Litter Dynamics and Phenology of Melaleuca Quinquenervia in South Florida T
Total Page:16
File Type:pdf, Size:1020Kb
J Aquat. Plant Manage. 40: 22-27 Litter Dynamics and Phenology of Melaleuca quinquenervia in South Florida T. K. VAN', M. B. RAYACHHETR"Yl,T. D. CENTER! AND P. D. PRATI' ABSTRACT and extended into the spring. Very little new growth was ob- served in melaleuca forests during the summer months, We monitored litterfall biomass at six different sites of from May to August, in South Florida. In contrast, the fall of melaleuca (Melaleuca quinquenervia (Cav.) S.T. Blake) forest- leaves and small wood was recorded in every month of the ed wetlands in South Florida from July 1997 to June 1999. year, but generally increased during the dry season with Annual litterfall of melaleuca varied between sites from 6.5 higher levels observed from February to April. Also, no sea- to 9.9 t dry wt ha-' yr! over the two-year period. Litterfall was sonality was recorded in the fall of seed capsules, which ap- significantly higher (P < 0.001) in seasonally flooded habitats parently resulted from the continual self-thinning of small (9.3 t ha-1yrl) than in non-flooded (7.5 t ha" yr') and perma- branches and twigs inside the forest stand. In planning man- nently flooded habitats (8.0 t ha-'yr'). Leaf fall was the major agement for perennial weeds, it is important to determine component forming 70% of the total litter, woody material the period during its annual growth cycle when the plant is 16%, and reproductive material 11%. Phenology of flower- most susceptible to control measures. These phenological ing and leaf flush was investigated by examination of the tim- data suggest that the appropriate time for melaleuca control ing and duration of the fall of different plant parts in the in South Florida might be during late winter and early litter traps, coupled with monthly field observations during spring, when the plant is most active. the two-year study. In both years, flowering began in October Key words:Litterfall; Biomass; Productivity; Biological con- and November, with peak flower production around Decem- trol; Paperbark tree. ber, and was essentially completed by February and March. New shoot growth began in mid winter after peak flowering, INTRODUCTION Melaleuca, often referred to as paperbark or broad-leaved 'u.s. Department of Agriculture, Agricultural Research SeIVice. Invasive paperbark tree, is native to Australia. It is highly invasive in Plant Research Laboratory. 3205 College Avenue. Fort Lauderdale. FL 33314. wetland habitats of South Florida where it has infested about 'University of Florida. Fort Lauderdale Research and Education Center. 3205 College Avenue. Fort Lauderdale, FL 33314. Received for publication 200,000 ha (Bodle et al. 1994). Vast areas of wetlands, includ- Q<;toh<:r30,2000 and in revised fonn November 13, 2001. ing parts of the Florida Everglades, have been converted 22 J Aquat. Plant Manage. 40: 2002. from marshes to closed-canopy melaleuca forests, with major environmental and economic impacts (O'Hare and Dalrym- ple 1997). As part of a biological control program, the Aus- tralian melaleuca snout beetle (Oxyops vitiosa Pascoe) and other natural enemies specific to M. quinqueneroiaare being imported into the U.S. (Center et al. 2000). This study was designed to obtain quantitative information on melaleuca productivity in terms of litterfall, and the relative propor- tions of the litter components, prior to the impact of the re- leased biological control agents. Litterfall collection is a standard non-destructive tech- nique for assessing productivity, phenology, and turnover of biomass in a forest (Newbould 1967). In general, net produc- tivity of woody trees such as melaleuca, is difficult to measure because a large proportion of the plant material exists in the trees for a long time before entering other parts of the eco- system. Much of the short-term input from a melaleuca forest to the ecosystem is as the litter falling from the tree canopy to the forest floor, and incorporated into the rest of the system by detritus feeders and decomposer organisms. This com- partment of net productivity is relatively easy to measure. In w:i Melaleuca Distribution particular, the amount of leaf material falling reflects a for- est's productivity. Rainforests, with their dense leaf canopies and high rates of leaf fall, are considered among the most highly productive plant communities (Proctor 1983). Annual litter production varies primarily with latitude, and ranges from 0.6 t ha'! yr! in the arctic tundra to 25.4 t ha'!yrl in a tropical lowland forest of eastern Australia (Lowman 1988). In addition to latitude, the amount of litterfall is affected by Figure I. A map of South Florida showing location of the six study sites and their grouping into three different habitats. The Holiday Park sites (l and precipitation and various attributes of the vegetation (Bray 2) were non-flooded, Krome (Site 3) and Clewiston (Site 4) seasonally and Gorham 1964). No information is available on litterfall flooded, and the Conservation Area sites (5 and 6) permanently flooded. or turnover of melaleuca forests in South Florida. Measure- ments in its native range in Australia (Finlayson et al. 1993, Greenway 1994) however, have indicated that annual litter- fall in melaleuca forests was the highest recorded among oth- of heavy rainfall; however, this flooding is usually of short er Australian temperate/subtropical forest types, suggesting duration and may not happen every year, and thus cannot that these woody wetlands form highly productive ecosys- be considered as seasonal flooding. Therefore, in this study tems. This paper presents data on litterfall biomass from six the term 'non-flooded' means 'not flooded seasonally' rather sites representing different hydrological conditions within than 'never flooded'. melaleuca forested wetlands in South Florida. The objectives Mean annual rainfall of the study area was 1420 mm were to report the first measurements of litter production by (NOAA).3 The area received most (75%) rainfall during the melaleuca in South Florida, and secondly to describe impor- months of May through October in the form of convective tant phenological events of flowering and growth of melaleu- storms. Mean rainfall during those months was 178 mm per ca through records of the seasonal patterns exhibited by the month. Mean monthly temperatures ranged from 19.5C in different components of the litter. January to 28.2C in August. Vegetation at all six selected sites consisted of dense, MATERIALS AND METHODS closed-canopy melaleuca stands similar to those described by O'Hare and Dalrymple (1997) as '75-100% dense, mature Six permanent study sites were established throughout the melaleuca (DMM)'. The forest under-story was generally geographic range of melaleuca in the Everglades in South sparse, and consisted of shade-adapted shrubs (e.g., Myrica Florida (Figure 1). The sites were selected to encompass dif- eenfera L., Myrsine guianensis (Aubl.) Kuntz., Baceharis sp.), ferent hydrological conditions, as hydro-pattern has been several ferns (e.g., Thalypterissp., Osmunda sp.), sedges (e.g., considered as a strong determinant of wetland species and CladiumjamaieenseCrantz.), and grasses. their performance (Mitsch and Gosselink 1986). The sites Four 10 by 10 m plots were established in a stratified ran- consist of typical 'glades' characterized by high organic dom design at each of the six study sites. In each plot, all (muck) soils, and were grouped into three habitat catego- woody individuals >1.3 m tall were identified, counted, and ries: seasonally flooded, permanently flooded, and non- flooded (Table 1). There were two sites for each habitat. It should be noted that non-flooded melaleuca forests in the 'National Oceanic and Atmospheric Administration (NOAA), 1997- Everglades can still be covered by flood water during periods 1999, Local Climatological Data Monthly Summary. J Aquat. Plant Manage. 40: 2002. 23 TABLE 1. ABOVEGROUND BIOMA<;S AND OTHER STRUcrURAL CHARA(''TERISTICS OF MELALEUCA FORE.'US AT THE STUDY SITE.<;. HOLIDAY PARK SJn~<; WERE NON- FLOODED, KROME AND CLEWISTON SEASONALLY FLOODED, AND CoNSERVATION AREAS PERMANENTLY FLOODED. DATA ARE MEANS:t STANDARD ERROR OF FOUR 100-M' PLOTS FOR EACH SITE, OR EIGHT PLOTS FOR EACH HABITAT. BIOMASS WA<;CALCULATED USING REGRE.<;SIONS ON DBH (VAN ET AI.. 2000) DBH Distribution Description < 10 em 10-20em >20 em Density Biomass By Sites: (Trees 100m" plot) (Trees ha") (t ha") Holiday Park 1 171 :t 49 19:t 6 3:t2 19225 :t 4559 141 :t 27 Holiday Park 2 192:t67 19:t 5 2:tl 21375:t 6349 146:t 25 Krome 238 :t 43 36:t 5 9:t4 28225 :t 3635 302 :t 53 Clewiston 82:t 41 2l:t 8 12:t 6 11450 :t 4394 294 :t 77 Conservation Area 2Bl 331 :t 127 25:t 8 8:t4 36275:t 11658 265 :t 50 Conservation Area 2B2 242 :t 65 34:t 3 8:t3 28400:t 6120 304 :t 30 By Habitats: Non-flooded 181 :t 39 19:t 4 3:tl 20300 :t 3641 144:t 17 Seasonally flooded 159 :t 40 29:t 5 1O:t3 19837:t 4125 298 :t 44 Permanently flooded 286 :t 68 30:t 4 8:t2 32337:t 6274 285 :t 27 their stem diameters at breast height (dbh) recorded. lations involved, it was assumed that daily litterfall rates did Shrubs and herbaceous cover were also enumerated by spe- not change during collection intervals. cies for use in vegetation inventory for the plots. Phenology of melaleuca was determined by examination Aboveground biomass and its distribution (wood, leaves, of the litter components coupled with monthly field observa- and reproductive parts) were estimated using least-squares tions. For example, different reproductive structures in the regressions on dbh. The regression equations for total bio- litter traps provided empirical data to support visual field ob- mass (Van et al.