Abstract Background and Aims Even when adapted to flooding environments, the spatial distribution, growing strategies and anti-herbivore defences of plants face stressful conditions. Here we describe the effects of flooding on carbon allocation on growth, domatia and leaf production, and the herbivory on the myrmecophyte domatia-bearing Tococa coronata Benth. (Melastomataceae) growing along river banks in the Amazon region. Methods In an area of 80 000 m2 of riparian forest along the Juruena River we actively searched for individuals of T. coronata. In each plant we evaluated the size of the plant when producing the first domatium and determined its best predictor: (1) plant total height; (2) size of plants above flood level; or (3) length of time each plant spent underwater. We also compared the herbivory, internode elongation, foliar asymmetry and specific leaf weight between T. coronata individuals growing above and below the maximum flooding level. The distance to the river and the height of the first domatium produced were compared between T. coronata and its sympatric congener, T. bulifera. Key Results We found that T. coronata invests in rapid growth in the early ontogenetic stages through an elongation of internodes rather than in constitutive anti-herbivore defences to leaves or domatia to exceed the maximum flooding level. Consequently, its leaf herbivory was higher when compared with those produced above the flooding level. Individuals with leaves above flood levels produce coriaceous leaves and ant-domatias. Thus, flooding seems to trigger changes in growth strategies of the species. Furthermore, T. coronata occurs within the flood level, whereas its congener T. bullifera invariably occurs at sites unreachable by floods. Conclusion Even in conditions of high stress, T. coronata presents both physiological and adaptive strategies that allow for colonization and establishment within flooded regions. These mechanisms involve an extreme trade-off of postponing adult plant characteristics to rapid growth to escape flooding while minimizing carbon allocation to defence. Ant–plant mutualisms, trade-off, fluctuating asymmetry, flood stress, plant strategies, submergence tolerance INTRODUCTION The Amazon Basin is formed by large rivers subjected to an annual flood pulse with an amplitude of up to 10 m and flood duration up to 270 d (Junk et al., 1989; Junk, 1997). This dynamic of inundation directly influences the vegetative and reproductive growth of riparian vegetation along these river banks (Junk et al., 1989; Waldhoff et al., 1998; Piedade et al., 2000). For the plants growing on the margin of rivers, the flooding imposes a seasonal stress, such as the reduction of soil oxygen to extremely low levels (Pezeshki, 2001; Jackson and Colmer, 2005; Mommer and Visser, 2005). Over evolutionary time, some plant species have developed different morphological and physiological strategies to escape and tolerate survival in flooded conditions (Kramer, 1963; Navas and Garnier, 2002; Parolin, 2009; Ferreira et al., 2010). In fact, there are species that have adapted and occur exclusively or primarily in floodplains (Foster, 1992; Wittmann et al., 2013). If vegetative parts of the plant become completely submerged during periods of inundation, one of the principal strategies employed by the plant is growth restriction and quiescence (i.e. unavailability of favourable conditions) (Fukao and Bailey-Serres, 2008; Bailey-Serres et al., 2012). However, some plant species allocate resources to the rapid elongation of stems (Fu, 2001; Nagai et al., 2010; Bailey-Serres et al., 2012) to escape inundation and maintain vegetative parts above water levels (Parolin, 2002). Through this escape strategy, the plants remain in contact with the air and can maintain the structures responsible for gas exchange, even when other parts of the plant are submerged during flooding (Bailey-Serres et al., 2012; Nishiuchi, et al., 2012). The use of different strategies to decrease the effects of flooding can generate stress which may be reflected in changes to its phenotype (Vierling and Kimpel, 1992; Dat et al., 2000). This is mainly due to a reduction in nutrient uptake and assimilation levels as a result of limitations imposed on the plants during flood periods (Louda, 1986; Renaud et al., 1990). Nutrient availability is a determinant in the allocation of metabolites for defence against herbivores (Coley et al., 1985). Plants growing in environments with low nutrient concentrations tend to invest in quantitative carbon-based chemical defences with consequent slower growth (McKey, 1988; Coley and Barone, 1996). In view of this situation, plants growing in floodplains may be confronted by the following challenge in terms of energy balance: while there is pressure to hold the nutrients within tissues against herbivory, there is also a pressure for rapid growth as an escape mechanism from floods (Parolin, 2002, 2009). In fact, some plants found in floodplain environments which are either completely or partially submerged during floods invest less in chemical defences and are more susceptible and less tolerant to herbivore damage (Stout et al., 2002; Chen, 2008). Therefore, could frequent flooding events influence energy allocation in plants that use biotic defence? Biotic defences in plants are extremely common in Amazonian environments, most of them provided by ants in the deterrence of potential herbivores (Benson, 1985; Vasconcelos, 1991). One example of biotic defence is myrmecophytes (or ant-plants). Mymecophyte plant species invest in the production of physical structures called domatia to house colonies of highly specialized ants. In exchange for the protective shelter provided by the domatia, ants defend and protect the host plant against herbivores (Benson, 1985; Agrawal and Dubin-Thaler, 1999). One myrmecophyte species often found near rivers and small streams of the Amazon is Tococa coronata Benth. (Melastomataceae) (Guarim-Neto and Asakawa, 1978; Benson, 1985). This plant species has foliar domatia consisting of a pair of small pockets in the leaf limb with individual entrances, with all domatia occupied by a single colony of ants (Benson, 1985). In myrmecophytes, the colonies’ effectiveness in protecting the plant depends on a minimum number of worker ants patrolling the leaves of the plant (Izzo and Vasconcelos, 2005). In areas subject to periodic flooding such as those in which T. coronata is found, the rise of the water column can lead to total submersion of the plant, promoting stress to the plant and causing the death of the ant colony. Since the production of domatia implies energy expenditure (Fonseca, 1999), and given the infeasibility of plant defence by small ant colonies (e.g. Izzo and Vasconcelos, 2005), this energy could be redirected to vertical growth and flood escape. A tool commonly used to measure the influence of external factors on the phenotype of an individual is called fluctuating asymmetry (FA) (Palmer and Strobeck, 1986). The FA measures the deviation from morphological symmetry in an individual’s ontogeny of organisms growing under constant stress (Møller and Eriksson, 1994; Wilsey et al., 1998; Graham et al., 2010). Here, we evaluated the effect of flooding on growth–defence trade-offs (Herms and Mattson, 1992) for the myrmecophyte T. coronata along river banks of the Juruena River located in the southern Brazilian Amazon. We approached the escape of flooding in the elongation of the internodes, the carbon allocation in defence against herbivores, herbivory and production of domatia. The FA was used to infer environmental stress on plants growing on total flooding conditions. In this case, we also compared some of these aspects with the myrmecophytic congeneric species Tococa cff. bullifera Mart. and Schrank ex DC (Melastomataceae) which inhabits the same riparian forests, but is commonly associated with non-floodable environments (Michelangeli, 2005). Specifically, we tested the following hypotheses: (1) the production of leaves with domatia by T. coronata occurs only after the plant surpasses the height of the river’s maximum flood level; (2) when growing in flood areas, T. coronata invests resources in the rapid growth and elongation of internodes, but less in the leaf, until reaching a height that has surpassed the maximum flood level; (3) the herbivory of T. coronata individuals is higher on leaves produced below the maximum flood level; and (4) plants growing below the river’s maximum flood level produce more asymmetric leaves (indicative of a stressing environment). MATERIALS AND METHODS Study area This study was conducted in three field campaigns in the beginning of the rainy season (2009, 2010 and 2012) in riparian forests along the margins and islands of the Juruena River at São Nicolau Farm (9º48’S, 58º15’W, elevation 254 m), located in the municipality of Cotriguaçu, north of the State of Mato Grosso, Brazil. The reserve area covers 7000 ha of dense rain forest in the Brazilian Meridional Amazon, surrounded by >50 000 ha of continuous forest. The climate is wet tropical (Am – according to Köppen classification) with an annual average temperature of 24 °C, 85 % humidity and 2300 mm precipitation (Camargo et al., 2010). The Juruena River is approx. 400 km in length and joins the Teles-Pires river to form the Tapajos river, one of the tributaries of the Amazon river. Dynamics of the water level in the Juruena River are driven by two distinct seasons, a high-water level season between November and April and a low-water level season between May and October. According to the Brazilian Water Agency’s System of Hydrological Information, the river’s range of flood elevation in our study area was 4.6–10 m between the years 1984 and 2014. An ANA station (Brazilian National Agency for Water Availability) is located approx. 500 m from the study site, which takes daily readings of the Juruena River’s water level. All data on the Juruena River’s water level are available at http://www.snirh.gov.br/hidroweb or at http://hidroweb.ana.gov.br/default.asp. Distribution of T. coronata To describe the life history and abundance distribution of the myrmecophyte species T. coronata, we surveyed a 2000 × 40 m (80 000 m2) stretch of riparian forest along the Juruena River. For each individual of T. coronata found, we measured: the base height of the individual from the ground that was not submerged on the day of observation; the height of the first domatium, when present, from the base of the individual; and the distance of the individual from the river bank. In addition, we also measured the river’s maximum water level from the previous year’s flood. This was done by locating the flood mark left on the trunks of trees near the sampled plants, then using a laser level to measure this flood mark in relation to the river level at the time of data collection. This measurement was compared with data provided by São Nicolau Farm, where the study was conducted. The heights of plants and first domatia in relation to flood level were obtained in two campaigns (2010 and 2011). Leaf fluctuating asymmetry (LFA) and internode elongation We used leaf fluctuating asymmetry (LFA) as a proxy for T. coronata stress against water saturation and compared LFA of plant parts above and below the maximum flood level observed in the year prior to this study. For this, we selected 30 individuals that had vegetative parts growing above the river’s maximum water level in the previous year, and 30 individuals whose vegetative parts would remain totally submerged during the high-water period. For the calculation of internodal symmetry we chose the first three leaves arranged in the first three internodes below the apical bud. In each leaf, the asymmetry index was calculated: AI = LA – LB/LA × LB) × 100, where LA = greatest distance from the central vein to the margin of one side of the leaf and LB = greatest distance from the central vein to the margin of the side opposite the one previously measured. In addition, we measured the length of the first six internodes from the base of individuals (n = 30 for each of the two categories), with the objective of finding possible differences in internode elongation between individuals with emergent parts and individuals of T. coronata that would remain totally submerged at peak flood. Specific leaf weight index To evaluate carbon investment in leaves of T. coronata individuals growing above and below the maximum flooding level, we used specific leaf weight (SLW). SLW is obtained by the division of an oven-dry mass of a given leaf by its area (SWF = M/A; where M = dry weight in g and A = leaf area in cm2). Higher SLW values represent a greater allocation of resources in carbon-based structural compounds and macromolecules and lower amounts of nitrogen (Dyer et al., 2001; Westoby et al., 2002). From each of 15 individuals growing under or above the flooding level, we obtained 27 cm2 of the leaf limb of three leaves (excluding the central and peripheral veins). These pieces of leaves were then dried at 60 °C for 48 h (until constant weight) and weighed using a 0.005g (Marte®, AM-220) precision scale. Leaf area removed To quantify the percentage of foliar herbivory in plants of T. coronata growing above and below the maximum flood level, we used 30 individuals for each of the two categories. In each individual, we measured the leaf area removed through the categorization proposed by Dirzo and Dominguez (1995): (1) 0 % (no loss of leaf area); (2) 1–6 %; 6–12 %; (3) 12–25 %; (4) 25–50 %; and (v) 50–100 % (large loss of leaf area). This method renders herbivory values as accurate as other quantitative methods such as image processing (Vásquez et al., 2007). In this study we consider herbivory to be any loss of photosynthetic area due to the action of chewing, scraping or sucking (necrotic areas) insects, miners and/or gallers. Comparison with another sympatric myrmecophyte To compare the height and spatial position of the first leaf with domatia produced in relation to the water between T. coronata and the sympatric and phylogenetically similar myrmecophyte Tococa bullifera, both species of plants and the ant species found in these myrmecophytes were collected for identification at the Community Ecology Lab (Universidade Federal de Mato Grosso, Brazil). Ant identification was carried out by comparison with the reference collection, while plant identification was obtained using a taxonomical key to the genus (Michelangeli, 2005) and confirmed with specimens from the herbarium at the Instituto Nacional de Pesquisas da Amazônia (Brazil). Data analysis All independent variables were first tested to determine any correlation. Evaluation of the three independent factors revealed that the length of time that each plant remained submerged, total plant height and the size of branches that were outside the maximum flood level in the previous year were clearly correlated (Pearson r > 0.4), and therefore could not be used in multiple models (Zar, 1996). We used three generalized linear models (GLMs) with binomial error distribution, where the occurrence of a developed domatium (i.e. capable of hosting an ant queen) was related to each factor: (1) total plant height (cm); (2) size of the plant found above the flood level (cm); and (3) length of time that each plant spent underwater (months). We selected the best of the three models to predict the occurrence of the first domatium using the best-fit model (explained deviance) and the ΔAIC (Akaike information criterion) of each GLM. All the statistical analyses were performed using the R software/environment (R Development Core Team, 2015). Comparison between the production height of domatia, as well as the straight-line distance from the river to the location of T. coronata and T. bulifera plants within the study area was evaluated using a t-test once the normal distribution was confirmed. The comparison between fluctuating asymmetry, elongation of internodes, specific leaf mass and percentage leaf area removed was conducted between plants growing above and below the maximum flood level also using Student t-tests. RESULTS The production of domatia by T. coronata appears to be closely associated with the flood level. Although production of the first domatium in T. coronata can vary greatly in relation to the height of the individual, we did not find a single domatium below the previous year’s flood level. In fact, the length of time plants spent submerged in the previous year did not show any correlation with the probability of the occurrence of domatia (z = –0.05, P > 0.05, AIC = 25.35). In contrast, both the total height of the individual in relation to the soil (d2 = 0.77; z = 2.57; P = 0.01; AIC = 19.25) and the size of the branches found emerged during the previous period of rising water (d2 = 0.82; z = 2.82; P = 0.004; AIC = 16.15) appears to be related to the production of the plant’s first domatium (Supplementary Data Fig. S1). However, the model containing the size of the emerged part shows a better fit to the data (ΔAIC = 3.1) (Fig. 1). Throughout the 80 000 m2 study area, we found only 11 individuals of T. bullifera, all of them with at least one mature domatium, while within the same section we found 33 T. coronata individuals with domatia. The number of individuals without domatia, however, was not counted, as juvenile T. coronata occur in areas that can contain tens to hundreds of individuals. All individuals found with domatia were colonized by some species of ant. In T. bullifera, 62.5 % of the plants were colonized by the Allomerus octoarticulatus ant, 25 % by Azteca sp1 and 12.5 % by Crematogaster sp1. For T. coronata, 30 % of the plants contained colonies of Allomerus octoarticulatus, 30 % with colonies of Crematogaster sp1, 20 % with colonies of Azteca sp1, 10 % with colonies of Crematogaster sp2 and 10 % with Hylomyrma sp1. Fig. 1. View largeDownload slide Probability of occurrence of the first domatia in individuals of Tococa coronata (Melastomataceae) found in the Juruena River riparian forest modelled as a function of the height of the plant found above the maximum flood level. The smoother line was the predictive result of a generalized linear model [presence of domatia = –4.74 + 0.17 × height above flood level (cm), GLM, family = binomial, link = logit] and the grey area indicates the confidence interval of the model. Each point represents the occurrence of domatia in a single plant. The presence of domatia is categorized as ‘1’, whereas absence is ‘0’; however, the function ‘jitter’ was employed to reduce the point overlap and showinged the high number of plants with no domatia when they are without any vegetative part above the maximum flood. Fig. 1. View largeDownload slide Probability of occurrence of the first domatia in individuals of Tococa coronata (Melastomataceae) found in the Juruena River riparian forest modelled as a function of the height of the plant found above the maximum flood level. The smoother line was the predictive result of a generalized linear model [presence of domatia = –4.74 + 0.17 × height above flood level (cm), GLM, family = binomial, link = logit] and the grey area indicates the confidence interval of the model. Each point represents the occurrence of domatia in a single plant. The presence of domatia is categorized as ‘1’, whereas absence is ‘0’; however, the function ‘jitter’ was employed to reduce the point overlap and showinged the high number of plants with no domatia when they are without any vegetative part above the maximum flood. Leaves of T. coronata found both above and below the previous year’s maximum river level presented a high degree of leaf asymmetry. However, leaves produced on branches below the previous year’s maximum flood level were more asymmetrical (mean ± s.d. 8.91 ± 3.22 %) than those on branches produced above that level (1.25 ± 0.86 %; Fig. 2A; t = 13.129; d.f. = 29; P < 0.001). Also, the parts of the plants produced below the maximum flood level presented internodes almost twice as elongated (5.76 ± 1.69 cm) in relation to plants with internodes growing above the maximum flood level (3.36 ± 0.32 cm) (Fig. 2B, t = 7.913; d.f. = 29; P < 0.001). Fig. 2. View largeDownload slide Box plots of the following variables: (A) leaf fluctuating asymmetry (%), (B) internode elongation (cm), (C) specific leaf weight (g cm–2) and (D) leaf area removed (%) in Tococa coronata (Melastomataceae) individuals found above and below the maximum flood level in riparian forests of the Juruena River, Southern Amazon, Brazil (n = 30 for each treatment). Fig. 2. View largeDownload slide Box plots of the following variables: (A) leaf fluctuating asymmetry (%), (B) internode elongation (cm), (C) specific leaf weight (g cm–2) and (D) leaf area removed (%) in Tococa coronata (Melastomataceae) individuals found above and below the maximum flood level in riparian forests of the Juruena River, Southern Amazon, Brazil (n = 30 for each treatment). Individuals of T. coronata found above or below the maximum river level presented different strategies for carbon allocation in their leaves (t = –13.617; d.f. = 14; P = 0.001). Leaves produced below the maximum river level presented a lower specific leaf weight (0.09 ± 0.01 g) than leaves produced above the flood level (0.2 ± 0.02 g) (Fig. 2C). In addition, leaf area removed by herbivores also differed (t = 9.507; d.f. = 29; P < 0.001), with individuals found below the previous year’s maximum flood level subject to greater leaf damage caused by herbivores (mean ± s.d. 28.42 ± 14.25 %) when compared with individuals found above this level (4.26 ± 1.99 %) (Fig. 2D; additional information can be found in Supplementary Data Table S1). All observed individuals found above the maximum river level produced flowers at the beginning of the rainy season (November 2011 and 2012). On the other hand, even allocating less carbon on the leaves, about 20 individuals among hundreds of small individuals of T. coronata found below the maximum river level individuals were observed producing flowers. Both myrmecophytes found in the area differ in place of occurrence along the river bank (t = –6.571; d.f. = 8; P = 0.001). Tococa coronata individuals occurred almost exclusively below the river’s maximum flood level (mean ± s.d. 52.1 ± 32.9 cm), whereas T. bullifera individuals occurred above the river’s flood level (176.9 ± 37.2 cm) (Fig. 3A). The two species also differed in the height of the first domatium developed (t = 5.507; d.f. = 8, P = 0.001). The height of the first domatium in T. coronata was 122.2 ± 45 cm, while for T. bullifera it was 33.3 ± 19.6 cm (Fig 3B). In addition, both myrmecophytes also differed in distance from the river (t = –5.590, d.f. = 8, P = 0.001), with an average distance of 131.9 ± 83.4 cm for T. coronata individuals, and 747.6 ± 278 cm for T. bullifera. Fig. 3. View largeDownload slide Box plots showing the median, quartiles and range (whiskers) of the following Tococa bullifera (Melastomataceae) and Tococa coronata (Melastomataceae) variables: (A) base height of individuals in relation to river level at the time of study (cm) and (B) height of the first domatium (cm) in Juruena River riparian forests, Southern Amazon, Brazil (n = 11 for T. bullifera and n = 33 for T. coronata). Fig. 3. View largeDownload slide Box plots showing the median, quartiles and range (whiskers) of the following Tococa bullifera (Melastomataceae) and Tococa coronata (Melastomataceae) variables: (A) base height of individuals in relation to river level at the time of study (cm) and (B) height of the first domatium (cm) in Juruena River riparian forests, Southern Amazon, Brazil (n = 11 for T. bullifera and n = 33 for T. coronata). DISCUSSION In this study, we demonstrated that individuals of the myrmecophyte T. coronata postpone a typical characteristic of the adult individuals, presenting the first domatia only when vegetative parts are emerged after the maximum water level is reached for the high-water period. In fact, it is easy to find plants with seedling characteristics up to 2 m high, depending upon its position in the riverside. In plants growing below the flood limit, we observed that several vegetative characteristics are modified, favouring rapid growth. Initially, T. coronata invests only in energy for growth through a disproportionately large stretching of their internodes. Rapid development requires a very characteristic trade-off, where smaller individuals that would be entirely submerged in the rainy season grow at the expense of the lower allocation of carbon in their leaves (Nagai et al., 2010; Maurenza et al., 2012). This low investment in constitutive defences is associated with greater rates of observed herbivory (Coley and Barone, 1996). The pressure to escape through rapid growth and the physiological changes by decreasing the anti-herbivore traits probably results in the observed higher herbivory. It also suggests that plants are growing under particularly stressful conditions, which is reinforced by the pronounced asymmetry observed in leaves produced below the maximum flood level. The myrmecophyte T. coronata presents two well-defined phases of life, with a set of physiological adaptations allowing escape from floods (Parolin, 2002) when in the ‘juvenile’ stage, and the development of domatia for ant nesting in the ‘mature’ stage (Benson, 1985). These stages are not related to the plant’s reproduction, but to their persistence in the area, since few ‘juveniles’ should produce flowers, but the major part die or lose part of their vegetative tissue when submersed. All the ‘mature’ plants studied here produced flowers at least once during the 4 years of study. The duration of submersion was not a good predictor for first domatia production by the plant, as no submerged plant produced domatia, regardless of time. However, with the size of the plant and the size of the emerged part at the time of flooding, the latter presented a better fit to the data. This explains why, analysing the distribution of T. coronata plants along the river planes, only larger plants, taller than 2 m, located near the river produce domatia. In contrast, small plants established in sites which are topographically higher have the ability to produce domatia, as is the case of plants smaller than 50 cm producing domatia in unflooded areas. Such a striking characteristic suggests that the maintenance of a photosynthetic part of the plant above water during the flood period may trigger the development of the first domatium. During flooding, the entire ant colony residing in a domatium would die. Studies dealing with ant–myrmecophyte interactions have shown that domatia are extremely important structures for myrmecophyte-associated ants, as the growth of ant colonies associated with myrmecophytes is dependent on the number of available domatia (Benson, 1985; Fonseca, 1999). As a consequence of host plant growth, the colony’s demand for proteins also grows (Fonseca, 1999). In fact, there are several examples within the literature indicating an increase in protection by ants in plants with more domatia (Duarte-Rocha and Bergallo, 1992; Del-Val and Dirzo, 2003; Izzo and Vasconcelos, 2005). Therefore, if production of domatia is associated with greater defence against herbivory, why does T. coronata not produce domatia below the maximum flooding level, even though there would be an annual mortality of the resident ant colony? First, the high-energy cost of producing domatia that would be submerged without ant protection would be a waste in the general energy balance of plants. Domatia located at this level may become unfeasible after flooding due to sediment accumulation. There are ant species that have the ability to ingest water and regurgitate outside of the colony after short-term flooding (Moog et al., 1997). However, we believe that such water-bailing behaviour would not be possible with fully submerged domatia during flood periods that may last months. Whole colonies residing in submerged domatia would quickly perish. The migration of ants from lower to upper domatia would not be viable in all cases. In fact, ant colonies in myrmecophytic plants are strongly limited by available nesting space (Fonseca, 1999). A large colony, even if not fully flooded, might not recover, and perish by losing the queen or a large number of workers. If colony death happens, a new colonization may take some time, since initial colony growth is slow because ant–myrmecophyte interactions are highly specialized (Dáttilo et al., 2013). During the period between colonization events, both young and mature plant individuals may remain unoccupied and unprotected against herbivores, until the colony was able to produce a sufficient number of workers for defence. During these periods, herbivory may be of such intensity as to reduce the plant’s reproductive success drastically or even lead to its death (Vasconcelos 1993). One strategy used by individuals without domatia to counterbalance the absence of anti-herbivory protection by ants would be to invest in quantitative defences in the leaves (McKey, 1988). Leaves with a higher specific leaf weight (i.e. higher carbon concentration), due to a higher concentration of cellulose, tannins and lignins, are harder, mechanically resistant and/or unpalatable to herbivores (Dyer et al., 2001; Westoby et al., 2002). However, smaller T. coronata individuals showed lower specific leaf weight compared with larger individuals. The possible reabsorption of chlorophyll in submerged leaves (Crawford, 1992; Dyer et al., 2001; Maurenza et al., 2012) cannot explain this difference in specific leaf weight, as all sampling took place during the dry period. Therefore, the observed phenomena resemble the grow–defence trade-off (Herms and Mattson, 1992). The difference in specific leaf weight and high herbivory levels in leaves growing below the maximum flood level may relate to a reduction in carbon allocation to sclerophyll and quantitative defences, leaving the available carbon for investment in growth. We suggest that, as a strategy to reduce the damage caused by flood seasonality (e.g. absence of domatia and carbon in leaves), T. coronata invests in rapid growth to surpass the maximum flood level by elongating the internodes, like several floodplain plant species using the flood escape strategy (Colmer and Voesenek, 1999; Fu, 2001; Parolin, 2002; Pierik et al., 2005; Nagai et al., 2010; Nishiuchi et al. 2012). Some plants can significantly increase in length, even daily, through internode elongation (e.g. rice: 25 cm d–1, bamboo: 114 cm d–1) (Vergara et al., 1976; Fu, 2001; Parolin, 2002). In T. coronata the difference in internode size, coupled with the drastic modification of tissue quality between leaves above and below flood levels, shows the escape of flood pressure. In conclusion, even under high stress conditions, T. coronata shows physiological mechanisms allowing the species to colonize and establish in flooded environments, a characteristic not observed in its sympatric congener. These mechanisms involve a trade-off, which maximizes plant vertical growth as an escape strategy while drastically minimizing carbon allocation to chemical and biological defence. We do not know what physiological/molecular mechanism triggers hormonal growth modification (Parolin, 2012), or what environmental signal is used by the plant to change its allocation strategy. However, the suppression of domatium production preceding floods is something totally novel in ant-plant studies, demanding ecophysiological and genetic studies focusing on the mechanisms inducing production of domatia on ant-plant species. SUPPLEMENTARY DATA Supplementary data are available online at https://academic.oup.com/aob and consist of the following. Figure S1: two branches of Tococa coronata Benth. (Melastomataceae) sampled in riparian forests of the Juruena rRiver, Southern Amazon, Brazil. Table S1: parameters measured in Tococa coronata plants growing below or above the maximum flood level of the Juruena River, Cotriguaçú-MT. ACKNOWLEDGEMENTS We are grateful to Miquéias Ferrão Silva Junior and Silvana Cristina Hammerer for their help during the fieldwork in the pilot versions. We also thank ONF-Brazil and the PPBio project for logistical and financial support and the staff of the Central Herbarium of UFMT and INPA-AM for identification of plant specimens. M.F.P. thanks PELD-MAUA (CNPq-FAPEAM) and Grupo Ecologia, Monitoramento e Uso Sustentável de Áreas Úmidas, INPA, Manaus. This is publication 65 in the NEBAM technical series. LITERATURE CITED Agrawal AA, Dubin-Thaler BJ. 1999. Induced responses to herbivory in the Neotropical ant–plant association between Azteca ants and Cecropia trees: response of ants to potential inducing cues. Behavior Ecology and Sociobiology 45: 47– 54. Google Scholar CrossRef Search ADS Bailey-Serres J, Fukao T, Gibbs DJ, et al. 2012. Making sense of low oxygen sensing. Trends in Plant Science 17: 129– 138. Google Scholar CrossRef Search ADS PubMed Benson WW. 1985. Amazon ant-plants. In: Prance GT, Lovejoy TE, eds. Amazonia . Oxford: Pergamon Press, 239– 266. Camargo FF, Costa RB, Resende MDV, et al. 2010. Variabilidade genética para caracteres morfométricos de matrizes de castanha-do-brasil da Amazônia Mato-grossense. Acta Amazonica 40: 705– 710. Google Scholar CrossRef Search ADS Chen MS. 2008. Inducible direct plant defense against insect herbivores: a review. Insect Science 15: 101– 114. Google Scholar CrossRef Search ADS Coley PD, Barone JA. 1996. Herbivory and plant defenses in tropical forests. Annual Review of Ecology and Systematics 27: 305– 335. Google Scholar CrossRef Search ADS Coley PD, Bryant JP, Chapin FSIII. 1985. Resource availability and plant antiherbivore defense. Science 230: 895– 899. Google Scholar CrossRef Search ADS PubMed Colmer TD, Voesenek LACJ. 1999. Flooding tolerance: suites of plant traits in variable environments. Functional Plant Biology 36: 665– 681. Google Scholar CrossRef Search ADS Crawford RMM. 1992. Oxygen availability as an ecological limit to plant distribution. In: Bergon M, Fitter AH, eds. Advances in ecological research . London: Academic Press, 93– 185. Google Scholar CrossRef Search ADS Dat J, Vandenabeele S, Vranova E, Van Montagu M, Inzé D, Van Breusegem F. 2000. Dual action of the active oxygen species during plant stress responses. Cellular and Molecular Life Sciences 57: 779– 795. Google Scholar CrossRef Search ADS PubMed Dáttilo W, Izzo TJ, Vasconcelos HL, Rico-Gray V. 2013. Strength of the modular pattern in Amazonian symbiotic ant–plant networks. Arthropod-Plant Interactions 7: 455– 461. Google Scholar CrossRef Search ADS Del-Val E, Dirzo R. 2003. Does ontogeny cause changes in the defensive strategies of the myrmecophyte Cecropia peltata? Plant Ecology 169: 35– 41. Google Scholar CrossRef Search ADS Dirzo R, Domínguez CA. 1995. Plant–herbivore interactions in Mesoamerican tropical dry forest. In: Bullock SH, Mooney A, Medina E, eds. Seasonally dry tropical forest . Cambridge: Cambridge University Press, 304– 309. Google Scholar CrossRef Search ADS Duarte-Rocha CF, Bergallo HG. 1992. Bigger ant colonies reduce herbivory and herbivore residence time on leaves of an ant-plant: Azteca muelleri vs. Coelomera ruficornis on Cecropia pachystachya. Oecologia 91: 249– 252l. Google Scholar CrossRef Search ADS PubMed Dyer LA, Dodson CD, Beihoffer J, Letorneau DK. 2001. Trade-off in antiherbivore defenses in Piper cenocladum: ant mutualist versus plant secondary metabolites. Journal of Chemical Ecology 27: 581– 591. Google Scholar CrossRef Search ADS PubMed Ferreira CS, Piedade MTF, de Oliveira Wittmann A, Franco AC. 2010. Plant reproduction in the Central Amazonian floodplains: challenges and adaptations. AoB Plants 2010: plq009. doi: 10.1093/aobpla/plq002 Fonseca CR. 1999. Amazonian ant–plant interactions and nesting space limitation hypothesis. Journal of Tropical Ecology 15: 807– 825. Google Scholar CrossRef Search ADS Foster JR. 1992. Photosynthesis and water relations of the floodplain tree, boxelder (Acer negundo L.). Tree Physiology 11: 133– 149. Google Scholar CrossRef Search ADS PubMed Fu J. 2001. Chinese moso bamboo: its importance. Bamboo 22: 5– 7. Fukao T, Bailey-Serres J. 2008. Submergence tolerance conferred by Sub1A is mediated by SLR1 and SLRL1 restriction of gibberellin responses in rice. Proceedings of the National Academy of Sciences, USA 105: 16814– 16819. Google Scholar CrossRef Search ADS Graham JH, Raz S, Hel-Or H, Nevo E. 2010. Fluctuating asymmetry: methods, theory and applications. Symmetry 2: 466– 540. Google Scholar CrossRef Search ADS Guarim Neto G, Asakawa NM. 1978. Estudo de Mirmecodomaceos em algumas especies de Boraginaceae, Chrysobalanaceae, Melastomataceae e Rubiaceae. Acta Amazonica 8: 45– 49. Google Scholar CrossRef Search ADS Herms DA, Mattson WJ. 1992. The dilemma of plants: to grow or defend. Quarterly Review of Biology 67: 283– 335. Google Scholar CrossRef Search ADS Izzo TJ, Vasconcelos HL. 2005. Ants and plant size shape the structure of the arthropod community of Hirtella myrmecophila, an Amazonian ant-plant. Ecological Entomology 30: 650– 656. Google Scholar CrossRef Search ADS Jackson MB, Colmer TD. 2005. Response and adaptation by plants to flooding stress. Annals of Botany 96: 501– 505. Google Scholar CrossRef Search ADS PubMed Junk WJ, Barley PB, Sparks RE. 1989. The flood-pulse concept in river–floodplain systems. Canadian Special Publication of Fisheries and Aquatic Sciences 106: 110– 127. Junk WJ. 1997. Structure and function of the large central Amazonian river floodplains: synthesis and discussion. In: Junk WJ, ed. The central Amazonian floodplain: ecology of a pulsing systems . Berlin: Springer, 3– 20. Google Scholar CrossRef Search ADS Kramer PJ. 1963. Water stress and plant growth. Agronomy Journal 55: 31– 35. Google Scholar CrossRef Search ADS Louda SM. 1986. Insect herbivory in response to root-cutting and flooding stress on a native crucifer under field conditions. Acta Oecologia, Oecologia Generalis 7: 37– 53. Maurenza D, Marenco RA, Parolin P, Piedade MTF. 2012. Physiological responses to flooding and light in two tree species native to the Amazonian floodplains. Aquatic Botany 96: 7– 13. Google Scholar CrossRef Search ADS McKey D. 1988. Promising new directions in the study of ant–plant mutualisms. In: Proceedings of the XIV International Botanical Congress . Koeltz, Königstein/Taunus, 335– 355. Michelangeli FA. 2005. Tococa (Melastomataceae). Flora Neotropica 98: 1– 114. Mommer L, Visser EJ. 2005. Underwater photosynthesis in flooded terrestrial plants: a matter of leaf plasticity. Annals of Botany 96: 581– 589. Google Scholar CrossRef Search ADS PubMed Møller AP, Eriksson M. 1994. Patterns of fluctuating asymmetry in flowers: implications for sexual selection in plants. Journal of Evolutionary Biology 7: 97– 113. Google Scholar CrossRef Search ADS Moog J, Drude T, Maschwitz U, Agosti D. 1997. Flood control by ants: water-bailing behaviour in the Southeast Asian plant-ant genus Cladomyrma Wheeler (Formicidae, Formicinae). Naturwissenschaften 84: 242– 245. Google Scholar CrossRef Search ADS Nagai K, Atori Y, Ashikari M. 2010. Stunt or elongate? Two opposite strategies by which rice adapts to floods. Journal of Plant Research 123: 303– 309. Google Scholar CrossRef Search ADS PubMed Navas M, Garnier E. 2002. Plasticity of whole plant and leaf traits in Rubia peregrina in response to light, nutrient and water availability. Acta Oecologica 23: 375– 383. Google Scholar CrossRef Search ADS Nishiuchi S, Yamauchi T, Takahashi H, Kotula L, Nakazono M. 2012. Mechanisms for coping with submergence and waterlogging in rice. Rice 5: 2– 14. Google Scholar CrossRef Search ADS PubMed Palmer AR, Strobeck C. 1986. Fluctuating asymmetry: measurement, analysis, patterns. Annual Review of Ecology and Systematics 17: 391– 421. Google Scholar CrossRef Search ADS Parolin P. 2002. Submergence tolerance vs. escape from submergence: two strategies of seedling establishment in Amazonian floodplains. Environmental and Experimental Botany 48: 177– 186. Google Scholar CrossRef Search ADS Parolin P. 2009. Submerged in darkness: adaptations to prolonged submergence by woody species of the Amazonian floodplains. Annals of Botany 103: 359– 376. Google Scholar CrossRef Search ADS PubMed Pezeshki SR. 2001. Wetland plant responses to soil flooding. Environmental and Experimental Botany 46: 299– 312. Google Scholar CrossRef Search ADS Piedade MTF, Junk WW, Parolin P. 2000. The flood pulse and photosynthetic response of trees in a white-water floodplain (várzea) of Central Amazon, Brazil. Verhandlungen der Internationale Vereinigung für Limnologie 27: 1734– 1739. Pierik R, Millenaar FF, Peeters AJ, Voesenek LA. 2005. New perspectives in flooding research: the use of shade avoidance and Arabidopsis thaliana. Annals of Botany 96: 533– 540. Google Scholar CrossRef Search ADS PubMed Renaud PE, Hay ME, Schmitt TM. 1990. Interactions of plant stress and herbivory: intraspecific variation in the susceptibility of a palatable versus an unpalatable seaweed to sea urchin grazing. Oecologia 82: 217– 222. Google Scholar CrossRef Search ADS PubMed R Development Core Team. 2015. R: A language and environment for statistical computing . Vienna, Austria: R Foundation for Statistical Computing. http://www.r-project.org Stout MJ, Riggio MR, Zou L, Roberts R. 2002. Flooding influences ovipositional and feeding behavior of the rice water weevil (Coleoptera: Curculionidae). Journal of Economical Entomology 95: 715– 721. Google Scholar CrossRef Search ADS Vasconcelos HL. 1991. Mutualism between Maieta guianensis Aubl., a myrmecophytic melastome, and one of its ant inhabitants: ant protection against insect herbivores. Oecologia 87: 295– 298. Google Scholar CrossRef Search ADS PubMed Vasconcelos HL. 1993. Ant colonization of Maieta guianensis seedlings, an Amazon ant-plant. Oecologia 95: 439– 443. Google Scholar CrossRef Search ADS PubMed Vásquez PA, Grez AA, Bustamante RO, Simonetti JÁ. 2007. Herbivory, foliar survival and shoot growth in fragmented populations of Aristotelia chilensis. Acta Oecologica 31: 48– 53. Vergara BS, Jackson B, De Datta SK. 1976. Deep water rice and its response to deepwater stress. In: Yoshida S, ed. Climate and rice . Los Baños, Philippines: International Rice Research Institute, 301– 319. Vierling E, Kimpel JA. 1992. Plant responses to environmental stress. Current Opinion in Biotechecnology 3: 164– 170. Google Scholar CrossRef Search ADS Waldhoff D, Junk WJ, Furch B. 1998. Responses of three Central Amazonian tree species to drought and flooding under controlled conditions. International Journal of Ecology and Environmental Science 24: 237– 252. Westoby M, Falster DS, Moles AT, Vesk PA, Wright IJ. 2002. Plant ecological strategies: some leading dimensions of variations between species. Annual Review of Ecology and Systematics 33: 125– 59. Google Scholar CrossRef Search ADS Wilsey BJ, Haukioja E, Koricheva J, Sulkinoja M. 1998. Leaf ﬂuctuating asymmetry increases with hybridization and elevation in tree-line birches. Ecology 79: 2092– 2099. Google Scholar CrossRef Search ADS Wittmann F, Householder E, Piedade MT, et al. 2013. Habitat specifity, endemism and the neotropical distribution of Amazonian white-water floodplain trees. Ecography 36: 690– 707. doi: 10.1111/j.1600–0587.2012.07723.x Google Scholar CrossRef Search ADS Zar JH. 1996. Biostatistical analysis , 3rd edn. Upper Saddlle River, NJ: Prentice Hall. © The Author(s) 2018. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For permissions, please e-mail: email@example.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
Annals of Botany – Oxford University Press
Published: Jun 7, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera