Anatomy of an invader

In the continental United States, the Florida Everglades ranks among the top locations in regard to the number and coverage of introduced invasive species. Among Florida’s non-indigenous invasive flora, a relatively recently introduced species, Lygodium microphyllum, also known as Old World climbing fern, is quickly becoming one of its most widespread. L. microphyllum first drew attention for escaping cultivation and spreading into natural lands in the late 1960s. However, little was done with this observation until the late 1990s, when active scientific research on its ecology and management, including biological, chemical, and mechanical control activities began in earnest.

L. microphyllum, a vine-like fern native to the pantropics of the Old World, is typically considered a benign constituent of its native plant communities. However, in its Florida introduced habitat it can develop a near monoculture, smothering the native understory under a mat of vegetation that can be thicker than 1 metre. The fern also climbs into the native overstory, eventually shading and killing the canopy trees and resulting in their collapse. In some cases, the fronds act as a ‘fire ladder,’ allowing fire into the canopy and resulting in the death of the host trees.

After an invasive species has been studied for a time, a Gestalt of that species usually begins to emerge. Unfortunately, this has been slow to develop in the case of L. microphyllum.

After more than a decade of study, the reasons for its success as an invader only now begin to become apparent. To truly understand why an invasive plant is successful often requires examining the species across a range of scales and in both its native and introduced habitats. This has been the strategy my colleagues and I have used to discover why L. microphyllum is such a successful invader in its de novo environment. We began by studying its reprodution, including its mode of reproduction, the seasonality of sporulation and its spore production. Our subsequent studies examined the fern’s community ecology and its growth, both in situ and under controlled conditions, with various environmental treatments, including the influence of different light and different hydrological environments on its growth.

Several of these field and controlled studies were linked with landscape scale studies to develop a spatial model of the fern’s potential spread across its introduced environment. Finally, crosscontinental controlled and common garden studies were carried out to compare the growth of the fern in its native versus its introduced environment. We learned much from our studies of L.microphyllum’s reproductive biology. The fern is homosporous-it produces bisexual gametophytes and, consequently, it can reproduce by three different reproductive strategies: intragametophytic selfing, intergametophytic selfing and outcrossing. Intragametophytic selfing allows for long distance dispersal because it only requires one gametophyte to establish new colonies, since sporelings (i.e., young plants) result from the union of an egg and sperm on the same gametophyte. The second and third strategies require two spores to land close together and to germinate into gametophytes, such that a sperm can swim from one gametophyte to the other. In the case of intergametophytic selfing, the two gametophytes are from the same parent plant, while with outcrossing the two gametophytes are from different parent plants. Typically, homosporous ferns are conservative in their reproductive strategy, in that they only reproduce by one of the three modes. We found that this is not the case for L. microphyllum, which we determined was able to reproduce by all three mating strategies. However, we also discovered an interesting invasive twist to its reproductive strategy.

When a gametophyte first germinates from a spore, it is almost always female and this female produces a pheromone that makes all younger nearby gameotphytes to become male, thus assuring outcrossing. On the other hand, if after a few weeks there are no nearby spores or gametophytes, the female will become bisexual by producing male organs thus assuring fertilisation. Collectively, these highly plastic reproductive strategies help explain the fern’s long distance dispersal. We followed this initial work with studies related to the fern’s seasonal spore patterns, spore production and community ecology. In its introduced environment, we found that dense sites of L. microphyllum produce spores year round, with a peak production in the wet season. We estimated a density of 15,000 spores per cm2 of fertile leaf area, resulting in approximately three billion spores present at each invaded site at any given time. Our community-level studies showed that the presence of the fern coincided with a wet but not permanently inundated environment, and the coverage of the fern was greatest in a low-light understory environment, which likely facilitates its early establishment and growth. Anecdotally, L. microphyllum can be found in the various light conditions that exist in Florida’s native ecosystems, from low to high irradiances. Later evidence in controlled studies showed that small plants of L. microphyllum have the same relative growth rate after 3 months whether growing in 20, 50 or 70% full sun light. The ability of the fern to maintain high growth under different light environments appears to be related to its ability to allocate carbon to stems in a highly plastic and favorable manner. For instance, in the highest light environment, the plants grow on average 2.9 m for every gram of carbon invested, while in the lowest light conditions the plants grew 4.2 m per gram of carbon. These findings were supported by 2 years of field observations where under high light conditions the height of the actual climbing mat (i.e., not individual stems) increased 1.43 m per year on host canopy trees.

We were able to combine the attributes of the fern from our studies of its reproduction biology and ecology with a series of aerial transects to develop a spatial model predicting the future spread of this highly invasive fern across the landscape. In 1978, a survey of known individual L. microphyllum sites was published. In 1993, the South Florida Water Management District began conducting aerial surveys every 2 years to monitor invasive species, including L. microphyllum, locations across the Southern Florida region. Combined with our earlier data, we were able to construct a cellular automaton landscape model by calibrating the model from the 1978 survey data stepped through to 1993. Then we validated the model independently from 1993 to 1999 using the aerial data. Finally, by initializing our model with the 1999 aerial survey data, we predicted the spread of the fern to 2014, showing that 37% of the 40,000 km2 -grid cells covering Southern Florida would have L. microphyllum present by 2014. In the absence of aggressive control measures, this model predicted that the coverage of L. microphyllum could exceed the coverage of the top five invasive species in Florida by 2014. Fortunately, aggressive biological, chemical and mechanical control for this species has been ongoing.

These early studies highlighted some of the attributes that help explain L. microphyllum’s ability to be invasive in its introduced environment. Thee same studies, combined with intensive monitoring efforts, allowed us to develop a spatially explicit predictive model that assisted in sounding the alarm, particularly to policy makers, since most land managers had already understood the dire consequences that ignoring this species would have on natural lands.

However, this series of studies was not sufficient to fully understand why this species is so successful in its new environment, while being apparently so benign in its native range. Moreover, to optimise management of this species, further studies were needed to provide additional information that could be used in both biological control and land management efforts.

Interestingly, from our earlier controlled studies under different light conditions, we found that L. microphyllum allocated 48% of its carbon to belowground tissues, while a native vine in the same study averaged only 27%.

Interestingly, from our earlier controlled studies under different light conditions, we found that L. microphyllum allocated 48% of its carbon to belowground tissues, while a native vine in the same study averaged only 27%. The latter is a typical investment for a vine belowground, given that vines characteristically maximise their allocation to above ground tissues in order to reach greater heights to optimise light interception. L. microphyllum, on the other hand, is able to grow to great heights, but it does it with relatively little investment in carbon. It was this observation that first led us to consider that, perhaps, L. microphyllum is such a successful invasive because it may be released from its natural belowground enemies. This would allow the plant to allocate substantially to belowground tissues, which may confer a competitive advantage in the low nutrient environment of southern Florida. We tested this hypothesis through a series of coordinated control and field common garden studies in its introduced range in Florida as well as in its native range in Australia. We examined the fern’s growth across soil treatments, including soil sterilization to eliminate below ground natural enemies, and nutrient amendments to examine the possible interaction of soil nutrient availability. In addition, in the native range, we used three different fern source populations; one from Florida, one from the study location in Southeastern Queensland, Australia, and the third from northeastern Queensland, Australia. This third source is the reputed original location of the plants introduced into Florida. All populations tended to have comparatively poor growth in unaltered soil, but growth for all was stimulated by nutrient amendment and sterilisation. The overall effect of sterilisation, however, was muted under high nutrient conditions, except for the population originating from the same region as the soil used in Australia. Regardless of nutrient treatment, ferns in this population grew significantly faster in sterilized than non-sterilized soil.

So what does this mean? It appears that overall these results supported our hypothesis that the invasiveness of L. microphyllum in Florida is in part mediated by release from soil-borne enemies, and given the different response of the Southeastern Queensland population in Australia as compared to the other populations, a region-specific response may be occurring.

In both its native and introduced environment, L. microphyllum is found in wetland environments, but appears to be able to tolerate a range of hydrological conditions. Therefore, we conducted another study to examine L. microphyllum growth response to three hydrological treatments; flood, drought and field capacity. Field capacity and drought showed no differences in growth, while flooded plants had significantly slower growth. However, it should be noted that, even after more than 2 months of continuous inundation, the flooded ferns demonstrated positive relative growth rate.

This apparent hydrological plasticity is likely another contributing factor to the fern’s widespread establishment across a range of plant communities within the greater Everglades ecosystem. The fern’s growth response to differences in hydrological conditions was largely explained by changes in its specific innae (i.e., leaf) area and photosynthesis: in other words, morphological and physiological determinants of growth, respectively.

From these numerous studies conducted at various scales under different environmental conditions, as well as from comparisons between introduced and native habitats, we eventually developed an understanding of why L. microhpyllum is such a successful invader, while at the same time a relatively benign presence in its native environment. In essence, L. microhpyllum possesses several life-history characteristics that may enhance its competitive ability in Florida, including plastic reproductive strategies, and proli!c and continuous spore production. This fern appears to optimise its morphological and physioological characteristics to maximise photosynthetic area and minimise carbon costs in tissue construction, thereby gaining the ability to grow rapidly across gradients in light and hydrology. On the other hand, in its native range, this ability to grow across environmental gradients and reproductive output may be constrained by natural enemies. Our cross-continental comparisons appear to support our hypothesis that release from natural enemies belowground may also help explain the success of the fern in its introduced environment and, provocatively, given the different source population responses, it appears that a regional-scale version of the enemy release hypothesis might play a role in L. microphyllum’s invasive success.

Our findings demonstrate the importance of a biogeographical approach to invasive species studies and have implications for the identification of potential biocontrol agents. Moreover, the numerous studies at different scales in the fern’s introduced environment allowed us to make specific management recommendations, provided clues to its invasiveness that could then be corroborated by comparative studies of its native versus introduced ranges, and allowed us to develop a landscape-level model that could inform scientists, land managers, policy makers and the public, of the imminent danger of this particular invasive plant.

John Volin is a Professor of plant physiological ecology and Head, Department of Natural Resources and the Environment at the University of Connecticut, Storrs, CT, USA. Mail at

This article is from issue


2010 Mar