Patterns of Mangrove Forest Structure and Soil Nutrient ... · 'The coastal forested wetlands in the southwest- ern Everglades of' Florida extend h-ore Whitewater ... wetlands in

Estuaries Vol. 22, No. 4, p. 955-970 December 1999

Patterns of Mangrove Forest Structure and Soil Nutrient Dynamics

Along the Shark River Estuary, Florida

RONGHUA CHEN

ROBERT [2~. TWILLEY 1

D e p a r t ~ t of Biogog 7 O%iverd(y of Louisiana at Lafayette R O. Box 42451 Lafayet< Louisia~za 7050~# 2451

.a=gSTRACT: T b e basM area and prodilctivi~y of mangrove wetlands are descr ibed in relation m selected sell propert ies m tmderstand die general pa~tern of optimum forest stature at lfie moutb of esmarles in lfie Everglades~ sucb as die Sbark River Slougb~ Florida (U.S@ The basal area of mangroves decreases from 40.4 m ~ ba i and 89.7 m 2 ha x at ~wo stadons 1.8 km and 4.1 km front tbe estuary moudl to 20.7 m ~ ha * and 19~6 m ~ ha i at two sites 9.9 km and 18.2 km from die moudb re~speetively; T b e gradient in b~sal area at illese four sites is mostly dm result of approximately 34 yr of gTowlfi since Hurricane Detain. Wood producdvity is bigber in dm lower estuary (10.7 Mg ba ~ yr ~ and 12.0 Mg ba yr x) than in lfie upper estuary (3.2 Mg ba * yr ~ and 4.2 Mg ba x yr ~). Porewater s~hd ty among lfiese four mangrove sites during seasonal saunplblg in 1994 and 1995 ranged from 1.6 g kg * to 33.5 g kg *~ wbile sulfide was generally < 0.15 mM at all sites. Tbe.se sell values indicate tbat ablodc stress cannot explain lfie decrease in forest structure Mong rills e~tuarlne gTadienL Concentrations of nitrogen (N) and pbosphort~s (P) are more closely related to pat terns of forest developmen% wilfi hlgber soil fertility at ille moutb of the estuary as indicated by bigher concentrations of extractable ammonium~ total sell P~ arid available E along witb hlgber ammonium production rates. T b e more fertile sites of tbe tower estuary at'e dominated by Iz~g~t~Marla raeemosa, wbereas tbe less fertile sites in the intermediate and upper estuary are dominated by Bhlzopbora *m*t~gle. Reladve N mineralization per unit of total N is higher in tbe lower estuary and is related posldvely to concentrations of available p. indicating the bnportance of turnover ra~es and nutrient interactions m soil fertility; Concentrations of Ca-bound P pe r volume soil in the lower estuary is 40-fold bigber d~an in die upper estuar}l and Mong widi an increa.~e in re, slduad P in the upper esmar N indicate a Mdft from minerM to organic P along lfie esmarine gradienL Mineral inpu~ to dm mou*b of Sbark Rh,er estuary from the Gulf of Mexico (raflmr/ban upland inputs) apparendy control the pat terns of mangTove structure and producdvity=

I n . e d u c t i o n

Mangrove wetlands are the dominan t intertidal communi ) " type of mar ine to h e s h water ecotones along tropical estuaries (Tomlinson 1986) An ecogeomorph ic classification describes the diverse par- terns of mangrove forest s t ructure and ~hnction based on the combinat ion of geophysical, geomo*- phological, and ecological processes within coastal envi ronmenta l se i tmgs (Twilley 1995, 1997) Envi- ronmenta l settings are de te rmined by the na ture of circulation, sed iment transport , regional topog- raphy, phys ica l p rocesses , and c l ima te ( T h e m 1982, 1984; Blasco 1984; Woodroffe 1992) The ecogeomorph ic classification scheme proposes a g~adient in the geophysical energies of a coastal life-zone will result in variat ion in the ecological proper t ies of mangrove wetlands, including productivi/y, nut r ient cycling, litter @namics, succe~ sion, and sedimenta t ion (Twilley 1995, 1997). Mangrove wetlands occur in diverse environrnental

x Corresponding author; tele: 937/482-6146; fax: 337/482- 5834; e-mail: rtwilley@louisimaa edu

settings, including de[te~s, lagoons, and estuaries in wet, humid , or & y climates, all of which vary in specific pat terns of biodiversity, energy flow, and material wcling. We propose that the ecological significance of the ecogeomorph ic classification system can be verified by compar ing pat terns of soil nut r ient b iogeochemis t ry and mangrove structure a m o n g emdronmenta l settings as a funct ion of past and present geomorph ic and geophysical processes (Twilley 1995) Mangrove ecotones in the southwest Everglades of Florida ( U S ) offer an interest ing compar ison of coastal settings influ- enced by carbonate lagoons and estuaries v~dth significant f?:eshwater d i s cha rge

The propert ies of mangrove communi t i es vary within an envi ronmenta l setting depend ing oil the location of mangroves along tile salinity gradient of art estuary and with distance inland h e m shore along the intei t idal zone, Intert idal soils can be quite he te rogeneous because many edaphic characteristics depend on deposit ional and erosional pat terns and on the inf luence of plant communi- ties, which add organic matte*; wcle nutrients, and

�9 1999 Estuadne Research Federation 955

9 5 6 R. Chert and R. R. Twilley

modify soil salinity and redox condit ions (Thorn 1967) Inputs f i o m terrestrial r u n o f f ( ter*igenous sediments) generally- are considered the mqior nutr ient source that supports mangrove forest developmen t (Calter et al. 197,8; O d u m et al. 1982; McIvoi et a l 1994) This hypothesis of terrestrial control predicts that nut r ient availability in mangrove soils vdll decrease f iom the landward zone to the seaward zone along the estuarine grad ien t of a tropical estuary-. Local pat terns of tidal inun- dation will also inf luence soil characteristics that control species zonat ion of mangrove wetlands (Watson 1928; C h a p m a n 1944, 1976; Walsh 1974) Patches and zones of fb~est composi t ion result fi-om complex gradients of hydroper iod and soil conditions, such as nut r ient l imitation (eg . , nitrogen and phosphorus; Boto and Wellington 1984; Feller 1995) and abiotic stressors ( eg . , salinily and sulfide; Cintrdn et al. 1978; Lugo 1978; Nickerson and Th ibodeau 1985; McKee et a l 1988; McKee 199S) Mangrove forest s t ructure in the Caribbean region is also inf luenced particularly by the fi-e- quency of hur r icane disturbance, emphas iz ing the impor tance of temporal variat ion in forcing flanc- tions to mangrove deve lopment (Lugo 1997)

'The coastal forested wetlands in the southwestern Everglades of' Florida extend h-ore Whitewater Bay to Ten Thousand Islands and include one of the lalgest mangrove areas in the Gulf of Mexico (Davis i[940; Smith et a l 1994). The stature of mangrove forests a long the m o r e mar ine zones of estuaries in this region is about 25 m; tree he ight decreases inland toward a mangrove-marsh ecotone inhabi ted by "dwar f " mangroves (Davis 1940; Lugo and Snedaker 1974; Smith et al. 1994) More than 50 yr ago, Davis (1940 p ,8,84) noted that the lower estuaries of this region contain "probably the opt imunl growth of mangroves in the Ameri- cas", in contrast to condit ions l imit ing mangrove growth jus t a few km inland as indicated by the presence of "dwar f " mangrove zones. There have been no synoptic surveys of soil condit ions and plant structure to test what specific envi ronmenta l condit ions control the productivdy of mangrove wetlands in this land-margin ecosystem. Mangrove wetlands along the Shark Rive~ estuary in the southwestern Eve~glades provide a natural trajec- tory expe*iment to investigate the ~elative association between soil nut r ient b iogeochemis t ry and mangrove deTeelopment in response to a whole- scale d i s t u r b a n c e ( D i a m o n d 198(3) H u i r i c a n e Donna swept th rough the Shark River reg ion in 1960 causing mortalilT~ of nearly 80% of the trees (Craighead and Gilbert 1962) We document the spatial pattern of mangrove forest r egenera t ion correlated with selected soil proper t ies following 34 yr of d e v e l o p m e n t We tested the hypothesis

that reduct ions in forest s t ructure along the estuarine gradient are associated with variat ions in soil fertility and stressors along one of the rnauoi sources of f iesh water to this coastal region, the Shark River estuary-.

Study Area

The Shark Rivei estuary is on the west coastal plain of Everglades National Park, one of the largest estuaries of the southwest Florida coast. 'The study area is located where water is t ransported f iom the Shark Rive* Slough to the Gu[f of Mexico (F ig 1 ) Mean annual air t empera tu re m the Fa,- erglades National Park is 94~ }~dth a seasonal *ange f i o m 19~ m J a n u a r y to 98~ in August (Thomas 1974) Avelage annual precipitation in the Everglades National Park is 1,390 ram, }~4th distinct wet (60% annual rainfall f iom J u n e to Sep- tember) and dry (95% annual rainfall f iom No- v e m b e r to April) seasons (Duever et a l 1994) Dis- charge of fi'esh water f'~om Shark River Slough to the estuary follows pat terns of seasonal and inter- annual iainfall; historically, changes have occurred in these flows (Light and Dineen 1994) Tides within the Everglades National Park are predominant ly diurnal and have a m e a n ampl i tude of 1.1 m (Pro- vost 197S) Seasonally h ighe r sea levels fi-om Ms}, to N o v e m b e r along the Gulf Coast of Florida (Pro- vost 1073) result in more h e q u e n t tides in mangrove wetlands horn August to October (Twilley 1985; Twilley and Chen 1998)

Mangrove forests consisting of three mangrove species, Rhiz@l~ora m~zgle L (red mangroves) , k v icenaia gerr~irzar~s (L) S t e a m (black mangroves) , and Lag'urtcugc#riaracer~osa ( L ) G a e r m (white mail- groves), fo rm a cont inuous band extending f i o m the Gulf of Mexico to the upland limit of per iodic salt water influence, a distance of about 10-20 k m Uorzoca,@us #recta L. is restricted where the salinity does not exceed 10 g kg 1 Four sites were located along a salinity gradient of the Shark River estuary, approximately 18 km, 41 km, 9 9 kin, and 182 km i n l a n d f r o m the e s tua ry m o u t h ( F i g 1) (25~ ", 81~ to 25~ 80~ These sites represent a hydrologic gradient f i o m tide- domina ted inundat ion near the estuary mou th to h-eshwate~ r u n o f f f iom the Shark River Slough in the upst ream r e g i o n Sites $ 1 8 and $41 (distances horn the mou th are used as nomenc la tu re for the individual sites) were in the mar ine zone of Shark River estuary-, $ 9 9 was in the mesohal ine zone, and $18.2 was in the oligohaline zone ( O d u m et aL 1984; Duke 1992). In J u n e 1994, we established four adjacent 10 m >:: 10 m plots (in a 20 m • 20 m block) about 10-20 m inland h-ore the shore at sites $1.8, $41 , and $9.9. Th ree smaller (5 m • 10 m) plots about 5 m fi-om shore were established at

Mangroves and Soil Nutrients 957

Fig. 1. Study site locations along the Shark River estuary in the Everglades National Park.

S18.9; the smaller size was due to the restricted zone of mangroves at this upsneam r e g i o n Man- grove forest structure and soil properties were measured at the fbui sites, Three randomly chosen stations within each block were fixed fbr repeated measures of interstitial porewater, soil nutr ient concentrations, and mineralization rates in August 1994 and January and May- 1995. Plots are the ex- perimental units for basal area and densipy of forests, while the individual stations are sampling units tested for differences in soil characteristics,

Materials and M e t h o d s

FOREST STRUCTIJRE eMND GRO%H'H

In Jarmary 1995 we tagged and measured all t~ees in each plot having a diameter at breast height (dbh) > 2.5 cm and re-measured them in February 1996 Twenty to S0 trees were chosen randomly in each block at all four sites to measure tree height (using

a range finder), which was used to develop a regre> sion between dbh and tree height fbr each species (Chen 1996) Forest structure is defined as densi b, and basal area of t*ees using equations of Cint*dn and Schaeffer-Novelli (1984b), including the calcu- lation of relatwe importance values Tolal aboveground biomass was calculated using published al- lomerlic equations fi-om Cintr6n arid Schaeffe>No- velli (1984a) for R. mang/e arid L rae~mosa, and fiom Day et aL (1987) for A gerrtdrt~a. Fhe annual net increase in biomass was estimated as the difference between the initial and final biomass of each individual tree using species-specific equations, We used the sum of' these differences to eslimate biomass production for each site The~e is ao allometlic equa- tion for C erecta, so the allometric equations for L racemosc~ were used to estimate the biomass of U e~: ecta because of' the similarity in growth forms of these two species

958 R. Chert and R. R. Twilley

SOIL PROPERTY DETERMINATION

htterstitial Porewater Analyses

Using a plastic siphon and syringe (McKee et al. 1988), porewater samples were collected fi-om three soil stations in each block at -90 cm depth in August 1994 and May 1995, and at 40-50 cm depth i n j a n u a r y 1995 (due to lowe~ water levels). One aliquot of porewater was assayed for salinity using a portable Labcomp Instruments Model SCT analyzer and fo* pH using a HANNA H19025 pH me- ter. A second aliquot of sample was filtered to re- d,ace erro~ associated with suspended sediments; the first 10 ml of each sample was discarded to avoid oxida t ion The next 5 ml of sample was immediately added to an equal volume of ant ioxidant buffer and brought to the laboratory where it was analyzed %r sulfide concenuat ions with a LAZAR Model IS-156 sulfide sensing electrode (McKee et a l 1988) A third aliquot of porewater was passed through a G F / F filter and stored fiozen until assayed for inorganic nutrients. Ammonimrl (NH4 ~ ) concentrations were measured by a colorimetric method (Parsons et al. 1984); and nitrite (NO~) , nitrate (NO~) , and phosphate (PO4 ~ ) were determined by ALPKEM Autoanalyzer using standard techniques (Strickland and Parson 197-9)

Soil Sampling

We sampled soils, using a 50-cm diam core, to 40 cni depth at three stations within each block Soil samples were dMded into 10-cm intervals and stored at 4~ m the laboratory prior to analysis. Recognizable plant litte~ arid coarse root material we*e ~emoved. Subsamples of soils were oven-dried at 60~ to a constant weight and ground with a Wiley Mill to pass through a -950-p~ni m e s h Cores with a 15.-9 cm diam were used to estimate soil bulk densib, because the wider diameter reduced com- pac t ion Carbon (C) and ni t rogen (N) contents were determined on two replicates of each sample with a LECO CHN analyzer using standard proto- cols T,atal P (TP) was extracted from soils with 1 N HC1 af|er igni t ion at 550~ (Aspila et al. 1976) and inorganic PO4 s was determined on filtered samples (Parsons et al. 1984). Available PO4 s was determined toy dilute acid-fluoride extraction of fi-esh soil san/pies (Olsen and Sommers 198-9) Amount of P extracted toy this method is generally used as the available P index in plant ecology (Binkley and Vitousek 1989)

P ~actio~zatior~

We used a sequential extraction scheme de~'el- oped by Hieltjes and L{iklema (1980), with modi- fications %r wetland soils (Koch and Red@ 199-9; Gale et al. 1994) and estuarine sediments (Zwols-

man 1994). Soil samples for P fractionation were collected Dora each station in August 1994. Fresh soil samples were extracted sequentially with 1 M KC1 (1:10 m/v) , 01 M NaOH, and 05 M HC1 by shaking (-950 rpm) samples for 9 h, 16 h, and -94 h with each extractant, respectively. Af/er each step, the suspensions were centrifuged and the su- pernatants were analyzed colorimetrically for inorganic PO4 S according to Parsons et a l (1984) A subsample of the NaOH extract was digested by the Kjeldahl procedure to determine I 'P m the extract Phosphorus concentrations in extracts of KC1, NaOH, arid HC1 solution are defined as labile P, Fe~M bound P, arid Ca-bound P, respectively (Gale et al. 1994) O~ganic P was determined as the difference between TP and inorganic P in the NaOH-extract arid is defined as hydrolyzable organic R Residual P is defined as the diffe*ence between the TP, which was determined by the igni- tion-acid extraction method (Aspila et al. 1976), arid the sum of all extracted forms of R

N and P Min~alization

Since no single method provides accurate estimates of N and P mineralization, all methods should be considered an index of nutr ient availability for comparative purposes (Binkley and Vi- tousek 1989). We used anaerobic incubations of soil cores in the laboratory to determine net N and P mineralization rates, which were used to compare soil quali/7/ among the four mangrove sites along the estuary. Gla~ss tubes (2.5 cm diam and 10 cm length) were used to recover intact soil sam- p ies The end of each tube was capped with a rub- ber sleeve immediately af/er collection to avoid aeration, arid the tube was stored on ice in a cool- er. Paired tube saxiiples at each depth of 0-10 cm and 10--90 cm intervals were collected at each sam- pi ing station and maintained at 4"C until incubations b e g a n One tube f iom each pair was extracted with 1 M KC1 (1:10 m/v ) soon after samples were transported to the laboratory; the other tube was incubated anaerobically in the dark in an en- vironmentMly controlled chamber at -95~ for 15 d, followed by the same extraction, The extract was centrifuged and analyzed for NH4 ~, NO 2 + NO z , and PO4 s using statidard techniques described above, Rates of net nitrification, ammonification, arid P mineralization for each soil depth were determined by subtracting the initial concentration of exchangeable NO 2 e NO z , NH4 ~, and PO4 s horn extracted concentrations at the end of the incubation period, Mineralization rates were con- verted to a soil area basis using respective bulk density values, Denitrification was not m e ~ u r e d but was assumed to be minimal given the NIt4 + and NO s concentrations measured in the poiewater at

all sites (Rivera-Monroy et al. 1995; Rivera-Monroy and Twilley 1996)

STATISTICAL ANALYSES

A repeated-measmes :~MNOVA was used to ana- lyze the effects of site and sampling date on porewater variables (salinity, sulfide, and inorganic nutrients) and soil variables (acid fluoride P, exchangeable NH4*, exchangeable NO 2 + NO s , N mineralization, and P minera l iza t ion) We also used linear contrasts to evaluate the differences in selected variables among sites on each sampling da te The differences among sites for components of the P fi-action and forest basal area were tested using one-way .~MNOVA All statistical tests were done using either original data or log-transformed data to meet assumptions of parametric analysis

Results

FOREST STRLCTURE AND GRO~#'FH

The landscape patterns of mangrove forest structure along the Shark Rive~ estuary exhibited dif- ferential ~ecovery f iom distmbance toy Hurricane Donna in 1960 Mean tree heights in the four sites decreased f iom 101 m in the lower estuary to 5 4 m in the upper estuary (Fig. 2A). R me~ngge, A. gerr~gnans, and L raceraosa were found in the lower estuary ($18, $41, and $9.9) A. germinans did not occur in the upper estuary site, which was the only location where C erecta was observed ($182; F ig 2B-D) R manggewas found in all sites and this species iepresented 357o to 90% of tota[ tree density among the stu@ sites (based on > 2.5 cm d b h ) Mangrove forests were mixed-species communit ies at $1.8 and $41, where L racemosa had the highest importance value and A g ~ i n a n s the least (Fig. 2 D ) Further up the estuary, R. r~ang-ge was dominant at $9.9 and co-dominant with 6' erecta at $182 (Fig 2C,D). Mean basal areas were similar in the t~o lowe~ estuary sites at about 40 m ~ ha ~, but the upper estuary sites at about 20 rn ~ ha t (Fig. 2C) were significantly lower (Tukey's test, p < 0.05).

The density of' larger trees (dbh > 10 cm) was 1,900 trees ha 1 in the lower estuary and only 334 trees ha 1 in the upper estuary (Fig. ,g) However, the mean density of' all trees with dbh > 2 5 cm was lower at the mouth of the estuary compared to the upper estuary (Fig. 2 B) There was a higher density (5,5,34 trees ha 1) of smaller trees (< 6 cm dbh) in the upper estuary, and tree density decreased exponentially with an increase in dbh at $1822 (Fig. S ) There also was a decrease in tree density- with increase in size class at sites $41 and $9.9 'The fi-equency distribution of size classes was bell-shaped in the lower estuary at $ 1 8 (Fig g).

Mean total aboveground biomass in 1994 was ap- proxinmtely 2 >:: h igher in the lower estuary (about

Mangreves and Soil Nutrients 95~

250 Mg ha 1) compared with the upper estuary (Fig. 2E). Rates of wood productMty (kg tree 1 yr *) were highest for L raeemosa and A germin~ns at $4.1 and decreased along the estuary in the upstream location (Tuble 1). Growth rates of R rge~n Jeva r i ed by only a factor of,g among the four sites and its peak rate also occurred in the lowe~ estuary. The annual increments in total basal area also decreased from the lower to upper estuary (Table l) . R. r,~angge contributed 78% ($9.9) and 86% ($18 2) of total wood productivity in the upper estuary, due to the higher density of R. mang& in these sites L racemosa dominated total wood pro- ductMty in the lower estuary. Wood productivity was higher in the lower estuary, approximately twice that of the upper estuaiy (Table 1).

SOIL, CFD~CFERISTICS

Salinity decreased f iom the lowe~ to the upper estuary on all sampling dates; peak salinity concentrations at each station occurred during August (Fig. 4-A) Salinities ranged f iom 335 g kg 1 ($18 in August) to 1.6 g kg t (SI&2 in January). The sites with the greatest temporal change m salinity were $41 and $9.9 (Fig. 4_A) Porewater sulfide concentrations were significantly different among the sites in August, with the highest concentration (mean of 0.46 mM) in the upper estuary at $18.2 and the lowest concentrations (< 0.15 raM) in the lower estuary at $1.8 (Fig 4B) No sulfide concentrations could be detected m porewaters at any of the sites in January when the ground water level was 40 cm below the smt:dce. Porewater pH wa~ near neutral at all sites, with mean values ranging [ lom 6 6 to 6 9

Mean porewater NOs and NO 3 concentrations were < 1 p,M and < 05 p,M, respectively, at all sites during the three sampling dates (Fig 4C,D). In JanuaD,, concent*ations of NO, were significantly h igher at $182 axed $99 than at $1.8 and $41 (Fig 4 C ) There were no significaElt site and date effects on NO~ (Table 2 ) Porewater NH4 + concent*a- tions were significantly different among sampling dates, with the highest mean in August (Fig 4E); there were no significant site effects on NH4 ~ concentrations (Table 2 ) MeeaJ porewater PO43 concentrations were < 2 aM at a[1 sites d in ing the three sampling dates, except at $4.1 in January (Fig. 4F). There was a significant effect of date and a significant interaction between site and sampling date, but there was no site effect on PO45 concentrations (Table 2).

The highest exchangeable NO2 + NOs in all fore sites was 0.06 g N m 2 (data are not shown), which indicated that most of the inorganic N in these mangrove soils was dominated by NH4* (Fig. 5A). Mean exchangeable NH4 ~ pools within the 0-

9 6 0 R. Chen and R. R. Twilley

s U: 09 I o~ 8 D-

s Oa

g

o~ e5

ffJ

�9 A. germinans [] L. racemosa [] R, mangle [] C. erectus ]

12

10.

8

6

4,

2

0- Sl.8 S4.1 $9.9 $18.2

Mangrove sites

ii ~ 30-

2o ,Oo II !i i/i

300

a: 200 0~

.o E 100 03

0

Sl.8 S4.1 S9,9 S18,2

Mangrove sites

iL~ili~ ii!

81.8 S4,1 S9,9 S18,2

Mangrove sites

8000

% 6000

6 Z

4000

2000

0

100

g 8 o -

o 60- CI_

~ 40-

0 ~" 20-

O-

$1.8 S4.1 $9.9

Mangrove sites

$18.2

$1,8 $4 .1 S9.9 $18,2

Mangrove sites

~'-'14 > ,

'T. 12- ~3 .C ~10-

~ 8- 0 o 6

s 4. n_

0 2- 0 ~: 0

S1.8 $4 .1 $9.9 S18.2

Mangrove sites

Fig 2. Mangrove forest stTucnH-e by species at sites along the Shark River estuary from the downstream location (S1.8) to the upst iemn location (S18 2) including: A) mean tree height, B) density of trees >2.5 cm diameter at breast height , C) basal area, D) importmlce values, E) biomass, and F) wood p roduc t ion Data are expressed as l&qean value of plots m A, B, C, m~d E

20 cm soil interval decreased significantly along the estuaty for three sampling dates (Fig. 5A). The extractable PO4 s pool was significantly greater at the mouth os the estuary- for all sampling dates but was not significantly dif'f~rent between $9.9 and SI&2 in August and May- (Fig, 5B).

The product ion of NO S [rom the remineralized NH4 + in the soil incubatious was negligible [or ~11 sites and dates (data are not shown). In January net NH4 + product ion was significantly h igher at S1.8 and $4.1 than at 59.9 and 518.2 (Fig. 5C). For all sites, no net NH4 + mineralization was ineasured

900- 1 ~-" ~ S1.8 ~ '05 ] 03

d ,5 z z

"0 "t~

}-- I--

900

600

300


2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 Size classes (cm)

0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0

Size classes (cm)

x: d d z z z - ._z-

I-

2867

2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 Size classes (cm) Size classes (cm)

[ ] L. racemosa �9 A. germinans [ ] C. crectus [ ] R. mangle

E[g. 3. Size class distribution of mangrovc, s by species along the Shark River e'slqmry from the' downstream location ($1.8) ro the- upstream location (S18 2). Data are means of plots for each sire

in August. Mean daily net ammonificat ion rates within the 0-20 cm depth for three sampling dates decreased t iom the marine to o[igohalme sites: 182 rag N m ~ d i 185 m g N m ~' d 1 46 mg N m ~ d t and 17 mg N m ~' d 1. Net P mineralization rates were more variable than N mineralization (Fig. 5D). I n j a n u a r y net P mineralization was significantly h igher in the upper estuary than lower estuary. Mean daily rates of' net P mineralization over the three sampling dates within the 0-20 cm soil were positive at $99 and $18.2, compared to negative values reflecting immobilization at $1.8 and $41 (Fig 5 D )

Both total C and N concentrations in the top 40

cm of mangrove soils increased linearly with dis- tahoe along the Shark River estuary fkom $1.8 to $18,2 (Table 3). P concentrations showed an in- verse relationship ~/th distance fkom the estuary mouth. Bulk densi/y also decreased ~i th increasing distance tkom the ulouth of the estuary (Table S). ~Then [mtrient pools were expressed as mass per uni t area, C and N pools were not correlated ~i th distance along the estuary (Fig. 6A,B). The consis- tent relation betwee[l C mid N concentrations in these mangrove soils was reflected in atomic C:N, which were between 22:1 and 25:1 along the sa[in- i ~ gradient of' the estuary (Fig. 6D). However, P content dec*eased exponentially from 152.2 g m ~

9 6 2 R. Chert and R. R. Twilley

TABLE 1. Average wood production per tree, tobql annum basal area increment, mid wood productivig~ per unit area of max> groves along the Shaxk River estuary, the Everglades National Park

; t u o n

SI 8 S4 ! ;59 9 ;18 2

Aveiage Wood Production (kg tree ~ yI x) Avicer~rda g'cr~r~ir~ar~s 403 481 037 -- La~r~c~laria race*~aosa 309 9.17 850 115 Rhiz@hora ~ar~g'ge 2.35 0.60 104 0.69 Cor~oceryt~s erect~ -- 0.11

Basal Area Increment (m 2 ha x yr x) Aviccf~rda g'e'~ir~af~s 024 0.06 001 -- Lag~r~ozloria race~z~osa 066 1.11 009 004 Rhiz@hara ~aar~g'& 0.51 0.12 0.36 071 Cor~ocarytzs erecta -- 007 Total wood productivi~e 1 40 1.29 046 083

at S1.8 to 26..3 g m ~ at S18.2 (to depth of 40 cm) (Fig. 6C). Accordingly, atomic N:P increased exponentially with distance f iom the estuary mouth, f iom 15:1 at $ 1 8 to 102:1 at $18.2 (Fig 6E)

The amount of labile P was significantly h igher at $18 than other sites but was not significantly different among $41, $99, and $182 (Table 4 ) The contribution of labile P to the pool o fTP ne~,- er exceeded 2% among all sites within the 0-20 cm depth (Fig. 7). Concentrations of the Fe/A1- bound P pools also decreased f iom $1.8 to $18.2 (Table 4). The contr ibution of Fe/Al-bound P to TP ranged flora 8.2% to 105% and did not exhibit a significant uend along the estuat-y (Fig. 7 ) The amount of hyd~olyzable-organic-P ranged f iom 2.39 g m ~ to 7.64 g m 2 and "#as significantly higher in the lower estuary than the upper estuary (Ta- ble 4). Except at $18, the hydrolyzable-organic-P fi-action was the second largest pool of TP among sites along the estuary. The concentrations of Ca- bound P exhibited a 40-fold decrease horn the lower to the upper estuary (Table 4). The peicentage of TP that was Ca-bound P also decreased significantly Dora g2.6% at $1.8 to ,g.g% at $18.2 (Fig. 7). The Ca-bound P pool was greate~ than the Fe / Al-bound P pool in $1.8 and S4A, while Fe-A1- bound P concentrations were greater than Ca- bound P concentrations in $9.9 and $18.2. Resid- ual P was the largest P fi-action at a[1 sites, v~dth a significant increase along the estuary f iom 455"/) of TP at $1.8 to 70% at $1&2 (Fig. 7). 'The amount of residual P, in g m ~, decreased significantly along the estuary (Fable 4).

D i s c u s s i o n

The Sha~k River estuary provides an opportunity to evaluate mangrove development along an estuarine gradient of soil fertility on a decadal time

scale (30-40 yr) using a specific starting point Dora a known disturb3nce, Hurricane Donna in 1960. The~e is a distinct change in community- compo- sition and productivity- of mangroves along the estuary-, particularly between 5 km and 10 km Dora the mouth of the estuary. Basal area, biomass, and wood p loduc t ion declined by neaily half m this region of the estuary and dominance shifled Dora mixed species dominated b y L raeemosct to nearly monospecific forests dominated, by- R. r~a~zg/e. In the upper estuary, the size class distribution is dominated by a higher density- of smaller trees, in- dicative of young regenerat ing [oiests. The basal ar-ea and biomass of the two sites in the lower e~ tuary are representative of more mature forest stands. If we assume a clear forest stand tbllowing Huri ican e Donna and use the average species~specific tree growth rate at each location measured in our study, the basal area would be about 45 m s ha 1 at the mouth of the Shark River estuary compared ,~ith 20 m s ha 1 at sites in the upper estuary fol- [ovdng ~g4 yr of growth. These values are similar to values measured in the field and suggest that each alea is repiesentative of environmental conditions and not different stages of maturity. \,re also tested this assumption in model simulations of these sites using growth rates in this study to demonstrate that differences in initial conditions are insufficient to create the gradient in forest s t ructme observed along the estuary (Chen and 'Y~dlley 1998).

Mangrove soils along the Shark River estuary can be characterized as only slightly reducing, with mean redox potentials of 214 mV and 117 mV at $1.8 and. $9.9, respectively (Koch 1996). Sulfide concentrations can also indicate the effects of f looding and abiotic stress on mangrove development. Porewater sulfide concentrations ranging fi-orn 1.5 mM to 4.1 mM have been described as significant to patterns of forest structure in Florida (Carlson et ab 198.g; McKee f99S), the Bahamaus (Nickerson and T h i b o d e a u 1985), and Belize (McKee et aL 1988; McKee 1995). Concentrations of sulfide (< 0.46 raM) were lower in our sites m the Shark Rive* estuary compared to these reported values, indicating that this stressor is not re- sponsible [br the decreased stature of mangrove wetlmids along the estuary-.

Hypersaliniff has been frequently suggested as one of the major ~;actors l imit ing mangrove forest stature and growth (Lugo and Snedake~ 1974; Cin- trdn et al. 1978). Threshold salinities above which there is an apparent reduction in forest structure have been estimated at 65 g kg 1 in Puerto Rico (Cintr6n et aL 1978) and 50 g kg t in Mexico (Day et al. 1987). Porewater salinity, in mangrove wetlands along the Shark River estuary during all three sampling dates did not exceed 35 g kg ~, be-


Go

1.0

o.8] "~' 0.6 63

z

August January May

C C.

0.4 a a a

0.2 a

0 ~

20

15- .E 10- O E " E < 5-

0

0.60

0.45

0.30

0.15

0

1.0

~ 0~

Z

B, c

j a undetectaNe a

i August January May

g .

a

August January May August January May

8 E. F.

_b

6-

tll a ~ 4-

S Q - r

a o 2- a a ~ a a a a

a ~ ] ~ a a a a aaa a ~

August January May August January May

Etg. 4. Spatial and te 'mporal variation in porewate 'r stressors a n d nu t r i en t s induc t ing saliniUz (A), sulfide (B), nitri~. (C), n i t ra te (D), a m m o n i u m (E), a n d phospha t e (F) m e a s u r e d a long the S h ~ k River estuary d~ r ing Augus t 1994, Ja rmary 1995, and Ma 7 1995 The values at each smIlpling da~e wifll the same lett~-r are n o t sigmficmldy d ~ f e r e n t a m o n g sites (p > 0.0~). Values are meat1 (n 3) ~ S E

low the critical value that influences mangrove complexity (Lugo et 61 1988) O u r surveys of porewater sulfide and salinity measuremel l t s examined relatively d iy (1994) and wet (1995) years in the Shark River watershed. Salinities in the u p p e i re- g ion of the Shark River estuary near $182 reached 18 g kg 1 in lun e 1994, compared with salinities in this reg ion < g g kg 1 th rough out 1095 The higher salinity and sulfide readings in August 1094 along the mangrove sites reflect this change in

f ieshwater input, Previous monthly- surveys of the Nor th River and Shark River estuaries also ~eport salinities < 30 g kg 1 in mangrove regions (Heald 1969; Powell et a l 1989) Forest basal a rea was lowest in the oligohaline region and increased with h ighe r salinity in the Shin& River estuary, indicating that forest deve lopment was not l imited by soil salinity.

In the absence of physiological stress by sulfide and salinity, soil nut r ient availability- is oflen impli-

964 R. Chen and R. R. Twilley

TABLE, 2. S u m m a r y o f F statistics f r o m repea t ed measm'es ANOV& of site' a n d samplLqg date' effc'cts on porewater chemis t ry and n u t r i e n t availability Statistical tests were c o n d u c t e d us ing tmivariate teats.

Source df S !ir!/W Sulfide 2~C) 2 NO2 N1~4" PO{ 2

Sire S, 8 1 6 7 8 **~ 3 1 6 2 ~** 1 2 2 " * 1 .3" 51" ' 2.7 ~*~ Dat~- 2, 16 445.7 ~** 50.82*** 3 5 4 * * * 0.9 ~ 3 0 3 ~ 5.17" SiN" X Dais- 6, 16 28.9 T M 1 3 3 0 ~*~ 7 8 * * * 0.9 ~ 2 1 ~ 2 9 3 *

Exchangeab le Exd langeab l e Acid Fluoride Somces ctf NO~ I NO3 NH4 ~ P O ~ Amm~ni f i ca t ion Pminerag tza t ion

Site S, 8 8.15"* 2 6 8 5 ' ' '< 108.5"** 7 5 3 * 4 1 0 " Dais- 2, 16 2 6 8 * 17.28 ~** 10.88 *~* 7 5 7 0 T M 8 3 8 * * SiW )( Da~e 6, 16 1 9 1 " 3.84* 4 1 0 " 7 8 8 * * * 8.83***

p < 0.05; ~* p < 0.01; *"~ p < 0.001; ns n o t significaut

cared as the principal factor deterrnmmg wt i a t ion in mangrove biomass and productivib,. Most concentrations of porewater NH4 + were < 10 p~_%([ in the study sites along the Shark River estuary and are lower than reported values for other mangrove softs, including Florida Bay (RosenfMd 1979), Mis- sionary Bay of Australia (Smith et aL 1991), Papua New Guinea (Alongi et a l 1993), and Belize (McKee 1995), which are all > 10 loom Our values

were compmable to [hose ~epox ted in Indian River, Florida (< 25 p,M, Carlson and Yarbro i[988) and in dwarf mangroves of Belize (1,0 10aM, Feller 1995) Inorganic PO4 ~ concenuat ions in porewaters in our study sites were < 2looM, similar to mangrove soils in Australia (Smith et aL 1991) and Pap- ua New Guinea (Alongi et aL 199.8), and less than those observed in other forests in the Cai ibbean (Feller 1995; McKee 1995) The more p~oductive

E Z

+

- r Z (D

e)

J~ o X

LU

August January May

3.0 a_ 2,4

& 1,8 0 13_ m 1.2

1D

O 0.6

< 0

B.

August January May

"o 600, ~ 2 0

t C, ~ z a_ 10. r~ 400 g g

�9 g .~ O 200

"K ~ b a ab ~'- o - -10-

bbb o...

- I00- "~ -20- Z August January May Z August January May

Fig 5 Extractable- a m m o n i t t m (A), extractable p h o s p h o r u s ( B ) n e t anamonificat4on (C), a n d p h o s p h o r u s z~Zt~e'ralJzation (D) vtid.in the" 0-20 c m soil a t f o / a mazagrow' siWs along the Shark ldvc.r c.stuary in Augus t l g 9 4 , j a n u a r y 1995, a n d Ma]~ 1995 Each value is thu m~-ma (+ SE) o f thx~-~- s ta t ions T h e v~ues f r o m difs si~.-s a t .-ach smnplirig da/u w i t h / h e s&t~c, let ter ar~- n o t signhqcaut[y d~fere'rlt among sites (p .'-- 0 0 5 )


TABLE 3. Soil carbon (C), n i t rogen (N), a n d p h o s p h o r u s (P) concen t ra t ions and bu lk density a long the Shark River e s t aa ry Data are m e a n s o f t h l ee sampl ing dates • S E Co:re la t ion coefficient~ (:) are be tween respective soil variables for each site a n d distance h o r n the m o u t h o f the Shark River estum-y

v riables Sl 8 $4 ! S99 i 82

Total C o (,o) 143 • 03 222 • 1.2 32.5 215 43.3 • 1 1 0.985* ~bta[N (%) 0.70 • 007 115 • 0.09 1.55 ~ 0.07 2.33 4 0.10 0.990** Total P (/~) 0.11 • < 0.01 012 • < 001 0.09 • < 001 005 • < 001 0960* Bulk density (gcm ) 0.462 0212 0.148 0135 0.751

* p < 0.05; ** p < 001 .

mangrove sites at the mouth of Shmk Rivet estuary had similar porewater NH4* and PO4 ~ concenua- tions as those forests m Belize with much less structure (Felle~ 1995) These comparisons of' mangrove forest structure among sites along the Shark River estuary and with othe* mm~grove locations indicate that porewater NH4 + and PO4 ~ concentrations, which are a minor fraction of the total N and P pools, are not a sensitive measure of soil fertili|y, Other forms of N and P, such as different pools of the exchangeable flactions, may be available fol plant uptake and may contribute to soil fertili/y,

Nutr ient tmnover rates may be more important in determining nutr ient l imitation than simply concent ra t ion Mangrove soils contain a high proport ion of' organic P, but most of this organic P tends to be ~ecalcitrant (Alongi et ab 1992) or ad- sorbed and fixed by minerals as different chemical forms (Hesse 196~2) Rapid i m m o b i l i z a t i o n of' amended inorganic P in mangrove soils suggests that P adsorption by- minerals can strongly immo- bilize inorganic PO43 (Hesse 1962; Alongi et al. 1992) Our study showed that Ca-bound P was higher in the lower estuary than in the oligohaline zones along the Shark River estuary, indicating the significance of mineral processes at the mouth of the estuary It has been demonstrated that carbonate can rapidly adsorb inorganic P released from mineralization of organic matter (Delianel and Morse 1978; McGlathery et al. 1994) We observed little net P mineralization in mangrove sites along the Shark River estuary and low concentrations of' inorganic PO4 ~ in soil porewaters Both chemical fixation and microbial immobilization can contribute to the absence of inorganic PO4 S in mangrove soils along the Shark River estuary

Net NH4* product ion also declined with P concentrations along the estuarine gradient of Shark River estuary Net ammonificat ion ~ates in mangrove soils along the estuary in .January (gS-2.g9 mg N m s d 1) and May (18-70 mg N m s d 1) are comparable to other macrophyte ~Tstems, such as subtropical seagrasses (28-294 mg N m ~ d 1, Boon et al. 1986) and fi-eshwater marsh (44 mg N m s d 1 Bowden 1984) However, net ammonifi-

cation was negligible at all mangrove sites in Au- gust, with evidence of" N immobilization; in Janu- aiy, when the water level was lowest, amrnonifica- lion lates were highest. Litter fall of all th~ee mangrove species peaks f rom .June to Augus t in southwest Florida, and most of this litter is defi- cient in N as indicated by C:N generally > 50 (Heald 1969; Twilley et ab 1986; Lawton-Thomas 1997), Nitrogen isotope studies of both lSNH4" and l~NOs have demonstrated the strong potential for N immobilization m mangrove sediments with similar N concentrations and C:N as those of the sites in our study (Rivera-Monroy et ab 1995; Rivera- Monroy and T~/dley 1996) There are indications fi-om om study that patterns of N immobilization may be seasonal in response to changes in hydro- period, as well as the quantity and quality of leaf litter,

Relative N mineralization (N mineralization per unit of total N) is a useful index to compare overall soil quality- as a factor de termining N minemliza. tion rates (Pastor et a l 1987) Relative N mineral- [zation decreased exponentially with distance from the mouth of the estuary (Fig, 8A) This indicates that the proport ion of ~ecalcitrant organic N in mangrove soils increases rapidly with distance flora the Gulf of Mexico toward the upstream mangrove- marsh ecotone, Nitrogen mineralization was not correlated with either total C, total N, or C : N This suggests that ammonificat ion rates in this man- giove region are independen t of total N concentration, Mean rates of' net ammonificat ion were significantly correlated with acid-fluoride extractable P concentrations (r ~ 0,94, p < 0,05), suggesting that N mineralization was associated with P availabili~- (Fig, 8B), Highly significant correlations of microbial activity with P components indicate that microbial activi 9 is pi obably P-limited in mangrove soils in New Guinea, which then may limit N mineralization (Alongi et al. 1993) These indicators suggest a very complex interaction between sub- strate quality- along with N and P mineralization rates that control the availabili b, of these two nutrients for mangrove productiviw.

The landward decline of mangrove basal area and biomass was associated with a decrease in sev-

9 ~ R. Chen and R. R. Twilley

3O

~" 20 t %

o~c~176 1 0 ]

1 o

A [ ] $1.8 [ ] $4.1

| [ ] 818.2

1.5

~v 1.0- Gl

E 0 .5 - Z

0 -

150~ C r2=0.95 (/:'<0.05)

s loo2

50~ K

0 - - - L________

- D z, 30-

._o 20- E _ o < 10-

02

80-

E ,9, ~ 4 0 - .<

0 0 5 10 15 20

Distance from river mouth (kin) Fig. 6. C~a~nges of soft carbon (C), n i t rogen (N), and phos-

p h o r u s (P) vdfllm 0-40 cm soft m mangrove fores t , f lora the m m i n e site ($18) to the o l igoha lme sit~- ($18.2) a long the Shark River e s m m y of the Everglades Nat ional Pat-k Data are m e a n o f / h r e e sampl ing dates * S E

eral indices of soil fertility along the axis of the Shark River estuary, The greater decrease in TP in mangrove soils compared ~/th more constant total N concentrations from the lower to the upper estuary results in N:P < 20 m the lower estuary, 40 in the inteHnediate zone, aqd > 80 in the uppe i estuary. The lower ielative availability of P per unit of N indicates P was a l imit ing nutr ient to mangroves in the oligohaline zone of this estuary. The higher concentrations of TP at the mouth of the Shark River estuary, along ~/th lack of stress flora salinity and sulfide, result in opt imum conditions for mangrove growth in this r e g i o n Fertilization studies in other mangrove forests (goto et al. 1984; goto and Wellington 1984; Feller 1995; T~/flley 1995), including the marl soils of the southeastern Everglades (Koch and Snedaker 1997), have re- vealed that P availabiliw can be a maior f,~cto~ lim- iting annual growth of mangrove forests in soils hav ing low salini ty or sulf ide concen t ra t ions . Changes in species composition and basal area along this soil fertility gradient in Shack River estuary indicates that L rc~ceftzosa is more dominant in the h igher nutr ient sites of the lower estuary-, wheleas R. f~angle dominates m the upper estuary at the lower nutr ient s i te Simulations of forest deve lopment using the relationship of species-specific response to soil fertility" indicate that most, but not all, of the shil l in community composition and basal area along the estuarine gradient can be ex- plained with relative soil 2~'P concentrations (Chen and Twilley 1998)

The relative bulk concentrations of P and their chemical fractions in mangrove soils contribute to our unders tanding of P dynamics along the Shark River estuary. 'There are two major sources of nutrients to this estuary: discharge flora Shark River Slough and sediment resuspension from the neai- shore boundary along the southwest coast of Flor- i da In our study; sequential extractions of P indicated a significant deciease in Ca-bound P and increase in residual P f~om the lower to the upper estuary. Significantly h igher Ca-bound P as a pe~ centage of TP near the mouth of Shark River estuary suggests that the greater bulk density concentration of' P at the marine sites may be related to sediment deposited ~iorn the Gulf of Mexico Shallow mangrove soils in the marsh-mangrove ecotone of the uppe~ estuary ~eceive little mineral sediment and have a greater proport ion of o~ganic R Increased soil concenuat ions of both total and Ca-bound P with salinity, along ~i th higher soil bulk density, were used as boundary conditions in simulations of soil processes at three sites along the Shark River estuary (Chen and Twilley 1999) These simulations, using Pb-210 estimates of sedi- mentat ion at $41, demonstrated that the Gulf of


TABLE 4. Phosphorus (P) fractionation at 0-20 cm depth in softs from mangrove forests along the Shark River estuary. Data are means + SE (n 3) Chmlges in superscript letu'rs wiih the mean P concentration for each fraction indicate significant differences among si/vs (Tukey's p < 005)

Soil P Components ( J " m )

!• o40:able Sire Labile P Fe/A/bound P ore anic ~ Ca.bound P k.es!dual P

S1.8 0 . 5 4 + 0 0 4 5 5 8 + 0 2 1 7 6 4 '~ + 0 .39 21 .33 ~ + 1.60 30 .36 "~ + 5 6 8 $ 4 1 0.261 ~ 0 .06 3 3 2 ~ ~ 0.37 6.55 ~ + 0 5 2 5.87 b + 1.21 23 .32 ~ ~ 2 .28 $ 9 9 0 3 0 b + 0 0 3 2 6 @ :~ 0 3 4 5 4 0 b + 0 4 0 0 9 2 ~ 0 0 4 16.14 "~ + 2 .07 $18 .2 0.29 b :~ 0 .02 1.24 ~ ~: 0 0 3 2 3 9 ~ + 0 1 6 0 5 0 ~ * 0.03 1 0 6 9 b + 0 .59 p 0 .003 < 0 0 0 1 I- 0 O 0 1 < 0O01 0 .013

Mexico was a majo* somce of P for mangroves m the lower estuary-. Surveys of water quality param- eters in the Whitewater and Florida Bay regions found distinct inne* and outer mangrove zones of similar influence (Boyer et a l 1997) Inne l mangrove zones had lower TP concentrations and higher TN:TP (TN is total N) than outer zones, matching the spatial characteristics of the mangrove soils in the mangrove zones of' our stu@ These water qualib, surveys, along with indicators of higher concentrations of P availabili/7 in the northwest region of Florida Bay (Fourqurean et aL 1992), suggest P is transported in dissolved and suspended forms to this region f iom the coastal boundary current that passes by the mouth of the Shark River estuary. Phosphorus transported along the southwest coast of Florida appears to regulate mangrove forest development and productivity at the lower regions of the Shark River estuary.

Accelerated nutr ient loadings, especially P, from different anthropogenic sources in the watersheds

1 O0

80-

6 0 -

r 4 0 - ._o

20-

0-

152.2 84.7 50.2 26.2 g m "2

81.8 84.1 $9.9 $18.2

Mangrove sites

�9 Labile-P [ ] Ca-bound-P [ ] Residuat-P

[ ] Fe/AI-bound-P [ ] HydrolyzableOrganic-P

Fig. 7. C h a n g e s i n p h o s p h o r u s (P) f r a c f i o n a t i o n s as a pc'> c e n t o f t u t a l P : d f t : i n t he 0 - 2 0 c m soft a t f b m m a n g r o v e ain-s a l o n g t h e S h a r k R i v e r e s t u a r y T h e c o n c e n t T a f t o n of t o t a l P a t e a c h sit~- is g i v e n a b o v e t h e respecfi ,~e bar.

of' south Florida can change the vegetation chat- acteristics of oligotrophic regions of the Everglades (Davis 1994), Soil P f'lom areas near the Everglades Agliculture Area (EAA) range Korn 0,12% to 0.16% (Craft and Richardson 1999a; De'gusk et al.

0.09/0 to 0.07% in regions 1994) and decrease to ~or south along the Shark River Slough (Davis 1991; Koch and Red@ 1992; Craft and Richardson 1993a, b; Red@ et al. 199.8; Degusk et a l 1994) In our study-. TP concentrations in the upper estuary site (0.06%) are similar to those concentrations in lower regions of Shark River Slough. }towe~Ter, P concentrations near the mouth of Shark River estuary (0.12%) are similar to the upper limits of total soil P in the marshes adjacent to eutrophic soils of EAA Based on bulk density values, P content per uni t volume of soil at the mouth of Shark River estuary (021 mg cm s and 0..38 mg cm s in $41 and $18, respectively) are even higher than those in the enriched area in Water Conservation Area 2A which is part of the Everglades, where P content ranges horn 012 mg cm s to 02 mg cm s (Koch and Red@ 1992; Red@ et al. 199`8; Crafl and Richardson 199`8a) These gradients in TP represent a gradual shift fi-om more mineral soil at the coastal margin to organic soil at the headwater ~egions of Shark Rwer S lough Fhe enr ichment of' soils with TP results in sensitive changes to wetland vegetation, demonstrat ing the potential effects of natural (e.g., sea-level rise and storm frequencT) and anthropogenic ( e g , i)rt i l ization and freshwater diversion) forcing on landscape vegetation dynamics in this coastal eco tone We hypothesize that upstream marshes in Shark River Slough ef- fectively ~educe surface water P concentrations. This leads to relatively low P inputs and contributes to the linfited mangrove structure in the o[igoha- line zone (along with possible effects of hydroper- i o d ) Marine inputs of P are a significant contri- but ion to the opt imum conditions described by Da- vis (1940) for mangrove forest development along the southwestern Everglades These patterns in vegetation dynamics and soil properties in the coastal zone suggest that the entire Everglades

968 R. Chen and R. R. Twilley

2.0

1.5 ._O

1.0

Z (D

"~ 0.5

s

0,(

A

i'~ T Y = e(O'859"O'156x )

1'0 1'5 20

Distance from the Estuary Mouth (km)

200

',.o

E 2: 150

g g '~ 100 to

0 E E 50 .<

z

0

r2_ 0.989 (p< 0,0 h

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8

Available P (g m -2)

Fig. 8 A) The relative ammoni f ica t ion rate of mangrove softs in re la t ion ~o locat ion of siles a long the Shm-k River esmary Relative earanonification is calculat~-d as ft~e n e t total a m o u n t of a~nr*onium minera l i zed (0-20 cm) du r ing three per iods of 15- d incubaticm as a pe rcen tage of mem~ total N p resen t fniftail v B) Net arnmonif ica t ion rate in zelat ion to the mr*ount of available p h o s p h o r u s in soft to d e p t h of 20 c m Values are m e a n (n

S stations) + S E

landscape front freshwater uplands to the coastal marg in are sensitive to the m a n a g e m e n t of' water quality and quantity, particularly those practices that inf luence the d isuibut ion of R

/~CKNO~vVLEDGMEN TS

We are grateful to Dr s 1 b i n 2u~r*entano, ~fbm S~fith a n d De- vd~ Smi th tb r assistm~ce in provicilng access to maa~grove sites in the Everglades Nat ional Pea& We thaxik Mike Radford, J o h n Foret , Robe r t Bourgeois, S tephanie Cogburn , Russel Carre t t , and C Coronadc-Moin~a for assistance in the field a n d / o r labora tor v. We t h a n k Joseph E Neigel, Samuel C. Snedake~; Paul L Klerks, Charles L Burras, Victor RiverwMonro?, Charley Hail ,

g i l d t w o a f l o n y l 1 3 o u s r e v i e w e r s f o r coxlsxlsents o n e a ~ e r v e r s i o n s ofo~x manusc r ip t 'tibia resem-ch was suppor t ed b v Un in -d Stat~,s Depm-tment of Interior . Nat ional Pm'k Ser~4ce, Cooperative A g r e e m e n t No. CA 5280-4-9019 with the Everglades Nat ional Pm-k

LITERATURE CITED

ALONGI, D M , K G BOTO, 2~N~J A I ROBERTSON [ 9 9 2 Nino- gen and p h o s p h o i o u s cycles, p. 2 5 2 - 2 9 2 ~rr~ A I. Rober t .on a n d D. M Alongi ( e d s ) , Tropical Mangrove Ecos?stems Amer ican Geophysical U n i o n , W a s h m g m u , D C

ALONOI, D M., R C}tRI~*cO~2eRs kN~) F. T~2~I)I. 1993. The inf luence of fbrest tj4,e on micl obialm u t r i en t re la t ionships in tropical m a n g r o v e sediments. Jo~,.;~aal of Experimerdd Marine Bi ologj)~ and Ecdog), 171:201-223

A.SPILA, K. I., H AOE~IaN, A~N'D A. S. Y CKAL 1976 A semi- au toma ted m e t h o d for the de~-rmilaation of i no rgamc , organic aa~d total phospha te m sedhamnts Anal?st 101:187-197

BIN~L~y, D. A_ND R VITC/LSEK 1989 Soil n u t r i e n t availabftib', p. 76-96 }r~ R. W Pearcv, j . Ehleffmger, H. A Moone v and R W Runde l (eds.), P lant Ph?siological Ecology--Fie ld Methods a n d InstruIr~entation C h a p m a n & Hall , London .

BL~;CO, F 1984 Clixnatic factors and the N o l o g ) of mangrove plants , p 18-35. lr~ S. C Snedaker a a l d j C. Snedaker ( eds ) , The Mangrove Ecos}stem: Research M e t h o d s Uni~-d Nat ions Educat ional , Scientific aald Ctt[tttral Organiza t ion , Pro-is

BOON, R I., D j W MorbidiTy, .~'~U R C SA~TION~ 1986 Rates of a r n m o n i n m tu rnover aud the role cff amino-acid demni- na t ion in seagrass (Zvstera capricorni) beds of More ton Bay, Anmai la . Marine Biology 91:259-268

BOTO, K C 2~\~)j t ~7F~L~,GTON 1984 Soil characmiist ics mid n u t r i e n t stants in a n o r t h e r n Austtaliaxl mealgrove fo res t E~ t~a~.ie~ 7:61-69.

BOTO, K C , J . S. Bt,~.'r, ~ \ ~ J T ~rELLINC-TON 1984 ~'miations in mangrove forest producftvii;~ in n o r t h e r n Austsalia a n d Papua New Gffmea IGtuari~,e. doastai and She~ &ience 19:321- 330.

BOWDEN, W. B. 1984 A n i n o g e n - 1 5 isotope di lut ion study of a r r ~ o n i n m p roduc t ion eald c o n s m n p t i o n in a m a r s h sedi- m e n t LimnoZogi and Oceanograph i 29:1004-1015.

BOY~R,J. N , J . W FOLW<qLR~'~% ~'~) R D. J o ~ s . 1997 Spatial chm-act~.rizatlon of wa~e'r quality of Florida Ba?, and White- wa~er Ba;~ b? m u l t i v m a t e anal?ses: Zones of silnftar i n f l u e n c e Estueries 20:743-758

C~d~LSON, R R. ~N~ L A Y~tBRO. 1988. Physical ea~d biological c o n e o l of m a n g r o v e pore wat~-r chemistry, p. 112-182. li~ D D Hook, W. H. McKee,JL, aud H. K Snfith ( e d s ) , 13ae Ecol- ogy a n d M a n a g e m e n t of Weftmads T imbe r Press, Por t land , O~egon.

g~'~aLSON, R R , L A Xs C. R Zff?&IER~NiA~\*~q, &N~)J R MON'> C-OJV~2RXZ 1983 Pore ;va~-r chemis t ry of an overw~sh mangrove i s l and Flu'ida &ientist 46:239-249.

CARIER, M R., L A BL*~\rS, T. R. ( 2 ~ E R , K R. DUGGER, P L FORE, D E HIOI~q, H C REVELL% A,\K~ A W SC&~IDT 1978 Ecosvstem 2malysis of the Big Qepress Swamp and Estuaries. EPA 90. t /9-74-002. Unit~-d Stat~s Env i ronmen ta l P~otection Agencv, Region 4, Adanta , Georgia

C~d-,~r ~' J. 1944 Cambr idge University expedi t ion m Ja. m a i c a I. A s tud v of the botanical processes c o n c e r n e d in the deve lopmen t of tl~e Jamaican shore-lh~e Journal qfthe Li~nean Soc ietj o fl London Bota n',~ 52:407-447

C>ta_p?~c~% V. J. 1976. Mangrove ~egetat ion. J. Gramer, Vaduz, Germany.

C~mN, R. H. 1996. Ecological anal?sis mad smaulation models of landscape pa t te rns in mmagrove forest deve lopmen t a~d soil chea-act~-t{stics a long the Shark River esmm'y, F lor ida P h D . Dissertation, University of Southwestern Loulsiea~a, Lafayette, Lo~dsiea~a

C>~x~, R. A~\~ R R ' I%r [998. A gap dy~mr*ic m o d e l of

m a n g r o v e forest deve lopmen t a long gxadients oF soft saKaity a n d n u t r i e n t r e sources Jou~nal oyEcvlo~? 86:37-52.

CHEN, R M\~) R R T~mLLEga 1999 A sinatflation m o d e l of ol- ganic matte~ and n u t r i e n t accumulat ion in mmagrove wet land soils Biogeoehemistr 3, 44 :93-118

CINTR6N, G , A E LUGO, D j POOL, INX) G MORRIS. 1978 Mm~groves of a d d e n v i r o n m e n t , in Puer to Rico and ad jacent islands. Biotr@ica 10:110-121.

CIN-KRON, G 2~ND Y S c a L ~ - N o ~ u t 1984a Caracteristicas y desm'~ollo estr~lctttral de los mang la res de Norh- y Sux Airier- ica. g~!'oKrama Regional de De:,arrolio Cient!fico y Tecnolog[co 25:4- 1 5

CIJ'.WR6?x, G ~N~3 Y gCr4AEF~ER-NO'~LLI 1 9 8 4 b Methods for studying mangxove attuctm-e, p. 91-115 }ra S. C Snedaker a n d J. G Snedaker ( e d s ) , The Mangrove Ecosystem: Re- search Methods. Uni/~.d Nat ions Educat ional , Sdenfif ic and Cultural Organiza t ion , Pavis

CRY@W, C. B. m'~q) C. J RiC}#'.~DSO~. 1993a Peat accret ion and phosphoxus accmrmlat ion a long an eu t rophica t ion grac~ent in the n o r t l l e r n Everglades Bioge~hemistrj 22:13.%-156.

CR~FT, C. BAWD C J. R i ~ D S O N 1993b Peat accret ion and N, R and organ ic C acc~lmulation in nu t r i en t - en r i ched and ut~enr iched Everglades p e a t l a n d s Ecological A~lications 3: 4 4 6 - 4 5 8

CRgd6HF~D, F C aND ~v C. GILBERT. 1962 The effect , of Hur- r icane D o n n a on the vegetation of sou the rn F lor ida The Quarierg] feelgmai of the F~offda Academy o/ &iences 25 :1 -28

DA\~% JR , J. H. 1940. The ecology at~d geologic role of mangroves in Florida. G~negle Institute o) Washbzgr Pnb~ication 517, 52:303-4[2.

HA\ms, S M 1991. Grow*h, decomposi t ion , m~d n u t l i n n t re ,on- d o n of Cgadium ~a*naicens Crants and 75~ha do'mingensi:, Pets in Florida Everglades Aquatic Botarg ~K):203-224

DA.aS, S M 1994 Phosphorus i n p u t , a n d vegetation senNtbi W i n / h e Everglades, p. 357-378. b~ S. M. Davis arMJ. C O g d e n ( e d s ) , Everglades: The Ecosystem a n d it. Res to ra t ion S t Lu- t ie Press, Delray Beach, ~]or ida

DAY, j W , W H. Co.xv~aR, F L ~ : L o u , R. H DA< .<~n A M. NAVARRO. 1987 ' I l le p~oductivi~7~ m~d composi t ion of mm~- grove forests, Lag~ma de Farnf inos , M e x i c o Aquatic Boto%, 27: 2 6 7 - 2 8 4

DeBLsK, W E , K R ~ D ~ ; M S KoCH, mN~O Y W ~ w o 1994 Spatial distr ibution of soil nut f ient~ in a n o r t h e r n Everglades marsll: Water Conservat ion M-ea ~ Soig &ience Societ] oyAme~ ica jo~srnai 58:543-552

DeK&N~L, J. ~N~ J. W MORSe 1978 The chemis t ry of orthc~ phospha/~" uptake f r o m seawater on to caldte a n d m-agonite. Geoehimica et Cosmo&i~nica Ae~a 42:1335-1340

Dma~oN~, J. 1985. Overview: Labora tory expe r imen t . , f ield ex- per iments , and na tura l experknen ts , p. 3-22. b ~ J D i a m o n d a n d rE J. Case ( e d s ) , Commani t ) , Ecology. Hm~per mad Rove, New Yolk

DUE~Y~R, M. j . , J. K MEKr)Ea% L C MEEDER, AND J. M MCCOLLOM. 1994 ' l]le d i m a t e of sou th Flovida a n d i t . role in shap ing the Everglades ecosystem, p 225-248 }n S M. Davis a n d J. C O g d e n ( eds ) , Everglades: "tile Ecos?stem and its Res tora t ion S t Lucie [hess, Delta? Beach, F lor ida

Du ty , N C 1992. Mangrove floristics a n d Nogeograph? , p 63 - 100 b~ A I. Rober tson a n d D. M Atongi ( e d s ) , 7~opical Mangrove Ecosystems Mr*evicma Geophysical Union , Wash- i ng ton , D.C.

F ~ A , I C. 1995. Effects of n u t r i e n t e m i c h m e n t on growth a n d herb ivory of dwar f r e d m a n g r o v e (Rhizqph~ra mangle) EcoZoglcai Monographs 54:477-505

FOL%QUREMN, J. w , J c . 7~EMAN, ~N~ G ~v N Po~ 1992 Phosphoinas l imitat ion of p r imary p~oduct ion in Florida Bay: E.ddence flc,m C:N:P ratios of the dominax~t seagrasa Thagassia teatudine~m. Limno&c] and Oeeanograph] 37:162-171.

GALE, P. M., K. R. R~DD~; m~,n) D A. GRa~TZ 1994 Phosphoi~ls

Mangroves and Soil Nutrients 9 6 9

r e t en t ion by wet land soils used for t rea /ed wasie~mWr dispos- al. J~lrnai qf Environmentai Qvaiit] 23:370-377

HEATH, E.J. [ 9 5 9 The p roduc t i on ,,f o rgan ic detri tus in a soutll Florida es tuaIy P h D Dissertation, University of Miami, Coral Gables, F lor ida

HESSE, P R. 1962. P h o s p h o r u s f ixat ion m matagrove swmrlp m u d s Nature 193:29~296.

H I E L I ~ , A. H M. iND L LI[KLFC,~J\ 1980 Fract ionat ion of re- o rgan ic phosphaWs in calcareous sediments. Jet.real WEnvi ron~n~tai Qualit), 9:405-407

KOCH, M S. 1996. Reso~trce a~ilabiliu~ and abiotic stress effect. on Rhiz@,hora ~.arlg'h L (Red mangrc,ve) deve lopmen t in sou th F lor ida Ph.D. Dissertation, Universi U of Miami, Coral Gables, Flovida.

K o c h , M S ANrJ K R RzroDz 1992. Distribution of soft mad p l an t n u t r i e n t , a long a tropistic g i s t / t en t in the Florida Ever- g lades Soil &ienee Socie~, o/ Amerlca,/~,t~al 56:1492-1499.

KOCH, M S MNX3 S C. SJN~DAKER 1997 Factors in f luenc ing P&i zophora mangze L seedling deve lopmen t in Everglades caxbon- ate soils. Aqvatic Boron',,, 59:87-98

LkWToN-TIIogLGS, L L 1997 Canopy re t tans locat ion and [itser immoMliza t ion oF n i t rogen and p h o s p h o i u s in tgm-ee mangrove spedes a long the Shark River estuary, F lor ida M.S. 1lie- sis, University of Southwestern Louisimla, Lafayette, Louisi- k~lla

LIofrl ' , S S ANDJ. W DLNEFN 1994 Water cont ro l in the E.~- erglades: An tdstorical perspective, p. 4 7 - 8 4 ~%~ S M Davis a n d J C O g d e n ( e d s ) , Everglades: The Ecosyst~-m a n d it. Res to ra t ion S t Lucie Piess, Deft-a? Beach, F lor ida

L u o o , A E 1978 Stress a n d ecos?stems, p. 62-101. ~%J. H. T h o r p and J. W Gibbons ( e d s ) , Energy a n d Env i ronmen ta l Stress DOE 771114 Depm' tmen t of Eneigy, Washing ton , D .C

Lo( .o , A E. 1997 Old-g~owth m a n g i o v e forests in the Unin-d States Ccmservati~r, Biology 11:11-20

L u o o , A E a ' ~ S C. S~.mD2~KER 1974 The ecology of rnm> g ioves Annual Review of Eccgog? and S']ste~natics 5 :39 -64

LUGO, A E , S BRO\VN, ,MN'D M. M BRmSON 1988 Forested wei- latlds in f reshwater a n d sah-waWr envh-omnents . Limnotog? and O~eanouaph i 33 :894-909

MCGkK2ItERy, K J., R MAR~NO, MN~) R ~W ~{O\~rARTI{ 1994 Var- iable rat~-s of phospha t e uptake by shaBow m a r i n e carbonate sedhlmnts: Mechanisms a n d ecological significmlce. Bioge(~ &emistr] 25:127-146.

Mclvoa , C C., J. J. LLu ~-~\~D R. D glOgK 1994 Chmages in f leshwater inf low f r o m the Everglades to Florida Bay i n d u d - in S effects on biota and biotic piocesses: A re.dew, p. 117 In S. M Davis a n d J C O g d e n (eds.), Everglades: The Eco~ysWm a n d it. Res to i a t i on S t Lacie Press, Delta? Beach, F lor ida

McKFz, K. L 1993. Soil ph?sicochemical pat teFns a rm mangrove species cftstriloution--Reciprocal effects? ~ae'~;"~al o/ EcologT] 81: 4 7 7 - 4 8 7

McKKE, K L 1995 Seedl ing recr~fi tment pa t te rns in a Belizean m a n g r o v e forest: Effect. of es tabf tshment ability m~d physico- chemical fac tors Oecol%ia 101:448-460

McKE~., K L , I A MEN~DELSSOHN, AND M ~#[ ~ T ~ J R 1988 Reexar~fination oF pore wat.-r sulfide concen t ra t ion m~d r e d o x pomnt i a l n e a r the aer iN ~oota of P&iz@hora man?tic m~d Avi tennis g'er~mina:os. Ame~{can Jou~nai of Botain, 75:1352-1359.

N I G ~ O N , H H. m~.~ F R. ~u 1985. Association between p o r e w-ater sulfide concent ra t ions a n d the *ftstdbution of m a n g r o v e s Biogeachevzistr] 1:183-192.

ODLN, W E , C C MCIvoR, a~\~ T J. S~aPa, III. 1982. The ecology of the mm~groves of south Florida: A c o m m u n i D' pro- f i l e FWS/OBS-81/24 U n i t e d States Fish a n d Wildlife Ser- vice/Office of Biological Services, Washington, D C

ODu~l, W E., T.J . SMrreL III , J. K H o o ~ , < \ ~ C M c l v o a 1984 The ecology oF tidal fresftw'ater marshes of the tidal f reshwater maxshes of tile Un in -d States East Coast: A corn-

9 7 0 R. Chen and R. R. Twilley

m~mi~ y profile. F W S / O B S - 8 3 / 1 7 bni te 'd Smte, s Fish and Wild- fife Service, Wastf ington, D C

O L S ~ , S R ,~\~ L E SoMbreros t982. Phosphorus , p 403-430. s L Page, R H. Miller, mad D. R. Keeney ( eds ) , Me~hods of Soil Anal?sis Par t 2 Chemica l and Microbinlogical P~op- e r t ies }krnerican Society of A g r o n o m y and Soil Science Soci- e U, of America, Madison, Wiscons in

P~u~SONS, 12 R., Y M2~rlA, A~\qD C M L~LLI. 1984 A Mmaual of Chemica l a n d Biological Methods for Seawater Anal?sis Per- gmr*on P~ess, New York.

P~uSTOR, J , M A S~LLX@i~, &\~ D ' I ~ A ~ . . 1987 Ni t rogen minera l iza t ion a n d nitr if icat ion ha fi)ur Minneso ta old f ie lds Oecdogia 71:481-485.

Powemr A B , D E Hoss , W F }t~Tr~ga, D. P~Th~S, ~NY~ S WAO~t~ 1989. Ab~mdax~ce a n d cfistfibution of ichrhyoplauk- ~on in Florida Bay a n d adjacent wa te r s B~g~etin o/Marine Sci once 44:3~-48

P~ovos% M. W 1973. Mema h i g h w~ter mark a n d use of tide- lands in F lor ida Ft.'ida Scie:4tist 36:60-66.

Rr~Dy, K. R . R. D DeI~m.~m, W F. DEBLSIr 2~N~3 M 8. KOCtI. 1998 Long- |o t to n u t r i e n t acctmaulation rates in the Ever- g lades 8oil &ie*~e~ S~iet I, o/ Amedca joez~r~ai b7:1147-1165

RIVtLRA-MoNNo~z; ~Y H Y'~N~D R. R. T~VIKZ~s 1996. The relative role of denitr if icat ion and immobi l iza t ion on the fate of inorgmaic n i t rogen in m a n g r o v e sedha*ent, of Te rminos Lagoon , Mexi- co. Li~ir~olog 1 arid Ocea~ograph}, 41:284-296

RIVERr V H., R. R I"WILIs R BOUST&N~Z, J. V~ 7. DAY, .JR, F V~R~-HERRER~., .~\~ M R&~mEZ 1995 Dh-ect denitt ifi- cation in mangrove seNmenCs in Te rminos Lagoon , M e x i c o Marine Ecoioffi Progress Serie~ 126:97-109

ROSF=~TeLB, J. K. 1979. Inters t inal waler and sed iment chemist t y of t~o cores f lo ra Florida Bay ,f~u~'uai or Sedi~e~ztacy l~ 49:989-994

S ~ T ~ , I I I , T J., K G BoTo, S D FRUSHEa, A~O R L GmDIN8 1991. Keystone species a n d mangrove forest dyaaamics: 'flue inf luence of bin-rowing by cxabs on soil n u t t i n n t status axed forest productivity. Estvarine. Cbastai and She{r &ier~ce 33:419- 432.

8IVdlT~I, III , '11 J , M. B ROBBLEE, } t R. \~7K\XAGb8, A~.D T W DOYLE. 1994 Mangroves, hur r icanes , a n d l i g h m i n g strikes. Bio8tierace 44:256-262.

STRtCI~r J. D f t ~\~0 'E R. PAP.SONS. 1972. A practical hmad- books of semw~ter axial?sis Fisheriea Re:,ear& Board o/ Canada 167:1-310

THO*,I, B. 1967 Mmagrove ecolog) a n d deltaic morphology: Ta- basco, Mexico. Jcez;~r~al o tEcology 55:301-343

1MOM, B. G 1982 Mangrove ecology--A geomoEgholog{cal perspective, p. 3 - 1 7 ~ B E Clough (ed.) , Mangrove Ecos~st.-ms ha Australia. Aus|Iafian Nat ional Uni.~ersi|7~ Piess, Canber ra , Aus t r aha

THOM, B. G 1984 Coastal landfo~rns mad geomo~phic proces> es, p 3-17. l~z S C Snedaker a n d J G Snedaker ( e d s ) , The MangTove Ecos?s~em: Reseaich M e t h o d s Uni ted Nations Ed- ucat ional , S d e n ~ c mad Ca lmra t Organf ra t ion , Paris

T~io*vL~.s, rE M 1974 A detai led analysis of cl imatological a n d hydrological records of south Florida with re ference to m a n ' s inf luence upon ecos?smm evdu t ion , p 8 2 - 1 2 2 }r~ R J . Glea- son ( e d ) , E n v k o n m e n t s of Sc,uth Florida, Pxesent mad Past, M e m o h 2. Mimrd Geological Socie~ b Miami.

TOM~SON, R B. 198i5. The Botany of Mangroves Cambr idge Universi |? Press, Cmr*bfidge

TWmUEu R. R. 198B "llle exchange of o rgan ic carbon m basin mangrove forests in a southwest Florida e smar> s Coastal arid Shr Scie~ce 20:543-557

'~/rlLLEY, a R. 199~. P~operties of m a n g r o v e ecosystems relat..d to the energy s ignantre of coastal e n . & o n m e n t s , p 43-(;2. }r~ C Hall (ed.) , MaxhT~mr* Power. University of C 4 o r a d o PIess, Boulder , Color ado.

'Ih~mxaag R R 1997 Mmagrove wetlaxlds, p 445-478. b~ M Me> sina a n d W C o n n o r (eds.), Sou the rn Forested Wetlands: E c 4 o g y mad Management . CRC Press, Boca Raton, Florida.

'I'WgLLEy, R R aND R C ~ L \ 1998 A water budge t a n d hycho f ogy m o d e l of a basin mangrove forest m Rookery Bay, ~lofida. A~st~aga~ Jo~'~rnag oy Fresi~water and Marine Rcsearch 49:309- 323.

'I 'wmL~Z R. R., A E L u o o , 2~*~ C. PATTeXSON-ZUOCA 1986 Pro- duct[on, s tanding crop, a n d decompos i t ion of fitter in basin mangrove forests in southwest F lor ida Eco&~2} 67:670-683.

W*~LSH, G E 1974 Mangroves: A review, p 5 1 - 1 7 4 b~ R Rei- t o d d aild W. Q u e e n ( e d s ) , Ecology of Halc,phytes Academic P~ess, New Yo~k

WAtsON, J 1928 Mmagrove forests of the Malay Peninsula. Ma. layan Forest Records b. F:aser & Neave, Ltd. , S i n g a p o r e

~gOODRO}'PE, C. D 1992 Mangrove sediments mad geomor- pholog}: p 7-41. b a A I. Robe I t . on a n d D M ~Mongi ( eds ) , Tropical Mangrove Ecosystems. Amer ican Geophysical lonion, Washing ton , D.C.

ZWOLS.~_w, J. G 1994 Seasonal rm'iabili b, a n d b iogeochemis t t y of p h o s p h o r u s ha the Scheld t esmar,~, south-west Nedte r lands EgZ~ari*~e. Uoast~i and She{/&lento 39:227-248

Rccdved tot considerafi~rr,. M@ 11. 1998 Acc@td p~r pez&iication. March 2. 1999

Documents

Patterns of Mangrove Forest Structure and Soil Nutrient ... · 'The coastal forested wetlands in the southwest- ern Everglades of' Florida extend h-ore Whitewater ... wetlands in