USVI Public Records

Document text

'- ; "\1 IMPACT OF STORM FLOODING IN THE MANGROVE LAGOON Final Report Prepared for: Department of Conservation and Cultural Affairs Government of the Virgin Islands By: Island Resources Foundation, st. Thomas, Virgin Islands July, 1979 IMPACT OF STORM FLOODING IN THE MANGROVE LAGOON by Maynard Nichols and David Olsen with Robert Franks and Galen Thompson Island Resources Foundation Box 4187, St. Thomas U.S. Virgin Islands Dr. Edward L. Towle, President Final Contract Report This work was supported by the Government of the Virgin Islands, Department of Conservation and Cultural Affairs, Division of Natural Resources Management under Contract #PC-CCA-129-78. July, 1979 ABSTRACT The tropical wave of October 5-9, 1977 produced a threefold stress on the Mangrove Lagoon and Benner Bay marked by high rainfall and stream inflow. As flood energy dissipated through the lagoon, it reduced salinity to 9.8 ppt and spread freshened surface water seaward 2 km. The large influx of suspended material and excess nutrients with consequent rapid generation of phytoplankton, raised turbidity three times average levels, and increased phosphorous 15 to 30 times. Flooding switched the salinity structure from a well-mixed to a highly stratified regime. The intense salinity strati- fication with consequent reduced vertical mixing, led to complete depletion of oxygen in deeper parts of the lagoon. Despite high runoff, sediment loads were surprisingly small. The flood mainly supplied high nutrient loads which over- fertilized the lagoon and stimulated phytoplankton growth within the system. The lagoon reco~ered from the shock of freshening and oxygen depletion in about 8 days while turbidity and phosphorous content remained high for more than one month. It is concluded that flooding triggered "worst case" con- ditions for water quality with potential ecological conse- quences extending for years. i i TABLE OF CONTENTS Page Abstract............................................... ii 1. Introduction........................................... 1 2. Methods and Procedures................................. 2 Water Qual ity....................................... 2 Biota............................................... 4 3. Storm Stress........................................... 5 4. Lagoon Res pons e. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Water Quality....................................... 7 Temperature...................................... 7 Salinity.......... ............................... 7 Dissolved Oxygen ................................. 12 P hos phorou s. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 13 Turbidity and Suspended Sediment ................. 14 Trace Metals ..................................... 15 5. Statistical Analyses................................... 24 Biota............................................... 24 Water Quality ....................................... 28 6. Recommendations ........................................ 42 Benthi c Sampl i ng. .. . . . . .. . . . . . . .. . . . . .. . . .. . . . . ... .. 42 Water Quality ....................................... 42 7. Referenc es. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 44 8. Acknowledgements....................................... 45 iii I. Introduction IMPACT OF STORM FLOODING IN THE MANGROVE LAGOON Although the Tropical Wave of October 5-9, 1977 was not an unusual storm of hurricane intensity, it produced "worst case" conditions for water quality with potential ecological consequences. The extreme storm runoff introduced more water into the lagoon during a few days of rainfall and runoff than during many months of normal conditions. Where then does the flood-borne water and sediment go? Is it flushed through the lagoon or trapped within the system? And how far seaward did the flooding effects extend? What changes occurred in the water quality and biota? And how long did it take the lagoon to recover? These are among the questions we attempted to answer at the onset of the investigation. Whereas a great many measurements and observations have been made on coastal waters at average conditions of wind, waves and runoff, little is known about the changes that take place when stability breaks down and extraordinary events begin to happen. The purpose of this study was: (1) to determine the impact of storm runoff on the water quality, sediments and biota of the Mangrove Lagoon; (2) to determine what changes take pl~ce in comparison to normal conditions which were recorded in an earlier study (Nichols and Towle, 1977); - 1 - (3) to recommend measures that can be taken to minimize the impact of future floods. 2. Methods and Procedures Water Quality. Stations were located to cover the horizontal gradient of water quality parameters. They include the same stations occupied during former IIbase ll surveys (Nichols and Towle, 1977). Additionally, stations were sited offshore in Jersey Bay to cover the extent of flooding effects as well as in salt ponds, guts and along Turpentine Run, the main stream entering the lagoon (Fig. 1). Field observations began near peak runoff in Turpen- tine Run, October 8, 1977. They were concentrated daily in the lagoon between October 9-14 and weekly between October 17 and November, 1977. Altogether, ten areal surveys consisting of 18 to 27 stations each, and one diurnal survey of three stations, were completed. Additionally, seven areal surveys were conducted during and following, the wet season between October 30, 1978 and January 30, 1979. Instrumentation and laboratory procedures for water sampling and analyses follow those used in the base survey. In brief, water temperature and dissolved oxygen content were measured with an ~ situ probe of a YSI model 54A oxygen meter. The pH was measured on fresh samples with a Beckman model 1009 pH meter. Water samples were returned to the laboratory for analyses of chlorinity and salinity by mercuric - 2 - 640153'W w 640153'W Figure 1. 64°152'W 466 , , ' .. .' .. Nazareth t:l /JOy +410A 1» ~ V 45A .6.20 t( ... , .. ~. ', .... .... 42A 4 21 Jersey Bay 425 426 423 422 4 24 STATION LOCATIONS o 100 200 300 400 500 METERS ,~ ___ J I 0--- . 0'.1 0'.2 0'.3 d.4 NAUTICAL MILES 4 Water + Benthos 640152'W Location of sampling stations for water quality and biotic quadrats. , . .. ~:, . ISO j9i 22A .:.6.1 N N nitrate titration. Suspended solids were analyzed by gravi- metric analyses and Milipore filtration using O.8~ pore size filters. Dissolved reactive ortho-phosphate and total hydro- lyzable phosphate were analyzed by the ammonium molybate and stannous chloride reduction method (U.S. E.P.A., 1976). Laboratory methods for trace metals and organic pollutants in stream water and sediments are given in Appendix 1 of Nichols and Towle (1977). Current velocity was measured with a tethered drogue and stop watch. Aerial observations were conducted October 9, 1977 to trace and photograph the extent of turbid plumes. Biota. Sampling of the benthic fauna and flora consisted of repetitively sampling O.25m 2 quadrats at five stations. The stations were established at the same locations as the former surveys, Figure 1, (Nichols and Towle, 1977). Sampling was performed one week and one month after flooding, October 15 and November 9, 1977 and also after the wet season, February 2 and May 18, 1979. The resulting data, Tables 5 - 9, together with data of the former surveys (Nichols and Towle, 1977) were analyzed statistically to assess their adequacy and to determine: (1) What distinct community associations can be recognized? (2) What temporal ~hanges occur and what is their significance? (3) Are lagoon biota deteriorating with time? Collection methods and procedural details follow those described for former surveys (Nichols and Towle, 1977). 4 Statistical analyses consisted of computing: (1) the sample diversity and analysis of variance, and (2) the community diversity with sub-analyses of sample size. The Shannon-Weaver sample diversity statistic H was calculated s according to the equation: ( 1 ) where Pi equals the proportion of the ith species in the collection and ns is the number of species in the collection. , The community diversity statistic, H , was calculated after the sample order was randomized, following Pielou (1966) and utilizing the equation: HI = ,,~c 1 ~~-s IP, 1 n (p. ) L 1 = Ll = 1 1 ( 2 ) where nc is the number of quadrats. The units are in nats when natural logarithms are used. The calculation involved creation of a 32 sample by 40 species matrix, calculation of relative species abundance, the Shannon-Weaver calculation for each species, addition of species abundances from sample to sample and recalculation. 3. Storm Stress The torrential rainfall of the Tropical Wave of October 5-9, 1977 produced 6.83 inches of precipitation over a five day period (U.S. NOAA, 1977), Figure 3A. The storm differed from others on record in that the bulk of the rain- fa 11, 5.3 inches or 78 percent of the tota 1 for the event" occurred within 24 hours, October 8-9. High winds, 25 to 40 mph and high wave action accompanied the passing storm. Total 5 precipitation was relatively uniform over a wide area of the island and it occurred more-or-less simultaneously over the watershed. Consequently, extreme runoff discharged into the lagoon first from nearby streams as Benner Run and hours later from distance streams as Turpentine Run which drains the upper watershed. The event was preceded by 4.3 inches of rainfall during a 16-day period. It was followed by a 3- inch rainfall, November 1, 1977 (Fig. 3). The flood crest passed down Turpentine Run about 16 hours after the main rainfall began. It topped the animal shelter bridge in Nadir by about 36 inches and inundated the lower flood plain. Wash lines along the channel where the Run enters the lagoon, indicate water levels reached 12 inches (30 cm) above high tide level. In the lower part of Pleasure Boats Run, washlines reached 30 inches above the main bridge. Elsewhere in the lagoon itself, no abnormally high water was reported. Although gaging stations were not in place, it is estimated from historical records at Mt. Zion on Turpentine Run (Jordan and Cosner, 1973) that a total of 165,800 m3 of water discharged into the lagoon during the main 24-hour rainfall. This amounts to about 86 percent of the lagoon tidal prism, i.e. the amount of water discharged by the tide in a tidal cycle (25 hours). In summary, the storm produced a three-fold stress on the lagoon; first high wind wave stirring, then high inflow from local drainage followed by extreme inflow from the main stream. 6 4. Lagoon Response During progress of the flood and for one week thereafter, lagoon water quality changed markedly with time, depth and with distance seaward. Response of the lagoon is revealed by: (1) the seaward extent to which changes - ... _ .. _9 GCt.!rre9'-____ (2 ) the de v i at ion of mea s u red va ria b 1 e s from mean values, and (3) the magnitude and duration of temporal variations. Comparison of changes in the stress and response variables provide a means to trace the dynamic link between inflow from the watershed and the response it caused in the lagoon. Water Quality Temperature. Temperature distributions shown in charlets, Figs. 5-8 ,indicate that flood waters were warmer o . than offshore water, e.g. by 3.3 C, October 10th. Highest surface water temperature, 31.9 0 C, was recorded at the mouth of Turpentine Run, October 10th. Flooding created thermal stratification with temperatures decreasing downward 0.4 to 1 .SoC per 2 meters of depth. Such a condition, when added to diurnal heating of surface water, tends to retard vertical mixing and buoy up flood water as it passes seaward. Salinity. Flooding produced a sharp decline in surface salinity with freshening extending seaward 2.0 km to Nazareth Bay and Red Point. Within 32 hours after the main rainfall October 8th, inner lagoon salinity was depressed 7 from 33.5 ppt to a minimum of 9.8 ppt. The distribution of isohalines, which serves as an indicator of water movement, Figure 2, indicates most freshened water was extruded through Bovoni Passage, out Benner Bay entrance~southward past Rotta Cay and eastward toward Nazareth Bay. Thus, reefs to the west around Cas and Patricia Cays were relatively free of freshening. The seaward extent of freshening is delineated by the 33 ppt isohaline which coincides with the seaward edge of brown water observed from the air. After inflow declined and salinity reached a minimum, October 9th, salinity began to rise at all stations except number 20 where ,the rise started October 10th. Time variations of salinity (Fig. 3) s.how salinity recovery was faster in Benner Bay, station 7, than in the inner lagoon, station 27B. Eight days after flooding most waters had regained near- normal salinity for the wet season. Flooding switched the salinity structure from a homogeneous well-mixed regime to a highly stratified regime. Figure 4 reveals a layer of freshened water overlying a layer of more salty water below mid-depth. Stratification was so intense at station 27B, October 11th, that salinity changed downward 18.3 ppt within 1.6 meters. Since the effects of freshening were detected as far as 2 km offshore in near- surface water, it is assumed that the freshened flood water overode more salty or residual lagoon water rather than pushed it out of the lagoon and into the sea. In this way benthic 8 <J ...... . ..... . N \3 .• 1 A <>.1 i 0.3 _ 0.4 NAUTICAL MILES SURFACE 64° 52'W 64° 52'W -..... . . :.':=. :: .. : .'. -' .. Jersey TURSI DITY~J T.U. o 100 zoo 300 400 SOO METERS ~ I I ! ! o O.l_.-9.,.~._..9L-_J~UTICAl. MILES SURFACE -<- Figure 2. Distribution of water quality parameters ncar the maximum extrusion of lagoon water, October 9, 1977. A. Salinity, ppt. B. Turbidity, J.T.U. and location of turbid fronts from aerial observations. 6 5 4 in 3 2 WATER T.otal 1--6.83-1 QUALITY TR·ENDS A RAINFALL Ft. Mylner Wintberg . "7':':':-:~ ,. • .. .A................. . ............. _. -'" - - 35 __ ~. /.Jerse y Bay, 20 ............................................... . 30 .:}}}\ , ..... '. .................. _"";::::0~~ Annual Average at 27B 20 12 10 8 JTU 6 4 ; , J I I 5 I I C TURBIDITY 10 15 20 25 5 10 15 OCTOBER NOVEMBER, 1977 .. ~--.. ~~--- --~----~-----------~-~ 20 Figure 3. Temporal variations of stress and response parameters between October 1 and November 20, 1977. A. Rainfall at stations in the Turpentine Run watershed. B. Salinity at three lagoon stations. C. Turbidity at three lagoon stations. , n 10 ems 5 Seaward 99 9 A S A U NIT Y. %0 o I km 279 99 9 • 5 5 SA 21 ."""",- 8 --- 6 -7 ·35.0 ~.-.8 .. -: -: .. 23 . 0 6 , ----- 6_ • B 01 SSOLVEO OXYGEN, mgtl 2 . -- ---~--------~-~-.-'--~--"':".- .... ---'----'---'------ ._._.~c _____ . __ .. ___ .:.._~ __ ~_+.--... ,. __ '" Figure 4. Vertical distribution of water quality parameters seaward from the lagoon head into Jersey Bay. A. Salinity structure, Oct. 9-10, 1977. B. Dissolved Oxygen content, Oct. 10-11, 1977. ! , 11- fauna in dee~er parts of the lagoon, as well as deep reefs off Rotta Cay, were partly buffered from the salinity shock. ' .. However, the pool of unmixed salty water retained in the lagoon, and the consequent stratification, were critical to reducing vertical mixing and thus.l the depl etion of oxygen. Oxygen. Concentrations of dissolved oxygen content in open near-surface water after flooding, display a broad gradient in the range of 6.8 to 8.3 mg/l (Fig. 5). Locally however, in embayments and backwaters, e.g. stations 7 and 27B.1concentrations reached as high as 11.4 and 13.5 mg/l during afternoon hours (Fig. 5). At the same time,near-surface water became supersaturated, 181 to 196 percent. Diurnal range~ October 13. were exceptional; for example at station 27B concentrations ranged from 7.45 mg/l in early morning to 16.45 mg/l in mid-afternoon. Since this condition was ass 0 cia ted wit h h i g h tot a 1 . p h 0 s P hClT 0 usn ear pol 1 uti 0 n sou r c e s , - .. -~ and with high turbidity produced by abundant phytoplankton, it seems probable the marked oscillations of oxygen content were caused by rapid photosynthetic production of plankton such as the blooms reported by Burkholder et.al. (1972). In deeper parts of the inner lagoon, oxygen content was markedly reduced after flooding. Complete anoxia occurred at station 27B, during the morning, October 13th while near- anoxia, less than 1.8 mg/l and 20 percent saturation, was measured at stations 7, 27B, 9C and 9B between October 11 and 14th (Fig. 2B). There is little doubt that the depletion 12 of oxygen is created by the intense stratification which deters vertical mixing and allows for rapid consumption of oxygen demanding organic matter in the bed and water. These processes are supported by the high total phosphorous content in near bottom water at stations where oxygen was depleted (Figs. 5, 7). Therefore, it is likely that the lowered oxygen content is created by a combination of: (1) reduced mixing, and (2) an influx of organic matter from varied sources, streams, sewage or production of plankton stimulated by the influx of nutrients. Oxygen depletion places a heavy stress on benthic fauna by reducing vigor and producing mortality. Phosphorous. Flood water supplied a large amount of phosphorous and nitrogen compounds to the lagoon. Con- centrations of total phosphorous in Turpentine Run reached 4738 ~g/l while total nitrogen reached 4630 ~g/l October lOth (Table 1). Potential sources of nutrient loading were traced by additional samples November 30, 1977 (Table 2), to two treatment plants in the upper drainage basin (Tutu). Moreover, very high concentrations of nutrients were recorded in pond water near the treatment plant at the lagoon head (station 68, Table 2). Besides these proximate sources, there are substantial amounts of nutrients in bordering mangrove detritus and in organic rich bed sediments (Table 3). It seems likely that storm waves and local flooding also released some nutrients into lagoon water from these sources. 13 Concentrations of total phosphorous were 15 to 30 times greater in the lagoon than in offshore water of Jersey Bay. Horizontal distributions (Figs. 5, 7, 8) vary systematically with distance seaward from sources in the inner lagoon. There is a corresponding seaward shift in nitrogen-phosphorous ratios from 2.8:1 in the lagoon to 12:1 in Jersey Bay. This indicates an excess of phosphorous was introduced by flooding. Of the total phosphorous in the lagoon more than 80 percent is in organic combination whereas the rest is present as dissolved or inorganic phosphate which has not been assimilated by production. With time the phosphorous gradient flatten; 15 days after flooding concen- trations were three times higher in the lagoon than in Jersey Bay. There is no doubt that flooding overfertilized the lagoon with consequent adverse affects in stimulating phyto- plankton growth which in turn, produced high turbidity that persisted for months. Turbidity and Suspended Sediments. Flooding increased turbidity most in the inner lagoon. Values exceeded the annual average three fold (Figs. 2,3). Discolored water spread seaward past Rotta Cay and eastward clouding water off Sprat Beach and Secret Harbor at levels of 1.6 to 1.8 J.T.U. (Figs. 2,5). Microscopic examination of material retained on filters indicates the turbidity is caused by phytoplankton and organic matter. Presumably these constituents were flushed from the lagoon during early stages of flooding. Horizontal 14 with copper (Tables 1,2). Concentrations of these constituents probably come from sewage effluents released into drainage of the Run. Extreme values of copper, more than 100 times off- shore water, were recorded from pond water near the lagoon treatment plant, stations 68,69. Additionally, lead and zinc in these ponds were higher than normal. Bed sediments from Benner Bay, stations 7 and 8, contain higher concentrations of copper and zinc than else- where. They are most likely supplied from boat repair facilities but local sewage inputs may also contribute. Lead, zinc, cobalt and oil and grease from pond sediments, station 68, were also much higher than elsewhere. Such con- centrations probably reflect sewage releases from the nearby treatment plant. 16 ".''-~:.., ... ",""",l,t<."~,,,,;;"". TABLE 1. Concentration of Trace Metals, Nitrogen and Phosphorous Compounds in Lagoon Water and Inflowing Stream Water, October 8-14, 1977. Ortho* Tota 1 * Station Sampling NH3-N TKN COD N03-N N02-N P04-P P04-P Cd Cu Pb Zn Number Date mg/l mg/1 mg/l ]JgA/l N IJgA/l N ]JgA/l P IJgA/l P IJg/l IJg/l IJg/l IJg/l 27B 9 Oct. 1977 0.19 0.92 30 ' 93 2.13 5.68 7.99 <0.5 7.2 <5. 1.6 9C 9 Oct. 1977 0.98 2.3d 55 220 29.10 35.20 36.80 <0.5 <0.5 <5. <0.5 7 9 Oct. 1977 0.20 9.69 53 96 1. 07 2.08 3.71 3.1 16.0 <5. 6.7 60 9 Oct. 1977 2.00 2.67 42 218 48.30 31.20 33.60 <0.5 5.4 <5. <0.5 lOA 9 Oct. 1977 0.06 cui 65 92 1.18 2.96 4.97 <0.5 . <0.5 <5. 32.0 70 14 Oct. 1977 0.35 1. 83 64 43 0.55 1.80 5.63 <0.5 1.8 <5. 17.0 N 66 10 Oct. 1977 Tr. Tr. 19 1,360 3.33 6.40 7.35 <0.5 3.6 <5. <0.5 t--I ~~ , 60 10 Oct. 1977 3.43 4.63 50 190 70.60 50.00 50.40 <0.5 3.6 <5. 1.6 20 12 Oct. 1977 0.02 0.65 26 12 0.33 0.50 1. 09 <0.5 1.8 <5. <0.5 23 12 Oct. 1977 0.02 .0.33 26 20 0.1 b 0.18 0.56 <0.5 <0.5 <5. <0.5 60 8 Oct. 1977 0.55 2.39 52 310 10.20 21.60 22.80 <0.5 9.0 <5. 5.0 61 8 Oct. 1977 0.08 1.49 52 670 10.50 10.40 11.40 <0.5 5.4 <5. 3.3 TABLE 2. Concehtration bf Tota1 Nitrogen, Total Phos~horbu~ and Trace Metals in Water Samples from lurpentihe Run Nov. 30 1977 and Salt Pond~. 49, 6S. 69, Oct. 11, 1977. Sample TKN * td Cu TP Ni Pb Zn Number mg/1 )JgA/1 ~ )J9/1 \Jg/l \.Ig/1 )Jg/1 )Jg/1 TRl 1. 04 6.3 TR2 0197 17.0 TR3 0.77 14.0 iR4 0.36 3.0 iR5 39.00 108.0 <0.4 E:016 <2.0 <:3 <1 'rR6 8104 41. 3 N IR1 4.12 39.4 N 64(8) 4.87 45.6 62(9) 2.71 ~6.0 ~O.4 11 .0 10.0 <3 38 60(10) 0.83 3.5 ~9 2.35 7.6 <0.4 110.0 22.0 27 34 68 9.82 31. 2 <:0.4 160.0 22.0 73 38 49 1. 47 5.4 'fABLE 3. Concentration of Trace Metals, Organic Constituents, Nitrogen and Phosphorous Compounds in Sediments and Certain Plants from the Lagoon Area, October 10-14, 1977. Sample T.S. V.S. TKN TP O&G C.O.D. Cd Co Cr Cu Ni Pb Zn Identifi ca ti on % % mg/kg mg/kg mg/kg % )..(g/g ).!g /g .ug/g )..lg/g J.lg/g llg/g llg/9 68 (sediment) 40.65 12.ob 2,930 425 3,300 8.61 0.23 7.6 12.0 11 3.9 110 108 66 67.53 8.59 1,480 265 300 5.21 0.23 3.9 2.0 16 1.6 51 60 7 42.69 11.90 2,250 ~80 . 330 10.90 <0.10 0.4 12.0 120 1.9 22 109 lOA (sediment) 33.26 13.02 3,300 198 190 7.59 <0.20 <0.5 9.4 11 4.8 10 14 65 61. 11 9.37 2,520 408 790 8.94 0.26 4.4 2.8 21 2.1 47 60 16 32.56 15.18 4,520 306 . 615 10.50 <0.20 1.4 14.0 26 2.7 12 29 27A 32.11 15.91 5,080 350 .1,450 12.30 "0.30 3.9 16.0 55 8.7 22 59 N.· 27B 30.13 15.85 3,310 298 420 10.60 <0.70 4.6 16.0 34 8.6 93 47 w 42 72 9A 39.85 12.30 3,180 300 950 9.92 <0.10 6.1 8.9 40 1.3 G 43.46 8.66 1,520 187 610 4.31 <0.40 1.5 17.0 69 <0.7 27 63 49 60 cm depth 80.16 2.75 145 80 50 0.38 <0.10 0.4 1.7 3 0.55 0.53 2 49 (surface) 68.06 8.02 1,140 160 95 2.98 0.17 4.5 5.3 22 1.8 4 11 67 35.29 17.09 7,550 250 1; 190 16.80 <0.10 2.1 3.3 15 1.7 15. 2 8 54.87 7.25 1,660 242 570 4.50 <0.30 <0.6 27.0 170 2.4 22 130 68 Leaves,mangrove <0.10 0.6 <0.5 2 1.1 0.69 3 10 <0.30 <0.6 14.0 15 <0.6 13 21 11 65.36 5.01 560 .. 103. 75. 1. 56 <:0.70 2.5 44.0 3 12.0 6 6.9 lOA Benthic Plants <0.10 <0.1 3.3 2 0.94 2 5.9 All sediment values on dry weight basis. 5. Statistical Analyses of Biota Table 4 summarizes a list of 40 plant species and their relative abundance in the lagoon. These were sampled from 32,0.25 m2 quadrats between January 1977 and February 1979. Table 51ists phytoplankton species and abundance for March 1979. The sample species diversity values calculated for all samples collected between January 1977 through February 1979, are given in Tables 6-9.· The average sample:diversity (Hs) is equal to 0.4539 nats. The coefficient of variation (C.V.) i.e. standard deviation divided by the mean and multiplied by 100, is 72 percent between sample sites and 102 percent between various sampling dates. The combined C.V. for all ·samples is 98 percent (Table 10). When the Hs values for repetitive samples at different stations were analyzed by the two-way analysis of variance, the variability between Hs values for the stations is significant with F equal to 4.72, p<O.Ol. However, the differences between different sampling dates is not significantly different, i.e. F equals 0.43, p is n.s. This finding indicates that statistically valid time series statements are not possible. When the average values are examined for the different study periods, the diversity is lower during the winter months than at other times. Using the same technique for analyses of variance in the wet weight of algae samples, Table 11 shows a signif- icant difference in the standing crop of autotrophic plant 24 species at different stations over the sampling period with F equal to 3.85, p<0.05. However, the difference between different dates within the stations is not significant; i.e. F equals 1.43, p is n.s. Results of the community diversity statistics are summarized graphically in Figure 9 . The H' statistic increases irregularly with the addition of each sample. However, after the twelvth sample the variations disappear and tend to decay exponentially approaching an asymptotic level in the last few samples. Consequently, the technique was modified to analyze parameters of the Von Bertalanffy growth equation following Walford (1946). In this modification the estimate of H~ co~munity diversity for sample i was plotted on the x axis and the estimate for sample i+l on the Y axis. These two values are equal at the point where the community asymptote occurs. By using a regression equation: , I H 't-l = 0.23H t ~ 0.93 (3) H~ is significant at the 0.001 percent level where r equals 0.96 and n comprises 17 pairs. / Furthermore, the H value for the autotrophic components in the Mangrove Lagoon was 3.1240 nats .. This value is close to that calculated for 1972 ichthyofauna from the lagoon~ 2.987 nats, (Olsen, 1978). High species diversity has long been considered to characterize stable climax communities. On the other hand, Connell (1978) showed that high species diversity arises from a mix of species. This occurs when species are present 25 N m HI 3.0~1------------------------------------------------------------------~ -,," ------ 2.5 ~ 2.0 1.5 Start of HI estimate calculation 1.0 0.5 o H (nats) • % Total Species -- H = -0.25 + .82 (Samples) 0° 0 5 10 15 2'0 2'5 30 36 Number of Samples Figure 9. Species diversity HI, and cumulative sP2cies of benthic plants collected from thirty-two 0.25 m quadrats. The dashed line is a least squares regression predictor from 100 80 If) cu o Q) 0- en 60 a ~ o 40 20 from various successional stages at the same time and before interspecific competition eliminates species from previous successional stages. Historical analysis of benthic community patterns surveyed in the Mangrove Lagoon between 1947 and 1977 (Nichols and Towle, 1977, Figures 39-40) indicate a transitional character of community associations. There were greater kinds of species associations in 1977 than in prior surveys. The H' estimate for the present communities is higher than the 95 percent upper confidence limit for the ichthyofaunal community (2.987 + 0.0215 nats). Although different statistical treatments wer~ used it appears that a greater portion of the lagoon is undergoin~ successional changes in 1977 than in 1972. To determine sample size required to obtain the H~ estimate within 95 percent of the actual value, the data were fitted to the equation: H' ~ a+b In(x) ( 4 ) whereby x equals the number of samples. The resulting correlation was significant at the 0.01 percent level (r equals 0.75). The results indicate that to obtain an H' estimate within 95 percent of the asymptotic value previously estimated, and in turn to discriminate between biotic associations, at least 55 samples per association, or over 500 samples per sampling periodJ are required. Betause of the limited sample size, natural vari- ability, especially variability produced by rapid successional changes, in addition to possible.errors introduced through 27 sampling and analyses by different investigators, biological changes between sampling periods of the present data can not be demonstrated by statistical analysis. The sampling variability and the annual variability tend to mask the short-term effects of flooding. Statistical Analyses of Water Quality Since benthic communities are often linked to water quality parameters, the long-term variations of these parameters were analyzed to determine spatial and annual differences between stations. The water quality parameters consist of temperature, salinity, dissolved oxygen concentra- tion, pH, turbidity, and fecal coliform bacteria. Data were taken from surface water at three stations, 26 (A), 27A, 278 (Figure 1)~ in the inner and polluted lagoon zones by the Department of Conservation and Cultural Affairs, Division of Natural Resources, between 1972 and 1978. To establish a comparitive baseline, similar monthly monitoring data were analyzed from nine stations occupied between 1976 and 1978 in open coastal bays along the south coast of St. Thomas, Stations 60, 7A, 78, 8, 25, 28A, 288, 29A, and 30. These bays are generally less restricted and more actively flushed than the Mangrove Lagoon. The parameters change in absolute values as well as in their variability. The absolute differences may create continuing stresses on the organisms while the variability may det~rmine the species occurrence by occa~ionql extreme occurrences. Some stations may have identical 28 Table 4 .Re1ative abundance of flora from thirty-two 0.25 m2 quadrats collected in the Mangrove Lagoon between January 1977 and February 1979. Values are percent of total wet weight in grams. Percent of Total Taxon Wet Weight Spermatophyta Tha1assia testudinum Ha1odu1e wrightii Ha1ophi1a bai10ris Syringodium fi1iforme Chlorophyta Penicillus capitatus Cau1erpa sertu1arioides C. cupressoides C. webbiana C. vickersiae C. ashmeadii C. pro1ifera C. fastigata C. racemosa C. mexicana C. sp. Cladophora howei C. s p. Halimeda incrassata H. monile H. opuntia Auranvi11ea nigricans Chaetomorpha sp. Udoteafl abe11 urn U. spinu10sa U. s p • Acetabu1aria crenulata Unident. filament. Rhodophyta Hypnea musciformes Acanthophora spicifera Spyridia fi1amentosa Ge1idium sp. Graci1aria cy1indria G. sp. Unident. Unident. Unident. Phaeophydeae Dictyota 1inearis D. indica D. divaricata Cyanophyceae Lyngbya sp. 29 21. 90 1. 90 6.30 1. 50 14.10 8.60 3.20 2.60 <0.01 0.90 2.20 <0.01 1. 80 9.00 <0.01 5.00 2.80 3.40 3.80 <0.01 0.10 o. 10 3.90 <0.01 <0.01 <0.01 2.4'0 0.60 1. 10 0.01 <.01 <.01 <.01 <.01 <.01 <.01 0.10 2.01 2.50 0.20 Table 5 .Com~osition and abundance of phytoplankton and microfauna in the Mangrove Lagoon and Benner Bay March 19, 1979. Values are estimated cells per ml. Taxon Crytophyta 8~ Chroomonas sp. 16~ Chroomonas sp. 5~ Cryptophyte Bacillariophyta Skeletonema costutum 14~ Nitzsthia sp. 25~ Nitzschia 10ngissima 60~ Pennate 5~ Cyclotel1a sp. Coscinodiscus lacustris Thalassiosira sp. Chaetoceros sp. Chlorophyta Pyramimonas sp. Pyrrophyta Katodinium rotunda tum Peredinium sp. Gymnodinium sp. Mi crofl agell ates 3~ flagellate 3~ bi fl agell ate 5~ flagellate Tot~l Cells Per ml 1 352 186 10 2 1 83 10 1 269 103 10 1013 30 5A 426 284 52 13 26 26 52 465 1306 STATION 7 827 52 1551 6 52 52 1 103 2 1396 310 52 4343 lOA 1396 776 155 103 1 2 4 1 5 103 2 1 1448 465 4084 27B 4292 3671 1706 207 1 1 52 6 3 5 310 155 92 1 1913 11 013 Table 6 .Composition of ~ sq. m. quadrats collected in Benner Bay and the Mangrove Lagoon October 1 5 , 1977 STATION Taxon 7 9 lOA 27A 27B FAUNA, Numbers of individuals, (Minfmum number of species) Annelida Polychaeta 6 ( 1 ) 10 ( 3) 1 (1) Oligochaeta 3 (1) Mollusca Bivalvia 2 ( 2 ) 24 ( 6 ) 7 (2 ) 22 ( 3) 13 ( 3) Total Number of Animals 8 ( 3) 37 (l 0) 8 ( 3 ) 22 (3) 13 ( 3) FLORA, 0~t w~ight, gms. (dry weight, gms.) Chlorophyta Caulerpa vickersiae C. ashmeadii C. cupressoides v. mamillosa C. cupressoides v. lycopodium 0.03 (0.01) C. prolifera f. obvata C. fastigiata Halimeda incrassata 9.70 (2.73) Udotea sp. U. spinulosa 3.30 (1 .49) Rhipocephalus oblongus Cladophora sp. 34.30 (6.70) 0.68 74.61 (0.13) (10.63) 0.04 (0.02) 3t 82.29 (17.11) 0.05 (0.02) 84.46 (13.90) 4.80 (1.56) 50.27 (15.27) 36.0 (7 .0) Table 6.(Continued) STATION Taxon 7 9 lOA 27A 27B Rhodophyta Acanthophora spicifera 0.03 (tr) Spyridia aculeata 0.34 0.04 (0.02) (0.01) Hypnea sp. 0.03 " (tr) Spermatophyta Halophila baillonis 0.89 0.01 (0.12) (tr) TOTAL WEIGHT OF PLANTS 13.92 35.43 330.52 (4.35) (6.87) (65.50) Number of Plant Species 4 7 8 0 0 (tr) means trace 32 Table 7.Composition of ~ sq. m. quadrats collected in Benner Bay and the Mangrove Lagoon November 9, 1977 Taxon STATION 7 9 lOA 27A 27B FAUNA, Numbers of individuals, (Minimum number of species) Coelenterata Anthozoa 0 0 0 0 0 Annelida Polychaeta 4 ( 2 ) 1 (1) 0 0 0 Oligochaeta 0 6 ( 1 ) 0 0 0 Mollusca Gastropoda 2 (1) 0 0 0 0 Bivalvia 1 (1) 6 (3) 8 ( 2 ) 1 (1) 5 (2 ) Echinodermata Ophiuroidea 0 0 0 0 0 Crustacea Paguridae 0 0 a 0 0 Majidae 0 0 0 0 0 Tunicata 0 0 6 ( 1 ) 0 0 Total Number of Animals 7 (4 ) 1 3 ( 5 ) 14 ( 3) 1 ( 1 ) 5 ( 2) FLORA, wet weight, gms. (dry weight, gms. ) Chloro~h,Yta 0 0 <0.01 0 0 Caulerpa spp. ( 2 ) (t r) '- 0 0 4.76 0 0 C. cupressoides (0.68) 9. 11 0 15.26 0 0 Penicillus capitatus (2.61 ) (4.89) 2.53 0 39.82 0 0 Halimeda incrassata (0.73) (10.13) 0 0 0 0 0 H. mon il e 0 0 5.62 0 0 Udotea flabellum (1.94) 33 50 1 (1) 22 (3) 0 19 (4) 1 (1) 4 (1) 2 (1) 4 (1) 1 (1) 54 ( 13) 0 0 0.13 (0.10) 8.41 ( 4. 00) 22.37 (11.99) 0.43 (0.30) Table 7 .Continued Taxon STATION 7 9 lOA 27A 27B 50 SQermatoQh~ta 0.85 (0.01 0 0 0 0 Ha1ophi1a bai110nis (0.31) (tr) Thalassia testudinum 0 0 0 0 0 865.00 (232.90) Tota 1 Weight of Plants 12.49 <0.01 65.46 0 0 896.34 {3.65 (tr) (17.64) (249.29) Number of Plant Species 3 1 6 0 0 5 (tr) means trace 34 Table 8.Composition of ~ sq. m. quadrats collected in Benner Bay and the Mangrove Lagoon February 2, 1979 STATION NUMBER Taxon 7 9 lOA 11 27A ANIMALS, Numbers of individuals, (Number of species) Mollusca Taegulus divisis 1 Macomapseudoramal Bulla striata 1 Chione cancellata 1 Planaxis nucleus 1 Ringicula semis~/ trata 1 Crepidula convexa Ectoprocta Annelida Po1ychaeta Sa b'e 11 i due tr means trace 2 23 1 2 (1) 9 (3) 35 2 11 (2) 27B 6 7 50 2 (1) Table 8 .Composition of ~ sq. m. quadrats collected in Benner Bay and the Mangrove Lagoon February 2, 1979. Wet weight, in grams, is presented with dry weight in parenthesis. STATION NUMBER Taxon 7 9 lOA 11 27A 27B PLANTS S~ermato~h.ltes Thalassia testudinum Halodule wrightii 20.72 (4.52) Syringodium filiforme Halophi1a baillonis 0.6 (0.12) Chloro~h.lcae Penicillis capitatus 4.7 31.06 4.47 (2.23) (9.15) (3.31) Caulerpa sertularioides Cladophora howel 140.91 7.82 (28.71) (1.54) C. sp. < .01 (tr) Halimeda incrassata 1 36. 77 .04 (36.08) ( .03) Avranvillea nigricans 3.48 (0.60) Chaetomorpha gracilis < .01 (tr) Udotea flabellum 59.27 (21. 80) RhodoQh.lceae Hypnea musciformes Acanthophora spicifera 13.82 4.03 (2.14) (0.45) Gelidium sp. 0.03 (0.02) Rhodophycea unident. < .01 (tr) Phaeo~h.lceae Di ctyota linearis 4.35 0.66 D. indica tr indicates trace 'iF> 50 26.6i1 (4.97) 0.42 (0.12) 56.51 (9.20) 10.76 (5.84) 50.16 (9.95) 0.58 (0.20) 7.1 (2.35) 22.77 (5.52) 91.06 (22.09) Table 9. Com p 0 sit ion of ~ sq. m . quadrats collected in the Mangrove Lagoon and Benner Bay, May 18, 1979 Taxon STATION 7 9 lOA 27A FAUNA, Numbers of individuals, (Minimum number of species) Annelida Polychaeta Mollusca Gastropoda Pelecypoda 3(250) l( 2) 1 ( 4) 2(9) 2(80) 1 ( 1 ) 2(13) FLORA, wet weight, gms. (dry weight, gms.) Chlorophyta Caulerpa cupressoides Acetabularia crenulata Halimeda incrassata Udotea flabellum Penicillus capitatus Rhodophyta Acanthophora specifera Gracilaria cylindrica Phaeophyta Dictyota divaricata Spermatophyta Halophila baillonis Haludule wrightii <0.01 «0.01 ) 2.80 0.21 (0.21) (0.03) LO.Ol 39.73 « 0 . 0 1 ) ( 6 . 4 0 ) 37 51.60 (6.36) . 01 « . 0 1 ) 84.78 (27.65) 44.73 (16.01) 1 3. 71 (4.45) 33.75 (4.40) 0.40 (0.35) 2 ( 1 6 ) 5(43) 27B 3 ( 1 3) 2(45) w 00 Table 10.Sample species diversity ~s from 0.25 m2 quadrats collected from the Mangrove Lagoon, St. Thomas, U.S.V.I., from January 1977 to February 1979. COLLECTION DATE February November October August April January Station 1979 1977 1977 1977 1977 1977 7 0.0133 0.7365 0.7819 0.5385 0.7333 0.5195 9 0.8648 0.0 0.1749 1.2045 0.7472 0.4947 lOA 0.8566 1.0446 1.6227 0.5078 1.0204 0.2100 11* 1 . 1 153 27A 27B 50* x C.V. o o 1 .6144 0.3469 135% o o 0.1752 0.3562 140% o o 0.5159 135% 0.2125 0.3622 0.6179 0.3046 0.0500 0 0.5536 0.5826 0.3684 70% 65% 69% *Deleted from Means and Analysis of Variance All samples combined 0.4934 (98%) ** C.V. is the coefficient of variation ..,. x 0.5538 0.5810 0.8770 0.1988 0.0591 0.8948 0.4539 Between Sites Between Dates (C.V.l** (52%) (77%) (56%) (127%) (206%) (113%) (72%) (102%) w 1.0 Table 11 .Wet weights of algae in grams collected from 0.25 m2 quadrats in the Mangrove Lagoon, St. Thomas, U.S.V.I. between January 1977 and February 1979. Station 7 9 lOA 11* 27A 27B 50* x C. V. February 1979 4. 71 221 .50 189.52 12.89 3.48 0 265.97 83.81 (133% ) November 1977 12.49 .01 65.46 0 0 896.34 18.41 (152%) October 1977 13.92 34.43 330.52 0 0 75.77 (189%) COLLECTION DATE August Ap ri 1 January 1977 1977 1977 155.42 11 9 . 42 1. 40 101.31 5. 10 56.50 470.31 3.91 342.10 93.08 5.10 218.80 1.1 0.90 0.60 164.24 26.89 123.88 (110%) (179%) (122%) *Deleted from Means and Analysis of Variance All samples combined 109.89 (170%) ** C.V. is the coefficient of variation x 51 .52 69.81 233.64 53.41 0.40 81 .70 Between Sites Between Dates (C.V.)** (133%) (119%) (77% ) ( 166%) (138%) (126%) (148%) statistical averages for the same parameter although the individual data may be highly variable and thus place extreme stress on the organisms. As water quality parameters, e.g. temperature, salinity and dissolved oxygen, vary and interact, even without one exceeding normal limits for organisms, they also may produce stress conditions. The results, presented in Table 12, show marked differences between stations. For example, at station number 26A on Figure 1, from Benner Bay, water temperature is higher than the south coast average and salinity is higher than at other stations. This trend differs however, inasmuch as the values at station 26 (A) are less variable than the lagoon stations, the variability being close to the south coast average. The high temperature and salinity at station 26 (A) is probably responsible in part, for the sub-saturation of dissolved oxygen occuring at this station. Stations 27A and 27B at the lagoon head (Figure 1) have large variations in measured parameters. Fecal coliforms are higher at 27A than elsewhere whereas salinity is lower, a trend that reflects discharge of freshwater and nutrients from the nearby sewage treatment plant. Diversity and standing crops of benthic plants are higher at 27A than at 27B. The low values at station 27B may be either produced by sub-saturation of dissolved oxygen content, or associated with the low plant populations. Station 27B has the greatest variability of fetal coliforms, a trend that may relate to nutrient conditions and the low plant populations found at that site. 40 Table 12.Comparison of average water quality parameters for three sampling stations in the Mangrove Lagoon with the average for all south coast stations. Parameter Temperature (oC) N C. V. Salinity (ppt) N C • V. Dissolved Oxygen (mg/l) N C. V. % Saturation pH N C. V. Turbidity (FTU) N C. V • Fecal Coliforms (pr 100 ml) N C. V. Average C.V. except fecal coliforms South Coast Average 41 27.0 231 5.2% 35.7 227 2.3% 6.52 232 7.1% 100.6% 8.25 231 1. 2% 0.59 225 62.9% 0.53 230 621% 15.74 26 A 28.3 41 6.9% 36.03 37 4.4% 5.75 40 10.1% 91 .0% 8.17 37 1. 7% 2.30 39 50.4% 14.9 41 161.8% 14.7% STATION 27A 27.8 42 7.7% 35.16 37 11 .3% 6.44 40 26.6% 100.5% 8.16 37 1. 6% 3.18 40 79.4% 63.6 42 167.2% 48.9% 25.3% 27B 27.6 37 7.4% 35.66 33 6.4% 6.09 35 14.4% 94.9% 8.18 32 1. 9% 3.38 34 71 .4% 23.7 37 203% 78.9% 20.3% 6. Recommendations Benthic Sampling 1. Future efforts in Virgin Island bays should be statistically designed to account for spatial and temporal variations utilizing foregoing results from the Mangrove Lagoon. . 2. Once the design is established it should be followed systematically in identification of species, positioning of sample locations and in laboratory analyses. 3. A chart of benthic community distributions and their boundaries resurveyed from time to time, demonstrates subjectively, the differences in community type's .and their changes with large environmental stresses. 4. For spatial detail, over 50 samples per community association are required. For long-term temporal changes, four repetitive samplings distributed throughout the year with approximately 15 to 20 stations in all subcommunities, are required. 5. To detect short-term changes, 100 samples per association may be required. Simple wet weight measurements from quadrat sampling may be adequate. However, for more cost-effectiveness water quality changes may be assumed to indicate biotic changes. Water Quality 1. Monitoring of water quality in Virgin Island bays could be improved by more frequent measurements of salinity, turbidity, phosphorous and dissolved oxygen, following periods of high rainfall. 2. Since flooding produces intense stratification with consequent oxygen depletion in near-bottom water, future measurements of water quality should include detailed vertical profiles from the surface to the bottom. 3. Salinity shocks on the lagoon should be reduced by reducing peak storm flows with flood detention structures in the watershed. A feasible storm- water management plan for Turpentine Run is provided by Black, Crow and Eidsness, CH 2M Hill (1979). 42 4. The influx of excess nutrients on flood-borne sediment and organic matter should be reduced by controlling sewage effluents at their source, i.e. in overloaded and malfunctioning treatment plants. As suggested by B, C and E - CH2M Hill (1979), inner canals of the race track could be used as settling basins to reduce sediment as well as nutrients in particulate form. 5. The long-term impact of floods on degrading water quality in the lagoon could be reduced by pro- viding better circulation and drainage between the salt ponds, lagoon and Jersey Bay. A complete list of recommendations for management of the lagoon is provided by Nichols and Towle (1977). 43 7. References Black, Crow and Eidsness, CH2M Hill, 1979. A sediment reduction program. Rept. to Gov1t of the Virgin Islands, 196 pp. Burkholder, P. R., Brody, R. W. and A. E. Dammann, 1972. Some phytoplankton blooms in the Virgin Islands. Carib. J. Sci. 12:23-28. Connell, J. H., 1978. The evolution of species diversity. Science, 199:1302-1310. Jordan, D. G., and O. J. Cosner, 1973. A survey of the water resources of St. Thomas, Virgin Islands. U.S. Geol. Survey, open file report, Caribbean District, 55 pp. Nichols, M. and E. Towle, 1977. Water sediments and ecology of the Mangrove Lagoon and Benner Bay, St. Thomas. Island Resources Found.; Tech. Rept. 1, 159 pp. Olsen, D. A., 1978. The ecology and diversity of· fishes in two U.S. Virgin Islands Mangrove Lagoons. Island Resources Found.; Occasional Paper No. 27, 42 pp. Pielou, E. C., 1966. The measurement of diversity in different types of biological collections. J. Theor. Biol. 13: 131-144. U.S. Environmental Protection Agency, 1976. Methods for chemical analyses of water and wastes. EPA 625-/ 6-74-003a, 298 p. U.S. National Oceanographic and Atmospheric Administration, 1977. Climatological Data. 23(10,11,12). Walford,L. A., 1946. A new graphical method of describing the growth of animals. Biol. Bull. 90(2):141-147. 44 8. Acknowledgements This study was supported by funds from the U. S. Environmental Protection Agency and the Virgin Islands Government through a contract (#PC-CCA-129-78) from the Virgin Islands Department of Conservation and Cultural Affairs, Darlan Brin, Commissioner; and the Division of Natural Resources Management, Pedrito Francois, Director. Mr. Robert Van Eepoel of the Department of Conservation and Cultural Affairs served as general program manager. We thank Judith Towle of Island R~~ources Foundation for an accurate fiscal accounting. Teresa Wilburn typed the report and data were reduced by Carl Cerco and Pamela Peebles. Water quality monitoring data for 1972 through 1979 were provided by the Virgin Islands Department of Conservation, Division of Natural Resources. 45