1 CIPOLLINI AND SCHRADIN: IMPACTS OF RANUNCULUS FICARIA
3 Guilty in the Court of Public Opinion: Testing Presumptive Impacts and Allelopathic 4 Potential of Ranunculus ficaria
16 KENDRA A. CIPOLLINI AND KELLY D. SCHRADIN
7 Wilmington College, Wilmington, Ohio 45177
9 ABSTRACT.—Information about invasive species can be based primarily on anecdotal evidence, 10 indicating the need for further information. Ranunculus ficaria is an ephemeral riparian plant
11 species that is presumed invasive in the United States, despite the lack of any published 12 information on its impacts. Mechanisms by which R. ficaria may affect native plant species
13 include competition and allelopathy. We examined if R. ficaria negatively affected the growth
14 and reproduction of the native Impatiens capensis and, if so, whether it is by allelopathy, nutrient 15 competition or some combination thereof. We performed a fully-factorial field experiment in the 16 field, manipulating the presence of R. ficaria, nutrients, and allelopathy with the use of activated 17 carbon. The presence of R. ficaria tended to negatively affect survival of I. capensis. In the
18 absence of carbon, R. ficaria significantly decreased seed production, illustrating the negative 19 impact of R. ficaria. In the presence of carbon, there was no effect of R. ficaria, suggesting that
20 carbon may have ameliorated the negative allelopathic effect of R. ficaria. Nutrient competition
21 did not show strong effects. Despite its widespread identification as an invasive species, this is 22 the first study to demonstrate the negative impact of R. ficaria on a native species and the
1 Corresponding author: Phone: 937-382-6661 ext. 367, Fax: 937-383-8530, e-mail: KAL143@alumni.psu.edu
23 possible role of allelopathy in its success. Further, the negative impacts of this ephemeral species 24 persist well beyond its early growing season, which calls into question previous assumptions
about R. ficaria exerting effects primarily on other ephemeral species. 25
30 Invasive species threaten biodiversity on a global scale (Wilcove et al., 1998; McGeoch et al.,
31 2010) and are defined as those species that cause or have the potential to cause economic or 32 environmental harm, weighed against their benefits (NISC, 2006). Most naturalized plants are 33 introduced through the horticultural industry (Mack and Erneberg, 2002) and some can still be 34 purchased in some instances despite their invasive status (Harris et al., 2009; Axtell et al., 2010).
35 However, only a portion of naturalized species are actually categorized as invasive (Milbau and 36 Stout, 2008). As a result, there is much interest in characterizing the species traits and 37 introduction routes that make a species invasive (Lambdon et al., 2008; Milbau and Stout, 2008;
38 van Kleunen et al., 2010). Yet, in many of these studies the methods by which species are 39 identified as invasive are vague and based on expert opinion and anecdotal evidence, with little 40 scientific evidence (Blossey, 1999). Further, even when there is some published information, the 41 impacts of an invasive species can be overstated by the popular press (Lavoie, 2010), which may 42 lead to inappropriate response strategies or undue focus. The lack of information on the impact 43 of invasive species and on the possible mechanism of impact is an obstacle to effectively 44 prioritizing the control of invasive species during a time of dwindling resources.
45 Invasive plant species can negatively impact native species through a variety of mechanisms 46 (Levine et al., 2003). Invasive species can simply outcompete native species for above- and/or
below-ground resources (e.g., Kueffer et al., 2007; Cipollini et al., 200847 a; Galbraith-Kent and
48 Handel, 2008). Enhanced nutrient acquisition can lead to invasive species success. For example, 49 Centaurea maculosa acquired more phosphorus than surrounding native species which may have 50 increased competitive success (Thorpe et al., 2006). Additionally, Leishman and Thomson (2005)
51 found that invasive species had greater responses to nutrient addition than native species, thus 52 outcompeting the natives in nutrient-rich environments. Another mechanism by which invasive 53 species affect native communities is allelopathy (Hierro and Callaway, 2003; Ens et al. 2009).
54 Most plants produce secondary compounds (Ehrenfeld, 2006) that can affect an adjacent plant 55 either directly (Cipollini et al., 2008b) or indirectly through changing soil ecology (Stinson et al.,
56 2006; Mangla et al., 2008). Some allelopathic chemicals that have no negative impact in their 57 native environment may have negative effects in an invaded community, a mechanism coined the 58 ―novel weapons hypothesis‖ (Callaway and Aschehoug, 2000; Callaway and Ridenour, 2004;
59 Callaway et al., 2008; Thorpe et al., 2009).
60 Discovering impacts due to allelopathy can be done with careful experimentation (Inderjit and 61 Callaway, 2003). Allelopathic effects have been studied using activated carbon (e.g., Ridenhour 62 and Callaway, 2001; Cipollini et al., 2008a). Activated carbon adsorbs organic compounds,
63 including allelochemicals (Ridenour and Callaway, 2001). Addition of carbon can also has 64 effects on soil properties and plant growth in potting soil (Lau et al., 2008; Weisshuhn and Prati,
65 2009). Addition of nutrients is thought to help ameliorate any fertilizing effects of the addition of 66 activated carbon (Inderjit and Callaway, 2003). Activated carbon may also serve as a restoration 67 tool to change soil conditions in invaded soils (Kulmatiski and Beard, 2006; Kulmatiski, 2010).
68 Ranunculus ficaria L. (Ranunculaceae), or lesser celandine, is a groundcover, native to 69 Europe (Taylor and Markham, 1978; Sell, 1994), which appears to be affecting native plants in
forested floodplains in many US states (Swearingen 2005). There are five known subspecies, all 70
71 of which are found in the United States (Post et al., 2009). Ranunculus ficaria was first recorded
72 naturalized in the United States in 1867 (Axtell et al., 2010). As it is still being marketed by the
73 nursery industry (Axtell et al., 2010), it was likely imported for horticultural purposes. It was 74 recognized as a naturalized species in the Midwest in the 1980’s (Rabeler and Crowder, 1985)
75 and in southern states more recently (Krings et al., 2005, Nesom, 2008). Ranunculus ficaria is
76 currently documented in at least 21 US states, the District of Columbia, and 4 Canadian 77 provinces (USDA, 2010). It has been identified as invasive in 9 states and the District of 78 ). Columbia and is banned in Massachusetts and Connecticut as a noxious weed (Axtel et al., 2010
79 Ranunculus ficaria emerges before most native spring species, which may provide it with a 80 competitive advantage. Once established, it spreads rapidly across the forest floor to form a 81 dense monoculture, which native species seemingly cannot penetrate (Swearingen, 2005). 82 Hammerschlag et al. (2002) reported that R. ficaria created a monoculture in the Rock Creek
83 floodplain in Washington, DC and few native species re-colonized after its removal. It is thought 84 that R. ficaria, as a spring ephemeral, has impacts primarily on other spring ephemerals 85 (Swearingen, 2005). However, most all information on R. ficaria as an invasive species is
86 primarily anecdotal in nature. Unpublished and preliminary data indicate that presence of R.
87 ficaria is associated with reduced abundance and richness of native species (Hohman, 2005). 88 Ranunculus ficaria exhibits direct allelopathic effects on germination of some native species 89 (Cipollini, unpublished data), indicating that R. ficaria may have negative effects beyond the
90 spring time period through lingering allelopathic effects.
91 For our study, we examined if R. ficaria negatively affects the growth and reproduction of the 92 native annual Impatiens capensis and, if so, whether it is by allelopathy, nutrient competition or
some combination thereof. In the field we performed a fully-factorial experiment with the 93
94 treatments of R. ficaria (present and absent), carbon (present and absent), and nutrient addition 95 (present and absent). We expected that the presence of R. ficaria would have an overall negative
96 impact, that addition of nutrients would have an overall positive impact and that addition of 97 carbon would have no overall effect on the performance of I. capensis. If allelopathy were
98 important, we expected to see a significant carbon by R. ficaria interaction and, if nutrient
99 competition were important, we expected to see a significant nutrient by R. ficaria interaction on
100 I. capensis response variables.
105 We performed the study in 2009 at Hamilton County Park District’s Winton Woods in
106 Cincinnati, Ohio in an area invaded by R. ficaria ssp. bulbifer, a subspecies that forms asexual
107 bulbils (Post et al., 2009). The study area was found in a floodplain along Westfork Mill Creek. 108 We set up the experiment on 24 April, choosing an area with a uniform coverage of R. ficaria.
109 We fenced the entire site using deer fencing to prevent trampling and/or herbivory. We used a 110 fully factorial design with the main factors of: presence/absence of R. ficaria, presence/absence
111 of activated carbon and presence/absence of additional nutrients, replicated 8 times (2 R. ficaria
112 levels x 2 carbon levels x 2 nutrient levels x 8 replicates = 64 experimental units). The treatment 113 combinations were haphazardly assigned to each plot and each plot was located approximately
114 25 cm apart. We tested the effects of these treatment combinations on the target plant Impatiens
115 capensis Meerb., jewelweed (Balsaminaceae). We chose I. capensis because of the overlap in
habitat and distribution with R. ficaria. Another advantage is that reproduction can be measured 116
117 in this annual species. Additionally, I. capensis has served as a model organism in previous
118 studies of invasive species effects (e.g., Cipollini et al., 2008a; Cipollini et al., 2009; Cipollini
119 and Hurley, 2009). All I. capensis seedlings were obtained in an immediately adjacent area free 120 of R. ficaria.
121 In the R. ficaria-present treatments, we removed R. ficaria and planted them back in place
122 while transplanting I. capensis seedlings. In R. ficaria-absent treatments, we removed the R.
123 ficaria completely before transplanting I. capensis. In activated carbon-present treatments, we
124 worked 10 ml of activated carbon (Aquarium Pharmaceuticals Black Magic Activated Carbon, 125 Chalfont, Pennsylvania) into the top 8 cm of soil of each plot. This ratio of activated carbon to 126 soil volume has been shown to mitigate allelopathic effects in previous research (Cipollini et al.,
127 2008a; Ridenour and Callaway, 2001). In nutrient-addition treatments, we worked the 128 manufacturer-recommended amount of 1.5 teaspoons of slow-release fertilizer (Scotts Osmocote, 129 Scotts-Sierra Horticultural Product Co., Marysville, Ohio) into the top 8 cm of soil in each plot. 130 We disturbed the soil in each subplot regardless of treatment combination to control for soil 131 disturbance effects. On May 1 we replaced any I. capensis transplants that did not survive,
132 presumably due to transplant shock. We measured the height of each seedling when transplanted. 133 The number of fruits, number of seeds and survival (days to death) of the seedlings were 134 recorded once each week. Height and stem diameter (measured directly beneath bottom node 135 with a digital caliper) was measured. Ranunculus ficaria had lost all of its foliage by 2 June
136 (week 6) and the leaf litter had decomposed by 12 June (week 8), leaving the bulbils exposed on 137 the soil surface. Measurements began on 5 May and ended on 28 August (week 18). 138 We performed a series of Analysis of Variances (ANOVAs) on the I. capensis response
139 variables, using the full model with fixed factors of nutrients (+/-), R. ficaria (+/-) and carbon
140 (+/-). We were unable to include all of the variables in a single MANOVA model due to the 141 missing values generated as plants died. For survival, we analyzed the number of days to death 142 for all plants. For total number of seeds, we included starting height as a significant covariate. 143 We removed those plants that had died within 8 weeks of transplant, as seed production was 144 essentially non-existent up to that time. For final height and stem diameter, we used a MANOVA 145 with starting height as a significant covariate for the 32 surviving plants. When significance was 146 found in the MANOVA using Wilk’s λ, we ran separate univariate Analyses of Variance
147 (ANOVAs) for each variable. For all statistical tests, model assumptions of normality and 148 heteroskedasticity were verified prior to analysis and transformed where appropriate. 149
153 For survival, there was a near-significant effect of R. ficaria (F= 3.75, p = 0.058), with I. 1,55
154 capensis tending to die sooner when R. ficaria was present (Fig. 1). For the total number of seeds
155 produced, there was a significant effect of nutrients (F= 20.53, p < 0.001) and a significant 1,47
156 interactive effect between R. ficaria-presence and carbon (F= 5.03, p = 0.030). The presence 1,47
157 of nutrients nearly tripled seed production (Fig. 2). When carbon was absent, the presence of R.
158 ficaria reduced seed production. When carbon was present, seed production was similar whether
159 R. ficaria was present or absent (Fig. 3). In the MANOVA, there was a significant effect of 160 nutrients on final height and stem diameter (F= 14.904, p = 0.000). In the univariate ANOVA, 2, 22
nutrients increased both height (F= 7.28, p = 0.013) and stem diameter (F = 30.34, p < 161 1, 23 1, 31
162 0.001) (Fig. 2).
167 We wanted to know if the putative invasive R. ficaria negatively affected the growth and
168 reproduction of the native I. capensis, and if it did, whether allelopathy or nutrient competition 169 played a causative role. Because R. ficaria has been identified as invasive (Axtell et al., 2010)
170 and has been associated with reduced native abundance and diversity (Homann, 2005), we 171 expected that the presence of R. ficaria would have a negative impact on the performance of I.
172 capensis. Our results show that R. ficaria does indeed negatively affect I. capensis providing
173 confirmatory evidence of its assumed impact on native species. Ranunculus ficaria showed a
174 tendency to have a negative overall impact on the survival of I. capensis with plants dying ~10
175 days earlier in the presence of R. ficaria. In the absence of carbon, R. ficaria decreased seed
176 production by ~50%. As expected, nutrients showed a significant effect on growth (in terms of 177 height and stem diameter) and seed production. If nutrient competition were a key factor in 178 inhibition of I. capensis by R. ficaria, we would have expected I. capensis to exhibit a release
179 from competition in the presence of added nutrients. There was no significant interaction 180 between nutrient addition and presence of R. ficaria, suggesting that nutrient competition was
181 most likely not the primary mechanism of impact of R. ficaria. We do however acknowledge that
182 our study does not rule out other forms of resource competition, such as competition for light or 183 water.
184 In the presence of carbon, there was no effect of R. ficaria; carbon therefore ameliorated the
185 negative effect of R. ficaria on I. capensis. In previous research, application of carbon to soils
186 has mitigated the effects of invasive species (Cipollini et al., 2008a; Ridenour and Callaway,
187 2001). Ridenour and Callaway (2001) found that the competitive ability of Eurasian grass 188 species was greatly reduced by activated carbon and suggested that the advantage of this species, 189 at least in part, was due to allelopathy. Similarly, one possible conclusion from our study is that 190 the addition of carbon may have attenuated the allelopathic effect of R. ficaria. However, when
191 using activated carbon as an experimental tool to test allelopathy, caution must be used in 192 interpreting results. The addition of activated carbon itself can change soil conditions in potting 193 soil (Lau et al., 2008; Weisshuhn and Prati, 2009), though it is important to note that these 194 studies used twice as much activated carbon compared to the amount used in our study. 195 Activated carbon itself can have a direct fertilizing effect (Lau et al., 2008). In our study, the
196 mitigating effect of carbon was significant even in nutrient-addition treatments, suggesting that 197 direct addition of nutrients by the activated carbon was most likely not the mechanism through 198 which carbon exerted its effects. Ranunculus ficaria is purported to have medicinal uses in
199 herbal medicine as a treatment for hemorrhoids and as an astringent (Chevallier, 1996) and 200 contains a number of secondary chemicals (Texier et al., 1984; Tomcysk et al., 2002). The
201 presence of secondary compounds may be a good predictor of invasive species impacts, 202 including allelopathy (Ehrenfeld, 2006). Further, because R. ficaria exhibits direct allelopathy on
203 some species in the laboratory and greenhouse (Cipollini, unpublished data), allelopathy is a 204 likely mechanism in the field, though clearly further study is needed (see Blair et al., 2009).
205 Whatever the exact mechanism of its effect, activated carbon nevertheless was clearly able to 206 negate the negative effect of R. ficaria, suggesting that, at the very least, it may serve as an
effective restoration tool to modify soil conditions in the field to the benefit of native species 207
208 (Kulmataski and Beard, 2006; Kulmatiski, 2010),.
209 Interestingly, we found that R. ficaria had lasting effects beyond its brief growing season.
210 We expected that effects on I. capensis would not be particularly strong, as it is thought that the 211 negative effects of this species are exerted primarily on spring ephemeral species (Swearingen, 212 2005). Further studies using spring ephemeral species are obviously needed to clarify the 213 comparative effect on this set of species presumed most sensitive. Even though R. ficaria had
214 completely senesced by week 6 of the experiment, it still significantly negatively impacted I.
215 capensis, which lived without the presence of R. ficaria for about two-thirds of its life span.
216 Other invasive species can have residual effects on native species, even after the removal of the 217 invasive species (Conser and Conner, 2009). The effect of R. ficaria well past its growing season
218 may be due to lingering effects, such as those due to allelochemicals or to other modification of 219 soil conditions. An alternative explanation may be that the seedling stage is the most vulnerable 220 stage of I. capensis, similar to the findings in Barto et al. (2010).
221 This is the first study to show the negative impact of R. ficaria on I. capensis and to point
222 towards a possible mechanism of success – allelopathy or other modification of soil conditions.
223 It is surprising that this is the first published study to investigate the invasive potential of R.
224 ficaria, given that natural resource agencies have recognized it as a species important to control 225 in natural areas since at least the year 2000 (Hammerschlag et al., 2002). Ranunculus ficaria has
226 already been banned in two states (Axtell et al., 2010) since the year 2006. Admittedly, there is a
227 balance between taking immediate action against an invasive species and waiting for time-