Ichthyological Studies: Reproduction Study

KSC TR 5 1-2, Vol. Ill, Part 2 July 1980

(HAS A -CH-163122) A CONTINUATION OF N80- 289kO BASE-LINE ST5 DIES FOR ENVIRONMENTALLY HOI(ITOR1 NG SPACE TPANSP3RATION SYSTENS AT JOHN P. KElr'N EDY SPACE CENTER. VOLUUE 3, Unclas PA9T 2: (Uaiversity of Central Florida) G3/U5 28167 NAS.4 Contract Report 1631 22

A Continuation of Base-Line Studies for Environmentally Monitoring Space Transportation Systems at John F. Kennedy Space Center

Ichthyological Studies: Sailfin Molly Reproduction Study

National Aeronaut~csand Space Admin~stration John F. Kennedy Space Center ,- - .. ~... .. i I 't

t\ .. . ;!

VOLUME 111: PART 2 OF THE FINAL REPORT TO THE NATIONAL AERONAUTICS AND SPACE ADMINISTHATION JOHN F. KENNEDY SPACE CENTER

A CONTINUATION OF BASE-L IN€ STUDIES FOR Ei.IV1KONMEhTALLY MONITOR IN SPACE TRANSPORTATION SYSTEMS (STS ) A1 JOHN F. KENNEDY SPACE CENTER

CONTFthCT NO. NAS 10-8986

VOLUME 11 I OF IV: PART 2 - ICliTtiYOLOtiICAL STUDIES, Sailfin Molly Reproduction Study PR INCiPAL INVESTIGATOR: F. F. SNELSUN, Jr.

UNIVEESITY OF CENTRAL FLORIDA - P. b. BOX 25000 URLANDU, FLOR IDA 32816

BIOMEDICAL OFF ICE BIOSCIENCE OPERATIONS CODE MD-B JOHN F. KENNEDY SPACE CENTER NASA PREFACE

This docwent is part of a University of Central Florida contract report, "A Continuation of Base-L ine Studies for Environnental 1y Monitoring Space Trdnsportation Systems at John F. Kennedy Space Center." 1

The entire report consists of four vol unles and an executive sulmllary, a1 1 identified as liSC TR 51-2; NASA CR 163122:

Volu~lreI : Terrestrial Comuni ty Anal ysi s Volume 11: Chemical Studies of Rainfall and Soil Analysis Volur~~e111 : Part I--Ichthyological Studies, Ichthyological Survey of tayoonal bters; Part 11--1chthyoloqical Studies, Sail fin Molly Reproduction Study Vol une IV: Part I--Threatened and Endangered Species of the Kennedy Space Center: I4arine Turtle Studies; Part 11--Threatened and tndangered Species of the Kennedy Space Center: Threatened and Endangered birds and Other Threatened and Endangered Forms Executive Sumnary TABLE OF CONTENTS

PAGE

Sai 1fin Molly Reproduction Study Introduction...... 1 Species Synopsis ...... 2 Nethods. Materials dnd Study Area ...... 3 Study Area Description ...... 4 Study Sites...... 5 Station 1 .VABI ...... 5 Stat ion 2 = TONY ...... 7 Station 3 .BCHRD ...... 7 Methods...... 8 Fish Sampling ...... 8 Environmental heasure~lents...... 11 Water Temperature ...... 11 Monthly Water Tanperature...... 11 Salinity ...... 11 Water Level Change ...... I1 Turbidity...... 12 Oissolved Oxygen ...... 12 0rtho.phosphate ...... 12 Nitrate-Nitrogen ...... 12 Phytoplankton Chlorophyll ...... 12 Periphyton Ch\orophyll ...... 12 Labordtory Methods ...... 13 Specimen Examination and Hand1 ing ...... 13 Sex Detern~ination...... 13 Reproductive Condition ...... 14 Counting and Staying Embryos ...... 14 Counting Unfertil ized Eggs and Abnonnal Ei~lbryos.... 15 Measuring...... 16 Data Handling...... 16 Results ...... 17 Molly Reproauction ...... 17 Length of Reproductive Season ...... 17 Frequency of Pregnancy ...... 17 Size of Reproduci ng Femal es ...... 19 Broodsize...... 24 Size Specific Fecundity ...... 24 Hepldcanent Index ...... 31 Environ~~ientalAnalysis ...... 34 Discussion...... 40 Sunmlary ...... 45 Conclusions ...... 47 Acknowledgwnents ...... 48 Literdture Cited...... 49 Appendix Tables ...... 53 LIST OF TABLES

TABLE PAGE

1 Sumnary and comparison of some environmental features of three sailfin molly study sites on Merritt Island, Florida. .... 9 2 The reproductive record of pregnant female moll ies from November samples...... 20 3 Monthly rainfall (in an.) measured at Kennedy Space Center, Florida, during the period 1973 through 1978. The 10-year average is based on the years 1968 through 1977. Data pro- vidcd by NOAA .National Weather Service...... 40

4 tkan monthly air temperatures (OC) measured at Kennedy Space Center, Florida (0 L C Building) during the period January, 1973 to August, 1978. The 12-year mean is for the period April, 1965 to Decenber, 1977. Data provided by NOAA - National Weather Service...... 41 LIST OF FIGURES

FIGURE PAGE

1 MapofMerritt Islandandvicinity, BrevardCounty, Florida showing the locations of three study sites used in the sai 1fin mol ly reproduction study ...... 6 2 The percentage of adult females pregnant in monthly samples of the sailfin molly (Poecil ia lati inna) from three study sites. Females were considered adu1-A- t 1 t ey exceeded 21.9 mm SL at VABI and TOW, and 23.9 mm SL at BCHRD...... 18 3 The mean length (mm SL) of pregnant females and the mean brood size for monthly samples of the sailfin molly (Poecilia latipinna) from the VABI study site...... 21 4 The mean length (mn SL) of pregnant females and the mean brood size for monthly samples of the sailfin molly (Poecilia latipinna) from the TOWY study site ...... 22 5 The mean length (mm SL) of pregnant females and the mean brood size for monthly samples of the sailfin mlly (Poecilia latipinna) from the BCHRD study site...... 23 6 Least-squares 1inear regression 1 ines for brood size on female length calculated from monthly samples of the sail fin molly Poecilia lati inna) collected at the VABI study site in 1977. Lines++ ase on a monthly sample size of less than 10 or where the regression was not statistically significant (p=0.05) are not shown. See Appendix Table 98 for statistical details ....25 7 Least-squares linear regression lines for brood size on female length calculated from monthly samples of the sail fin molly (Poecil ia latipinna) collected at the TOW study site in 1977. Lines based on a monthly sample size of less than 10 or where the regression was not statistically significant (p=0.05) are not shown. See Appendix Table 98 for statistical details. ....26 8 Least-squares linear regression lines for brood size on female length calculated from monthly samples of the sail fin molly Poecil ia latipinna) collected at the BCHRD study site in 1977. LinesC-, ased on a monthly sample size of less than 10 or where the regression was not statistically significant (p=0.05) are not shown. See Appendix Table 98 for statistical details ....27 ? Least-squares linear regression lines for brood size on female length calculated from monthly samples of the sail fin molly (~oecilia lati inna) collected at the VABI study site in 1978. L a=---on a monthly sample site of less than 10 or where the regression was not statistically significant (p=0.05) are not shown. See Appendix Table 99 for statistical details. ....26 LIST OF FIGURES ( Continued)

FIGURE PAGE

10 Least-squares 1i near regression 1i nes for brood size on female length calculated from monthly samples of the sailfin molly (Poecilia lati inna) collected at the TOWY study site in 1978. ~-e 7?- on a monthly sample size of less than 10 or where the regression was not statistically significant (p=0.05) are not shown. See Appendix Table 99 for statistical detail s). ...29 11 Least-squares linear regression 1ines for brood sire on female length calculated from monthly samples of the sai i fin mol ly (Poecilia lati inna) collected at the BCHRD study site in 1978. wea-- on a monthly sample size of less than 10 or where the regressit n was not statistically significant (p=0.05) are not shown. See Appendix Table 99 for statistical details. ....30 12 Least-squares 1inear regression 1ines for monthly and pooled samples of the sailfin molly (Poecil ia lati inna) collected at three study sites in 1977. Ten-r-%mIX a e 100 for statistical details...... 32 13 Least-squares 1inear regression 1i nes for monthly and pooled samples of the sailfin molly (Poecilia lati inna) collected at three study sites in 1977. nend~x. - -%r a e 101 for statistical details...... 33 14 Replacement indexformonthlysamplesof thesailfinmolly (Poecil ia latipinna) from three study sites...... 35 15 Graph of average median surface water temperature, salinity, and water level measured monthly at the VABI sailfin molly study site, October, 1976 to December, 1978. Salinity was measured at the time of fish sampling. The highest water level measured was recorded as zero, and other monthly readings are recorded as cm below zero. The average median water temperature was for the 15-day period preceding the monthly fish sample. Breaks in the graph li nes represent missi ng observations...... 16 Graph of average median surface water temperature, salinity, and water level measured monthly at the TOWY sailfin molly study site, October, 1976 to December, 1978. Salinity was measured at the time of fish sampling. The highest water level measured was recorded as zero, arid other monthly readings are recorded as cm below zero. The average median water temperature was for the 15-day period preceding the monthly fish sample. Breaks in the graph 1ines represent missi ng observations...... LIST OF FIGURCS (Continued)

F IGURE PAGE

17 Graph of average median surface water temperature, salinity, and water level measured monthly at the BCHRD sail fin mol ly study site, October, 1976 to December, 1978. Salinity was measured at the time of fish sampling. The highest water level measured was recorded as zero, and other monthly readings are recorded as cm below zero. The average median water temperature was for the 15-day period preceding the monthly fish sample. Breaks in the graph lines represent missing observations. The circled dot is the median water temperature for September, 1978...... 38 SAILFIN MOLLY REPRODUCTION STUDY

Int roduct ion

In recent years, ecologists have emphasized that deleterious chdnges in the envirorunent of a,? organism do not always result in overt responses such as mortality or mass exodus from an affected area. In Inany cases, the organism responds in more subtle ways, through changes in behavior patterns, physiological nlechani sms, denlographic parameters, or niche requirements . Such subtle changes might go unnoticed by the casual observer, but may have ii~si- dious consequences for the species in the long term.

Certainly, one of the best studied and nust infatilous examples of the matter in question deals with the recent world-wide decline in populations of many species of raptorial and fish-eating birds (see Stickel, 1973, for a review). The accumulation of residues of DDT and other persistent chlorinated hydrocarbon canpounds in the envi ronnlents seems to have re1at i vely 1i ttle overt effect on these birds once they are fledged. As free-living predators, however, they ingest these compounds in their food, and the concentration of residues in the adult often is sufficient to cause insidious alterations in the physiological process (pri~narilycalcium rnetabol isin) of egg shell fornia- tion, resulting in a she1 1 much thinner than normal. When the adult bird attmpts to incubate its eggs, the thin eyg shells are often broken by the weight of the parent, and the embryos die. The consequence of this phenorne- non, and possible related changes in nesting behavior, is a dramatic reduction in the frequency of successful fledging in many populations.

There currently is great interest in developi ny techniques to detect, moni tor, and assay the subl ethal responses of organi sms to environmental alteration. A few examples frorn fishes wi 11 illustrate recent trends. Foster --et al. (1966) studied how the behavior of the flagfish, Jordanella floridae, W~Smdified by sublethal concentrat ions of a1 kylbenzene sul fonate constituent of "hard" detergents. Tkcy demonstrated a fundamental change in feeding behavior after as little as four days exposure. Later, they found a dramatic suppression of egg production with (calculated) concentrations of /US as low as 11.5 ppln in static aquarium bioassays (Foster et al., 1969).

The extreme toxicity of the insecticide Dieldrin to aquatic organisms has been widely recorded. Hdrri nyton and Bidl ingrnayer (19%) documented the devastating effect of an aerial Dieldrin treatment on fishes and invertebrates in a Florida salt a~arsh. More recently, chronic effects of very low Oieldrin concentrations have become evident. Lane and Livingston (1970) demonstrated total rriortal ity in Poecil ia lati inna from bfeldrin treatments of 12 and 6 b. Over hdlf of their experi~nenta-5- fish survived treatments of 1.5 and 0.75 ppb, but the growth arid reproduct ion of the survi vors was affected adversely. Cai rns and Scheier (1964) reported that the sunfi sh Lepomi s gi bbosus, exposed to 1.7 ppb Dieldrin for 12 weeks, had higher oxygen consumption and poorer swirrmir~gability than control fish.

Fina; * y, the sublethal effects of hedvy rnetals on fishes have received considerdbl e at tent ion in recent years. For exarnpl e, Brungs (1969) exposed fdthead III~nnows (P in~ephales pron~elas) to subl ethal concentrat ions of zinc. He showed that the various test concenrrations had no effect on growth, survival, or nlaturatioli, hut d'lrnost total ly inhibited reproduction. Spawni fly frequency was greatly reduced, and the nualber of eggs laid per fernale was drarnatically lower in those fish exposed to zinc. Benoi t and Holcolnbe (1978) later showed that zinc severely reduced egg adhesi verless and incredsed the f ragi1 i ty of the chori on membrane, both factors contributing to unsuccessful reproduction.

The overall objective of this study was to learn as much as possible about the reproduction of field populations of a test fish in the KSC area. These data would constitute a "before" base1 ine on reproductive performance in the selected species. By moni toriny reproduction during and after the initiation of spdce shuttle operations, it was hoped, it wduld be possible to compare "after" reproductive patterns in order ta 'Iurntify any st1 bt1e del er- terious envi ror~rrlental changes.

The sailfin molly, Poecilia lati inna, was selected as the "test" species for several reasons. (l>)n*Ilowed that this species was abundan: on Merritt Island and distributed in all types of habitats and, thus, was "representative" of the Is1and's rather depauparate f ish fauna. (2) A1 though not I~UC~was known abwt the details of reproduction in the species (Hubbs, 1964; Snelson, 1976), it is a inernber of the well-studied falllily Poecil iidae. (3) Because of its srnall size, abundance, ease of capture, and adaptability to the labordtory, the species was convenient to work with loyistically. (4) Finally, the live-bedriny habit makes it possible to gather nuch more detailed infomation on reproduction of the molly than would be possible for an ovipar- ious species.

Spec ies Synopsi s

Ttle sai!f!.~ molly, Poecilia latipinna, is a member of the fish family Poeci 1' idae. The species is distributed in coastal environ~~~erltsal~~lost con- tinuously from South Carol ina to the Yucatan Penisula of Mexico (Rosen and Bailey, 1963). It is a srnall fish, rarely exceedi~agthree inches in total length. It inhdbits a variety of fresh and brackish w.~ter environinents, but is rrrost characteristic of shallow, low-sal ini ty ditches and rilarshes, where it often is abundant.

P. latipinna feeds primarily on detritus and periphytic alyae, and does not oTten take anirnal food (Harrington i.na Harrington, 1961). The rek 9duc- tive biology of the species is sirnilar to other generalized nleli~bers of the family (hllloroso, 1960; Rosen dnd Bailey, 1963; Thibaul t and Schul tz, 1978). Feinal es outnurr~ber 1i1a1es in the adult popul at ions , but the sex ratio arnong neo- nates is approxirnately one-to-one (Snel son and Wetheri nyton, in press). Ma1 es have the anal fin highly modified and transformed into an intrornittent organ, the yonopodi uni, to effect inseriii nation of the femal e. Once inse~ninated , a fe- rl~dlecan store spenll for several months, produci ng several successive broods f rorn one 111ati ng. The sperril are stored in special ized parts of the ovary, where they are nurtured (Jalabert and Billard, 1969).

As a clutch of eggs [natures, they are ferti 1ized and undergo develop~i~ent in their follicles. In some species the developiny enibryo is heavily dependent on the maternal systern for sufficient nutrients to complete development. In other species, including P. lati inna, all or niost of the energy required for complete embryological dTvere op~nent is present in the eyg yolk, and the e~i~bryoprobdbly depencls on the maternal system only for yas, waste, and ion exchange. Ueoendi ng on the species and envi ronmental conditions, embryo1 ogi - cal development may take fro11120-50 days, with mean interbrood intervals for -P. latiyinna rangi ny from 26-36 days (Snelson, 1976). Young are born in an cdvanced stage of development and are given no parentdl care after parturi- t ion. Under favorable envi ronrnental conditions, a second clutch of eggs begins to mature i~nmediatelyafter parturition. At maturity they are fertilized and ernbryol ogical devel op~nent proceeds.

Brood size is positively correlated with female size, both within a species and between species. Small mosquitofish (Gambusia affinis) ar~dsail fin rr~olliesoften have broods of less than five and occasio~ayproduce only one or two young. Large individuals of the two suecies rnav have broods of over 300 anh lo0 young; respectively (Krumhol z, 1948 for -G". affinis; personal observat icns for -P. latipi nna).

The only major studies on the field reproduction of P. lati inna are, tlubbs (1964) and Sheinbaum (1979). Other aspects of the repra$- uctive and population biology of this species haw been treated by Baird (1968, 1974), Grier (1973), Sirnanek (1978), and Snel son and Wetheri rlgton (in press).

Methods, Materi a1 s and Study Area

Study Area Description

A detailed description of the aqudtic habitats on Merritt Island was given by Snelson (197G), and a sumnary is presented here.

Merritt Island originally was bordered along 111ost of its western and northwestern shore by extensi ve bracki sh-water Spart ina ri~arsh. On the interior of the Island there probably were few perrnanent bodies of fresh surface water. A1 though ephemeral runoff creeks and flooded catct~rnent basins lilay have forr~iedduring rainy periods, they would have coristituted re1ati ve! j unir~~or- tant fish habitats.

Two events drastically a1 tered the natural aquatic ecology 011 Merri tt Island. The first was the cons+ructior~in 1958 of earthen dikes (levees) around nearly the entire periphery of the Island, converting S artina and mangrove r~rarshinto a series of shallow "mosqui to-control" impoun%--- nents (Provost, 1959, 1973). The second rnajor a1 teration was the digging of borrow pits and ditches to provide fill for the construction of roads, and later, the install- ations for NASA's Kennedy Space Center in the early 1960's. These borrow dreas dnd ditches created perrnanent surface reservoirs of water on the interior of Merritt Island. In addition, road and other construction altered the natural courses of soroe existing waterways.

The sdl inity of the ditches, ponds, and ilr~poundrnents on Merri tt Island vdries corisiderably, rdnyiny from nearly fresh to levels in excess of 30 ppt, depending on location, water source, and man's activities (see later). The rldture dnd arllount of dquatic vegetation varies according to salinity. 111 low sdl inity waters, -Chard, Ndjas, ,Hu,j/i, Ceratophyl lunl, and Utricularia are ctlaracteristic subrllerged aquatic PI-:?ts, and usually are abundant. Sayi ttaria, Typhd, and Ludwi yia art. ~hardcteristic mary inal enieryents. At higher sal ::!iti es, s~ibr~lergedvegetation is 1ess abundant, usually confined to li~nitedstdnds of -Chdrd, Najas, and Ku Typicdl enleryent 111dnts dre vdrious species of Sf artina drid salt9- to erdnt varieties of Tyuha. In sorue areas, "salt grassesi&- such as Distichlis and Paspalunl may form mats that erlcrodch varying distdnces out into the water. Manyrove trees typically line the bariks and dikes around brackish ditches and irnpoundments. In high sal inity ie~poundrnents, subrrierged aquatic plants are 1irnited or absent, particularly in areds where bottom sediments are flocculent and unconsol idated. In extreme cases, the only vegetation in an inpoundrnent may be scattered clunips of Spart ina --'bakeri apparently remnants persisting from the marsh that existed prior to impoundment.

The rllost characteristic feature of the aquatic habitats on Merri tt Island is their great vari abi 1ity in physico-cherilical features. Because all the waters dre shal low, temperatures fluctuate dramat ical ly, both on a dai ly and annual cycle. Winter temperatures drop 1ow enough to car*se occasional hypotherrrial rnortal ity (Snelson and Bradley, 1978), and oi 'en reach 35°C at the surface in surmner. Furthermore, because there is a pronounced annual pattern of rainfall in central Florida, aquatic habitats on Merritt Island usually undergo a drari~atic floodi ng-dessication cycle. Water levels nonilally are highest in fall and early winter, at the end of the rainy season. During the dry spring period, water levels drop quickly, usually reaching low stages between April iind Jcne. In early sumnier, the initiatior~of tile rainy period drid ~i~an'sassociated activities (see below) again cause water levels to rise. Mdny other physical, chemical, and biological features of the aquatic environnlents fluctuate dramatically, some (e.g., dissolved oxygen) in a pattern associated wit'; tenlperature and water level, and others (e.g., turbidity) seerni nyly independently.

The r~lostdrdlnat lc on-goi ng influence of riian's dcti vities on Merri tt Island's aiiuat ic habitats is a vigorous riiosqui to-abaterllent program carried out by an Fgency of the county yovernment (Provost, 1959, 1973). Among the several activities of the mosquito control agency, the most dramatic for fish ecology is "pumping". In order to inhibit the breeding of sal t-marsh ~nosquitoes(Aedes taeniorhynchus, 4. sollicitans) on exposed bottoln sedin~ents, it is desiramo maintain a rather high and constant water level in impoundments duri ng the breedi ng season. Depending on rainfall, temperature, and other conditions, the irnpoundrnents and ditctles reach niost severe dessication between April and June. Some tircle during this period, usua'lly coinciding with ini ti ati on of heavy summer rains, niosqui to control personnel pump vast quantities of water fror~~bays and ctllnnels directly connected with the Indian River lagoon system into the impoundments. This pumping is effected by means of both portable arid permanently-i nstal led diesel pu~rips. The water level in the irnpoundnients increases abruptly, often reaching rnaxirrial level s within a few days. Lagoon water typically is of much higher salinity than the inipoundriie~~ts,so a dra~natic increase in sal ini ty results. Turbidity a1 so increases substanti dlly auri ng pumping, and inipoundment wdters may remain "cloudy" for weeks after pumping ceases.

Pur~ipingtook place at VABI between May 16 and 25, 1977 and again between Auyust 30 and Septer~iber3 of that year. VABI was not pumped during 1978. Pumpiny did not take place at either of the other two study area during the ti~rleof our study. Overall, the aquatic habitats on Merritt Island are harsh environments for 111ost fishes. Snelson (1976) listed orily 35 fish species inhabiting the fresh and brackish waters on the interior of Merritt Island. The most characteristic groups were the fami 1ies Poeci 1i idae and Cypri nodont idae, both notoriously hardy and to1erant groups.

Study Sites

The terms of the contract proposal st2.~5 that two field populations of moll ies would be sampled nlonthl) for reproduc~ive analysis. tlecauze of prior experience (Snel son, 1976), we planned for the possi bi1 i ty that unforeseen circumstances could terminate study at a locality and, therefore, initiated this study using three, rather than two, study areas.

The fol lowing criteria were considered important in choosi ng study areas: (1) they should be within the security perimeter; (2) they should provide contrasting environn~ents; (3) at least one of the sites should be relatively free from man-induced variations; (4) they should be readily accessible; (5) large populations of moll ies should be present; and (6) the areas should be large enough and contain sufficient deep water to insure that the mlly population would be able to survive through the annual low water period of later winter and spring. In addition, it was considered desirable, but not mandatory, to locate one study station near Launch Complex 39A or 398, in case launch operations should prove to have an adverse effect on the study species.

It was determined early that one study site would be the "VAB In~poundment" used during previous studies (Snelson, 1976). It was felt that this station was acceptable by all selection criteria, and it had the additional advantage of providing continuity in the investigations. An extensive field survey was conducted during October, 1976, to locate additional study stations. Two study populations which met most or a1 1 criteria eventually were selected. The geographic locations of the three study areas are shown in Figure 1. . ey are described below. Station 1 - VAB Impoundrnent (VABI). This is the salile site studied for three years durinq our earlier grant phase and desiqnated as the "VAB ~m~ounhn~ent"(~neison, 1976). This long, finger-1 iie impoundment, dei gnated on some maps as T-37, is located on the northeast side of the State Road 3, iwediately northwest of the Banana Creek bridge (TZZS, R37E, Sec. 7 and T22S, H36L, Sec. 12). Its center lies approximately 1.0 kilor~~eter northwest of the vehicle Assenibly Bui lding (VAB). The impoundn~ent is bordered on its southeastern tip by the Banana Creek dike, along its southwestern border by S.R. 3, jnd along its northeastern side by an elevated railroad bed. The northwestern end is not diked. Estimated surface area is 30 acres, This irnpoundment appears to cornr~cc,-,icate freely with the adjacent inipoundnient northeast of the rai1 road dike, through an open water-I eve1 control cui ver-t.

The water depth is shallow, averaging less than one-half 111eter over {nost of the impoundment. Oeep water is confined to dredged areas. One such hole, circular in outline, is in the center of the impoundnlent. A small dredged ditch and several small holes are located along the southwest side, adjacent to S.H. 3. The major dredged area is the dike-side canal along the northeast edge of the impoundment adjacent to the elevated railroad bed. Under normal conditions, these dee~erareas average 2-2.5 m. Figure 1. Map of krritt Island and vicinity, Brevard County, Florida, showing the locations of three study sites used in the sail fin molly reproduction study. Shal low areas, exposed during parts of the year, are vegetated heavily with a variety of aquatic or semi-aquatic plants such as Syartina (cord .grass), Dist ichli s, and Pa; alum (salt grasses). Deeper areas, particularly where the bottom is relative+ y irnl, support mats of Chara (musk grass) that became thick at certain times of year. Dredge holes withexcessively flocculent bottom sediments usually are not invaded by plants. Except for a few mangrove trees growing near the southeast end, the perimeter of the impoundment is re1at i vely free from overhangi ncj vegetation.

During periods of severe low water, Brevard Mosquito Control District personnel position a portable, diesel-powered pump on the dike between Banana Creek and the impoundment, and pump t~rbid,high-salinity water from the Creek into the impoundment. Salinity and turbidity of water in the impoundment increase drastically as water levels rise. Station 2 - Ton-way Ditch (TONY). To accomodate the activities of the upconring space shuttle program, a runway-landing strip was constructed west of and parallel to S.R. 3, north of Banana Creek. A ton-way, extending from the runway to the VAB, provides a route by which the reusable orbiter transported to the VAB complex for refurbishing. During construction of the tow-way, several mosquito control impoundnlents were disrupted. Impoilndment T-18 was most drastically affected. It was split approximately in half, and the majority of its eastern end was drained or eliminated.

Our TOMY study station is in the disrupted eastern end of old impoundment 1-18 (T22S, R36E, Sec. 12). It is an approximately triangular-shaped area, bordered oy S.R. 3 on the northeast, by the shuttle tow-way on the south, and by a dike on the northwest. Its center is approximately 2 km west by northwest of the VAB. The surface area is 7-10 acres during normal water stages. Shallow niarsh areas less than one-half meter deep are relatively limited, confined primarily to one spot. Deep water is relatively extensive in the long canal adjacent to the tow-way and in a sniall dredged hole in the northeast corner of the area. The shallow zones never dried completely during this study, and always maintained open comunication with the ditch, even during extreme low water.

Shallow 2r.d exposed areas are densely vegetated with a variety of aquatic and semiaquatic plants such as S artina (cord grass), Uistichlis, and Paspalum (salt grasses). The portion which--I usua ly is submerged, particularly where the muck bottom is finn, supports a dense growth of -Chara (r~iusk grass) and Najds. The dredged hole in the northeast ccrner has a deep, organic bottom redinent and supports no vegetation. During the second year of study, plants characteristic of standing fresh water became prevalent along the canal, particularly duckweed (Lernna) and cattail (T ha). There are a few scattered patches of mangrove located along the south-% si e, and the dike on the northwest side supports a moderate stand of Spartina. Otherwise, the perimeter of the area is relatively free of overhanging veqetation. Station 3 - Beach Road Inpoundment (BCHHDL. This study station is located immediately west of Beach Road, and lies less than 0.4 km from the ocean (TZlS, R37E, ~ec.28). Its center is approximately 4 km northwest of the intersection of Beach Road and Saturn Parkway, and 0.8 krn north of launch complex 398. It is bordered by Beach Road on the east and by a series of narrow dikes over the remainder of its periphery. The sampling area consists of two shallow borrow ponds which represent only a small section of an extensive network of interconnected ponds that drain the region adjacent to LC 398. The finger-like basins comprising the stati~nlie parallel, about 15 meters apart, in a southeast-northwest orienta- tion. The eastern basin is the smaller, with an estimated surface area of five acres. The larger western basin has an estimated surface area of ten acres. The two ponds are separated physically along most of their length by a low-elevation bridge of land. They are interconnected at their northern ends by a narrow ditch. A canal approximately 5 meters wide and 30 meters long extends eastward from the southeastern corner of the larger pond, toward the road.

Water depth averages less than one-half meter. Deep water is confined to dredged areas. These consist of the aforementioned ditches and dike-side canals that skirt the entire station, except for a short segnent of the rtorth- east corner of the smaller basin- The deeper areas have an average depth of 1.5 to 2.0 meters.

The salinity is higher at th~s station than at any other study area. Considering the proximity of the ocean, superficial salt deposits and/or sea water intrusion are likely.

The station is nearly free of rooted aquatic vegetation. Aquatic plants, including those tolerant of moderate sal inities, have not establ ished them- selves well over the soft bottom. The substantial turbidity also may act as a deterrent to submerged plant growth. In marginal zones and in scattered open water areas where the bottom is firm, characteristic plants are Chara and Ru ia (widgeon grass). The station's primeter is bordered by an extensive 9%stan of mangrove, but little of it overhangs or grows in the water. The entire area apparently was rechannelized and rediked a short time before initiation of the study, and gives the impression of recent disturbance.

Table 1 sumarizes and compares major quanti tatibe and qua1 itati ve physical, chemical, and biological parameters of the three study sites.

Methods

Fish Samplinq

Each study area was surveyed carefully and 12 permanent sampling sites were selected. These sites were marked with pieces of PVC pipe approximately 20 cm in length and designated by number. An attempt was made to incl ude a? 1 ~nicrohabitat types. Since approximately 300 moll ies per station per month were required to yield an adequate size subsample of adult females, it usually was not necessary to take fish samples fran all twelve sites. Therefore, one site was selected at random and sampled. If the appropriate nunber of fish was not caught, a sxond site was selected at random and sampled. This procedure was continued until approximately 300 moll ies were collected. If this quantity was not obtained from a1 1 twelve possible sites, r~petitive sarnpl ing became necessary. This was accmpl ished by repeating the above procedure.

Sampling began in October, 1976 and continued through December, 1978. Fish samples usually were taken during the first week of every montt~. E TT m 0 . .-- -r XCW ct m-r CJ a EEE 2

I Q, Q CnC WCW mmaJ LLE c, Table 1. Sumnary ti-d comparison of some environmental features ef three sailf ;n aolly study sites on Merri tt fsland, Florid*, (Cont f n,ed;.

VAB I TOWY

Phytopl ankton Chlorophyll (mg/p3) range 0.000 - 21.528 0.411 - 10.908 0.788 - 13.831 mean 3.357 2.886 4.976

Periphyton Chl orophyll (mg/cm2/day) range 0.0002 - 0.0082 0.0010 - 0.0143 0.0006 - 0.0101 mean 0.0017 0.0044 0.0026

Submerged Vegetation Abundant Abundant Scarce to absent

Rainfall About same as TOWY, About same as VAtlI, Less than both VABI and probably close and probably close and TOWY. to figures in to figures in Table 3. Table 3.

Impzct of Dessication Severe ; deep-water Minor ; deep-water Very severe; deep- refugia re1ati vely refugia relatively water refugia very 1imi ted. extensi ve. 1imi ted.

Production Base Probably peri phyt ic Like VABI. Probably plytoplank- algae, macro-algae, ton, detritus and rooted vascular pl ants.

Impacted by Pumping Yes The standard collecting gear employed was a "comnon sense" minnow seine 3.7 m long, 1.2 m deep, with a 4.5 mn bar mesh. Seining was quantified by timing. One seiner or an assistant operated a stop watch. The watch was started when a seine haul began and stopped when the forward progress of the haul ceased. After each seine haul every specimen was preserved in a solution of 10-15 percent formalin. Each site collection was assigned to a separate container, and the time a~rationwas recorded.

Envi ronmental Measurements

The following parameters of the aquatic environment at each study station were measured in the fie1d, concurrently with fish sampl ing. In each case, we tried to utilize a technique combining simp1 icity, re1iabil ity, and appropriate sensitivity, as suggested by Strickland and Parsons (1972) and EPA (1974).

Water Temperature. Water temperature was determined at each site from which a fish sample was taken. It was measured approximately half-way between the water surface and bottom, using a shaded stem thermometer cal ibrated in degrees cent, grade.

Monthly Water Temperature. Surface water tempeature was continously recorded at each study site by a submersible thermograph (Ryan Industries Model G). These instrunents were suspended from and shaded by wood and styrofoam floats. A continous temperature record was scribed on a 15 day chart by a stylus responding to temperature changes. The clock motor was battery driven. The chart was cal ibrated in degrees Fahrenheit (OF) which was converted to degrees Centigrade (OC). From the charts, we determined the temperature characteristics of the 30-day period preceding the fish sample day.

Salinity. Surface salinity was determined in the field, using either an American Optical Company optical refractometer or a Ye1 1ow Springs lnstrunent Model 33 battery-powered, sal inity-conducti vity-temperature meter.

Water Level Change. A water level monitoring gauge consisted of a piece of white PVC pipe approximately three meters in length. The pipe was marked with shallow saw cuts at two cm intevals along its entire length, and was permanently lodged in the substrate at one of the deepest spots in each study basin. The distance between the uppermost mark at the top of the pipe and the water surface was recorded monthly, to the nearest cm.

The following chemical water parameters were determined monthly from water samples taken either concurrently with, the day before, on the day following taking of fish samples. Water samples were taken from the deepest area at each station, using a Kemnerer water sampler. The sampler was operated fron a canoe or small boat to prevent disturbance of the bottom sediment. Usual ly, one 2.3 1 iter Nalgene bottle and two 301) mi BOD bottles of water were collected fron each station. Water samples were kept on ice and in darkness and returned imnediately to the laboratory for analysis. Turbidit . Turbidity was rneasured with a for~l~azin-standardizedHach d1 OA turbidirneter. Results were expressed in NTU ' s (Nephelon~etric Turbidity Units). NTU's are directly comparable to previously reported FTU's (Formazin Turbidity Units) and JTU's (Jackson Turbidity Units).

O issol ved Oxygen. Dissol ved oxygen was determined f rorn sa~nple water in 00 bottles, using a Ye1 low Springs Instrureet~tModel 51A oxygen meter dnd stirring probe. The instrument was calibrated before each use by a Winkler titrdtion determinati~nof dissrlved oxygen in one of a nidtched pair of sae~ples. Results were expressed in parts per million (ppm). fi. pH was determined usi ng an Orion Model 399A pH meter, cal ibrated using a stock buffer solution of pH 7.00.

Ortho-phosphate. Ortho-phosphate concentration in a water sample was detenrlined by usiny the phospho-molybdate standard method (Strickland and Parsons, 1972). Absorbance of the test solution was read with a spectro- photometer (Beckman Mdel 26 ar American Optical Spectronic Model 88) at 800 ndnometers, iind concentragion was calculated in nlg/liter fran pre- deternli ned standard curves.

Nitrate-Nitrogen. Nitrate level was evaluated in water samples by using the brucine sulfate method (EPA, 1974). Absorbance of the test solution at 410 nanometers bas determined spectrophoto~netrically,dnd the concentration in nlg/liter was calculated usiny a previously determined standard curve.

The standing crop of phytoplankton and periphyton was determined by chlorophyll analysis, and these measures were used as indicators of habitat productivity. Phytoplankton Chl orophyll . A 1i ter sample of water was f i1 tered through a Gelman glass filter, and the chlorophyll pigments of the trapped phytoplankton were extracted with acetone. The extracts were held in dark re- frigeration for 24 hours and then read in a spectrophotorneter at 665, and 630 nanometers. Concentrations of chlorphyll a, b, and c were determined by using the Richards and Thompson equations (Strickland and Parsons, 1972). Results were expressed in mg/m3.

Periphyton Chlorophyll. The standing crop of periphyton grown on a standard substrate was detemined at each locality. Floats constructed of wood and styrofoam were anchored in place in deep water at each station. Froan each float, four pieces of nominal 1/4" diameter braided nylon rope, approximately 45 cm lony, were suspended down into the water. The ropes provided an dttdchr~lentsurface for the growth of algae. the ropes were left in place for the approxiiliately 30-day period preceding the date of fish san~pling. At the tiriw of sampl ing, the ropes were rmioved, their precise lengths were measured, and they were transported to the 1aboratory in plastic bottles.

quantification of the attached plant matter was made by the extraction arid aieasurement of chl orophyll pigr~ientsds described above under ptiytopl ankton. Results were expressed in rng/crn2/day. Laboratory Methods. Specimen Examination and Handlin . All fishes collected in the field were preserved in 10-15 percent !orma1 in and returned to the laboratory for examination. The samples were sorted and identified to species, and dl1 individuals were counted and recorded.

Fran the moliies sampled at each station, a subsample of 150 adult fenlales 018 mn~ standard length) was reilioved as follows. The sample was pldct?d ina large bucket containing alcohol. The fluid was swirled and aixed by hand agitation until the specimens appeared to be rando~iilyniixe.! ibi the fluid medium. A small tea strainer then was dipped into the bucket, catching 20 to 30 specimens each dip. This process was continued, with constant agitation, until the number desired in the subsample had been renloved. These fish were used for reproductive analysis. The lower size 1 imit of 18 rm~was chosen because previous work (Snel son, 1976) had shown 18 mn to be the rninimum size at which females become sexually iiuture.

Sex Detenilination. As in a1 1 poeciliids, mature male P. lati inna have the first several rays of the anal fin elaborately traxsfon+ into an intromittent organ, the gonopodium (Cumnings, 1943). We classified males as rrlature when the gonopodium was judged to be fully differentiated on the basis of three features: (1) fleshy palp (also cdlled hood or pre- puce) along anterior (ventral) margin of ray three erilaryed and pendu- lous; (2) fin rays clear and translucent as opposed to cloudy and opaque; and (3) tenninal hook of ray three and retrorse claw of ray five well fornled and clearly projecti ng beyond surrounding cutis (terminology of Rosen and Gordon, 1953). In addition, large mature males usually could be readily identified externally by other secondary sexual features, especially body shape, fin size and shape, and color pattern (see Hoese and Moore, 1977, for color photograph). Inmature males often could be identified externally on the basis of incipient elongation and modifica- t ion of the anal fin rays.

Fish without gonopodi a1 development were sexed by dirwt exaini nation of the yonad. Gravid or pregnant females were identified by the pr ~sence of egys or enbryos in the enlarged ovary, even though they usually could be recognized without dissection by their bulging abdomens. The testis of imnature males is unpaired and elongated, appearing as a white, ribbon-like organ attached along the posterior-dorsal midline of the peritonea1 cavity.

The il~lnlatureovary and testis were easily distinguished in the small- est specimens examined, about 15 mn standard length (SL). In a fish this size the testis is thin and almost transparent. It can be destroyed cas- ily or overlooked in carelessly dissected specimens. The peritoneum coveri t?g the testis is devoid of inel anophores, and the organ appears to have a homogeneous texture. The ovary, in contrast, is a relatively large, conspicuous organ, even in very srnall fish. The shape usually is distinctive, cyl indrical to bluntly oval and tapering abruptly pcster- iorly to the narrow oviduct. Owiny to the presence of primary oocytes, the ovdry appectrs granular in texture under high nlagnification; and the invest iny peritoneurn contai ns numerous me1 anophores. Reproduct ive Condi tion. Femal es were assi gned to one of four reproduc- t ive conditions: (1) with undeveloped ovaries, \2) with developing ova, (3) with nlature ova, or (4) pregnant. Conditions 1 and 2 were readily determined fran the --in situ ovary. The undeveloped ovary contained only whitish, unenlarged ova. Developing ova, on the other hand, were noticeably enlarged and at least sl ightiy golden or yellowish, indicating that yolk formation had begun. Usually there were several eggs in the ovary in a similar developmental stage. However, the presence of a single developing ovum was sufficient to classify the female in this class. When --in situ examination revealed an ovary containing embryos or greatly enldrged eggs, the organ was excised. The ovary was teased apart on the ~nicroscopestage and each individual egg was renloved frav its foll icle and examined. If embryos were present on the eggs, the female was classified as yreynant. Usually, several e~nbryos in a similar stage of development were present, but the presence of a single embryo was sufficient to classify the female as pregnant. If the eggs had reached the maxirnum size but no embryos were visible, they were considered nature.

In using this classification, one occasionally nust make the trouble- some decision as to whether ova are in late stages of development (stage 2 above) or whether they are matiirc (stage 3). This decision was based on size of ova. Although eggs were not neasured. ,he experienced worker developed a good "feel" for the inaxirnum size attained by a mature egg. In addition, a1 1 eggs in d rilature clutch are similar in size. Eggs late in developlnent but not fully mature are smaller than mature eggs, and there usually is a substantial amount of size variability between eggs in the same clutch. It should be pointed out that mature eggs, as classified here, could be either unfertilized or fertilized. There is a pri- od, probably lasting about 48 hours after fertilization, during which enibryonic develooment proceeds but would not be recogni zabl e under the dissect ing rni croscope (Hopper, 1943; Tavol ga, 1949).

Counting and Staging of Embryos. Each mature egs was counted and the total clutch was recorded. When embryos were present, each was counted and the average stage of development for the clutch was determined. Embryos were assigned to stages 1, 2, or 3, corresponding to edrly, intermediate, or advanced phases of development . There is no pub1 ished study on the detailed embryological development of P. latipinna, so our stages were detenni ned arbi trarily after surveyi ngcJthe ent ire cont inuuni of anbryoni c development. Stage 1 begins when an embryo first becomes clearly recognizable. Staye 1 terminates at that point in development when a few melanophores are present on the top of the head and on the anterior part of the back. The caudal fin rdys are not clearly developed, and there is no pigment on the tail. Stage 2 begins when pignient is well developed on the head, down the rnidl ine of the back (usually in two parallel rows of nlel anophores), and down the sides of the caudal peduncle. The caudal rays are clearly formed, and a few black flecks of pigment are present on the fin. Stage 2 ends when the embryo has pigrnent we1 1-developed on all areas of the body. The pigment , however, is present as unifor~nlyscat- tered melanophores showing no pattern, especially on the head dorsum and posteriorly on the sides of the body. Stage 3 beyins when pignient begins to organize into a pattern characteristic of free-living young. This is first noticeable on the head dorsun and on the sides of the body as melanophores tend to organize along the margins of scales. We have assumed that the development of the sail fin mol ly is similar to that of Xi ho horus maculatus (Hooper, 1943; Tavolga and Rugh, 1947; Tavolga, 1a-is basis, we estimate that stage 1 lasts about 5.2 days, stage 2 about 5.7 days, and stage 3 about 5.5 days.

Since the mol ly does not normally superfetate (Turner, 1940a; Scrimshaw, 1944a), all embryos in a brood are in approximately the same stage of developnent. However, the exact timing of fertiliration may vary by as much as 48 hours among eggs in a single brood (tiooper, 1943; Tavolga, 1949). Consequently, all embryos in a brood are never of identical age, and there are occasi-ns when part of a brood will be in late phases of one stage (e.g. late stage 1) while other brood-mates are in early phases of the succeeding stage (e.g. early stage 2). In such cases, all embryos were recorded as being in that stage represented with greatest frequency in the clutch.

Counting Unfertil ized Eggs and Abnormal Embryos, If a single stage 1 embryo was observed among a clutch of eggs, am eggs present were enumer- ated as stage 1 embryos, wheiher or not an embryo was visible on every egg. The assu~nptionwas that all eggs in a clutch were fertilized if one embryo was clearly evident. If fertilization time varies by as much as 48 hours within a clutch, the oldest zygotes might be visible embryos whereas youngest zygotes still could be in indiscernible stages of developnent (Hopper, 1943). In most cases, even in very early stage 1, several embryos can be discerned within a clutch.

After a brood passes into stage 2 development or beyond, it occasionally is apparent that a few eggs within a clutch never were fertilized or, if ferti1 i zed, ceased development before a recognizable embryo was produced. Such eggs have a milky ye1 low color, often bear discolored spots, and usually are atypically hard and brittle. They were recorded as "unfertil ized eggs" among a developing brood in stage 2 or 3,

Hubbs (1964) defined a condition of "partial pregnancy" among the Texas populations he studied. Partially pregnant females were those that carried broods in which a significant (unspecified) number of the eggs were unferti1 ized. Hubbs attributed partial pregnancy to relatively low ma1 e frequencies resulting in inf requent inseminat ions and femal es con- sequent ly depleting stored sperm. We have observed no such conditions , and only rarely have encountered broods with more than one or two unfer- ti1 ized eggs. We $0 not, therefore, recogni ze part ia1 pregnancy.

In addition, abnormal embryos usually are easily distinguished from iormal broodmates by the time stage 2 development is reached. Abnormal anbryos exhibit any one of several anomalies, but usually are imnediately evident to the practiced observer. They usually are smaller than their broodmates, and are in earl ier stages of development. The embryo often is discolored; the head or eyes often are misshapen or atypically formed; pigment is poorly developed or appears in atypical locations; the caudal half of the body may be unusually shortened, and the caudal fin may not form; the spine may be abnormally flexed; often the embryo appears to be buried within its yolk rather than lying arol~ndits periphery. Abnormal en~bryoswere counted and recorded when they occut-red in stage 2 or 3 broods. Measurin . A11 specimens were lneasured to the nearest 0.1 ~rmstdndard _7Ylength SL) with vernier dial calipers according to the procedures of Hubbs and Lag1er (1958). Total length (TL) occasional ly was recorded. It was measured fran the tip of the lower jaw (niouth closed) to the tip of the longest caudal fin rays on a steel 1ni1limeter rule iltounted on J plex iglass holder in the fashion of a standard f ist1 r~lcasuring board. Total length was read to the nearest 0.5 rim. Data Handlin . Data were stored, retrieved, and analyzed by co~~iputer. l-ir-2he ata co ing formats and procedures are detai 1ed el sewhere (Snel so,^, 1977a). Data analysis used standard statistical procedures, co111bi ni ny "canned" software of .SAS and SPSS and progrdlns written speci fjcdl ly for the molly study.

The "rep1 aces~ent index", ca1 cul ated f ronl each rnontllly sample fro111 each 1ocal ity was derived as fol1 ows:

RI - 1/2 total normal embryos - tota1 adu 1t f einal es

One-half of the embryos was used because it is apparent that the sex ratio at birth in these populations is very near ~le-to-one (Snel son and Wetherington, in press,. Thus, half of the embryds in each monthly sample are assunled to be fernal es. For VAB I and TOWY, any femal e greater than 21.9 rnm SL was considered an adult; at BCHRD, the size for adulthood was greater than 23.9 mm SL.

Colnparisons among regression 1ines for brood size on female length were based on Bartlett's test for homoyenei ty for variances, followed by covariance analysis and, when required, the Student-Newman-Keuls test for mu1 t ipl e compari sons among slopes and intercepts (Zar, 1974). When usi ng untransfonned data, Bart1et t 's test a1 ways revealed heterogeneity anlong the variances in any given set of lines. In addition, plots of residuals revealed that the variance around the 1ines generally increas6.d as the independent vari abl e (femal e lenyth) increased. In order to rsemove these objectionable features, various transformat ions were appl ied to the data. In the f inai cornpari son, no transformat ion consistently resu 1ted in any greater improvenlent than simple loglo transformat ion of the dependent variable (brood size) . The log trinsformation virtually el irninated the tendency of the variance to increase with increasing female size. In all cases, this transfor~~~ation a1 so significantly reduced heteroscedast icity amoily the 1i nes, but in no case was complete homoscedasticity achieved as measured by the Bartlett's F-test (~(0.05). One pccul iar feature of the log transfornaion is that the procedure tends to make slopes more unifonn and intercepts rrrore dissirnil ar, virtually opposite the pattern in the un- trdnsfor~nedddta. In this report, the linear graph lines shown are com- puted fro111the untransforn~eddata, but a1 1 statistical interpretations and tests were based on the transformed data. The significance level used was p<0.05. Results

Molly Reproduction

The basic data from the mol ly reproduction study is summarized in Appen- dix Tables i-26.

Length of Reproduct ive Season.

The first month of study, October 1976, was the last month of reproduction for 1976. In both 1977 and 1978, reproduction had begun by early April. No fish were pregnant in early March sam?les. However, the high frequency of pregnancy in some of the April samples (Figure 2), for example 61 percent at BCIiRD in April 1978, suggests that active reproduction began in some cases in middle to late March. In all cases, pregnant females in the April samples czrried broods in the earliest phases of development, stage 1 or, rarely, stage 2. This is additional confirmation that actual reproduction had begun only a few days prior to the sampl ing date in April. There is only one exceptional instance when a central Florida population is know to have initiated reproduction prior to April. Snelson (1976) reported that his RRCD population, from a ditch on the western side of Merritt Island near the junction of State Roads 402 and 406, began reproduction in February, 1974.

The end of the reproductive season was somewhat more variable; but, i n general , successful reproduction terminated in middle to late Octo- ber. In 1976, there were no pregnant fish in the early November samples at VABI or TOWY. At BCHRD, 4 of 92 adult females (4%) were pregnant. In 1977, early November samples at VABI and TOiJY contained no pregnant fish. In early November 1977 at BCHRD, 7 of 156 adult fernales were pregnant. In 1978, the November sample at TONY contained no pregnant females. At VABI and BCHRD, a very few fernales in the early November, 1978 sample were pregnant (3 percent and 2 percent, respectively). A11 carried broods in late stages of development. Snel son (1976, 1977a) a1 so noted the cessation of successful reproduction in October.

The few females that protract reproduction into November may be rela- t ively unsuccessful at producing viable progeny (Tab1 e 2). In several cases all the gametes carried by a female were either abnormal embryos or unfertili zed eggs.

F-equency of Pregnancy.

The percent of adult females pregnant in each monthly sample is shown graphically in Figure 2, and the basic data are given in Appendix Table 87. Since only one month of re~roductionwas included in the 1976 study, only the data for 1977 ad 1978 are shown.

In 1977, the frequency of pregnancy curves showed similar trends for a1 1 three sites, with s; ing or early summer peaks, a middle sumer dc- pression, and a late summer or fall peak. The first peak occurred in May (TOWY, BCI1RD) or June (VABI). Actually, the peak in May at BCHRD was only 18 percent with 22 of 122 adult females pregnant. Pregnancy percentages at BCHRD were less than 20 percent until September in 1977. The mid-summer sl unlp in pregnancy occurred in July dt VAllI drld IICililD, and irl both June and July at TOWY. At all three sites, rei~roductionrecovcrcd dramatically after the inid-sununer @riod, and peaked again in August and September (TOWY ), Septeilber and October (BCHRD) or October alone (VABI).

In 1978, the general trendc of the 1977 curves were repeated at TOMY and BCHRD. At TOWY, peak pregnancy percentages occured in June ar:A August, with only a modest depression in July when 31 percent of the adult females were pregnant. At BCHRD, there was a dramatic surge of reproduction in early spring (April and May), a near-cessation of reproduction in July (2 percent), and a dramatic recovery in September and October.

The pattern of pregnancy percentages at VABI in 1978 was atypical. Reproductio~,dropped from 30 percent in April to 17 perSce?t in May, the low f: .- the season, and then gradual ly cl imbed to reach a peak in September.

;ze of 2cproduci ng Females

Hollies at BCHRD reached a larger size than at TOWY and VABI, and the mean si ze of pregnant females was r,~uch1 arger (Appendix Tables 88-95). The moll ies at TONY and VABI were sirtil ar in si ze; and during 1977, the distributions of mean sizes for reproducing females nearly were identical. For 1978, reproducing females at TO\JY were somewhat larger than in 1977, and were considerably larger than those at VABI (Figures 3-5).

The mean size of pregnmt females re1ates both to the rnaxirnurn size attained dnd to the size at which reproduction begins. At VAGI and TOWY, reproduction normally began at about 22 rrm SL (Appendix Tables 94-95). For both sitcs over the two years of study, only 5 of 1,213 pregnant females were wnller than 22 mm SL. At BCHRD, fenal e; generally were sl ightly larger, 24 mm SL or greater, when they began to reproduce. Of the 558 pregnant fema:es examined frorn BCHRD, only 3 were smaller than 24 mm SL.

In 1976 and 1977, pregnant females at VABI and TOWY rarely exceeded 40 nm SL, whereas females at l3Ct;RD were as large as 50 mm SL. 117 1978, the distribution at VABI and BCtiiiD essential ly was unchanged, but at TONY there were a small number of pregnant fern; les between 40 and 48 mcl SL.

During the 1977 season, the trends in ,nean size of pregnant females were similar at all three sites. The first females to begin reproductian in April were the largest of the overwintering females. The liiean dropped in May as small overwinteri~p females became pregnant. As the overwintering females grew, the rneal. si ze of pregnant fenial es increased, peaking in June or July. The mean si ze decl ined abruptly again in the late summer or fa1 1 (August or September). This apparently was due to a pcnnbina- tion of dying-off of large females from the population and recruitment of small females, born in April and flay, into the breeding population.

For 1978, the trends for meaii size of pr gnant females were unl ikc the 1977 trends. At 13CHR3 and TOWY, the 1978 trends are nearly Table 2 The reproductive record of pregnant female moll ies from November sampl es.

$ Brood Nomal Abnormal Unferti 1 ized Local ity-Date SL Stage L~~~br!to~ imbryos & VAB I 9 Nov, 1978 27.8 3 1 2 1 32.2 3 5 0 1 32.5 3 3 3 0 33.9 2 12 0 0

BCHRD

5 Nov, 1977 19.0 2 4 1 33.1 - 0 5 33. e - o 2 34.8 - 0 3 35.6 2 8 7 37.2 2 6 0 40.7 - 0 11 42. 1 3 10 1 FEMALE LENGTH *-----BROOD SIZE

V) W - aJ 40- E W LL -35 I- % 35- -30 Za 4 Wa Q

-10 \ 1 -5

I 1 I 1 I I 1 1 I I 1 1 1 1 AMJJASO AMJJASO 1977 1978

Figure 3. The mean length (m SL) of pregnant females and the mean brood size for monthly samples of the sailfin molly (Poecilia latipinna) from the VABI study site. MEAN BROOD SIZE , ? d' ., , , . . . .,.. ..rrr. rnr .-.,"Cls-r.~w.= w--mmm.FAa. rF

451, -FEMALE LENGTH +---4 BROOD SIZE

AMJJASO AMJJASO 1977 1978

Figure 5. The mean length (mn SL) of pregnant females and the mean brood size for monthly samples of the sailfin molly (Poecilia latlpinna) fran the BCHRD study slte. identical. %an size was virtually unchanged fron April to June, increased abruptly in July, and gradual ly decl i ned through October. At VABI in 1978, the mean size of pregnant females fol lowed an erratic trend.

It is noteworthy that in every case; the mean size of pregnant females peaked during the mid-sumner depression in pregnancy percentages, In every case, the few females that continue to reproduce during the mid- sunmer slmp are large in size. Small females, normally sexually active, cease reproduction during the sl unp period. --Brood Size. Over the two years of study, brood sizes raged fi-ccn 4 tninirnun c, - t.9 a xaximun of 54, The largest brood si zes di scovered at the three *ites were as follows: VABI, a 388 nm SL female with 47 embryos, col- iected June, 1978; TWY, a 39.4 nm SL female with 41 embryos, collected Arril , 1978; BCHRD, a 41.6 m SL female with 54 embryos, collcted June, 1977. Mean brood sizes were generally similar at VAGI and TONY, generally in the range of 5-15, At BCHRD, brood sizes were substantially 1arger, generally in the range of 12-26 (Figures 3-5; Appendix Tabl e 96).

There is a general psiti ve correlation between female size and brood size; i.e., larger females have larger broods (see details later under Size Specific Fecundity). Thus, it is not surprising that the curves for mean pregnant female size and man brood size track one another rather closely, Thus the temporal and spatial differences in mean brood sizes must be interpreted in 1ight of wan female sizes. Appendix Tabl e 97 shows that when a comparable size c1 ass of females is compared, many of the dramatic differences between months and between sites disappear.

Size Swcific Fecundity

The linear regression relationships for brood size on female length in monthly samples taken at the three study sites are shown in Figures 6-8 (1977) and 9-11 (1978). Statistical data for these graph 1ines are given in Appendix Tables 98 and 99. For months where the regression was not significant (p >0.05) and/or the nunber of pregnant females was less than 10, the regression line was not graphed.

Covariance analysis wzs performed on the monthly regressions from a given site in one year, using a log10 transformation of brood size to reduce heterogeneity of variances among the 1ines (see i4ethods and falaterial s). In a1 1 six year-station comparisons, the F-test was highly significant (p <0.01), indicating that there were significant differences ainong the monthly regressions. The monthly regressions then were sub- jected to a Newnan-Keuls mu1 ti ple comparison test for direct comparison of slopes and intercepts to determine rrh.ich 1ines were significantly divergent. This test successful ly identified divergent 1ines in all cases except the BCHRD, 1977 and BCHRD, 1978 data sets. Although the covariance F-test was highly significant in both cases, indicating significant variability among the monthly lines, the tieman-Keuls procedure failed to identify which lines were divergent. Although somewhat JUNE

I1ff,,II11I,IJ,,, 21 23 25 27 29 31 33 35 37 39 41 43 FEMALE SIZE (mm SL)

Figure 6. Least-squares 1inear regression 1ines for brood si ze on female length calculated from monthly samples of the sailfin molly (Poecilia lati inna) collected at the VABI study site in 1977. Lines based on a monthly ram- -fP-o oss than 10 or where the regression was not statistically significant (p=0.05) are not shown. See Appendix Table 98 for statistical details. #-&V) =ma U L -> mL -J=zaJO UO+ Lcc 6- 0 ccc U C1.C m 3?-r3 c +' al 0- r-C)Ln aala -c 5'- -ma, x Ec, La- w U 00 Y-w c -0 w c-74 a 00 a u e< w IU NnCa4 Q)aJ - 1 v, C V, OCT

I 1r1lr111lllrlllllrll111~l~ 21 23 25 27 29 3, 33 35 37 39 41 43 45 47 FEMhLE SIZE (mm SLI

Figure 8. Least-squares 1inear regression 1ines for brood si ze on fenal e length calculated from monthly. samples of the sail fin molly (Poecil ia lati inna) collected at the BCHRO study site in 1977. Lines based on a monthly Sam- +-o ess than 10 or where the regression was not statistically significant (p=0.05) are not shown. See Appendix Table 98 for statistical detail s. - 5 C. T +J To Q) a4 Q)F -u LD mw* 5+ %& x Ew L.r 'Q)'+a w OWc -0 a, C-i-i a 00 3 v c=z aJ " N-x 0 .rmC,aJ V)C m

L V) 0 "I" ce.- Nc,

c me- 0 ? mLt 7 ? 0 V).- h V) 0-0 WEE II L wzL mc Cv w.7 0 L+ Ec, - c L-- rg rn am w alV) c.- C 0% - aJ- - a Ln amam ~af* EaJ X aJU L.r * U ou w C C-0 w 0-4o Pa aJ U ca N rCI

0 a- .r *mLn mc 0 V)c 2, w 0-0 L ~r 11 er, @a aJ C C- L-rr 0 LF Eg rCI am Y c m C.- -r 0 lc- BCHRD

L lllllIlfl~llll~firlillllrl~l 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 FEMALE SIZE (mm SL)

Figure 11. Least-squares 1 inear regression 1 ines for brood si ze on female leng~hcalculated from monthly samples of the sail fin molly (Poecilia lati inna) collected at the BCHRD study site in 1978. Lines based on a monthly sarnp:e size 77P-o ess than 10 or where the regression was not statistically significant (p=0.05) are not shown. See Appendix Table 99 for statistical detail s. unusual, this problem is not extraordinary. It results fro111the fact that the F-test is a much more "powerful" procedure than the Newnan-Keuls tect (Zar, 1974).

After applyi ny the Neman-Keuls test, the monthly regressions showing no si yni f icant differences among themsel ves were pool ed and a colmlon reyression was calculated. Siynificaritly divergent 1ines were yraphed independently. For the BCHRD, 1977 and BCHRD, 1978 data sets, there was no alternative but to pool all the monthly lines into one, even though covari ance analysis had reveal ed signi f icant differences among the 1i nes. The results of these procedures are shown in Figures 12 and 13 and Appendix Tables 100 and 101.

It is noteworthy that the regression 1ines for monthly sa~rlplescan vary siynificantly. Size specific fecundity is not a stable relationship, but can vary both spatially and temporally. This means that size- specific fecundity is a third and potentially powerful mechanism for ad- justi ng reproductive output.

The monthly pattern of variability in regression 1ines followed no regular trend but often was associated with the pattern of variation of replacement index (RI) (Figure 14). For example, the June and October, 1977 peaks in RI at VABI corresponded to the two steepest reyression lines (Figure 6). Likewise, che dramatic April and May, 1978 peak in RI at BCHRD coincided with steep slopes for those two monthly regressions (Figure 11). However, there are some cases (e.g., September and October, 1978 at BCHRD) where RI peaks were not associated with unusual size- specific fecundity (Figure 11). Furthermore, in some cases, outstanding regression 1ines, such as June, 1978 at VABI (Figure 4) and June, 1977 at BCHRD (Figure 8) were associated with low to inoderate performance in RT (Figure 14). Finally, there was no clear pattern of seasonality in the regressions. Either high or low-slope regression lines were 1ikely to occur at any time during the reproductive season.

Pool ing the monthly regression 1i nes and con~paring among years, the most noteworthy observation is that at TONY and BCHRD, the overall regressions for the two years are nearly identical, whereas, at VABI, the 1977 line has a much higher slope than the 1178 line (Figures 12 and 13). Only the month of June, 1978 approached the steepness of the pooled 1977 regressions.

Rep1 acement Index

The replacemer~t index is a single number yenerated fron~c.ach monthly data set that reflects, in a static sense, the reproductive output of the population. Since it is ir~fluenced by the interactiun of preynancy percentages, size distribution of pregnant females, and size-specific fecundity, different reproductive strategies could result in similar index values. This needs to be kept in mind when evaluating replacernerlt index.

In all '.hree study stations, replacement index exhibited spring or early sulrlnier peaks, mid-su~rnier slumps, and late sumler or fa1 1 peaks in reproductive output during the 1977 season (Figure 14). At VABI, 1979 VABl ---- TOWY ...... BCHFO POOLED ...-• POOL

R ,, OCT.

FEMALE SIZE (mm SL)

Figure 12. Least-squares linear regression lines for monthly and pooled samples of the sailfin molly (Poecilia lati inna) collected at three study sites in 1977. See Appendix Table 100 for statistis+k- eta1 s reproduction was modest in April, May, August, and September, and was dramatically elevated in June and October. The Zune peak coincides with a peak in pregnancy prcentage, a high size- speci f ic fecundi ty re1ati on- ship, and a high (but not maximal) mean pregnant female size. The Octo- ber peak coincides with a high size-specific fecundity relationship, and a large mean pregnant female size.

At TONY in 1977, the peaks in replacement index were ver) modest, and fell in flay and August. The mid-sumner slump was about equally expressed in June and July. Despite a high pregnancy percentage in lky, the repl~cementindex was a modest 2.1, since the mean pregnant fernale size ana the size-specific fecundity re1ationship were not very large. In August, the replacement index (2.2) was nearly identical to ?by. However, it was achieved through a moderate presnancy percentage, a large mean size of pregnant females, but size specific fecundity relationship nearly identical to May.

At BCHRD in 1977, the replacement index reached a very modest "peak" in May and June, was minimal in July, and peaked dramatically in Septem- ber. In May a replacement index value of 1.2 was achieved with a modest pregnancy percentage, a modest si ze-speci fic fecundity re1ationshi ?, and a very low mean pregnant female size. In June r;he value of 1.3 was achieved despite the dramatically high size specific fecundity and moder- ately high mean pregnant female size. Both of these features were coun- teracted by a low pregnancy percentage. The September peak in replacement corresponded to a dramatic peak in pregnancy percentage and a high size-specific fecundity, but a low mean pregnant femal e size.

!n the 1978 data, gmeral trends in the rep1 acement index curve are less consistent. At BCHRD, there is a dramatic April-May peak, an equally dramatic July slump, and a second peak in September-October. At VABI, there r ere May and October low-points, and modest replacement through the remainder of the year, with an unspectacular "peak" in September. At TOHY, there is a hint of rlay-June and August peaks, with a sl ight depression in July, and a gradual 1 inear decl ine in September and October.

Environmental Analysis The data for the physical, chmical , and biological parameters measured at each molly study site are presented in Appendix Tables 102-116 and are sunmarized in Table 1. Temperature, salinity, and water depth are displayed graphical ly in Figures 15-17.

The 1977 year had much less rainfall than 1978, especially during the winter and early spring (Table 3). The data are incomplete for 1978 (due to the termination of data collection by the weather service); but from January through thy of 1377, 19.25 cm was recorded; whereas 33.19 cm was recorded in 1978. The ten-year avcrage for those four months is 25.37 cln. Thus, the winter and spring of 1977 was a little drier than average, and the same period in 1978 was much wetter than average.

In 1977, water levels reached their lowest point in early May at all three sites, and there was a partial recovery in June. After June, water

Figure 15. Graph of average rncdian surface water temperature, sal inity, and water level masured monthly at the VABI sail fin molly study site, October, 1976 to December, 1978. Salinity was measured at the time of fish sarnpl ing. The highest water level measured was recorded as zero, and other monthly readings are recorded as cm below zero. The average median water temperature was for the 15-day perfod preceding the monthly fish sample. Breaks in the graph 1ines represent n~iss ing observations. - TEMP ...... SALINITY .-- - LEVEL

Figure 16. Graph of average median surface water temperature, sal inity, and water level measured monthly at the TONY sail fin mol ly study site, October, 1976 to December, 1978. Salinity was measured at the time of fish sampling. The highest water level measured was recorded as zero, and other monthly readings are recorded as an below zcro. The average median water temperature was f~rthe 15-day period preceding the monthly fish sample. Breaks in the graph 1 ines represent missing observations. MEDIAN WATER TEMP (C) Sncucuaom0 w 5&S% aC Wh m .- aJO\WL .TI 7S4waaQ) L WW- E L -=ma WLcn EU 4JQ)IrJQ)QL %a artno u-GE+=* u- w-r c Q)aJ .-r an > O't-c, aL 0 -0- a ?U w#- 3- L --m a, 35% EhWOCF .POI a? 0 0 -43Q)EL a n - @A rc, aiC1 L w E= -aJa#vc,Q) an= z L omwCn+ 3 0.- rn c &US 'r 50 TIQ L wwam -xu v C %aJkLa,O Eu OL-r a.- uCLo otn*aJ a CnLU > :2.E aJ .: & o3-LLW moLa an 3w~no IrJ w aJ crrhen a#-r--u-rUe- CaC 'toww l @A 5E~c4-mm w . mc Lay .- 't- c't- ohrom m- -4Jc.4 a Q) - -a=+rLw x*.r'roaL- o O't- L Q,

cnxr L a~a, muow3 c, tn n k*oZ>amOLC8 a, Q) m= 3 -- c,UD+'- L 9- aJca 0 Ow L mLL- m3 an I V) amma Phm0 ELL a- aJ L a en3 ~r~ac,o ow N am Cw LIL omwac,g E 3 me fi mcE 4w1- 3.c3 a,c,aJ aJ L-ru r w L3tLa5AL 3 w.r 0.- ma, cnmc u Table 3 Monthly rainfall (in an.) measured at Kennedy Space Center, Florida, during the period 1973-1978. The lo-year average is based on the years 1968-1977. Data provided by NOAA National Ueather Service.

January 15.01 0.51 1.83 2.i6 4.95 5.38 4.19

February 3.66 1.07 1.09 0.30 5.69 14.40 4.75

April 3.61 2.84 3.02 0.56 1.80 0.43 2.16

June 13.82 29.03 27.97 25.22 846 14.35 20.54

July 8.69 27.69 23.47 5.31 9.40 38.94 12.87

August 23.65 21.49 6.38 11.40 11.68 4.55 13.67

September 9.50 12.17 18.42 21.13 11.89 12.62 11.72 October 8.05 12.90 8.53 3.35 4.70 - 11.83 November 1.88 2.06 3.45 9.83 16.41 - 6.21 December 5.97 6.17 0.43 11.66 11.23 - 5.86

Total 108.20 124.28 108.36 108.23 93.01 108.07 levels generally continued to rise throughout the remainder of the reproductive season. in 1978, levels also were lowest in May, but dessication was not nearly so severe as in 1977. At all three sites, water levels rose back to "normalu levels in June, 1978, and generally remained high throughout the remainder of the 1978 reproductive season.

Water temperatures followed a predictable annual pattern and generally were similar at al 1 three sites. Lowst water temperatures are reached during the period from mid-December to mid-February. Average water temperatures of around 30°C or higher were maintained through the mid-smer period, from mid-May to mid-September, depending on site. Unfortunate mechanical problems with the recording thennographs resulted in many missing data points, and detailed comparisons between sites can- not be drawn. It does appear that BCHRD was the warmest site in 1977, and that VABI and TOMY had similar temperature regimes during the sumner of 1978.

Discussion

Of the several physical , chemical and biological parameters inezsured during this study, quantitative and qualitative evaluation suggests that two, temperature and water level, are of major significance in explairling the pt- terns of reproduction observed.

Water temperature, especial 1y in conjunction with photoperiod, has been shown to be a dominant factor control1ing the initiation of reproduction, both in fishes in general (de Vlaming, 1972) and in P. lati inna in particular (Crier, 1973). The effect of high temperatureson-%r repro uction of the sail fin molly was first noted by Snelson (1976) and is confirmed herein and by two fortuitous laboratory experiments. -P. latipinna ceases to reproduce at some temperature between 3Z°C and 35°C. Laboratory fish held at 31°C and 32°C in two separate experiments reproduced and grew normally. 14ollies held at 35°C grew at a very reduced rate, ceased to reproduce completely, and exhibited unusually high mortal ity.

The stress temperature of 35°C and above was recorded regularly at a1 1 three study sites during mid-sunmer mofi'hs. The temperature lines in Figures 15-17 show the average daily median suriace temperature. Reference to Appendix Tables 102-116 shows that average daily maximun temperatures nearly always exceeded 32°C in mid-sumner. In order to achieve average maxima ranging from 32-35"C, many readings above this range were included, some as high as 39°C.

The determination of a specific surface water temperature at one spot in a large habitat clearly does not reflect precisely the temperature regime that a free-ranging fish is exposed to. Temperatures vary in a micro scale throughout a body of water. The response of many fish species in laboratory thermal gradient experiments demonstrates concl usively that fish are capable of discriminating between temperatures and selecting comfortable ranges (Murray, 1971).

In this context, water temperatures probably interact in a complex fashion with other variables to produce stressful conditions. In the three molly study areas, the most importtnt interaction appears to be with water Table 4 Mean monthly air temperatures (OC) measured at Kennedy Space Center, Florida (0 & C Buildfng) during the period January, 1973 to August, 1978. The 12- year mean is for the period April, 1965 to December, 1977. Data provided by NOAA - National Weather Service.

12-Year -1973 -1974 -1975 -1976 -1977 -1978 Mean January 16.4 21.2 19.0 13.6 10.8 13.1 16.7 February '4.0 16.3 19.3 16.6 13. 4 12.2

April

3 fky June

July

August

September

October

November

December depth. When water stages are high, habitat space is increased, often dramatically, by the flooding of extensive areas. During high water, confi tions for moll ?'es appear to be optimized in three respects. (1) There is a greater . , opportunity for the fish to avoid extreme temperatures (high or low) because of the greater range of temperatures avail able between exposed water vs shaded water, shallow water vs deep water, and sdrface zone vs bottan zone. 72) In addition, the expandchabi tat space under flooded coxitions should reduce densities of fish and moderate the effects of competition and potentia'ly stressful social interactions ammg individuals. We are just beginning to understand the influence of social factors on poecill id reproductive systeriis (Baird, 1968, 1974; Borowsky, 1973; Martin, 1975; Sohn, 1977). (3) Finally, food resources for moll ies are increased dramatically under high water conditions. Mollies primarily are "herbivores", feeding on a combination of periphyton, detritus, and algae, with the rare inclusion of sorne animal matter (Harririgtc!! and Harrington, 1961). Fl oodi ng prornotes the production and availability of the priixry food sources of P. lati inna. Numerous studies have shown the importance that food res~urce?p re?; ay in t e control of fish reproduction (e.g., Hester, 1964; Bagenal, 1969; Wooton, i9?2),

Much of the variation in rnolly reproduction observed in this study and in earlier work (Snelson, 1976) can be interpreted as a result of the interplay between temperature and habitat vol me.

The interaction between temperature and water depth probably explains the annual pattern of spring and fall reproductive peaks, with a depression of reproduction in mid-summer. This pattern was observed in 5 out of 6 cases in this study (Figure 14) and in 2 out of 4 cases by Snelson (1976).

A1 though reproduction normally began in April, only the 1argest overwintering females reproduced in the first month. In May and Jufie, the full size range of adult females reproduced. During those months, median water temperatures were in the middle to upper 20's C and water levels either were at their lowest or recently had increased abruptly, due to heavy June rains or pumping. Low watw conditions in the spring may not greatly stress the mrllies for two reasons. First, water temperatures are low and optimal for reproduction, not compl icating the crowded conditions and adding to the possibil ity of dissolved oxygen stress. Secondly, the level of reproductive output in the spring and early summer months [nay not be determined by food availability at that time. Although low water and crowded conditions would be expected to increase competition for food, the fertil ity and fecundity exhibited in spring months may be more dependent on energy reserves accumulated during the winter than on food resources immediately available.

The mid-summer depression in reproduction ranged from sl ight to dramatic in 5 of the 6 cases reported herein, and in 2 of the 4 cases reported by Snelson (1976). The depression normally was most severe in July, but occasionally was expressed as early as June or as late as Aigust (Figure 14). The period of lowest reproduction coincided with the period of ~naxirun tenpera- tures. Even though the moll ies may have some abil ity to avoid extreme temperatures, it is likely that many fish may be inadvertently "trapped" in areas where temperatures exceed 3Z°C, especially during the day. The effect of high temperatures would be most dramatic in low water conditions, where the few remaining bodies of water may be shallow and uniformly heated from surface to bottom. Even if water levels were high during the reproductive slump, the depressed reproduction might bc a consequence of low water levels a month or two prior, when resource iimitations resulted in subsequent reduction in fecundity or fertility. It is significant that mean size of pregnant females goes up dramatically during the peak of the sumer slwnp; i.e., only large females continue to bear young during the period of minimum reproduction.

A late summer or fall peak in reproduction was expressed weakly to strongly in all 6 instances recorded herein, and in 2 of the 4 cases reported by Snelson (1976). The peak occurred in August, September, or October. A late-season peak as early as August or September is not easy to resolve in terms of an interaction between temperature and water level. Temperatures usually remain at summer-time high levels through the middle of September. Perhaps the fact that water levels are characteristically high in late sumner and fall is significant, ?emitting fish to avoid stressful temperatures. It is noteworthy that August "peaks" a1 1 are minimal. Dramatic late-season peaks all occured in September and/or October, after water temperatures had begun to drop. In all cases, the fall peak(s) in replacement index is correlated with peak(s) in pregnancy percent. This reflects that the late-season peak is associated with resumption of reproduction by those individuals inhibited during the mid-summer depression, not by any dramatic shift in size-specific fecundity or size of the reproducing fish.

To a considerable degree, the differences in reproductive perfonance between sites and between years can be explained by the temperature-water 1eve1 hypothesis. Following the record-breaking cold winter of 1976-77 (Tab1 e 4), the spring of 1977 was unusually dry (Table 3), causing water levels to fall rapidly to extremely low levels. Because of the differences in basic configuration, low water conditions have a dramatic impact on VABI and BCHRD, especially the latter. Both sites have extensive surface area where depths are .2 to .5 m in depth under normal conditions. When water levels drop, all the fish from these extensive "flats" are crvwded into a few remaining deep holes and ditches. At TOWY, by contrast, lowered water level s have rninin~um impact. This site has only one shallow marginal "flats" zone connected to a very long, deep ditch, "U-shaped" in cross section. Lowered water levels cause the marginal shallow area to become more shallow, but the zone never dried completely during our study, even in May, 1977. Thus this area never became unavailable as fish habitat, even if some fish did vacate the area because of its extreme shallowness. Those fishes that were forced from the shallow zones moved into the deep adjacent ditch, whose surface area was not noticeably reduced during low-water conditions. As a result of basin configuration, TOWY is a more stable habitat than VABI or BCtiRD, and fish are less 1ikely to be stressed by over-crowding, competition, heat, or low dissolved oxygen concentrztions during dry periods.

The winter of 1977-78 was colder than average (Table 4), but was not nearly as severe as the preceding winter. In addition, the winter and spring of 1978 were very wet, compared to 1977 (Table 3). As a result of reduced springtime dessication, the water levels remained higher at all three sites in 1978. At TOWY, there was virtually no change in available habitat in 1978. At VADI and BCHRD, there was noticeable habitat restriction, but it lasted only about 30 days. VABI was pumped imnediately after the May dessication period in 1977, causing water levels to rebound quickly. With this exception, the water levels at TOWY and BCHRD rose to normal levels by June in 1978, and they remained high during the entire summer. In contrast, the water levels at both sites remained relatively low through mid-sumner in 1977, and did not reach normal levels until September. This difference is related to rainfall. In 1977, on13 41.43 cm fell at KSC between June and September, whereas in 1978 the value was 70.46 cm (Table 3).

In keeping with the more stable nature of the habitat, the r-productive output at TOMY waj much less variable, compared to the other two sites, both in 1977 and 1978. There were no dramatic peaks. Although reproduction was depressed noticeably in June and July of 1977, there was only a very minor depression in 1978. Finally, a higher water level through the sumner of 1978 seems to have resulted in generally higher levels of reproduction in 1978 than in 1977 (Figure 14). At VABI and BCHRD in 1977, both populations had minimal reproduction in July, recovered slightly higher in August, and then displayed a dramatic fall peak in October (VABI), or September and October (BCHRD). The two sites, however, differed conspicuously in the spring pattern of reproduction in 1977. At VABI, there was a dramatic peak in June, but there was no noteworthy spring peak in BCHRD (Figure 14).

In 1978, VABI had a much more uniform level of reproductive output than in 1977. There was a depressed period in May, but generally moderate reproduction continued throughout the sumner, with a small peak in September. (Figure 14). At BCHRD, the 1978 replacement index curve shows much more exag- gerated variation than in 1977. There was a dramatic April-May peak (not present in 1977), followed by a virtual cessation of reproduction in July, and a second peak period in September and October (Figure 18).

This analysis leaves many questions unanswered, and raises many new questions concerning the causes of variation between sites and between years. For example, what causes differences in the level of reproductive output at comparable times, such as at BCHRD in the spring of 1977 and the spring of 1978 (Figure 14)? Why was overall replacement index at TONY higher in 1978 than 1977 while at VABI the reverse occured? Finally, why should reproduction at a relatively stable, favorable site such as TOWY not achieve uniformly high levels approaching the peaks demonstrated at VABI and BCHRD? The answers to these and other questions require additional research and a more extensive period of field monitoring. Sumnary

The objective of this study was to learn as narch as possible about reproduction in a representative fish species characteristic of the waters around . Kennedy Space Center. The data gathered would constitute a "before" baseline on reproductive perfor~r~ancein the selected fish. By continuing to monitor reproduction during and after the initiation of space shuttle operations, it was hoped, it miyht be possible to identify sub1 etha? environmental changes adversely af fecti ny normal reproduct ion in the species. A1 tered reproduct ive performance cod1 d constitute an early warni ng of subtle environmental changes that eventually rnight prove detrimental for the fish comunity as a whole.

The sai lfi n in01 ly (Poeci 1i a 1at i inna) was chosen as the test species because it is widespreadmndant-?-Ti on erritt Is1and. Furthermore, because of its livebeariny habit, it is relatively easy to collect detailed data on reproduction of this fish.

Populations of sailfin moll ies in three contrasting habitats were sa~npled monthly from October, 1976 to December, 1978. Preserved females were autop- sied, and detai 1s of thei r reproductive :tatus were recorded. Various physi - cal, chemical, and biological habitat parameters were measured monthly at each study station.

The major features of reproduction in the sailfin molly were similar in the tnree study populations. The reproducti ve season extended from April to October. Females become sexually [nature at 22-24 mn standard length. The number of young in a brood was correlated positively with female size, and the sex ratio at birth was approximately 1:l.

Beyond these generh, lties , the detai 1s of reproducti ve output varied among the three popul ations studied, and a1 sc: varied frail month-to-month and between the two years of study. Overall reproductive output was measured by .i repl aceinent index. Rep1 acernent index was inf 1 uenced by three major aspects df a popul at ion's reproductive strategy: (1) the percentage of adult females that were pregnant, (2) the size distribution of the pregnant females, and (3) the size-specif ic fecundity. Sirnil ar repl acenient index values could be achieved by different combinations of responses in these three measures of reproductive performance.

The three study popul dti ons extii bited si~ni1ar monthly trends for repl ace- ~nentindex in 1977. In general, all three sites exhibited spring or early sunniler peaks, mid-sununer depressions, and late sunwiler or fall peaks in reproductive output. In 1978, the general trends in replacement index were less consistent. At the BCHRD site, spri ng and fall reproductive peaks persisted. At the VABI and TOWY site?, replacelllent was rather inodest end uniform throughout the 1958 season, without dramatic peaks.

The te~~~poraland spat ia1 vari ati on in reproduct ive perforniance reported in this study can be explained in part by variations in import.ant environmental conditions. Fluctuations in water temperature and/or habitat availabil ity resulted in conditions for reproduction that varied fran optimal to detrimental. Laboratory experiments have shown that P. 1ati pinna ceases to reproduce when water temperatures exceed 32OC. Surface water temperatures grea-er than 32 C were recorded re~ularlyat the three study areas during summer months. This temperature factor may be partial ly responsible for the mid-sumner depression in reproduction that was so conspicuous in 1977.

Rainfall in the Kennedy Space Center area usually is heavy from mid- summer through late fall, and minimal from late winter until early sumner. As a consequence, the impounded waters on Merritt Island undergo drastic desiccation in spring and early summer, and the shallow, marsh-like habitats of the sailfin rnol ly shrink dramatically. During this period the fish may be stressed due to crowding and competition for food. During late summer and fa1 1, heavy rains cause water levels to rise, and moll ies again are free to expi~lcitheir habitat into newly flooded areas. Populations are less dense, and food availabil ity is increased. However, during sumner, the populations are exposed to very high water temperatures.

Sprf ng peaks in reproduction, when evident, occurred during low water periods. However, spring water temperatures were optimal for reproduct ion, so the fish were not thermally stressed. The crowding of moll ies into peroansnt waters in the spring n~ightnot adversely affect reproduction if the level of spring re7roduct ion is control 1ed by overwi r~terenvi ronmental candi tions, rather than by spring-time cond~tions(e.g., food avai lability).

The period of depressed rdeproduction ir mid-sumner occasionai ly coi ncided with low-water conditions as well as elevated tez.?eratures. In most cases, however, a slump was evident even though water levels had begun to rfse to "normal " level s. This suggests that even under conditions of expanded habitat availability, the fish could not always escape the limiting effects of high temperatures. The fa1 1 peak in reproduct ion typical ly was the most dramatic. It usually occurred in either September or October, when water levels were high and temperatures had begun to coderate. Because ,: differences in basin configurations, the three study sites were not subject to the same degree of desiccation. In general, TOWY was the most stable study site. Its surface volume did not decrease dramatically as water levels fell, and there was a large, deep ditch system that constituted a more-than-adequate refugi um during stressful periods. At VABI and BCHRD, especially the latter, low water conditions resulted in severe loss of habitat, and fishes were crowded into small pockets of permnent water.

These differences in the three sites are partly reflected in the pattern of reproduct ion. Dramatic peaks and val leys in rep1acement index were most conspicuous at BCHRD and VABI, the two most variable sites. At TOWY, more stab1 e environmental conditions generally resulted in more stable reproducti ve performance throughout the season. These differences a1 so were reflected in the comparison of 1977 and 1978 reproductive trends. Less rainfall in 1977 resulted in much more severe desiccation conditions than in 1978. At VABI, the less stressf!~lconditions in 1978 resulted in much Inore uniform reproductive output than in 1.977.

The ability to utilize reproductive performance of the sail fin mol ly to biomonitor environmental conditions is dependent upon the abi 1i ty to differen- tiate between changes in reproduction caused by unnatural perturbations and those that are part of the natur$l response of the fish to vaned ecological conditior,~. This study has showi~that there is a vast ar~iour~tof teqnoral and spatial variation in reproductive performance of the sai 1fin nlol ly under "normal" conditions. Not only did each study population behave differently, there were dramatic differences between the two reyroduct ive seasons. Furthenaore , the observed gatterns of variation r~idy represent only a sn~allpart of the var- iabi 1i ty that could be exhibited umier other environ~nental circuaistances. A1 though the observed spati a1 and temporal variation is partly in acc~rdance with a hypothesis invo: ving water temperature and habitat conditions, there is no cor~venientway to tcst directly this idea. Furthennore, sone of the observed variation is not easily explained in the context of this hypothesis.

As a result of these 1imitations, reproductive performance of the sail fin rnolly could be utilized in the context of environrriental -.\nitoring only in a very general way. It would be impossible to differeritiatci between oatural and unnatural factors as the cause for site-to-si te, month-to-mont h, or year- to-year vari atil,n in percent pregnancy, size of pregnarlt females, or size-specific fecundity.

Concl usions

(1) Keproductive output in the saS lfin molly responds in a sensitive way to changes in environmental cmdit ions. (2) The three popul at ions studied varied in reproducti ve performance in different ways, to different degrees, and at different times. (3) At least so!ne of the variability in reproductive perfor~~~anceseems to be re1ated to varyi ng cnvi ronmental factors, especi a1 ly water temperature and habitat avai 1abil ity. Howcver, rnuch variation in reproductive output is not explained by this hypothesis. (4) Reproductive performance in the sailfin 1il01ly is not a suitable tool for eva1udt;i ng sublethal changes in er~virorlri~ental qua1 ity. At the present level of und~rstanding, it would be in~possiblcto distinyuish ~ariations caused by natural environmental fluctuations and thosc ir.luced by ~nan-made perturbations. I ow special thanks to the many students and technicians who have assisted in all facets of this project: Ken Bradley, Richard Byles, Tim Clabaugh, John Galluzzo, Helen Large, Larry Neebe, Jane Provancha, Bruce Stanley, Boyd Thompson, Jim Truner, Jeff Wetherington, and Sherry Williams. The data hand- 1 ing, management, and analysis was handled primarily by Jeff Wetberington, with assi stacce from Bruce Stanley and Phi 1ip Thompson. Jim Turner, He1 en Lase, Jane Provancha, and Jeff Wetherington examined the bulk of the 20,000 (+) moll ies utilized in this investigation. Helen Large drew all the graphs and Don Sctmitz assisted with art work. Rosa1 ie Creamer served as secretary and Mary Ann Johnston served as administrative assistant for the project. Special thanks to Ralph Gunter and Drs. Lc 'ie Ellis, David Vickers, and Jack Stout for helping with various administrative and contract problems. 3r. Yickers nas been especially encouraging, and has granted special favors OF^ rare than one occasion.

Mr. Jack Salnela, Brevard Mosquito Control District, kept me advised of punping activities, and discussed with me his substantial knowledge of Merritt Island natural history. The administration and staff of the Merritt Island National Wi ldlife Refuge have provided coomrati on, logistic and material support and many additional favors during the course of this work. I acknowledge especially Robert Yoder, Stephen Vehrs, and Dr. James Baker. The Fish and Wild1 ife Service, Florida Game and Freshwater Fish Commission, and Florida Department of Natural Resources have granted the permits necessary to conduct this work.

Finally, I acknowledge financial support from NASA's John F. Kennedy Space Center fw the support of this and earlier work on P. lati inna (Snel son, 1976). Drs. Paul Buchanan and William Knott ana Mr *m ice have acted a> NASA's 1iaisons and have helped solve a variety of problems. Mr. Lee p. ovided substantial encourage3ent by taking a special interest in the mol ly work. Literature Cited

Amoroso, E. C. 1960. Viviparity in fishes. Symp. 2001. Soc. London 1: 153: -181. Bagenal , 1. B. 1969. The relationship between food supply and fecundity in brown trout, -Salm trutta. J. Fish Biol. 1: 167: 182. Baird, R. C. 1968. Aggressive behavior and social organization in 1401 lienesia latipinna (LeSueur). Texas J. Sci. 20: 157-176.

Baird, R. C. 1974. Aspects of social behavior in Poecilia latipiqna (LaSueur) . Rev. Bi01. Trop. 21: 399-416. Benoit, D. A. and G. U. Holcmbe. 1978. Toxic Effects of zinc on fathead minnows (Pimphales promelas) in soft water. J. Fish Biol . 13: 701-708. browsky, I?. L. 1973. Social control of adult size in males of Xiphophorus variatus. Nature 245: 332-335. bungs, U. A. 1969. Chronic toxicity of zix to the fathead minnow, Pimephales promelas Rafi nesque. Trans. her. Fi sh. Soc. 98 (2): 272-279.

Cairns, J., Jr. and A. Scheier. 1964. The effect upon the punpkinseed sun- fish, Le omis gibbosus (Linn.), of chronic exposure to lethal and subletTi+- a concentrations of dieldrin. Not. Natur,, Acad. Natur. Sci. Phila. No. 370: 1-10.

Cmings, J. 0. 1943. Morphogenesis of the gonopodiun in Mollienisia latipinna. J. fbrphology 73: 1-17. de Vlaming, V. L. 1972. Environmental control of teleost reproductive cycles: a brief review. J, Fish Biol. 4:131-140.

EPA. 1974. Methods for chemical analysis of water and wastes. EPA-625-/6-74-003. U.S. Environ. Protect. Agency, Off. Tech. Transfer, Wash , D.C., 298 pp. Foster, N. R., 3. Cairns, Jr., and R. L. Kaesler. 1969. The flagfi sh, Jordanel 1a f loridae, as a 1aboratory animal for behavioral bioassay studies. ~md.tbtur. Sci. Philadelphia 121(5): 129-152.

Foster, N. R., A. Scheier, and J. Cairns, Jr. 1966. Effects of ABS cn feeding behavior of flagfish, -Jordanel la foridae. Trans. her. Fish. Soc. 95(1): 109-110.

Crier, H. J. 1973. Reproduction in the teleost Poecilia latipinna, an ultra- structural and photoperiod investigation. Pnsertat~on,Univ. South Fla., Tampa, 180 pp.

Harrington, R. We, Jr. and W. L. Bidlingmayet-. 1958. Effects of dieldrin on fishes and invertebrates of a salt marsh. J. Wild1 ife Management 22 (11: 76-82. Harrington, R. We, Jr. and E. S. Harrington. 1961. Food selection among fishes invading a high subtropical salt marsh: from onset of flooding through the progress of a mosquito brood. Ecology 42(4):646-666.

Hester, F. W. 1964. Effects of food supply on fecundity in the female guppy Lebistes reticulatus (Peters). J. Fish Res. Bd. Canada 21 :757-764.

Hoese, H. 0. and R. H. More. 1977. Fishes of the Gulf of Mexico- Texas A & M Univ. Press, Ccllege Station, Texas.

Hopper, A. F., Jr. 1943. The early embryology of Platypoecilus maculat~s. Copeia 1943(4) :218-224.

.Hubbs, C. 1964. Interactions between a bisexual fish species and its gynoge- netic sexual parasite. Bull. Texas Mem. Mus. 8:l-72.

Hubbs, C. L. and K. F. Lagler. 1958. Fishes of the Great Lakes region, rev. ed. Bull. Cranbrook Inst. Sci. No. 26, 213 pp.

Jalabert, B. and R. Billard. 1969. Etude ultrastructurale du site de conser- vat ion des spermatozoi des dans 1'ovai re de Poecil ia reticul ata (Poi sson Tel eosteen). Ann. Biol. him. Bioch. Biophm:273-280.

Krunholz, L. A. 1948. Reproduction in the western mosquito fish, Gahusia affinis affinis (Baird L Girard), and its use in mosquito control. Zcol. mnogr. 'm-43.

Lane, E. C. and R. J. Livingston. 1970. Some acute and chronic effects of dieldrin on the sailfin molly, Poecilia latipinna. Trans. her. Fish. SOC. 99(3) :489-495.

Martin. Re C. 1975. Sexual and aqqressive behavior. density and social siructure in a natural population of mosqui tofish, ~ambusiaaff inis hol brooki . Cowi a 1975:445-454. Murray, R. W. 1971. Temperature Receptors, pp. 121-134. In W. S. Hoar and D. J. Randall (eds.), Fish Physiolow, Vol. 5, ~ensoryTystemsand Electric Organs. Academic Press, New York, N.Y.

Provost, M. W. 1959. Impounding salt marshes for mosquito control and its effects on bird 1 ife. Florida Natur. 32:163-170.

Provost, M. W. 1973. Salt marsh management in Florida. Ann. Proc., Conf. on Ecological Animal Control by Habitat Management, Vol. 5, not paginated.

Rosen, D. E. and R. M. Bailey. 1963. The poeciliid fishes (Cyprinodontiformes); their structure, zoogeography, and systematics. Bull. her. Mus. Natur. Hist. 126(1):1-176.

Rosen, D. E. and M. Gordon. 1953. Functional anatomy and evolution of male genitalia in poeciliid fishes. Zoologica 38:l-46. Scrimshaw, N. S. 1944. Superfetation in poecil iid fishes. Copeia 1944 (3) :180-183.

Sheinbalm, Steven. 1979. Reproductive cycles, sex ratios, brood si zes and adult sizes of estuarine and freshwater populations of the sailfin molly Pc~cilia lati inna (LeSueur) , in the Tampa Bay area. N.A. Thesis, Univ. -.*64 -.*64 pp.

Simanek, D. E. 1978. Genetic variability and population structure of Poecilia latipinna. Nature 276:612-614.

Snelson, F. F., Jr. 1976. A study of a diverse coastal ecosystem on the Atlardic coast of Florida: Ichtyological Studies, Vols. I and 11. Final Grant Report, Grant NGR-10-019-004- NASA, Kennedy Space Center, FL.; 19 Feb., 1976.

Snelson, F. F., Jr. 1977. ichthyological Studies, pp. 245-349. In: D. H. Vickers, --et al. Semi-Annual Report: A continuation of base-Tine studies for environmentally monitoring space transportation systems (STS) at John F. Kennedy Space Center. NASA Contract No. NAS10-8986; 11 January, 1977.

Snel son, F. F., Jr. and W. K. Bradley, Jr. 1978. Mortal ity of fishes due to cold on the east coast of Florida, January, 1977. Fla. Sci. 41(1):1-12.

Snelson, F. F., Jr. and J. D. Wetherington. 1980. Sex ratio in the sailfin mol ly, Poeci lia latipinna. Evolution, in press.

Sohn, J. J. 1977. Socially induced inhibition of genetically determined maturation in platyfish, Xiphophorus maculatus. Science 195:199-201.

Stickel, L. F. 1973. Pesticide residues in birds and mamnals, pp. 254-312 -In: C. A. Edwards (ed. ) , Environmental pollution by pesticides. Plenum Press, New York, NoY.

Strickland, J. D. H. and T. R. Parsons. 1972. A practical handbook of sea- water analysis. Fish. Res. Bd. Canada Bull. 167, 310 pp.

Tavol ga, Id. N. 1949. Embryonic developnent of the platyf i sh (Pl atypoeci lus) , the swordtail (Xi ho horus), and their hybrid. Bull. Amer. Mus. Natur. Hist. 94(4) :16*

Tavolga, W. N. and R. Rugh. 1947. Development of the platyfish, Platypoeci lus maculatus. Zoologica 32:l-15.

Thibaul t, R. E. and R. J. Schul tz. 1978. Reproductive adaptations among viviparous fishes (Cyprinodontiformes: Poecil iidae). Evolution 32(2) :320-333.

Turner, C. L. 1940. Superfetation in viviparous cyprinodont fishes. Copeia 1940(2) :38-91. Uoonton, R. 3. 1973. The effect of size of food ration on egg production in the female three-spined stick1eback , Gasterosteus aculeatus~L. J. Fi sh. Bi 01. 5: 89-96. tar, 3. H. 1974. Biostati stical analybi s. Prentice-Hall, Inc., Englermod Cliffs, N.J., 620 pp. APPENDIX TABLES ..

0) rl r4 r( dOU oual 0 > 0) u :5 0) P GCO 0 +I rl 0) U(Dd O0)M U G tnSd 5 gcQ) WaJW WIdQ

m Q>U 4 0 d G hnnnnhnhn rl COOOOCO 0-Q) uw---\rwE+E w w 0)Ib o\om pt r- ' u aJm0 dW0 2e.k 0) Wbb 0) aJ na 0Ea uzo 0 h +JQL a4.n u Q) 4 B

C: OA* zz:lu 0 u9)rl

B w m m a

a 2 a3rp rl ;f t Q)hU u-4 VI I 8 Lh Oh9) WlL a cP 2 Ud71m~lo w 0 k L le*9) P rl ug9) UCP 3 a V) 0 3 L 9) 0.d 9)WPi L .* w h 0 h W m 0 Sr 4 L L

mmc

C d 00- 4UB U

Q) cw B I 4 P

*ale Ubrl 6 I d P a acb 446 d 4mu ow> 1 P 8 d k6u0 4) 8 d 9) 6 =E EhUQO -10 h C b r- OIQ) Vttnr w 0 0 I4 U&'u 6 0 'uro 9) r 0) D P 3sr uca) 0 Q. 3 '0 WdPD 0Q)o bdh auc r Q) 0 b .- ti u h obrcu d 0 h h h hhhhh h4mhInohO- $ m 4 rl 0 N N m rn qE VWY~VWWVY- 2, - O .- LI 3-Q 'aha OOIE

h Ohhhhh OhQ\OIOonnnn rlIDw4,,q- -m -a0 -o -m -4 zz z C 2g. uao UdC woo 404 d U OOQ U Z;u .* at2 rlcm + n c 0 **a0 (D urn4 QOW -55u C (D B O Oh9) & & YlXQ Table 13. Sumnary of reproductive data for female ~'Bcilialatipinn. from Station 1; collection date 8 October 1977; field number PLRS-77-106. Values in parenthesir in the colunn labeled mean number of embryos per brood are not means but are single obrervatione.

Undevel . Developing Mature Total Normal Size Group Ovaries Ova Ova Nonpregn , Pregnant Embryor /Brood@ m SL N (XI N (X) m w u X C 18-19 6 (100) 0 (0). 0 (0) 6 (100) 0 (0) C

Total 0 c s OU. rl ID UGC 6-60 U rl WIDU r( a t!:t $22

nnnnnnnn n 0 nnnnnnnnnnnn n 0 0 0000000 0 WV~VWWWV\*U~~w

000000000000 0 C 9) 35iU aco Url4 V) U e Q 6-43

Yc 0

.rl (ID oarWJ al 4 l a d d dSsU .a n 0 horn 9)zg $28 9)r-u w I 8 i39) -Bib Q Q r uala 0 430LC >0)C.n rtar u4P) Oal& I rl QWWJ 0 0 Ld .m goI? *2 ,B w O 22

rnhC w 0 w a 0 EZL gap:P aJ cnmc tcc 0

m m

0 0 d l-l

n 0 m w m

-0 m w

n 0 w

n 0 '",

n 0 u

0 d nnnnnnnnn n oo~~-*m0,0,SsIn ;aw w w z~'J~W~0 h) :'8 "I V4 u

m-tmL-bm~mm~m4 r-~?~~~rnmoa u CIhJNCJNmC?C?m 8 rn' nnnnnnnnnnn A oo~~ummoocm(3 A m-1 SSZ ~~WWWWW~e4 J2fs1 - w V V bO nnnnnnnnnn n C~~~"C~J~V\UOQI r( ad- w

nnnnnnnnnn n O~~~NNU~~Ou r;(d~OmQJCObIn*l-mI- nnnnnnnnnnnnn E3!z~www~~cc~~moo ;"1a w

al nnnnnnnnnn ~66Zoo3 a ,,,,-EccZZEZ!2 LZ-V

CCC VI40

hhh 000 WWV

CCC w40

hhh CCt- 000 000 uww nnnnnhhnnhn~n C zz~3c~55czs~zzw

OOOOOCOOOOOO 0

> rcr 4 ** 0 U \D or-b 3419) 0 b b ""Q8C 3 Q) 5 hd u0 Ca '4409) ooe hOIW " 9) d

a (d UY '0 l-l I4 4: H 0 E-J &d$$fig&Rd""*mmm I 8(Y~~hlNmmm*8 00"Ad.&&2Ad.w vr I

dmmbmdmmb 1111111 .!girlm N N N N N mmm

LDF. d OYIP dcuNNCVcu* mmm I I nnnnn n nnnfieQO\drl ~~2;5ZWU,WdVIsaosX nnnnn n OOOOCnn 0 OOOOO~~~O wwwVwwuwwu

h n OOhhh no0nnnnn rlvw~~~~ww~z~~~z2z rl rl 3 4 nhhhnwmmmnhnnnn W~"VVW.#2~ZZ

h n n nnnn

nnnnnnnnn OOOOvlOOOO VwwVwwYwv EZlooooooooooo

'0 0 0 UJ b4 T'P8 :: 8 t PIX B oo~mm~c~\p~m~rl~rl~ drldrl FL)

t: Q) 4 k m m 0) U 2 3 nnnnn a wwwszOrnh * 2a Y; 2' u rl u cb 2 h 10) t4 3.P.l a 0 k 0 9) t4 P P '3 * E Q)a 21 m 0) 0 -d h ru k Pa a3 Q) h m w rlo

4 UJn aJ aJN > .d v 9) & w (d nnnnnnnnnnnnn m *ham zzzww-w~"w

nnn nnnn hhhn 9 Y Q U WYYW

nnnhnnnnnnnnn n U3NNWGNOU~400~V) N Ic)N-N m - wS22=5&=5,- u w 0 (10

nnnnnnnnnnnnn .T 0,h,z2z2~zzez3,vr a0 a0 u w w w w w w LJ Q P)a 0) I4 +do> 0 fn u'u* U l.4 nnn aoEg1 0- P zzzzw,-00"s Id m aI4w 0) 0 LC - k bl u0)P) OPP

rn 0) wins

- .- m NeC

rtca0 U 04s 0 l.4 Yea tnwm a P U 0 ruedha. 0 00 )r c a 4 nn-nnhnhhhhnn Q m OOOOOOOOOOOOC Q a) d ~zzs~szs~~~~z a iqe wwwwwVwwwwVww U l.4 00 8 u 13

P.hnnhh~nhhnhh "=U r) 870 c~1z""88555"555C3 fa- C

b4 d UPID 0 0 Who 84

& '0 C C) Q

4 9

1 m . .g

b4 b4 84 Vc 8

3 0 Z 2 fa

.n CCO 0 .rl 4 0 urn4 a0ao U I c h h -.hhhhhhhhhLIh P- 4 0 FOP WwC, P F W'nmm \O V v w0 wwwwvuvvw w O'gggggzgggk@r la.1'

Table 60. of reproductive data far femle Poecilia latipima fran Statiorr 3; oollectim date 2 March 1977; field nmber -77-43.

UJKmel. -1wins mture Size Grarp Ovaries Ova Ova mnSL N (8) N (8) --N ($1 18-19 3 (100) 0 (0) 0 (0) 20-21 8 (100) 0 (0) 0 (0) 22-23 8 (100) 0 (0) 0 (0) 24-25 17 (100) 0 (0) 0 (0) 25-27 16 (100) 0 (0) 0 (0) 28-29 17 (89) 2 (11) 0 (0) 30-31 23 (92) 2 (8) 0 (0) 32-33 19 (90) 2 (10) 0 (0) 34-35 19 (100) 0 (0) 0 (0) 36-37 6 (75) 2 (25) 0 (0) 38-39 3 (100) 0 (0 0 (0) 40-41 6 (100) 0 (0) c (0) 42-43 5 (100) 0 (0) 0 (0)

44-45 1 (33) 2 (67) 0 (0)

46-47 1 (100) 0 (0) 0 (0) 48-49 0 (0) 1 (100) 0 (0)

Total 163 152 (93) 11 (7) 0 (0) 163 (100) 0 (0) hhhhhhh 0- h rltnom OV)r)R%~wQOmO oooom I nnnn n

nnnn n n *. n OOOnCnhhnOO C h OOO~OOOOO~~~~0 w hhnnn 0003000 n P h OhhtnnUO

X b*% ,-I BasI cu tv rn Q 4

onnn ~ornrnnnnhn~nn~. Table 66. Summary of reproductive data for female Poccilia latipinn8 from Station 3: collection

date 6 September 1.977: field number PLRS-77-93. Values in parentheris in the column

labeled mean number of embryos per brood are not means but are ringle obaorvationo (continued).

Undevel , Developing Mature Total Size Group Ovaries Ova Ova Nonpregn . Prognan t Embryos/Broods mrn St -N X (X) N (XI lo (X) N (X) X s

Total I-. m I 9 m d*mwl.-OOggggggg~g~ -nnn - nnnnnnnn Table 68. Sunmary of reproductive data for female Poecllia latipinno ftam St84tiam 3; w3lectiola. d.Le.

5 November 1977; field number PLRS-77-111. (Continued)

Undevel . Devel oping Mature Total Normal Size Group Ovaries OIJ~ Ova Nonpregn . Pregnant Embryos/Broode rn- SL N (Z) N (X) N (X) N (%) X s

ui= 3 Total 158

* Only 3 of 8 pregnant females carried broods that were normal. Five females had broods with unusually large

percentages of abnormal embryos or unfertilized eggs. Overall, out of 68 total embryos sampled, only 28 (41%)

were normal. Thus, no statistics were calculated. .- ,4 m o U C 0 rl I u5. u-nnnnnnn-~nn- cP (0 5": 000 0000 UCC cd 2.4WUc-UW-2Z m -ri 0 - C M a, r-r z OOOOOOOOOOCO a hhhhhhhhh85E333558b~ nD Un nnnnnnnnnnnn eewwwwwsVw-w00000 0000 Ei:d Table 71. Summary of reproductive data for female Poecilia latipinna from Station 3; collection date

G February 1978; "ield number PLRS-78-128. (Continued)

Undevel. Developing Mature Total Normal Size Group Ovaries Ova Ova Nonpregn. Pregnant Embryos/Broods mm SL -W N (%) N (X > N (%) N (X) N (%) X S

1 1 (100) 0 (0) 0 (0) 1 (100) 0 (0)

0 0 (0) 0 (0) 0 (0) 0 (0 ) 0 (0)

1 0 (0) 1 (100) 0 (0) 1 (100) 0 (0)

1 1 (100) 0 (0) 0 (0) 1 (100) 0 (0)

Total nnnnnnnnO~~i3CoaUY" w 2 z z

nnnnnnnn o~o~-ommo w haacvrlw wwwww

n 0 w

n 0 0 d w

cC)

n 0 u

n 0 0 d w r'l r'l " I rim .I la 1 0 * Z bl 01x1 d 1

'0 0 0 rn & 4 rn 82 1, 0 0 h z 1, PIX d

nnnn 00-0 u0100

nnnn 0 G 0 0 a-mw V V

nnnn EZZZ

- hnnn oh0 oorlo uc alo awrlw P #4 QEO

mm-m d Q & uuuu 8 . n mom F1mCU w nnn moo - m3m m WWV

P, 0 E$LI-'. Table 82. Surnrary of mthly wllections at three sailfin mlly study staticns.

V,W Study Statim (Staa Octotrr 1976 March 1977 April 1977

No. of sites scincd 3 3 4 No. of seine hauls 3 3 6 Total selnlnq tm (min.) 2.22 1.70 1.85 Avq. IK). fls! caimht mr min. 315.73 31 1.73 53.93 95.14 22.47 24.32 23.03 45.95

No. of sites seined 3 2 No. of seine hauls 4 2 Total scming tmt? (m.1 3.34 1.42 Avg. no. f lsti cauglt per min. P2ci!la L~tipl~a 106.29 646.40 Gzf~l!si;iaffinis 164.07 516.20 Lucm1,i p-w3 ------. 5.09 100.70 c2>rifi~lonvnr iega t us 4.49 14.79

1Icach Ro;ld Sttily Station (Sta. 3)

No. of sites sch~cl 3 2 2 No. of seme hauls 3 4 1 Total seining ture (min.) 4.23 8.17 2.54 Avq. no. flsh cauqht- pcr- min. Poecllla latlpinna 174.00 51.77 518.50 Gwbusla aff rnis - 131.82 71.65 Lucanla pa - 7.22 10.63 %rmmbn vx~egatus - 53.24 190.16 C .d E wL - a c '4 ",- E -3 u m YV) aJ L, .F "5 c .J~"V: w * .d bOh ::.5. U.d GO .d aJ .4 c VjvIaJ ffl w rulr aL 0 Od m m L .Ual do03 zzw4 rn r-i 3 .2- a' C -4 Ycn w 0 z0 drld IncuOCJ Ncn cn u 4 uuln -404 \DO 0 Table 87. The fzequency ,i pregnancy among adult female ---Poecilia la9inna from three study etat'.)ns for 1977 and 1978. Adult femala.; are

those eqtx.1 to or exceeding 22.0 mm SL for VABI and TOWY, and

equal or exceeding 24.0 mm SL at BCHRD.

Month, Year VAB I TOWY BCHRD

Apr., 1977

May, 1977

Jun., 1977

Jul., 1977

Aug., 1977

Sep., 1977

Oct., 1977

Nov., 1977

Apr. , 1978

May, 1978

Jun., 1978

Jul., 1978

Aug. , 1978

Sep., 1978

Oct., 1978

Nov., 1978 "'I9 du G)

n4 i*1 2 21

m.. m.. Table 89. Standard length frequency dirtribution for mn-prclnrnt lrti~innbin monthly #amplo# from Station 2.

St.~tion2 Oct.. 1976 8 SJV., 1976 17 kc.. 1976 16 Jan.. 1977 21 Fzb., !977 2 0 Ym., 1977 81 t\?r., 1977 18 >lay, 1977 - J~ne. 1977 10 Jiily. 1977 22 AStg., 1977 8 Sspt.. 1977 15 Oct., 1977 16 nw., 1977 17 b.c., 1977 11 Table 30. Stadrrd length frequency dirtribution for non-pre~nrnt Poecillr lrtlninnrr in monthly rrmples from Strrlon 3.

Station 3 Oct., 1976 Nov., 1976 Dcc., 1976 Jan.. 1977 Frb.. 1977 Mar.. 1977 Apr.. 197; )lay. 1977 June, 1977 July, 1977 hug., 1977 Sept. ,1977 nct., 1977 ?= xov.. 1977 Dec.. 1977 Table 91. Standard len~thfrequency distribution for non-pregnant Poecilia lacl~innala loothly smplea from study a:aClOm 1 (vuf), 1978

Station 1 Jan., 1978 Fcb., 1978 ILar., 1978 Apr., 1978 nay, 1978 June, 1978 July, 1978 Aug., 19i8 Sep., 19 78 Occ., 1978 Nov., 1978 Dcc., 1978 Table 92. Standard length freqwncy dirtribution for mn-prasnant Poecilia leti~innain monthly ramples from study rtatfon 2 @OW), 1978.

Station 2 Jan., A978 Feb., 1978 Uar., 1978 Apr., 1978 Hay, 1978 June, 1978 Jl~ly,1378 Aug., 1978 Sap., 1978 oet., 1978 Nov., 1978 ?'w Dec., 1978 QI Tabie 93. Standard lenath frequency distribution for non-pregnant Poecilia lrtipinnr in monthly ramplem Crm 8tudy statfan 3 (mm), 197s.

Scaiior. 3 Jan.. 1978 Feb., 1978 Yar., 1978 Apr., 1978 Hay. 1978 June, 1978 July. 1978 Aug. , 1978 Scp.. 1978 Oct., 1978 Nov.. 1978 Dcc., 1978 P 1 e+ Nu1 n ruI I rnn I NIIIrlP11 4 N 0 4 4 E 4 e 4 GI 021 I I.Dd In I Innu I l I I0c')l rn 4 4 d Z

- - .- ...... r.. r.. I.... I." rr...- C.. . .>>r.r.c.. . -LI*.u.C:. . . .PI>..., . . r* oY>bxc~ZCY OW >: XC.-.S.CU C.1 >L.XC-i+CU > UbOrL4- UU.-UC na=aaw~~-.u.~npz~z~.- =c&.rz- zrz;.:i.;l

S:.?cion 1 tipr.. 1978 ?k1y, L9 78 June, 1'178 July. 1978 ,lug., 1975 Scp., 1978 Oct.. 1'178 Sov., 1978 Starion 2 Apr.. 1978 Zlay. 1378 June. 1978 July, 1978 Aug.. 1978 tor Sep.. 1978 w Oct., 1978 Xov., 1978

Scstiolh 3 Apr.. 1978 Hay, 1978 June, l?78 July, 1978 Au;., 1978 S*p. . 1378 Oct.. 1978 SOY.. 1978 Table 96. The mean number of normal embryos in broods of pregnant

females from three study areas, 1977 and 1978.

VABI maY BCHRD fi S.D. I S.D. N ji S.D. N

Apr., 1977 15.0 6.75 5 35.0 7.00 3

June, 1977 17.2 9.64 93

July, 1977 10.0 2.97 6

Aug., 1977 11.6 3.05 28

Sep., 1977 12.5 7.81 30

Oct.. 1977 20.8 9.64 45

Apr., 1978 14.5 5.71 33 17.1 8.17 27 21.5 7.46 84

May, 1978 6.7 1.49 25 10.8 5.22 85 21.8 8.64 100

June, 1978 15.5 13.05 28 10.1 4.03 89 17.3 8.68 19

July, 1978 8.3 2.18 57 13.3 5.08 46 14.5 0.71 2

I Oct., 1978 5.7 2.52 28 9.4 5.14 19 14 .O 6.97 74 , Nov., 1978 5.2 4.73 4 - - - 14.2 3.54 4 Table 97. The mean number of normal embryos in brood6 of pregnant

females 28.0-33.9 am St from three study areas, 1977 and 1978.

- VABI - TOWY - BCHRD -X -N -X -N -X -N Apr., 1977 19 .O 2 10.6 23 - -

June, 1977 15 .O 6 2 11.4 8 - - July, 1977

Aug., 1977

Sep., 1977

Oct., 1977

Apr., 1978

May, 1978

June, 1978

July, 1978

Aug., 1978

Sep., 1978

Oct., 1978

Nov., 1978 mmuawrl-' oort(uu0 ooa01uo 00dhnNO O????? doooood8

OhnNaOmmrD ???"l?"29 OOOOOOO Table 99. . Statistical data for linear regressions of brood size on female Size for each monthly sample .:f Poecilia Latipima at three study areas in 1978.

Coefficient of Slope Intercept Determination Significance -I

Station 1 (VABI)

April May June J~Y August Septelober Cctober

Station 2 (TOWY)

April

June July August September October

Station 3 (BCHRD)

Apri 1

June July August September October Table 100. Statistical data for linear regressions of brood size on female site for monthly and pooled samples of Poecilia latipinno at three study areas in 1977. These data correspond to the graph liner shown In text Figure 12.

Coefficient of Slope Intercept Determinat ion -N Stat ion 1 (VABI)

Augu a t 0.78 -13.10 0.67 4 5 Pooled (~pr., Nay, Sun. , 1.60 -32 99 0.73 2 27 Sep., Oct.)

Station 2 (TOWY)

October 0.53 -10.03 Pooled (Apr., May, Jul., 0.86 -16.36 Aug., Sep.)

Station 3 (3CHRD) Pooled (May, Jun ., Aug ., 1.18 -24.04 Sep., Oct.) Table 101. Statlatical data for ].inear regrereions of brood sizc on (. .! famale alee for monthly and poolei -smples of Poecilla 1b:i~inna at three study areas in 1978. These data correspond to the graph lines shown In text Figure 13.

Coefficient of Slope Intercept Determination -N t Station 1 (VABI) June 2.10 -46.33 0.57 55 October 3.59 -11.51 0.74 30 Pooled (Apt., Nay, Jul., 0.79 -14.35 0.61 334 Aug., Sep.)

July 1.03 -24.06 Pooled (Apr ., May, Jun. , 0.85 -16.08 Aug., Sep., Oct.)

Station 3 (BCRRD)

, . Pooled (Apr ., ~y,Jun. , 1.32 -28.86 ; d Aug., Sep., Oct.) t E , I US , 'I 7, " , 7 , ,,. , " " ,,, .. , ' P.' I .- .- , "t k ,\ ; , , , , ,.lp ,' . . . . , I,, % w, ,, , . . . - . . . -...... *. - . . , . . . .*. .. -.-.*--- -,.--.,-,-- --.--- ...-. -..------. --"- . . ,- -. - .,. ,- + . .-...... < ..- *- .--, , . .- . . . . -.- .. I

Table 10 2. SIllnrsry of avirrrrmnccll ~aramtemat tJn Vm Study St.& (St.- 1).

Salinity 9.0 ept 9.5 opt 8.0 ppt 7.0 ppt 7.0 ppt

Wter level Oumge 0 -4 cm 40 0 an -4 an

7 Dissolved a pH

Nitrate Nitrogen Table 183. Sunmary of Environmertal Parametere at the VAB Study Station (Station 1). ?he July periphyton

chlorophyll mean is bast:d on three dateminations, not fo~rr. B1anlro in the 3.t~ rerulted from equipment failures.

---Parameter May 1977 June 1977 July 1977 Augurt 1977 September 1977

Water Temperature 24. OOc 30. IIOC 30. OOC 32. O°C 29, O0c Monthly Water Temperature

Average Daily Min. 22. OOC 24.g°C 26. OOC 30. O~C 28. OOC Average Daily Max. 28. O~C 28. 13'~ 32. OOc 32. O0C 30. 0.C Average Daily Median 24.0 C 26 .13Oc 30. OOC 31.0 C 29.0 C Salinity 10.0 ppt 35.0 ppt 28.0 ppt 29.0 ppt 30.0 ppt

Water Level Change -18 cm +4t3 cm -6 cm -12 cm +42 cn

Turbidity 23.0 NTU 4. L3 NTV 4.3 Nn; 7.9 NTU 3.2 NTU -I Dissolved Oxygen 2.3 ppm 2.:; ppm 2.2 ppm 7.9 ppm 3.4 ppm S 8.1 -..------8.0 7.3

Nitrate Nitrog. n ---- 0.235 mg/l ---- 0.065 mg/l 0.119 q/l Total Chlorophyll 2 Periphyton* 0. 0025mg/cm2/day 0.0014111g/c1n~/da~0. 0026mg/em2/day 0.0063mg/cm /day 0.001&~/cm~/da~ Phytoplankton 5.9 16 mg/m3 1.304 mg/m3 0.761mg/rn3 21.628mg/m3 2,980 mg/m3

* Mean of four determinations , 3 I.. ,, , ., , 1. . .

. .... -...... -- . -,.-_..,..., _*._ . _^ _

Table 104. Su~araryof Environmental Parametera at the VAB Study Station (Statfon 1). B&ah in the data resulted from equipment failurer.

Parameter Oct. 1977 Nov. 1977 bc. 1977 Jan. 1978 Feb. 1978 Mar. 1978 Water Temperature 25.OOC 26. OOc 20 .PC 17.WC 14.0% 18.WC Monthly Water Temperature

Average Daily Min. 26. OOC 22 ,OOc 20. 0% lZ.O?C 12.OoC 13,WC Average Daily Max. 30.O0c 26. OOc 22. 0% 14 .o"C 14.OOC 1S.WC Average Daily Median 28. OOc 24. OOc 21 .OOc 13.00C 13.WC 14 .OOC Salinity 28.0 ppt 30.0 ppt 19.0 ppt 21.0 ppt 38.0 ppt 15.0 ppt

Water Level Change ------14 cm 0 cn -4 cm +5 aa Turbidity 2.5 NTU 15.0 NTU 1.9 NTU 1.3 NTU 1.1 WTU 1.2 m7J

? Dissolved Oxygen Ppla 7.2 ppm d 4.0 ppm 5.2 PP 5-8 PPm 9.2 ppm 9.2 8.6 8.2 7.6 8.0 8.2 7.8 Ortho-phosphate 0.070 mg/l 0.092 mg/l 0.014 mg/l 0.070 mg/l 0.052 mgll 0.023 mg/l Nitrate Nitrogen 0.088 mg/l 0.108 mg/l 0.042mg/l 0.082mgll 0.001 mg/l 0.092 mg/l Total Chlorophyll

Periphyton* O.OOIZ.~/ 0.0$15.~/ 0.0 lo rnS/ 0.0 05 mg/ 0.0404 mg/ 0.0 08 mg/ cm2/day rn /day cm? /day cm 4Ida3 cm /day cm 8/day Phytoplankton --- 2. 565mg/m3 1.69 7mg/m3 0 mg/m 1. 804mg/r3 1.8 12mg/m3

* Mean of four determinations Table 105. Summary,of Environmental Parlmetera of the VAB Study Station (Station 1).

Parameter Apr. 1978 May 1974 June 1978 July 1978 I Water Temperature 25. o°C 27 .oOc 30. 5'~ 32. OOC Monthly Water Temperature Average Daily Min. 19. 0:C 23 .O;C 28.0;~ - Average Dai ly Max. 23.O0C 28. O0C 33.OOC - Average Daily Median 21.0 C . 26.0 C 31.0 C - Salinity 15.0 ppt 21.0 ppt 18.0 ppt 17.0 ppt

Water Level Change , -13 cm -18 cm +16 cm -4 cm Turbid1 ty 4.1 NTU 5.7 NTU 1.7 NTU 1.8 NTU 7 Dissolved Oxygen 3.2 ppm 1.5 ppm 6.8 ppm 2.6 ppm a0 PH 8.0 8.1 8.6 8.7

Nit rate Nitrogen 0.026 mg/l 0.000 mg/l 0.000 mg/l 0.000 mg/l

Total Chlorophyll Periphyton* 0,0007 mgl 0.0022 mg/ 0.0019 mg/ 0,0009 tug/ cm2/day cm2/day cm2/dey cm2/day Phytoplankton 1.604 m8/m3 0.750 mg/m3 6.308 mg/m3 1.686 m8/m3

*Mean of four determinations. Table 106. Summery of Environmental Paruneterm at the VAB Study Station (Station 1). Blanks in the data resulted from equipment failures.

Parameter Aug. 1978 Sept. 1978 Oct. 1978 Nov. 1978

Water Temperature 3O.O0C 31 .O°C 27.0°C

Monthly Water Temperature Average Daily Min. 28.0°C Average Daily Max. 32 .O°C Average Daily Median 30.0°C

Salinity 12.5 ppt 13.0 ppt 14.0 ppt 12.0 ppt 12.0 ppt Water Level Change +12 cm

Turbid1ty 1.0 NTU 1.5 NTU 1.2 NTU

Dissolved Oxygen 4.8 ppm 2.0 ppm 4.4 ppm 8.1 ppm 3.6 ppm

8.5 8.6

Ortho-Phosphate 0.011 mg/l

Nitrate Nitrogen

Total Chlorophyll Periphy ton* 0.0005 mg/ 0.0007 mg/ cm2/day cm2/ day

Phytoplankton 1.069 mg/m3

* Mean of four determinations hhla 107. Sunnuy of Onvimrm?ntal Parastcrs at the ltw-way Study Statim (Station 2).

Pamr'.otcr Cktxher 2976 Novarber 1976 Dsoerber 1976 January 1977 ?&nary 1977

water mmyerature 23.0% 20.0"~ 15. OOC 16.e 16. @C

mmly Water Ttqieraturo Average Daily Min. - Avorage OaiPr Max. - Average Daily wan - Salinity 3.0 ppt 3.0 ppt 3.0 ppt 4.0 ppt 2.5 ppt mter 1-1 change 0 -2 an t4 an 0 an 0 an > Table 108. Summary of Environmental Parameters at the Tow-way Study Station (Station 2). Blaakcr in the data resulted from equipment failures.

Parameter -May 1977 June 1977 July 1977 wust 1977 September 1977

Water Temperature 25. OOC 26. 0°C 30. O0c 32. OOC 29. OOC Monthly Water Temperature

Average Daily Min. 24. OOC -- Average Daily Hax. 29. OOc -- Average Daily Medirin 26. OOC -- Salinity 3.0 pp~ 5.0 ppt 2.5 ppt 6.0 ppt 7.0 ppt

Water Level Change -24 cm +16 cm +12 cm -8 cm +42 cm

Turbidty 2.0 NTU 10.0 NTU 2.2 NTU 4.1 NTU 2.3 NTU

Dissolved Oxygen 4.1 ppm 5.3 ppm 4.0 pym 3.6 ppm 7.2 ppm

PH 8.7 -- -- 8.4 7.8 Or tho-phocphate 0.442 mg/l 0.171 mg/l 0.048 mg/l 0.306 mg/ l 0.062 mg/l Nitrate Nitrogen -- 0.237 mg/l -- 0.062 mg/l 0.122 U!g/l Total Chlorophyll

reriphyton* 0.0023mg/eq2/day 0. 0010mg/cm2/day 0. 0010mg/cm2/day 0.001 1rng!c32/day 0.0011mg/cm2/day ' Phytoplankton 0.835 mg/m 2.525 ing/m3 3.132 mg/m3 0.916 mglm 2.390 mg/m3

* Mean of four determinations ., ,, ,., . , .i., .' , a,,, ,& . . . . I .,,.. . .,,: .,. ... . ,. .I .- . . . '...i... ,A l ,, , , . ,..,,,*A< , I., . I .. . . #. , ,. .lj. , .,,_._,. ., ~~MwwwH.w~NY'~W~*IU*@PI*~,W-(~~)~**WII~~--.-~ . _ , . 8

, Table 109. Sunwary of Environmental Parametere at the Tow-way Study Station (Station 2). Blenb in the data resulted from equipment failures.

Paramet e_r_ Oct. 1977 Dec. 1977 Jan. 1978 Feb. 1978 Water Temperat t:e 25.0oC 20. O0C 18.00~ 7.0% Monthly Water Temperature

Average Daily Min. 28. O°C Average Daily Max. 32. OOc Average Daily Median 30. O°C

Salinity 10.0 ppt 12.0 ppt 7.0 ppt 6.0 ppt 7.0 ppt 2.5 ppt Water Level Change --- -8 cm -8 cm +8 cm Turbidity 2.8 NTU 3.2 NTU 12.0 NTU 1.4 NTU 1.0 NTU

Dissolved Oxygen 6.1 ppm 7.2 Ppm 4.6 ppm 8.5 ppm 10.3 ppm 8.0 ppm pH 9.2 7.9 7.9 8.3 8.2 ,I-I W Ortho-Phosphate 0.145 mg/ 1 0.083 mg/l 0.092 mg/ 1

Nitrate Nitrogen 0.079 mg/l 0.119 mg/l 0.148 mg/ 1 Total Chlorophyll

Periphyton 0.0043 mg/ cm2/daya Phytoplankton --- a mean of four determinations b mean of three determinations c mean of two deteminations Table 11O.Summary of Environmental Paramtters at the Tow-way Study Station (Station 2).

Parameters Apr. 1978 May 1978 June 1978 July 1978

Water Temperature 25. OOC 28. OOC 30. OOC 30. OOC

Monthly Water Temperature Average Daily Min. 21 .O:C 24. O~C 29. O~C 28. O~C Average Daily Max. 24.0 C 27 .OoC 31. OoC 31 .O0C Average Daily Median 22. OOC 26.0 C 30.0 C 30.0 C

Salinity 2.0 ppt 4.5 ppt 5.5 ppt 3.0 ppt

Water Level Change -10 cm -14 cm +22 cm +6 cm

Turbidity 2.8 NTU 3.8 NTU 1.9 NTU 1.1 NTU

Dissolved Oxygen 4.9 ppm 2.4 ppm 2.1 ppm 1.6 ppm 7 d 7.8 A PH 7.8 8.0 7.9 P Ortho-phosphate 0.000 mg/l 0.052 mg/l 0.294 mg/l 0.052 mg/l

Nitrate Nitrogen 0.071 mg/l 0.018 mg/l 0.097 mg/l 0.000 mg/l

Total Chlorophyll Periphyton* 0.0019 mg/ 0.0081 mg/ 0.0052 mg/ 0.0036 mg/ cm2/day cm2/day cm2/day cm2 / d ay Phytoplankton 3.622 mg/m3 3.454 mg/m3 2.520 mg/m3 1.510 mg/m3

*Mean of four determinations. Table 111. Sumnary of Environmental Parameters at the Tau-way Study Station (Statfoa 2). Blanks in the data resulted from equipment failures.

Parameters Aug. 1978 Sept. 1978 Oct. 1978 Nov. 1978 Dec. 1978

Water Temperature 30 .O°C 29.0°C 26.S°C 23.0 'C 24 .O 'C

Monthly Water Temperature Average Daily Min. - Average Daily Max. - Average Daily Median - Salinity 1.0 ppt 0 PPt 0 PPt 1.5 pyt 0 PPt Water Level change +13 cm -27 cm +5 cm + 8 cm -9 cm

Turbidity 1.8 NTU 2.8 NTU 0.9 NTU 0.8 NTU 2.0 mu P Dissolved Oxygen 0.5 ppm 2.5 ppm 3.0 ppm - 6.9 ppm 4 d VI PH 7.7 7.8 7.9 9.4 8 .O Ortho-Phosphate 0.126 mg/l 0.011 -11 0.113 mg/l 0 .OlS mg/l 0.003 mg/l Nitrate Nitrogen 0.021 mg/l 0.033 mg/l - 0.071 mg/l r ,'5 mg/l Total Chlorophyll Periphyton* 0.00 30 mg/ 0.0024 mg/ 0.0 44 mg/ 0.0027 mg/ 0.0048 mg/ cm2/day cm2 /day cm S/day cm2/day cm2/day

Phytoplankton 6.574 mg/m3 10.908 mg/~3 3.351 mg/m3 2.543 mg/m3 2.002 mg/d

* Mean of four determinations

Table 113. Sumcoary of Environmental Parameters at the Beach Road Study Station (Station 3). Blanks in the data resulted from equipment failuree.

Parameter May 1977 June 1977 July 1977 August 1977 September 1977

Water Temperature 21.5~~ 29. o°C 26. o°C 30. o°C 30. OOC Monthly Water Temperature

Average Daily Min. 20. OOc 21.0~~ 28. o°C 29. OOC 28. OOC Average Daily Max. 31. OOc 34. o°C 34. O°C 35.0:~ 33. OOC Average Dai l.y Xedian 25. g°C 28. OOc 32. o°C 32.0 C 30. OOc

Salinity 20.5 ppt 75.0 ppt 27.0 ppt 35.0 ppt 33.0 ppt Water Level. Change -12 cm +22 cm +2 cm t2 cm +14 cm

Turbidity 48.0 NTU 42.0 NTU 14.0 NTU 24.0 NTU 10.1 NTU

Dissolved Oxygen 0.3 ppm 2.6 ppm 1.9 ppm 3.8 ppm 4.2 ppm 7A A 4 pH 7.5 -- -- 8.7 8.0

Nitrate Nitrogen -- 0.263 mg/l -- 0.065 mg/l 0.062 mg/l Total Chlorophyll

Periphyton* 0. 0016mg/cm2/day 0.00171c~/e$/day 0.00191n~/cm~/da~0. 0020mg/cm2/day 0. 00100g/em2/day Phytoplankton 12.935 mg/m3 6.667 mg/m 4.223 mg/m3 2.589 mg/m3 3.383 lag/rn3

* Mean of four determinations Table li4. Slrmmary of Environmental Parameters at the Beach Road Study Station (Station 3). Blanks In the data resulted from equipment failures.

Parameter Oct. 1977 Nov. 1977 Dec. 1977 Jan. 1978 Feb. 1978 Mar. I978

Water Temperature 24. o0C 24.5OC 19.0~~ 19. OOC 6.5Oc 17 .oOc Monthly Water Temperature

Average Dai ly Min. 29. o0C 21 .o°C 20. o0C 16.0'~ 11.0~~ 13.0'~ Average Daily Max. 32. o0C 26. o0C 22. o0C 19 .O~C 14. OOC 16. O~C Average Daily Median 30.0~~ 24.0'~ 21.0~~ 18.0 C 13.0"~ 14.0 C Salinity 40.0 ppt 38.5 ppt 27.0 ppt 28.0 ppt 29.0 ppt 16.0 ppt

Water Level Change ------+4 cm +4 cm +2 cm -6 cm

Turbidity 18.0 NTU 3.3 NTU 1.7 NTU 19.0 NTU 9.5 NT'J 9.3 NTU 3= --I Dissolved Oxygen 4.8 ppm 4.6 ppm . 4.2 ppm 3.4 ppz 9.4 ppm 3.6 ppm Q,

Ortho-Phosphate 0.385 mgll 0.140 mg/l 0.140 mg/l 0.101 mg/l 0.011 mg/l 0.044 mg/l

Nitrate Nitrogen 0.093 mg/l 0.062 mg/l 0.050 mg/l 0.062 mg/l 0.002 mg/l 0.094 mg/ 1

Total Chlorophyll

Periphyton*

Phytoplankton

* Mean of four determinations Table 115.Summary of Environmental Parameters at the Beach Road Station (Station 3).

Parameters Apr. 1978 May 1978 June 19z July 1978

Water Temperature 25. OOC 27.5Oc 30. OOC 31 .oOc

Mor.:hly Water Temperature Average Daily Min. 20. o:.: 21.02 23.0:~ Average Daily Max. 24. OoC 29. O0C 34. OoC Average Daily Median 22.0 C 25.0 C 28.0 C

Salinity 18.0 ppt 20.0 ppt 26.0 ppt 23.0 ppt

Water Level Change -8 cm -16 cm +6 cm +11 cm

Turbidity 13.0 NTU 28.0 NTU 15.0 NTU 13.0 NTU

Dissolved dxygen 7.8 ppm 1.9 ppm 3.8 ppm 2.8 ppm

Nitrate Nltrogen 0.033 mg/l 0.018 mg/l 0.000 mg/l 0.009 mg/'

Total Chlorophyll Periphy ton* 0.0022 mg/ 0.0020 mgj 0.0020 mg/ 0.0048 mgj cm2/day cm2/day cm2/day cm2/day Phytoplankton 3.910mg/m3 2.578mg/n3 10.474mg/m3 2.563rng/m3

*Mean of four determinations. Tabie 116. Summary of Envftonrntal Parameters at tt.2 Beach Poad Study Station (Stat:- 3:. Blanka In the data reR*J red from equip men^ fallurea.

-Parameters - -Aug. 1978 Sept. 197! -Oct. 1978 Nov. 1978 Doc. 1978 3ater Tempt?rature 29 .ObC 28. I0C 25.SbC 25.0.C 24 .OeC

Montnly Watez Temperature A,verrge Daily Min. - Average Daily Mxx, - Averago Caily Median - Salinity 19.5 ppt 20.0 ppt 22.0 ppt 26.; ppt 25.0 ppt

Water Level Change +8 cm -6 cm -8 cm +12 cm -6 cm

Turbidi tv 17.0 NTU 18.0 NTU 1.3 NTU 2.8 NTU 1.5 NTU

Di8.w lved Oxygen 3.2 ppm 4.1 ppm 3.8 ppm 7.8 ppm 5.2 ppm

PH 8.4 8.4 8.1 8.3 9.5

Nitrate Nitrogen 0 mg/l 0 mg,'l - 0.023 r (11 0.092 mg/l

Tea; Chlorophyll fiph:pton* 0.0031 mg/ 0.0023mg/ 0.0041mg/ 0.0046 mg/ 0.0036 mg/ cm2/day cm2/day r- ?!day an2/ day em2 /day Phytoplankton 1.707 rne/m3 5.121 mg/m3 2 * md/m3 0.788 mg/n3 1.341 -/a3

*Mean uf four determinati~na I STANDARD TITLE PACE I / 2. Gemrum Accesa~onNs. 3. R.ctp~.n?*s Cotolog N- '-'m &- CR-163122 1 Vol. 111. Part 2 I I 4. Tith rJ With S. Rw?l Omto Vol. 111, Part 2: A Continuation of base-Line Studies for Enviro~ntallybnitoriny Space Trans- 6. phlrine oqmnixetiw CA portation Systems at john F. Kennedy Space Center i. AV()U(s) 8. P&ly Qgonix.t~onR-t No. F. F. 3nelson. Jr. TH 51-2. Vol. 111. Part 2 9. Per(m~q-ia.tion Now -4 Address 10. WdUnl? No.

Depar-nt of Uioloyical Sciences -- - - ~ - - University of Central Florida 11. Conttoc? or &on? No. P.O. Box 25000, Orlando, Floride 32816 NAS 10-8986 13. Type of R.poc* ond Puiod Corerod

12. -?wing A-V ~~ll..and A~~OSS Environmental Base Line Nat I onal Aeronautics and Space October 1976-December 1978 Adurini stration Uashington, D. C. 20546 14. IIonaonnp A~CVCode HD-B IS. Absne Volume 111, Part 2: Ichthyoloyical Studies, Sailfin Molly Reproduction Study. Part 2 reports the results of a study conducted to determine the appl icabi 1i ty of nmitorinq populations of the siilfin ntolly and its reproductive efforts as re1iable indicators of envi rorunental effects of aerospace activities in the knnedy 3pace Center area. The report concludes that the sail fin nlol ly exper- iences drastic f 1uctuat ions in popul ation and reproductive success and is nGt an appropriate factor for ~nonitoringto establish perturbations of the epl*i- rorrnent due to aerospace activities. *r,plenu?ntary note: Prepared under the sponsorship of the Bioinedical Office, p science Operations, Dr. Ui11 iam M. Knott and Dr. Paul Buchanan.

16. Kqwords Reproduction; Indicator Specicss; Poccil ia Latiyinna; pregnancy; embryos; eggs

)TAR ~ateyory51 Unl imited

--- - I - 19. Securmtv Clars.?.(of rhtr report! j 10. Suur,tr Claaslf.(of the: pope) 11. No. of Pop.. 21. Proc* br1~1as~ltied ! None