York. York. 210 pp. Passano, L. M. 1961. The regulation ofcrustacean metamorphosis.Am. Kleinholz, L. H. 1976. Crustacean neurosecretory hormones and phys- Zool. 1: 89-95. iological specificity. Am. Zool. 16: 151-166. Smith, P. A., A. S. Clare, H. H. Rees, M. C. Prescott, G. Wainwright, Knowlton, R. F. 1973. Larval development of the snapping shrimp. and M. C. Thorndyke. 2(1(10. Identification of methyl farnesoate in Alphcus heterochaelis Say. reared in the laboratory. J. Nut. Hist. 7: the cypris larva of the barnacle. Balaam amphitrite, and its role as a 273-306. Knowlton, R. E. 1974. Larval developmental processes and controlling Waijnuwvreingilhet,hoGr.m,onSe..GI.nsWeecbtsBtieorc,heMm.. MCo.l.WiBlikoli.ns3o0n:,8J8.5-S8.9C0h.ung, and factors in Crustacea, with emphasis on Caridea. ThalassiaJugosl. 10: H. H. Rees. 1996. Structure and significance of mandibular organ- 138-158. Knowpolsttolanr,vaRl.sEt.ag1e9s88o.f thEeffescntasppoti'negyesshtarlikmpextAilrpphactuisonhoentelraorcvahlaealnidseSaaryl.y tinihhiobrimtoinnaglhorergmuolnateioinnotfhegcrroawbt.hCaanndcerreppraogduurctuison.iJn.vBoilovle.mCenhtcmi.n m2u7l1:- 12749-12754. Am. Zool. 28: 94A. Knowlton.R.E. 1994. Effectsoflar\alcyestalkextirpationonmorpho- Welsh, J. H. 196(1. Neurohumors and neurosecretion. Pp. 281-311 in genesisandmoltinginthesnappingshrimp,AlplieusheterochaelisSay. The Phvsiolo^v ofCrustacea. Vol. 2. T. H. Waterman, ed. Academic J. E.\p. Zool. 270: 162-174. Press, New York. l.aufer.H..andI).VV.Borst. 1988. JuvenilehormonesinCrustacea.Pp. Williamson,I). I. 1982. Larval morphologyanddiversity. Pp.43-110in 305-313 in Eiulocrinoloi;v ofSelected Invertebrate Types, H Lauter The BioloK\ ofCrustacea. Vol. 2. Embryology, Morpholoifv. andGe- and G. H. Downer, eds. Alan R. Liss, New York. netics, L. G. Abele. ed. Academic Press, New York.