RESEARCH ARTICLE ABSTRACT: Lygodium microphyllum (Cav.) R. Br. is a non-native invasive fern that has become a serious problem in many habitats in southern Florida. The effectiveness of fire and/or triclopyr ester in killing L. microphyllum, the time and .amount of herbicide required for inspections and re-applications, and • the effects of these treatments on a southern Florida pine flatwoods community were examined. These treatments were: (1) herbicide application with bimonthly inspection and re-application if necessary, (2) herbicide application with biannual inspection/re-application, (3) prescribed fire to reduce L. microphyllum Using Fire and biomass followed by biannual inspection and herbicide application, and (4) untreated controls. All fire and/or herbicide treatments killed standing L. microphyllum, and the prescribed fire reduced by about Herbicide to one-half the amount of subsequent herbicide, but not the time, required to kill regrowth. No treatment prevented L. microphyllum regrowth, and every treatment had at least one new frond at the end of the three-year study. Fire and/or herbicide treatments did not permanently decrease native species cover, Control Lygodium richness, evenness, or diversity (Shannon's H'), and native species cover increased following biannual herbicide and firelbiannual herbicide treatments. Two-month inspection/retreatment intervals were not more effective than six-month intervals. Lygodium microphyllum can return to former amounts of biomass microphyllum and and cover within a few years of burning. Waiting too long to inspect and retreat negates the benefits of Effects on a Pine using fire to reduce L. microphyllum biomass. Index terms: fire, herbicide, invasive plant, Lygodium microphyllum, Old World climbing fern Flatwoods Plant Community in South INTRODUCTION lum only very infrequently covers more than small patches of ground (R. Pember Lygodium microphyllum (Cav.) R. Br., Old ton, research entomologist, United States Florida World climbing fern, is an indeterminately Department of Agriculture, pers. comm.), twining fern found throughout much of and does not dominate its plant community the Old World tropics (Pemberton 1998), (Goolsby et al. 2003). In Florida, however, Randall K. Stocker1, 4 where it is frequently described as "weedy" the rapid spread of the species has been (Tagawa and Iwatsuki 1979; Singh and Raymond E. Miller Jr.2 compounded by the establishment of large Panigrahi 1984). It was first reported as nearly monospecific stands. The extensive 2 David w. Black a naturalized species in Florida (Martin indeterminate growth of single compound Amy P. Ferriter3 County) in 1965 (Beckner 1968), and con fronds of L. microphyllum can create dense Daniel D. Thayer2 cern about the spread of L. microphyllum in accumulations of light-blocking biomass central Florida was voiced as early as 1978 covering and killing herbaceous and shrub 1 IFAS Center for Aquatic and Invasive (Nauman and Austin 1978). This species layers, and can even lead to the death of Plants and Department of Agronomy has now become a recognized threat to na mature trees as the fern twines into the University of Florida tive plant communities throughout southern overstory canopy (Pemberton and Fer 304 Newell Hall Florida (Pemberton 1998; Pemberton and riter 1998). Plant communities in Florida PO Box 110500 Ferriter 1998). Extent of L. microphyllum infested by L. microphyllum include bald Gainesville, Florida 32611-0500, USA in southern Florida was calculated (aerial cypress (Taxodium distichum [L.] Rich.) survey) to be 11,213 ha in 1993, 15,892 swamps, pine flatwoods, wet prairies, 2 South Florida Water Management ha in 1997, 17,410 ha in 2001, and 48,878 saw grass (Cladium majaicense Crantz) District ha in 2005, and estimated to be 74,090 marshes, mangrove stands, Everglades tree P.O. Box 24680 ha in the entire southern Florida region islands, and disturbed areas (Pemberton West Palm Beach, FL 33416, USA in 2005 (Ferriter and Pernas 2006). The and Ferriter 1998). exponential nature of this increase can be 3 Idaho State Department of Agriculture seen by plotting the survey data collected While several herbicides are able to kill 2270 Old Penitentiary Rd. by Ferriter and Pernas (2006) and in a more L. microphyllum (Stocker et al. 1997), Boise, ID 83701, USA focused examination of spatial distribu the rapid spread of this species makes it tion of L. microphyllum over time in the unlikely that herbicide-only programs will • Everglades region of southern Florida (Wu be able to contain the plant, nor lead to et al. 2006). The potential distribution of its eventual control. A statewide manage 1 Corresponding author: L. microphyllum in the Everglades (Volin ment plan recommends biological control [email protected] et al. 2004) and more generally in North combined with herbicide treatments as the and South America has been examined as best strategy for long-term management well (Goolsby 2004). and control (Pemberton et al. 2006). The Natural Areas Journal 28:144-154 biological control program is well under In its native Old World range, L. microphyl- way (Goolsby et al. 2003), and the first
144 Natural Areas Journal Volume 28 (2),2008 biological control agent (Austromusotima control and regrowth of L. microphyllum; replicates (6) of each of four treatments. camptonozale [Lepidoptera: Crambidae; and the effects of those treatments on a pine Data collection was repeated in November Yen et al. 2004]), a frond-feeding mus flatwoods plant community, one of many 1998 and 1999. otimine moth, was released in 2005. It is plant communities in Florida that require possible that several different biological frequent burning to maintain desired native Treatments consisted of: (1) initial use control agents will be necessary, and until plant diversity and habitat for threatened of herbicide to kill L. microphyllum and a more diverse biological control program and endangered plants and animals (Lan herbicide treatment of regrowth at two is implemented, herbicides will remain the geland and Hutchinson 2006). Additional month intervals (termed bimonthly herbi most effective readily available manage objectives included comparing two differ cide), (2) initial use of herbicide to kill L. ment tool. ent frequencies of herbicide re-treatment microphyllum and herbicide treatment of (every two or six months) and determining regrowth at six month intervals (biannual The importance of fire in the evolution time and herbicide requirements for the herbicide), (3) initial use of fire to reduce of Florida's plant communities and the tested treatments. L. microphyllum biomass followed by use of prescribed fire to manage habitats herbicide treatment at six month intervals for native plant diversity (Wade et al. . (fire/biannual herbicide), and (4) untreated METHODS 2000) suggest that fire might offer an control plots. The final interval in biannual additional management tool to control herbicide and firelbin herbicide plots was The study site is a wet flatwoods dominated L. microphyllum, except for three very four months instead of six. Each treatment by Pinus elliottii Engelm. (slash pine), important factors: (1) fire alone does not was replicated in six plots in a completely located in Palm Beach County, Florida, on kill L. microphyllum (Maithani et al. 1986; randomized design. It is recognized that the south side of Indiantown Road approxi Roberts 1996; Stocker et al. 1997), (2) L. bimonthly application of herbicide is mately 1.6 km west ofInterstate 95. A fire microphyllum regrows very quickly after impractical for most resource inanage in 1995 killed many Taxodium distichum fire (Goolsby et al. 2003), and (3) fire may ment situations, but L. microphyllum had and P. elliottii. Lygodium microphyllum re provide open sites for establishment of new recovered quickly from some herbicide ap covered very rapidly following the fire, and populations (Langeland and Hutchinson plications in demonstration plots, and it was by 1997 covered much of the herbaceous 2006). Fire can, however, reduce both desirable to know if even an impractical and shrub layers and many of the stan4ing living and dead standing L. microphyllum bimonthly frequency' would successfully trunks of dead trees. At the start of the biomass, possibly reducing the amount of kill this species. study (December 1997), L. microphyllum . herbicide required for subsequent treat was estimated to occupy approximately ment of regrowth and unburned plants, Fire was applied to firelbiannual herbicide 73 ha of the 176 ha total property area. and make access to sites much easier for plots on 14 January 1998. Mter initial Since the property was scheduled to be management personnel (Langeland and ignition of biomass, the plot was allowed developed into a Palm Beach County Hutchinson 2006). In small demonstration to burn without further assistance unless natural area and recreational facility, the plots, spot fires have been used to burn re-ignition was needed to continue the research was conducted in parallel with a L. microphyllum climbing into overstory burn. Lygodium microphyllurn has very larger restoration effort. trees. Herbicide (glyphosate) applications thin pinnae, and fronds readily ignite and to regrowth three to four months after the support rapid, intense burning. Fire was On 3. December 1997, a 20-m x 40-m fire were effective, with little regrowth seen suppressed with shovels and water when area was divided into 24 plots, each 2-m (J. Hutchinson, graduate student, Univer it reached the plot borders. No heat ef x 8-m, with I-m strips between plots on sity of Florida, pers. comm.). While spot fects, such as curled or scorched leaves, their long axis and 2-m strips between plot burning followed by herbicide may not be were observed on plants in neighboring ends. Three 0.25-m2 square plot frames suitable for larger infestations, it may be unburned plots. were randomly located in each plot, from an effective tool in nascent populations which L. microphyllum biomass samples (Langeland and Hutchinson 2006). J'riclopyr ester (PathfinderTM, 0.09 kg L-l were collected and dried to constant weight active ingredient) was used for all herbicide in a forced air drying oven. Complete Since fire is an important and frequently applications because it had been effective floristic lists, and cover estimates for each used tool for vegetation management in killing L. microphyllum as a directed understory species, were recorded in five. (Wadeetal. 2000), and since wildfires spray in unreplicated demonstration plots. 20-cm x 50-cm sub-plots spaced at I-m occur regardless of management plans, All green parts of L. microphyllum were intervals centrally along the long axis of research is needed to learn more about the treated in a spray-to-wet application by each plot. Estimated cover was reported interaction of fire and herbicides and the ef backpack (initial treatments) or hand-held as one of six unequal size Classes (>0-5%, fects of these management tools on invasive sprayer (retreatments). Initial herbicide >5-25%, >25-50%, >50-75%, >75-95%, species like L. microphyllum and on the was applied on 13 and 17 January 1998. and >95-100%; Daubenmire 1959). Mean native plant communities. This study was Care was taken to minimize contact to cover for each species was calculated from conducted to assess the effects of herbicide, non-target species and avoid drift to ad the total (30) of all sub-plots (5) in all and a combination of fire and herbicide on jacent plots. No herbicide effects, such as
Volume 28 (2), 2008 Natural Areas Journal 145 browned or killed tissue, were observed on RESULTS tively (Figure 1). Cover for these species plants outside the treated plots. in treatment plots did not differ (P=0.15) Pre-treatment mean cover of L. microphyZ from control plots during the course of the At the scheduled intervals (two or six times Zum across all sub-plots (71.9%; SE=2.7) study. One non-native herb, Urena Zobata per year), L. microphyllum regrowth in was not significantly different among L., increased with fire and/or herbicide all plots (except untreated controls) was treatments (P=0.94), and was 56.4% of treatments (Table 1). treated with herbicide. The time required the total vegetation (all species) cover for inspection and application, as well as of 127.5%. (SE=3.7). Mean cover of L. Effects of treatments on all species amount of herbicide used, were recorded microphyllum in control plots was initially combined for each plot. Final assessment and treat 79.8% (SE=2.2; Figure 1) and did not ment were conducted on 2 November 1999 change significantly throughout the study When native and non-native species cover - 22 months after initial treatments. On (P=0.09). Cover of L. microphyllum in were combined by growth form (Table 1), the following day, three 0.25-m2 biomass control plots was 62.6, 67.7, and 61.4% cover increased for grasses from the first samples were collected from all plots of total cover for all vegetation in these to the final year of all three treatments, containing L. microphyllum for dry weight plots in 1997, 1998, and 1999, respec but decreased in control plots. Combined determination as before. tively. Pre-treatment mean dry weight of cover of herbaceous species increased in L. microphyllum for all sub-plots combined all treatments from beginning to end of the was 181.3 g m-2 (SE=15.3), and did not Data analysis and statistics study, but with about twice the increase differ among treatments. Final dry weight (114%) in firelbiannual herbicide plots, of L. microphyllum was zero in biannual To analyze the effects of the treatments on after a reduction in 1998, compared to herbicide and firelbiannual herbicide plots, the resident plant community, cover class other treatments (54%, 48%, and 58% in 4.5 g m-2 in bimonthly herbicide plots, and mid-points for each plant species, by sub control, bimonthly, and biannual herbicide 98.8 g m-2 in control plots. plot, were entered into PC-ORD (McCune plots, respectively). Combined cover of and Mefford 1999) for calculation of mean ferns reflects the intentional removal of cover, species richness, evenness (Pielou Effectiveness of treatments in L. microphyllum. Combined shrub cover 1969), and Shannon's Diversity Index (H'; controlling L. microphyllum decreased under all treatments, but signifi Shannon 1948). Cover class values (1-6) cantly more in firelbiannual herbicide plots. were then used to calculate an end-point Ten months after treatments were initi Changes in vine cover were not significant (beginning and ending years only) repeated ated, mean cover of L. microphyllum in for either treatment or time. measures ANOVA with treatment as the plots for all three fire and/or herbicide "between" factor and sampling year as the treatments had been reduced from 69.2% Impacts to native plant community "within" factor, using Wilk's Lamda sta (average for all pre-treatment fire and/or tistic and an all-years ANOVA of contrast herbicide plots) to 0.5% or less with no Native species cover for all sub-plots variables to compare the second and third significant differences among the three combined averaged 51.0% (SE=3.6) at the year data with pre-treatment data (SAS treatments (Figure 1). After 22 months, beginning of the study, which was only 1985). Fisher's Protected Least Significant mean L. microphyllum cover in fire and/or 40.0% of the total cover (127.5%; SE=3.7) Difference test was used to separate means herbicide plots was 0.7% or less and all for all species. In 1998, native species cover where the ANOVA F statistic showed treatments were significantly lower than was significantly lower in firelbiannual significance (P<0.05). Separate analyses control plots. However, even with regular herbicide plots than in other treatments were conducted on all species, native inspection and herbicide application, some (Figure 1). By the end of the study, native species, non-native species other than L. L. microphyllum was present at the end species cover was significantly higher in microphyllum, and L. microphyllum alone. of the two years in at least one plot of biannual herbicide and firelbiannual her Native species categorization followed each treatment, although not always in a bicide plots than in bimonthly herbicide Wunderlin and Hansen (2004). sampled area, and usually consisting of a and control plots. Cover of several native single small frond. species varied with time in control plots To analyze the time and herbicide require (Table 1). ments for the three fire and/or herbicide Effects of treatments on non-native treatments, one-way analysis of variance species other than L. microphyllum Since there was only one non-native (ANOVA; SAS 1982) was performed graminoid, SaccioZepis indica (L.) Chase, separately on total time required to inspect Mean cover of non-native species other which was a very minor component, the and re-treat plots and on total amount of than L. microphyllum was 4.7% (SE=1.4) cover of native graminoids was no different herbicide used. for all sub-plots combined in 1997 (pre than for combined native and non-native treatment), and was 3.5 (SE=1.7), 7.0 graminoids (Table 1), with increases under (SE=2.8), and 10.2% (SE=3.6) in control all fire and/or herbicide treatments. The plots in 1997, 1998, and 1999, respec- situation for herbaceous species was very
146 Natural Areas Journal Volume 28 (2), 2008 • Native D Non-native (minus L. m.) III L. microphyllum
. 100· December 1997 (pre-treatment) A 80
60
40 20 o
100 November 1998 A ~ 80 '-'... <1l 60 ;> 0 U 40 B 20 C C C 0
100 November 1999
80 B
60
40 20 C C C 0 Control Bimonthly Herbicide Biannual Herbicide Fire/Biannual Herbicide
Figure 1. The effect of removing Lygodium microphyllum with fire and/or herbicide (triclopyr ester) on mean cover for native species, non-native species (Lygodium microphyllum removed), and L. microphyllum in a Florida pine flatwoods understory treated in 1997. Within each year, means sharing the same letter are not significantly different (p different, with the fire and/or herbicide Species Richness Species richness dropped in all treatment treatments having no effect on herb cover. plots, with significant differences among Native ferns (Blechnum serrulatum Rich. A total of 101 species was recorded during the treatments, one year after treatments and Osmunda regalis L. var. spectabilis the study: 32 graminoids, 52 herbs, 3 ferns were initiated (reductions of 7.8, 27.8, [Willd.] A. Gray) decreased in control as (including L. microphyllum), 9 shrubs, and 26.8, and 38.8%, for control, bimonthly, well as fire and!or herbicide treatment plots 5 vines. Of these, nine were non-native spe biannual, and firelbiannual herbicide plots, during the study, with no significant effect cies, including one graminoid (Sacciolepis respectively), and recovered the following of treatment. Native shrubs were unequally indica), four herbs (Cuphea carthagenensis year to pre-treatment levels. distributed throughout the treatments at (Jacq.) J. F. Macbr., EmiliafosbergiiNicol- the start of the study, with greater cover in . son, Phyllanthus urinaria L., arid Urena Native species richness also dropped dur control plots than in the other treatments. lobata), one fern (Lygodium microphyl ing the second year (Table 2) and then Change over time was not significant, how lum), and three shrubs (Ludwigia peruviana returned to pre-treatment levels by the end ever, nor was there a significant interaction (L.) H. Hara, Psidium cattleianum Sabine, of the study, but only for the firelbiannual between treatment and time. Treatment and Schinus terebinthifolius Raddi). Initial herbicide treatment. Other treatments were and time were not significant effects on species richness (1997; Table 2) for all not significantly different than control native vine cover. species combined varied nonsignificantly plots. For non-native species other than L. from a low of 4.8 (biannual herbicide) microphyllum, neither time nor treatments to a high of 5.4 (bimonthly herbicide). were significant effects . . Volume 28 (2),2008 Natural Areas Journal 147 Table 1. The effect of removing Lygodium microphyllum with fire and/or triclopyr ester herbicide on Florida pine flatwoods understory species cover, for life-form groups, and the twenty taxa with the largest summed cover for years and treatments combined_ Data collected in December 1997, and November 1998 and 1999. Non-native species bolded. P values are for Wilks' Lambda z statistic in the repeated measures ANOVA. The first P value (within each set of parentheses) is for treatment effects; the second is for time (sampling year to sampling year) effects. There were .el. ...c no non-native vines. e!.. ~ Treatment ~ Control Bimonthlyherbicide Biannual herbicide FirelBiannual herbicide '-o c 1997 1998 1999 1997 1998 1999 1997 1998 1999 1997 1998 1999 ...::s e!.. ------, ~cover ------Graminoids All species (P= Herbs All Species (P=0.03/0.006) 13.2 19.7 20.3 19.7 21.7 29.1 25.9 27.4 41.0 15.0 7.1 32.1 SE= 3.6 5.0 3.9 4.3 4.7 6.0 5.1 5.4 7.5 3.5 2.6 5.6 Native spp. (p=0.42/0.54) 12.5 17.9 12.7 19.5 lOA 11.3 22.2 15.3 18.8 14.5 2.2 17.9 ~ C SE= 3.5 5.0 2.7 4.3 2.7 2.4 4.6 4.8 4.4 3.5 0.9 3.4 3 ('I) Centella asiatica 0.7 0.4 0.4 3.9 0.0 0.1 2.2 0.0 0.7 5.3 0.1 0.7 ~ co Eupatorium capillifolium 3.8 0.3 0.1 2.7 0.0 0.7 2.9 0.0 0.1 0.3 0.1 1.1 ~ o~ continued o CO ~ Table!. Continued c I 3 rD I\J O:l Treatment ~ Control Bimonthlyherbicide Biannual herbicide FirelBiannual herbicide I\J 0 1997 1998 1999 1997 1998 1999 1997 1998 1999 1997 1998 1999 0 O:l ------, % cover ~------. Herbs (continued) Lachnanthes caroliana 0.1 1.3 1.1 1.5 5.1 4.8 0.2 5.3 5.6 0.1 0.0 1.1 Ludwigia microcarpa 1.7 3.0 0.3 0.4 0.6 0.8 0.8 5.3 3.9 0.8 0.6 1.4 LycopodieUa appressa 0.0 0.0 0.0 0.0 0.0 0.0 1.3 2.3 0.5 10.3 0.0 0.0 Urena lobata 0.7 1.8 7.6 0.2 11.3 17.8 3.8 12.1 21.9 0.5 4.4 14.0 Ferns All Species (P= Vines (P=0.14/0.48) 0.6 2.1 1.0 0.2 0.0 0.5 0.0 0.0 0.0 1.1 0.0 0.0 SE= 0.6 2.1 0.7 0.1 0.0 0.5 0.0 0.0 0.0 0.7 0.0 0.0 Z c~... a > rD... III rn '- 0 ...c a= .... .j;:o, 1.0 Neither time (P=0.41) nor treatment .~ = "CS mell § f0: (P=0.07) were significant effects on na- .a'" -8 eQ. tive species evenness. Mean evenness ~ 8 ·c 0\ tn ...... ' 150 Natural Areas Journal Volume 28 (2), 2008 80 A ·70 Bimonthly herbicide 60 •0 Biannual herbicide Fire/biannual herbicide B 50 B '""'(Ij ::E • til.. ::l ..c::0 40 '-' Q) S ~ 30 20 10 0 0 2 4 6 8 10 12 14 16 18 20 22 Total Months After Treatment Initiated Figure 2. Time required to inspect and apply herbicide (tricIopyr ester) to Lygodium microphyllum regrowth in Palm Beach County, Florida. Means in the Total column that do not share the same letter are significantly different (p 50 A ~ t- 40 Bimonthly herbicide •0 Biannual herbicide Fire/biannual herbicide ..c::~ 30 bb • C "0 II> B :::>'" II> "0 20 :e'13 ::r:II> 10 o o 2 4 6 8 10 12 14 16 18 20 22 Total Months After Treatment Initiated Figure 3. Amount of herbicide (triclopyr ester) required to treat Lygodium microphyllum growth and regrowth in Palm Beach County, Florida. Means in the Total column that do not share the same letter are significantly different (P Volume 28 (2),2008 Natural Areas Journal 151 DISCUSSION burn was included in Figure 2. Prescribed plots and was not related to fire and/or burn logistical difficulties and costs would herbicide treatment effects. It is not known Both fire followed by herbicide, and be an important consideration in an actual why native species cover was lower at the herbicide-alone treatments were effective management program. end of the study in bimonthly herbicide in removing almost all L. microphyllum plots than the biannual herbicide and from the pine flatwoods community - and In the year following the prescribed burn, firelbiannual herbicide treatments. Even without lasting negative effects on the na native species cover was significantly though care was taken to avoid damaging tive species cover, richness, evenness, nor lower in firelbiannual herbicide plots than plants, it is possible that the more frequent diversity. The use of fire, however, resulted in other treatments, probably due to the traffic in the bimonthly herbicide plots re in some very different patterns than the less selective effects of a fire versus spot sulted in the lower native vegetation cover. herbicide-alone treatments. Regrowth of herbicide treatment than to any negative Bimonthly herbicide plots used the same L. microphyllum in firelbiannual herbicide response by native species to fire in this about of herbicide as the biannual herbicide plots was vigorous following the prescribed fire-maintained plant community. It is pos plots, so the effects were not due to differ fire treatment, and the first herbicide appli sible that the timing (winter) of the burn ences in amount of herbicide applied. cation visit (six months after the prescribed could have affected native plant response. fire) alone required the equivalent of 30 A comprehensive review of fire in southern The bimonthly herbicide schedule obvious hours ha- I of inspection and treatment time. Florida ecosystems (Wade et al.) points ly required more frequent visits, contribut The prescribed fire did reduce standing out that it is frequently assumed that pre ing to greater total time, although each visit biomass of L. microphyllum, but the sub European influence fires in Florida were was of very short duration. Since the final sequent new growth of many other species predominantly wet season (summer) fires. biomass of L. microphyllum was the same made finding regrowth of L. microphyllum While lightning in Florida is much more in both bimonthly and biannually inspect in the understory very difficult. common during the wet summer months; ed/treated plots, there was no benefit from and lightning-initiated fires peak at this the more frequent visits. It is likely that a Regrowth of L. microphyllum in burned time (Snyder et al. 1990), little is known longer interval than biannually would be plots was much less synchronous than about the frequency of dry-season fires set more cost effective, although observations in herbicide-only plots. Following initial by Native Americans (Wade et al. 2000). suggest that nearly complete regeneration herbicide application in herbicide-only Indirect evidence for anthropogenic dry of L. microphyllum biomass following fire plots, almost all regrowth emerged within season fire (Myers and Peroni 1983; Snyder or partially effective herbicide application two months. Regrowth in burned plots 1991; Robbins and Myers 1992; Kjellmark can occur within two years. started within two months, but new growth 1995, 1996) does suggest that Florida's fire emerged in several bursts over the next year. dependent plant communities, such as the It is important to note that even after in Herbicide application may kill L. micro pine flatwoods in this study, were subject to spection and retreatment for 22 months, phyllum fronds back to healthy rhizomes fire at any time of year for many centuries, some, albeit small, amounts of L. micro more uniformly than fire, resulting in less and thus the winter fire conducted for this phyllum remained in the treatment plots. sporadic re-emergence. The asynchronous study may not have presented an atypical This was new growth connected to existing regrowth in firelbiannual herbicide plots, disturbance. rhizomes, not new plants. This suggests however, did not negate the approximately that periodic inspection and retreatment 50% reduction in herbicide used compared The more important factor related to this may be necessary for quite some time after to the herbicide-only treatments, and the prescribed burn was probably its slow rate operational management programs begin. total time required to inspect and retreat of travel through the plot and the associ While the regrowth in this study could be firelbiannual herbicide plots was the same ated higher temperatures of the slow-mov shown to come from existing rhizomes, the as biannual herbicide plots without .fire. ing fire. Larger-scale management burns possibility of establishment of new plants is would have moved much faster through a always present. It is not known how far L. Under typical conditions, a crew of five pine flatwoods because of air movement microphyllum spores travel, and under what can burn about 120 ha of L. microphyl generated by a larger fire and because the conditions they are able to germinate, but lum-infested pinelands in a single lO-hour fire prescription would have called for they are very small and can be expected to day (S. Smith, research scientist, South appropriate wind strength and direction to be wind-borne for considerable distances. Florida Water Management District, pers. manage the dispersal of smoke. Very low Lygodium microphyllum is able to repro comm.), which equates to approximately wind speed was a required condition for duce by intragametophytic selfing, which 0.08 hours ha-I for five people, or about the prescribed burns in this study to reduce means that only one gametophyte, from one 0.4 hours ha-I person-I. Because the per the chances of fire damaging adjacent spore, is required, an obvious advantage in ha time for burning areas larger than the non-burn plots. long-distance dispersal (Lott et al. 2003). small study plots would be expected to be Regional approaches to L. microphyllum minimal compared to time requirements Cover of non-native species other than L. management will no doubt be necessary for ground-based herbicide applications, microphyllum increased during the study, to reduce the possibility of reinfestation no estimated value for time required to but the increase also occurred in control of previously cleared areas. 152 Natural Areas Journal Volume 28 (2), 2008 ACKNOWLEDGMENTS LITERATURE CITED Pemberton, RW. 1998. The potential ofbiologi cal control to manage Old World climbing We would like to thank Donny Forgioni and Beckner, J. 1968. Lygodium microphyllum, fern (Lygodium microphyllum), an invasive weed in Florida. American Fern Journal John Street, Palm Beach County, and Steve another fern escaped in Florida. American 88:176-182. Smith, South Florida Water Management Fern Journal 58:93-94. Pemberton, R.W., and A.P. Ferriter. 1998. Old District, for conducting the plot burns, Daubenmire, R. 1959. A canopy-coverage World climbing fern (Lygodium microphyl Ken Langeland, Brian Smith, Dorothy method of vegetational analysis. Northwest Science 33:43-64. lum), a dangerous invasive weed in Florida. Brazis, and Shelley Stocker, University of American Fern Journal 88:165-175. Florida, for providing field assistance, and Ferriter, A., and T. Pernas. 2006. An explosion in slow motion: tracking the spread of Ly Pemberton, R, J. Goolsby, and T. Wright. two anonymous reviews for very helpful godium microphyllum in Florida. Wildland 2006. Biological control. Pp. 26-33 in J. comments on a previous draft. Weeds 9:7-9. Hutchinson, A. Ferriter, K. Serbesoff-King, K. Langeland, and L. Rodgers, eds., Old Goolsby, lA. 2004. Potential dis'tribution of the World climbing fern (Lygodium microphyl invasive Old World climbing fern, Lygodium lum) management plan for Florida. Florida Randall Stocker is a professor at the Uni microphyllum in North and South America. Exotic Pest Plant Council, Lygodium Task versity of Florida. His research interests Natural Areas Journal 24:351-353. Force, 2nd ed. Available online Volume 28 (2), 2008 Natural Areas Journal 153 Distributions 10:439-446. Wade, D.D., B.L. Brock, P.H. Brose, J.B. Grace, G.A. Hoch, and W.A. Patterson, III. 2000. Chapter 4: Fire in eastern ecosystems. Pp. 53-96 in J.K. Brown, and J.K. Smith, eds., Wildland fire in ecosystems, effects of fire on flora. General Technical Report RMRS-GTR-42-Volume 2, Rocky Mountain Research Station, Ft. Collins, Colo. Wu, Y., K. Rutchey, N. Wang, and J. Godin. 2006. The spatial pattern and dispersion of Lygodium microphyllum in the Everglades wetland ecosystem. Biological Invasions 8: 1483-1493. Wunderlin, R. P. and B. F. Hansen. 2004. Atlas of Florida Vascular Plants 154 Natural Areas Journal Volume 28 (2),2008