Could biological invasion by Cryptostegia madagascariensis alter the composition of the arbuscular mycorrhizal fungal community in semi-arid Brazil ?

To understand more about the effects of invasion on arbuscular mycorrhizal fungi (AMF) community, current research has focused on feedback between AMF communities and invasive species (Zhang et al. 2010), in the host-fungal pairing (Klironomos 2003), and how AMF can impact the performance of invasive species (Stampe & Daehler 2003). This work is the first field study to examine the impacts of C. madagascariensis on AMF community composition along an invasion gradient and to investigate how the changed-AMF community may contribute to create a “window of opportunity” (Johnstone 1986; Agrawal et al. 2005) and consequently influencing plant invasions (Blumenthal 2005; Kulmatiski & Kardol 2008). Plant-AMF specificity affects plant invasion (Mummey & Rillig 2006; Parkash & Aggarwal 2009; Zhang et al. 2010) and may in some cases strengthen invasive ability of plants through complex feedback loops (Richardson et al. 2000). Plants provide AMF with energy while AMF provides water and nutrients such as phosphorus and nitrogen to the plant host (Smith & Read 2008; Hodge & Storer 2014). Plants have affected the AMF community structure through influences on soil resource availability, like available P (Alguacil et al. 2012; Ji et al. 2012; Beauregard et al. 2013; Lekberg et al. 2013; Geel et al. 2015; ABSTRACT Biological invasions pose a serious threat to native semi-arid areas of Brazil, especially in areas of the state of Ceará that are typically invaded by Cryptostegia madagascariensis, an exotic plant species from Madagascar. However, how this biological invasion influences the composition of the arbuscular mycorrhizal fungal (AMF) community and how this affects further invasion by C. madagascariensis is not well known. Here we tested how inoculation with species of AMF affects the development of this invasive plant. We analyzed and compared the AMF community composition of four different stages of biological invasion by C. madagascariensis, and examined the effects of inoculation with these four AMF communities, plus a dominant AMF species (Rhizoglomus intraradices) on plant dry biomass, root colonization, plant phosphorous concentration, and plant responsiveness to mycorrhizas of plants of C. madagascariensis. We found that all studied treatments (except the inoculum from the native plant root zone) promoted the growth of C. madagascariensis and lead to a higher P concentration. Our results demonstrate that the invader might be altering the composition of the AMF community in field conditions, because inoculation with this community enhanced invader growth, root colonization, and P uptake.


Introduction
To understand more about the effects of invasion on arbuscular mycorrhizal fungi (AMF) community, current research has focused on feedback between AMF communities and invasive species (Zhang et al. 2010), in the host-fungal pairing (Klironomos 2003), and how AMF can impact the performance of invasive species (Stampe & Daehler 2003).This work is the first field study to examine the impacts of C. madagascariensis on AMF community composition along an invasion gradient and to investigate how the changed-AMF community may contribute to create a "window of opportunity" (Johnstone 1986;Agrawal et al. 2005) and consequently influencing plant invasions (Blumenthal 2005;Kulmatiski & Kardol 2008).
Plant-AMF specificity affects plant invasion (Mummey & Rillig 2006;Parkash & Aggarwal 2009;Zhang et al. 2010) and may in some cases strengthen invasive ability of plants through complex feedback loops (Richardson et al. 2000).Plants provide AMF with energy while AMF provides water and nutrients such as phosphorus and nitrogen to the plant host (Smith & Read 2008;Hodge & Storer 2014).Plants have affected the AMF community structure through influences on soil resource availability, like available P (Alguacil et al. 2012;Ji et al. 2012;Beauregard et al. 2013;Lekberg et al. 2013;Geel et al. 2015;

Site description and sampling methods
We selected study areas near Ibaretama, Ceará, Brazil (06°51'11.3''S,35°55'51.5''W,and average altitude 109 m).The climatic conditions of studied areas is Bsh (Köppen), hot semi-arid with hot summers and mild to warm winters, annual precipitation and temperature of 600 mm and 30 °C respectively.Rainfall in these sites is scanty, unpredictable and irregular (Alves et al. 2009).The soil type was classified as an Eutric Vertisol (WRB 2006).
The invasive Cryptostegia madagascariensis Bojer ex.Decne.and the native Copernicia prunifera (Mill.)H. E. Moore, which co-occur in the "Copernicia forest" in the Brazilian xeric shrubland from Ceará State, Brazil, were selected.The seeds of the invader were collected directly from the studied areas.
We collected samples from the root zone of both plants in four different stages of biological invasion by C. madagascariensis: Exotic -areas that contains only C. madagascariensis plants (We collected samples from root zone of young plants with an average height between 0.5 and 0.9 m); Native -areas that contains only C. prunifera plants (We collected samples from root zone of plants with an estimated average height of 12.0 m); Invasive Stage 1 (Invasive1) -areas that contains C. madagascariensis plants (We collected samples from root zone of plants with an average height between 1.5 and 3.0 m) climbing over C. prunifera plants with an estimated average height of 10.0 m; and Invasive Stage 2 (Invasive2) -areas that contains C. madagascariensis plants (We collected samples from root zone of plants with undetermined average height) climbing over dead C. prunifera plants with an estimated average height of 8.0 m).
We established for each stage of biological invasion (four stages -Native, Exotic, Invasive1, Invasive2) by C. madagascariensis forty plots of 100 m 2 according to Fortin & Dale (2005).We collected samples from forty plants in each plot that were pooled within each plot.We selected plants with the following characteristics: plant with diameter near soil surface > 3 cm and without any individual from a different plant species at a distance lower than 3 m to the sampling point (Daubenmire 1968;Fortin & Dale 2005;Caifa & Martins 2007;Costa & Araújo 2007;Durigan 2009).
Root zone samples (including soil and root fragments) were collected out near the drip line and beyond (0-20 cm deep), at the beginning of September 2012 during the dry period.The root zone samples from each species in each plot were bulked, mixed and stored at 4 °C until host-plant bioassays.Thus, we analyzed 40 composite samples for each invasion stage (N=160, 40 composite samples x four stages).The sampling period was done during the dry season because fungal sporulation is expected to be higher at this time in semi-arid environments (Silva et al. 2014).The root zone samples collected from the field was divided into portions intended for: chemical soil characterization, AMF community assessment, and mounting trap cultures.
For chemical characterization we analyzed soil pH, total organic carbon, total nitrogen and available phosphorus from field soil.Soil pH was measured in a suspension of soil and distilled water (1:2.5 v:v, soil: water suspension) (Black 1965).Total organic carbon was estimated according methodology described by Okalebo et al. (1993).For nitrogen, the soil samples were digested and total nitrogen was quantified according to Kjeldahl (Black 1965).Available phosphorus (Olsen's P) was determined colorimetrically on spectrophotometer at 882 nm by extraction with sodium bicarbonate for 30 min (Olsen et al. 1954).
AMF communities were selected from each studied biological invasion stage.These communities naturally exist in the root zone of both studied model plants, where we classified as a changed-AMF community the community of arbuscular mycorrhizal fungi from Exotic, Invasive1 and Invasive2 areas; and as an unaltered-AMF community the community of AMF from Native areas.So, a trap culture of each AMF community for each condition was established from field soil (500 g) in plastic pots (2 kg).
The AMF communities of trap cultures were propagated on a common host-plant (Zea mays L.).This species is a standard host-plant used for mycorrhiza trap cultures and inoculum potential assays according to INVAM (http://invam.caf.wvu.edu/).The plants were grown in sterilized sand: field soil (3:1 m:m) in a greenhouse condition at University of Coimbra for four months until sporulation.The host-plants received weekly amounts of a modified nutrient solution (Hoagland & Arnon 1939) containing (in mg L -1 ): 554.0 KCl, 200.0 NaH 2 PO 4 .H 2 O, 2.24 MgSO 4 , 520.0 CaCl 2 .H 2 O, 1.7 MnSO 4 , 0.25 CuSO 4 .5H 2 O, 0.30 ZnSO 4 .7H 2 0, 5.0 NaCl, 3.0 H 3 O 3 , 0.09 (NH 4 )6Mo 7 O 24 .4H 2 O, and 32.9 Na-Fe EDTA.A portion of soil (500 g) from each studied condition was then used to create the initial AMF-inoculum (The soil, roots, and spores were used as inoculum) used for the experiment with C. madagascariensis plants.Another portion of soil (100 g) was taken from each studied condition for spore extraction and subsequent identification of AMF species for evaluation of the ecological index.The average temperature in the greenhouse was 28 °C, ranging from 20 °C to 35 °C, an irradiance of up to 70 % of full sun, relative humidity ranging from 65 to 75 %, and photoperiod of 16:8 h L:D.
Spores were extracted by the wet sieving technique (Gerdemann & Nicolson 1963) followed by sucrose centrifugation (Jenkins 1964).Initially, extracted spores were examined in water under a dissecting microscope.They were separated based on morphological characteristics.After it, they were mounted in polyvinyl alcohol lacto-glycerol (PVLG) with or without addition of Melzer's reagent (Walker et al. 2007).The identification of species was based on the descriptions provided by Schenck & Perez (1990), publications with descriptions of new genera and families (i.e.Oehl et al. 2008), and by consulting the international culture collection of arbuscular mycorrhizal fungi database -INVAM (http://invam.caf.wvu.edu).In this work we adopted the classification proposed by Oehl et al. (2011), including recently described new taxa (i.e.Goto et al. 2012 andSieverding et al. 2014).After AMF species identification, provide directly from field, we calculated the following ecological index: diversity index (H) proposed by Shannon & Weaver (1949), and dominance index (C) proposed by Simpson (1949) for each studied AMF community.In addition, we assessed spore abundance (total number of spores), spore abundance of AMF species (number of spores of each AMF species recorded), and the species occurrence frequency (FO i ) of each AMF species.FO i was calculated using the following equation: , where n i is the number of times an AMF species was observed and N is the total of AMF spores observed from each stage of invasion.We classify the results of FOi based on the Zhang's frequency of occurrence classification: dominant (FO i >50%), most common (31≤FO i ≤50%), common (10≤FO i ≤30%) and rare (FO i <10%) (Zhang et al. 2004).The AMF species were also classified as generalists (present in all areas), intermediate (present in three or two areas) or exclusive (present in only one area) (Stürmer & Siqueira 2011).
The bioassay experiments involved a completely randomized design with six treatments: non-inoculation (Control) and five inoculation treatments with inoculum of AMF communities from: Native = Inoculum from C. prunifera areas; Exotic = Inoculum from C. madagascariensis areas; Invasive1 = Inoculum from Invasive Stage 1 areas (beginning of invasion); Invasive2 = Inoculum from Invasive Stage 2 areas (invasion successful); and inoculum composed by Rhizoglomus intraradices (N.C.Schenck & G.S. Sm.) Sieverd.G.A. Silva & Oehl (Rhi), which we found as a dominant AMF species into changed-AMF communities from Exotic, Invasive1, and Invasive2 areas.The control received 500 ml of filtrate from 500 g of AMF-inoculum with no mycorrhizal spores obtained by sieving through a 25 μm mesh, and 500 g of sterilized mixed AMF-inoculum (a mix of all AMF-inoculums by model plant), thereby controlling for potential mineral and non-mycorrhizal microbial components of the AMF-inoculum.Each treatment had 30 replicates that were coming from 30 independent replicates at trap culture stage.
The experiment was also conducted in a greenhouse at University of Coimbra with average temperatures of 25/16 °C (day/night).Seeds of C. madagascariensis were surface Tancredo Augusto Feitosa de Souza, Susana Rodriguez-Echeverría, Leonaldo Alves de Andrade and Helena Freitas Acta Botanica Brasilica -30(1): 93-101.January-March 2016 sterilized in 10% sodium hypochlorite for 10 minutes, and then thoroughly rinsed with deionized water.Seeds were then were germinated in trays containing autoclaved sand (Twice at 121 °C for twenty minutes in two consecutive days).Seven days after emergence, seedlings (plant height varying 2-4 cm) were selected and individually transferred to plastics pots containing 2000 g of substrate (1500 g of autoclaved sand mixed with 500 g of soil containing AMF-inoculum, with the number of infective propagules per gram of soil ranging between 5.35 and 6.18.The number of infective propagules was determined following the protocol described by Habte & Osorio (2001).All pots were covered with aluminum wrap around the seedling to prevent dehydration and external contamination, and the plants were watered with tap water as necessary and fertilized with the same nutrient solution used on trap culture (Hoagland & Arnon 1939).
Five months after planting, the plants were harvested.Roots were separated from shoots, and immediately the fresh roots were weighed.A total of 0.5 g of each fresh root sample was used for determination of root colonization.The remaining material and shoots were oven-dried (72°C for 48h) and used for measurement of shoot-, and root dry biomass.The amount of colonization was estimated using a grid line intersect method with examination of 100 intersects under a compound microscope at 200X magnification (Phillips & Hayman 1970;Giovannetti & Mosse 1980).
The plant responsiveness to mycorrhizas (PRM) was calculated using dry biomass of plants in the AMF-inoculum treatment and the control (Janos et al. 2007), using the formula: PRM = [(B AMF -B control )/B AMF ] x 100, where B AMF is dry biomass of model plants in the AMF-inoculum treatment and B control is dry biomass of the control treatment.
A known mass of the grounded material was digested in a digestion flask containing a triple acid mixture [HNO 3 :H 2 SO 4 : HCl (60%), with a ratio of 10:1:4, respectively] to analyze the total P. Phosphorous was analyzed using the vanadate molybdate colorimetric method (Jackson 1973).We also calculated the AMF inoculum effect on plant P content (IEP), using the formula: IEP = (P AMF -P Control )/(P AMF + P Control ), where P AMF is plant P content in AMF inoculum treatment (adapted from Armas et al. 2005), assuring that the AMF inoculum effect can only range from -1 to +1, with positive values indicating positive effects of AMF on plant P content, and negative values indicating negative effects.
The relationships between the AMF community structure and soil properties were examined using correlation analyses by the Pearson correlation coefficient (r).Oneway ANOVA was used to test the effect of AMF inoculum on plant dry biomass, and plant P content for C. madagascariensis plants.Data sets were transformed (arcsin square root for percentage variables and logarithmic for other variables), but the results were presented in their original scale of measurement (means with stander deviation) (Zar 1984).For mycorrhizal colonization, PRM and IEP, the Control treatments were excluded from the analyses since the results were null, and were analyzed using Tukey's HSD post-hoc comparison test (p<0.05).Significant differences among AMF inoculum treatment on plant dry biomass and plant P content were also determined using the Tukey's HSD post-hoc comparison test (p<0.05).One way ANOVA and correlation analyses were conducted using SAS 9.1.3Portable and the ecological indexes were conducted using MVSP 3.1 (Kovach 2007).

AMF community and soil properties from field samples
The samples from Exotic and Native areas had the largest number of AMF identified species, with 16.80 and 16.00 identified species respectively, followed by Invasive1 (11.70) and Invasive2 (8.50).The number of spores was 2.74 spores/g soil in Exotic; 5.17 spores/g soil in Native; 1.16 spores/g soil in Invasive1; and 0.86 spores/g soil in Invasive2.Diversity (H, Shannon's index) and dominance (C, Simpson's index) in the Exotic and Native was about 2.42 and 0.89 respectively, but in the Invasive1 and Invasive2 these values changed to 1.76 and 0.76 respectively (Tab.1).
In Soil pH, total organic carbon, total nitrogen and available phosphorous were higher in the Invasive2 than other studied root zones, but no significant differences were observed between Exotic and Native root zones for any of the soil properties analyzed (Tab.3).
There were significant effects of AMF inoculation on the root colonization (F 4, 149 = 89.67, p<0.0001).No differences were found between Rhi, Invasive2, Invasive1, and Exotic treatments, but when compared with the Native results, these treatments significantly enhanced the root colonization of C. madagascariensis plants.We did not found any AMF colonization structure inside of roots from control, so we excluded it from the statistical analyses (Fig. 2).
The AMF inoculation significantly influenced plant P content (F 5, 179 = 13.62, p<0.0001).The inoculation with AMF community from Native root zone had lower plant P content than other five treatments.C. madagascariensis plants had more plant P content when growing with AMF inoculum from Invasive2 root zone, but the exotic plant also experienced positive effects growing with Rhi, Inva-sive1, and Exotic (Fig. 3).
Differences in AMF inoculation treatments were also detected in PRM (F 4, 149 = 4.59, p<0.001) and IEP (F 4, 149 = 4.59, p<0.001).Plant growth responses (in terms of the plant responsiveness to mycorrhizas and inoculum effect on P uptake) of the C. madagascariensis plants to AMF   Could biological invasion by Cryptostegia madagascariensis alter the composition of the arbuscular mycorrhizal fungal community in semi-arid Brazil?
inoculation differed among the five treatments and both variables were lowest with AMF from the native root zone .Analyzing the effects of other AMF inoculation treatments we found positive effects on PRM and EIP of C. madagascariensis plants growing in the Rhi, Invasive2, Invasive1 and Exotic inoculation treatments.

Discussion
The study of field samples shows that the biological invasion by C. madagascariensis changes AMF community composition (Tabs.1-2) and soil properties (Tab.3) along the different studied stages of biological invasion in the Brazilian semi-arid, supporting our first hypothesis.The results obtained in the bioassays demonstrate that these altered AMF communities, and especially R. intraradices, the most abundant species in the invaded area, promote the growth of the invasive C. madagascariensis, supporting the remaining original hypothesis of this study.
AMF sporulation, AMF diversity were higher in the Native root zone compared with the results observed in the root zones from different stages of invasion (Invasive1, Invasive2, Exotic) (Tab.1).The invader altered the composition of AMF communities promoting species from Order Glomerales (e.g.Rhizoglomus intraradices) at the expense of AMF species from Order Diversisporales (e.g.genus Dentiscutata and Scutellospora) (Tab.2).Our results about AMF diversity and dominance agree with the work done by Oehl et al. (2010) that studied the dynamics of AMF sporulation and found that the AMF community was more diverse in natural fields and with other studies on invasive exotic plants (Hawkes et al. 2006).The presence of C. madagascariensis changed AMF composition in roots of C. prunifera along the invasion gradient, and when coexisting with the native species, the invader was able to kill the palm tree during its growth, changing the abundance of spores, and AMF diversity (Tab.1), and the frequency of occurrence of different AMF species, like Acaulospora, Dentiscutata, Entrophospora, Gigaspora, and Scutellospora genus (Tab.2) The invasion also altered the values for soil pH, total organic carbon, total nitrogen, and available phosphorous in the root zone (Tab.3).AMF species from Order Diversisporales (genus Acaulospora), and Gigasporales (genus Dentiscutata, Gigaspora, and Scutellospora) are common in semi-arid sites (Dandan & Zhiwei 2007;Silva et al. 2014), and in acid soil with low phosphorous availability (Ramos et al. 2008).However, disturbances and changes in soil properties, such as soil pH, total organic carbon and available phosphorous can be favorable for Glomerales species (Oehl et al. 2010;Jansa et al. 2014;Carneiro et al. 2015).Therefore, changes in soil properties induced by the invasive plant may contribute to constrain the occurrence of some AMF species, reduce AMF sporulation, and decrease AMF diversity and AMF dominance (Tab.1).Silva et al. (2014) in a study performed in an environmental gradient in the Brazilian semi-arid showed a strong correlation of AMF community composition with soil attributes and soil properties are influenced by plant community on a local scale, affecting the community structure even in areas that are geographically close.In our study, the presence of high density of C. prunifera dead roots might have stimulate the sporulation specific AMF species and might accelerate chemicals processes in the rhizosphere soil that changed   Studies also showed that plant neighbors were important in structuring AMF communities (Hausmann & Hawkes 2009), the presence of invasive plants changed AMF composition in roots of their native neighbors (Hawkes et al. 2006), and soil properties also can have influence on the AMF community composition (Oehl et al. 2010;Jansa et al. 2014;Silva et al. 2014).
In our bioassay, all inoculation treatments were capable of root colonization of C. madagascariensis plants.Plant dry biomass and root colonization (Figs.1-2), indicated that the host-plant responded differently to the six AMF treatments.The results from these two variables also indicated a significant responsiveness (around 52 %) of C. madagascariensis to Rhizoglomus intraradices inoculation.In fact, we found a positive effect of Rhizoglomus intraradices in the invasive plant, which might promote the success of the invader .
Our work increases the understanding of the ecological mechanisms underlying how exotic species can alter the AMF community composition in field conditions; and how modified AMF communities can affect the growth of the invader.Invaders can encounter certain AMF species that can facilitate their establishment in new environments (Reinhart & Callaway 2006;CMR Pereira et al. 2014) or the invader can alter AMF composition so as to reduce native species' competitiveness (Vogelsang & Bever 2009;Stinson et al. 2006).Our results indicate that, under field conditions, C. madagascariensis can change important AMF species either due to a direct preference for some AMF species or indirectly by changes in soil properties.This shift in the AMF community generates positive feedback with the invasive plant as demonstrated in the bioassay experiment.Based on these results, we suggest that the strategy used by C. madagascariensis to occupy new environments is to increase the abundance of Rhizoglomus intraradices which in turn has a highly positive effect on the invader´s growth.

Table 2 .
Frequency of occurrence (FO i ) of AMF species of each studied rhizosphere soil Stürmer and Siqueira (2011)s the number of times an AMF species was observed and N is the total of AMF spores observed from each studied condition; 2 Classification of AMF frequency of occurrence proposed byZhang et al. (2004): R -rare (FO < 10%); C -common (10 ≤ FO ≤ 30 %); MC -most common (31 ≤ FO ≤ 50); and D -dominant (FO > 50 %); 3 Species classification proposed byStürmer and Siqueira (2011): generalists (present in all areas), intermediate (present in 2 or 3 areas) or exclusive (present in only one area).Tancredo Augusto

Table 3 .
Soil properties of the area of each invasion stage by C. madagascariensis (mean ± SD, N= 40) for each analyzed parameter.

Tancredo Augusto Feitosa de Souza, Susana Rodriguez-Echeverría, Leonaldo Alves de Andrade and Helena Freitas the
soil properties, like soil pH, total organic carbon, total nitrogen and available phosphorous (Tab.3).