Wetlands Ecol Manage https://doi.org/10.1007/s11273-019-09676-1 (0123456789().,-volV) ( 01234567 89().,-volV) ORIGINAL PAPER Effects of soil flooding, sunlight and herbivory on seedlings of Annona glabra and Pachira aquatica in a tropical swamp Dulce Infante-Mata . Patricia Moreno-Casasola . Teresa Valverde . Susana Maza-Villalobos Received: 3 October 2018 / Accepted: 20 June 2019 Ó Springer Nature B.V. 2019 D. Infante-Mata  P. Moreno-Casasola Instituto de Ecologı́a A. C, P.O. Box 63, 91000 Xalapa, Veracruz, México results showed that the survival of both species was high and was not affected by soil flooding, sunlight and herbivory. However, these factors affected plant growth rates. In general, the highest growth rates were observed in the treatment with high sunlight, mesic soil and herbivore exclusion. Both species displayed higher leaf biomass allocation under closed than under no canopy. Furthermore, under closed canopy conditions both species produced relatively more slender and taller stems, which may allow them to intercept light more efficiently. Also, both species showed low belowground biomass allocation in flooded soils, probably as a consequence of a high anoxic condition. Our results confirmed that soil flooding, sunlight and herbivory are important factors that influence the growth patterns of A. glabra and P. aquatica seedlings, but they do not affect seedling survival. This information may help resource managers to identify high-quality sites that deserve to be protected. Also, the knowledge on species responses to different environmental conditions may be useful in restoration programs for tropical swamp forests. T. Valverde Departamento de Ecologı́a y Recursos Naturales, Facultad de Ciencias, Universidad Nacional Autónoma de México. Ciudad Universitaria, 04510 Ciudad de México, México Keywords Biomass allocation  Morphological traits  Plant functional traits  Relative growth rate  Seedling survival  Tropical wetland Abstract Wetland seedlings, in addition to dealing with the effects of flooding, must gain access to sunlight and avoid herbivore damage in order to establish. Understanding the effects of environmental factors on seedling growth and how plants modify their functional traits in response to them, is a challenge of wetland ecology. We evaluated the effects of different conditions of soil flooding (flooded and mesic), sunlight (closed and no canopy) and herbivory (presence and absence) on the survival, growth, and morphological traits of Annona glabra and Pachira aquatica seedlings, two dominant woody species of Neotropical swamps. We had eight experimental treatments with five replicates each. Our D. Infante-Mata El Colegio de la Frontera Sur, Unidad Tapachula, Carretera a Antiguo Aeropuerto km 2.5, 30700 Tapachula, Chiapas, México S. Maza-Villalobos (&) CONACYT - El Colegio de la Frontera Sur, Unidad Tapachula, Carretera a Antiguo Aeropuerto km 2.5, 30700 Tapachula, Chiapas, México e-mail: smazavm@gmail.com 123 Wetlands Ecol Manage Introduction Only those individuals that possess traits that allow them to adapt to the filters imposed by the environment may survive and grow in it (Diamond 1975). These environmental filters vary among different ecosystems. The allocation of photosynthetic carbohydrates is an important determinant in the survival and growth of plants exposed to environmental stress (Poorter and Nagel 2000; Sultan 2000). In wetlands, flooding is the main environmental constraint for plant establishment (Keddy 2000). Flooding creates a hypoxic or anoxic soil environment (Pezeshki 1994; Kozlowski 1997), and some plant species have undergone adaptations that enable them to establish under these extreme conditions (Tiner 2016). Among these adaptations is the development of large root systems, which improve oxygen diffusion from aerated to belowground plant parts (De Oliveira and Joly 2010). Plant growth changes with resource availability, and biomass allocation to above or below ground surface, is determined by resource availability (Grime 2007); therefore, in flooded environments where oxygen is limited, one may expect a higher biomass allocation to roots than to leaf tissue (Kotowski et al. 2010). The availability of sunlight in wetlands also plays an important role for plant survival and growth, probably similar to what happens in tropical rain forests (Keddy 2000). Plants in a shaded environment (e.g., under a closed canopy) tend to have a larger leaf area, higher shoot:root ratio and/or higher height:diameter ratio (i.e. slender stems) than those in a sunny environment (e.g., no canopy), which allows them to reach the upper canopy and capture sunlight (Poorter and Nagel 2000; Kotowski et al. 2010). Another factor that has been reported to affect plant growth and survival in wetlands is herbivory. There is evidence that some wetland plants are highly palatable to herbivores, thus herbivory may cause high levels of mortality among seedlings and juveniles (Barton and Hanley 2013). In general, the herbivory carried out by invertebrates (i.e. insects) can be substantial, and the most frequent damage is foliage consumption (Batzer et al. 2007). The loss of foliar biomass represents a reduction in photosynthetic area, which may reduce growth and survival (Costa et al. 2017; Canelo et al. 2018). In natural conditions the different environmental factors operate simultaneously and may act synergistically on plant survival and 123 growth (Lucas et al. 2013; Gattringer et al. 2018). Knowledge about biomass distribution among leaf, stem, and root tissue is crucial in the assessment of plant responses to environmental stressors (Poorter and Nagel 2000, Niinemets 2010a). Many tropical swamps are periodically flooded (Keddy 2000; Infante 2011), thus there are alternating episodes with mesic and flooded soils that strongly influence plant establishment (Lucas et al. 2013). In seasonally humid climates, mesic and flooded soils occur during the dry and rainy season, respectively. During the rainy season plants grow actively and the canopy closes causing a reduction in sunlight availability at the understory level. However, during the dry season the canopy opens and sunlight availability at the understory level increases (Infante 2004). Also in the rainy season, when the soil is flooded and the canopy is closed, some herbivores are abundant (Tsindi et al. 2016; Wantzen et al. 2016). During the floods it is expected that larger plants will show higher survival probabilities. Variation in certain morphological traits, especially those concerned with absorptive surfaces linked with the acquisition of oxygen and solar radiation, may be crucial in determining their access to these limited resources (Sultan 2000; Grime 2007). Contrastingly, during the dry season when the soils are mesic, the sunlight increases, and herbivores are less abundant (Tsindi et al. 2016; Wantzen et al. 2016); under these conditions it is expected that seedlings will display a higher survival probability, higher growth rates, and a reallocation to their absorptive surface in response to varying resource availability (Sultan 2000; Grime 2007). Considering these seasonal changes in tropical swamps and the potential effect of herbivory on the survival and growth of seedlings and juveniles, we studied the effects of soil flooding, sunlight availability and herbivory on the survival, growth and morphological traits of seedlings of two dominant tropical swamp tree species widely distributed in Mexico: Annona glabra and Pachira aquatica. Tropical swamps dominated by A. glabra, P. aquatica and/ or Pterocarpus officinalis are amongst the most common in the Americas (Moreno-Casasola et al. 2012a). However, the area covered by these tropical forested wetlands has been greatly reduced due to land use change to grasslands for cattle ranching (MorenoCasasola et al. 2012b). Wetlands Ecol Manage The objectives of our study were: (i) to evaluate the effects of soil flooding, sunlight availability and herbivory on the survival and early growth of A. glabra and P. aquatica seedlings; and (ii) to assess how plant morphological traits respond to different levels of these environmental factors. For the first objective, our hypothesis was that both species would have a higher survival, absolute growth rate (AGR), and relative growth rate (RGR) in an environment with mesic soil, no canopy and when herbivores were excluded, rather than in conditions of flooded soil, closed canopy and in the presence of herbivores. For the second objective, we considered sets of related traits and their possible explanatory factors to establish the following hypotheses. As a response to low sunlight availability and in the absence of herbivores, we expected higher values of leaf area ratio (LAR), leaf weight ratio (LWR), and specific leaf area (SLA), compared to conditions of no canopy and herbivory. Under sunlight limitation we expected plants to exhibit higher height:diameter ratio than in no canopy conditions, which would allow them to steadily reach the upper canopy and gain access to direct sunlight. Finally, as a consequence of oxygen limitation, we expected higher root biomass allocation and higher root:shoot ratio in flooded than in mesic soil, which would increase the oxygen absorptive area and counteract the oxygen deficit. Methods Study area This study was carried out in a lowland tropical swamp forest that covers * 2 ha (Castillo-Campos and Medina 2002) surrounding an interdune lake locally called La Laguneta, close to the La Mancha lagoon in Veracruz, Mexico (19°350 4500 N and 96°230 0500 W), which is part of Ramsar site 1336. The annual rainfall at this area is 1300 mm, 80% of which falls from June to September, and mean annual temperature is 25 °C (Gómez-Pompa et al. 1972). Given the highly seasonal rainfall pattern, two marked seasons are distinguished in this tropical swamp: the dry season spans from February to late June, and the rainy season goes from July to January. During the dry season the depth of the water table is - 20 to - 30 cm, and soil redox potential values (Eh) reach 443 ± 19 mV (indicating the presence of oxygen; López and Tolome 2009; Pezeshki and DeLaune 2012), while during the rainy season the water level fluctuates between 20 and 60 cm, and soil Eh is 92.3 ± 5.2 mV, indicating a total lack of oxygen (Infante and Moreno-Casasola 2005). The tropical swamp forest at our study site has two strata, i.e. trees and shrubs. The main tree species are Acrocomia aculeata, A. glabra, Attalea butyraceae, Diospyros digyna, Ficus insipida, Ginoria nudiflora, Inga vera, and P. aquatica. The tree layer is approximately 12 m in height and A. glabra is the most abundant species. Among the shrub species are Acacia riparia, Bravaisia integerrima, Caesalpinia pulcherrima, Cestrum scandens, and Piper auritum. There is also a sparse herbaceous layer dominated by Crinum erubescens, Justicia spicigera, and Pistia stratiotes (Castillo-Campos and Medina 2002). Study species Annona glabra (Annonaceae; also known as pond apple) is a shrub or tree that may reach up to 15 m in height. It produces roundish yellow-green fruits, apple-shaped, up to 15 cm long and 9 cm wide. The seed coat is hard and impermeable, germination is epigeal and cotyledons break the seed cover to emerge above the ground. It is found in tropical swamps, mangroves and along river edges (Orozco and Lot 1976; Novelo 1978; Rico-Gray 1982; Lot and Novelo 1990) mainly in Florida (USA), Mexico, the Caribbean islands and Central America, as well as tropical Africa, Asia, and Australia. It has been reported as an invasive species in Australia (Agriculture and Resource Management Council of Australia and New Zealand, Australian and New Zealand Environment and Conservation Council and Forestry Ministers 2001). Its occurrence on well and poorly drained soils suggests that this species has a broad tolerance to different levels of flooding. A. glabra germinates extensively under parent plants, but after the floods seedlings are rare; this species does not form conspicuous seedling banks. Pachira aquatica (Malvaceae; money tree or Guiana chestnut) is a tree native to Central and South America that reaches up to 18 m in height. It has palmate leaves and a soft greenish cortex. Flowers bare long yellow petals and anthers are hair-like, yellow to orange in color. Seeds are crypto-viviparous, 123 Wetlands Ecol Manage and exhibit hypogeal germination in which the hypocotyl-radicle, which has a well-developed terminal bud, is completely covered by the cotyledons (Infante-Mata et al. 2014). This species has a more limited distribution than A. glabra; it is found in tropical areas that are subject to frequent flooding (Lot and Novelo 1990; Lot 1991; Infante 2004), sometimes mixed with mangroves with freshwater influence (Menéndez 1976). In Mexico it is more abundant than A. glabra (P. Moreno-Casasola, personal observation). In contrast to A. glabra, P. aquatica forms persistent seedling and juvenile banks (i.e. seedlings and juveniles in the understory level that endure over relatively long periods of time) under parent plants and close to waterways; even though their leaves may show considerable herbivore damage, these seedling banks persist. Seed germination and seedling growth Seeds of both species were collected at the study site on July 2002, and germinated to produce seedlings (for more details see Infante and Moreno-Casasola 2005). These seedlings were kept in a greenhouse for 5 to 6 months, the time required for them to develop leaves and to be able to withstand physical damage during the subsequent transplant (see below). The A. glabra seedlings had a mean initial height of 20.95 ± 0.31 cm (all data presented as mean ± standard error) (n = 80) before the initiation of the field experiment, while initial P. aquatica seedling height was 58 ± 0.25 cm (n = 80). Just before the start of the field experiment, ten seedlings per species were randomly harvested to obtain initial plant dry biomass (i.e. mean total dry weight per plant; A. glabra: 0.59 ± 0.05 g, and P. aquatica: 7.19 ± 0.70 g) and leaf dry biomass (i.e. mean dry weight of leaves; A. glabra: 0.25 ± 0.02 g, and P. aquatica: 1.16 ± 0.22 g), which would be used to calculate growth rates and SLA, respectively (see below). The in situ growth experiments described below started in January 2003. Experimental design Seedlings were transplanted to pots (25 cm in diameter 9 30 cm in height), putting two seedlings of the same species per pot and using wetland soil from the site. The seedlings were planted such that the root 123 collar was situated * 20 cm below the pot rim. Because seedling stems were thin at the time of transplanting (A. glabra: 3.19 ± 1.7 mm; n = 80, and P. aquatica: 8.42 ± 0.15 mm; n = 80), the level of burial into the soil (* 10 cm) of the whole root system was sufficient to keep the seedlings in a vertical position. For each species a factorial experimental design was used which included three experimental factors, with two levels each: (1) soil flooding (flooded vs. mesic), (2) sunlight (closed vs. no canopy), and (3) herbivory (presence/no exclusion vs. absence/exclusion). Herbivory exclusion treatments targeted both vertebrates and invertebrates species. The combination of these factors and levels resulted in a total of eight treatments per species (1. Flooded–closed canopy–herbivore presence; 2. Flooded–closed canopy–herbivore absence; 3. Flooded–no canopy– herbivore presence; 4. Flooded–no canopy–herbivore absence; 5. Mesic–closed canopy–herbivore presence; 6. Mesic–closed canopy–herbivore absence; 7. Mesic– no canopy–herbivore presence; 8. Mesic–no canopy– herbivore absence). Each treatment had five replicates (i.e. pots). Thus, for each species there were 40 pots and 80 seedlings. These were established in situ at La Laguneta interdune lake in two plots (50 9 20 m). Because La Laguneta spans both forest and open areas (i.e. no canopy), one plot was situated under closed canopy and another one in no canopy conditions. La Laguneta has a water depth that varies from 1 to 2 m, in the dry and rainy season, respectively (Vázquez and Legaria-Moreno 2006). For the flooded soil treatment, in each plot 20 pots were completely submerged in the water (35 cm from the pot base to the water surface). To maintain the flooding treatments (i.e. the root collar situated * 25 cm below the water), pots were placed on wooden platforms that were fixed to structures made of PVC pipes (5 cm in diameter) secured with wire. The height of the platforms was adjusted as the water level changed naturally with the yearly cycle. For the mesic soil treatment, in each plot 20 pots were placed on the edges of the lake, submerging only the first 5 cm of the pot base (for more details see Infante 2004). The soil redox potential (Eh) was measured in March, following the method proposed by López and Tolome (2009). The soil redox potential in the flooded treatment was - 40.93 ± 22.16 mV (n = 6 pots) and in the mesic soil treatment it was 147.87 ± 24.41 mV Wetlands Ecol Manage (n = 8 pots). For these measurements, pots were systematically selected in order to span the spatial heterogeneity; the number of pots selected in each plot was equally distributed among study species. With respect to the sunlight treatments, two sunlight levels were used which corresponded to the two established plots (see above): one under closed canopy forest composed mainly of A. glabra trees (i.e. shaded condition) and the other under no canopy (i.e. direct sunlight condition). Solar radiation was measured in five randomly selected subplots within each plot on the same clear day in March 2003, from 11:00 to 14:00 h, using sensors LICOR-LI 190SA connected to a Data Logger (LICOR-1000). Mean solar radiation in the closed canopy plot was 75.6 ± 9.88 lmol seg-1m2 (n = 5); and in the no canopy plot it was 840.3 ± 87.56 lmol seg-1m2 (n = 5). Within each plot, 40 experimental pots were randomly allocated to the different flooding (n = 20) and herbivory (n = 20) treatments. The herbivore exclusion treatment consisted in protecting the seedlings with a mosquito mesh attached to the pots (150 cm in height, 30 cm in diameter, 1 mm pore size). The mesh was supported by thin metal rods to prevent it from interfering with seedling growth. The mesh pore size excluded vertebrate and invertebrate herbivores. Seedlings lacking protection by the mosquito mesh were exposed to herbivory. To evaluate the effects of flooding, sunlight, and herbivory on survival and growth, seedlings were left to develop and grow under the different treatments for 4 months (from January to April 2003: dry season months). Growth variables were derived from plant measures obtained twice: once at planting (January 2003) and once at the final harvest. For each surviving plant we registered its height (from the soil level to the uppermost meristem), and stem diameter (at the soil level). In April 2003 all surviving seedlings were harvested to obtain the growth rates and the morphological traits described in Table 1. Immediately after harvesting, seedlings were separated in their different components (i.e. roots, stems and leaves). Total leaf area per seedling was obtained (LI-COR leaf-area meter 3100). Seedlings were oven-dried at 80 °C for 48 h (Imperial V800) and separate components were weighed (OHAUS scale CT200). Seedling survival per species and treatment was simply the number of surviving plants after the 4 month study period (120 days). For each seedling the absolute growth rate [AGR = (final dry biomass– initial dry biomass)/(120 days)], and the relative growth rate [RGR = (ln final dry biomass–ln initial dry biomass)/(120 days)] were calculated. In these cases, the initial dry biomass was the average dry weight of the ten seedlings harvested per species at the beginning of the experiment. The value of each response variable (number of survivors, AGR, RGR, root biomass, root:shoot ratio, height:diameter ratio, LAR, LWR, and SLA) was the mean of the two plants per pot. Data analysis For the first objective, we used a complete Generalized Linear Model (GLM) to evaluate the effect of the three independent variables (soil flooding, sunlight, and herbivory) and their interactions, on survival (Poisson distribution, log link function). A Multivariate Analysis of Variance (MANOVA) was carried out to test the effect of the experimental factors on the dependent variables. To evaluate the effect of the three experimental factors on AGR and RGR, we created a multivariate response variable by binding together the two relevant dependent variable (AGR, RGR) and applied the MANOVA to it. For the second objective, to assess the effect of the environment on the morphological plant traits of the studied species, we evaluated how the latter responded to the experimental factors as follows: (a) we used a two-way MANOVA to assess the effects of sunlight and herbivory on a multivariate response variable compounded by the three foliar dependent variables (LAR, LWR, SLA); (b) we carried out a Student t test to evaluate the effect of sunlight on the height:diameter ratio; (c) we applied an one-way MANOVA to evaluate the effect of soil flooding on a multivariate response variable compounded by root biomass and root:shoot ratio. For the multifactorial and two-way MANOVA, we calculated the type-III or type-II sum of squares when the interactions between factors were or were not statistically significant, respectively. The statistic used in the MANOVAs was the Pillai’s trace test. We calculated Eta-squared to estimate the proportion of the variance explained by the different experimental factors. When the response variables did not fit a normal distribution, they were transformed. All 123 Wetlands Ecol Manage Table 1 Description of the response variables evaluated along with their abbreviations and units of measurement Variable Survival Number of seedlings surviving Root biomass Root dry biomass (g) Root:shoot ratio Root dry biomass/shoot dry biomass Height:diameter ratio Plant height/stem diameter Leaf area ratio (LAR) Leaf area/shoot dry biomass (cm2 g-1) Leaf weight ratio (LWR) Leaf dry biomass/shoot dry biomass (g g-1) Specific leaf area (SLA) Leaf area/leaf dry biomass (cm2 g-1) statistical analyses were conducted with R 3.4.3 software, and the significance level for all statistical analysis was p B 0.05. Results The survival of A. glabra and P. aquatica over a 4 month experiment was high (100 and 99%, respectively). The environmental factors (flooding, sunlight, and herbivory) and the interaction between them did not have a significant effect on the number of survivors for A. glabra (v2 = 0.000, d.f. = 1, p = 1) or P. aquatica (v2 = 0.013, d.f. = 1, p = 0.91). The MANOVA model showed that flooding (Pillai1 = 0.35, F2 = 8.04, p = 0.001), sunlight (Pillai1= 0.65, F2 = 27.27, p = 2.1e-7), herbivory (Pillai1 = 0.39, F2 = 9.62, p = 0.0006), and the interaction flooding 9 sunlight (Pillai1 = 0.26, F2 = 5.2, p = 0.01) had significant effects on the growth (AGR and RGR) of A. glabra. The growth rates of A. glabra were higher in an environment with no canopy and mesic soil (Fig. 1a). For P. aquatica, the growth rates were also affected by sunlight (Pillai1 = 0.65, F2 = 28.94, p = 8.1e-8), herbivory (Pillai1 = 0.80, F2 = 62.4, p = 1.35e-11), and the interaction flooding 9 sunlight (Pillai1 = 0.32, F2 = 7.31, p = 0.002). The highest growth rates for P. aquatica were in an environment with no canopy and herbivore exclusion regardless of flooding condition. However, the growth rates of P. aquatica were influenced by flooding only under no canopy but not under closed canopy. In sites with no canopy, the growth of P. aquatica was higher in mesic than in flooded soils (Fig. 1b). For both A. glabra and P. aquatica, the growth rates were higher in the herbivore exclusion treatment, regardless of sunlight level and flooding (Fig. 1). 123 Description Sunlight, herbivory, and the interactions flooding 9 herbivory and flooding 9 sunlight 9 herbivory had significant effects on the foliar traits (LAR, LWR, and SLA) of A. glabra (Table 2a). In general, the foliar trait values for A. glabra were higher in environments with mesic soil, closed canopy, and herbivore exclusion, so their leaves were broader and thinner under these conditions (Fig. 2a). For P. aquatica, all foliar traits displayed higher values in closed canopy regardless of herbivory and flooding. Under the herbivore exclusion treatments, regardless of canopy and flooding, the foliar traits of P. aquatica were higher than when herbivores where not excluded (Table 2b; Fig. 2b). The height:diameter ratio of both A. glabra and P. aquatica was affected by sunlight (t = - 5.36, d.f. = 36, p = 4.9 e-6; and t = - 5.45, d.f. = 38, p = 3.1 e-6, respectively). The height:diameter ratio was higher in the closed than in the no canopy treatments for A. glabra (mean values: 61.32 and 37.14, respectively), and for P. aquatica (60.21 and 46.73, respectively). That is, in closed canopy conditions seedlings were taller and had more slender stems compared to those grown in no canopy conditions. The root biomass and root:shoot ratio changed in response to the flooding level in both A. glabra (Pillai1 = 0.15, F2 = 3.14, p = 0.05; Fig. 3a) and P. aquatica (Pillai1 = 0.20, F2 = 4.9, p = 0.01; Fig. 3b). The root biomass was higher in mesic than in flooded soils in both species (A. glabra: 1.76 ± 0.35 g vs. 0.85 ± 0.16 g; and P. aquatica: 5.34 ± 0.79 g vs. 3.50 ± 0.35 g). A similar pattern was observed for the root:shoot ratio, which displayed higher values in mesic than in flooded soils in both species (A. glabra: 0.56 ± 0.04 vs. 0.46 ± 0.05; and P. aquatica: 0.41 ± 0.02 vs. 0.31 ± 0.02). Wetlands Ecol Manage Fig. 1 Effects of flooding, sunlight and herbivory on the absolute growth rate and relative growth rate of a A. glabra and b P. aquatica. Each line represents the different factors and interactions. Ellipses indicate the errors (p B 0.05). Lines that extend outside the error ellipse indicate that the factor or interaction had a significant effect on the growth rates. See text for a description of the levels of each factor Discussion Soil flooding, sunlight, and herbivory influenced the growth of A. glabra and P. aquatica; however, these environmental factors did not affect their survival, which was very high in our experiment. These species occupy flooded forested environments; yet, their growth rates tended to be higher in environments with mesic soil and in no canopy conditions. Similar results have been found in other tropical and temperate wetland species (Battaglia et al. 2000; Anderson et al. 2009; Li et al. 2011; Lucas et al. 2013). Conservative growth in flooded soils may allow plants to preserve resources under stressful conditions, such as the anoxic environment characteristic of wetlands and swamps. Moreover, high growth rates in mesic soils may suggest that at least some wetland species tolerate rather than require high flood levels. These species depend to a certain extent on the periodic occurrence of mesic conditions to grow, while enduring the flooding periods. Indeed seedlings and saplings of some temperate wetland species, such as Taxodium distichum and Salix nigra, require low-water periods for their regeneration (Day et al. 2006). Regarding sunlight, evidently no canopy conditions are associated with high sunlight availability, which in combination with mesic soil, would predictably result in higher growth rates. This was indeed the case for our study species. The results for A. glabra and P. aquatica indicated that sunlight explained more variance in growth rates than flooding, as in the study of Lenssen et al. (2003), but in contrast with the results 123 Wetlands Ecol Manage Table 2 Results of the MANOVA that show the effects of flooding, sunlight, herbivory and the interactions (9) on the foliar trait* of A. glabra and P. aquatica *Foliar trait is a multivariate response variable obtained by binding together: leaf area ratio, leaf weight ratio and specific leaf area. The significance probability value is p B 0.05 (a) Annona glabra Pillai’s trace Aprox. F d.f. p \ 2.2e–16 Intercept 1 0.994 1612.25 3 Flooding 1 0.117 1.24 3 0.314 Sunlight 1 0.547 3 4.96e–05 11.3 Herbivory 1 0.353 5.09 3 0.006 Flooding 9 sunlight 1 0.096 1.00 3 0.408 0.0369 Flooding 9 herbivory 1 0.258 3.24 3 Sunlight 9 herbivory 1 0.204 2.39 3 0.090 Flooding 9 sunlight 9 herbivory 1 0.257 3.22 3 0.037 (b) Pachira aquatica d.f. Pillai’s trace Aprox. F d.f. p Intercept 1 0.992 1201.58 3 \ 2.2e–16 Flooding 1 0.115 1.30 3 0.292 Sunlight 1 0.308 4.46 3 0.010 Herbivory 1 0.425 7.39 3 0.0007 Flooding 9 sunlight Flooding 9 herbivory 1 1 0.086 0.145 0.94 1.69 3 3 0.434 0.190 Sunlight 9 herbivory 1 0.131 1.50 3 0.233 Flooding 9 sunlight 9 herbivory 1 0.079 0.86 3 0.470 of Kotowski et al. (2010), where flooding was a stronger filter than sunlight. However, it is important to consider that in our results the effect of sunlight observed in both species interacted with flooding. The flooding 9 sunlight interaction amplified the difference in growth rate between plants growing in closed and no canopy conditions. In flooded soils growth rates were similar in the closed and no canopy plots; however, in mesic soils plants grew more rapidly in no canopy than in closed canopy conditions. Mesic soils are not oxygen deprived, as is the case with flooded soils, and thus favor plant growth compared with the latter. Similar effects of the interaction between flooding and sunlight have been observed in other wetland species such as Lindera melissifolia (Lockhart et al. 2018). In flooded soils stomata are more prone to closing; thus, limiting CO2 acquisition and eventually leading to reduced growth rates (Pezeshki 1994; Pezeshki and DeLaune 2012). In general, our results confirm that flooding and sunlight are crucial for seedling growth in these two wetland species (Lucas et al. 2013). This, together with their high levels of tolerance to different environmental conditions, which reflect in their high survival rates, would explain their success as dominant species in tropical swamps. A. glabra and P. aquatica 123 d.f. are able to colonize open canopy areas, both within canopy gaps and on swamps edges, which have proved to be important features for the success of ecological restoration programs (Sánchez 2018). On the whole, foliar traits reached higher values in the closed canopy plot and when herbivores were excluded. In shaded environments, a greater investment in leaf biomass (i.e. high LWR) and an efficient leaf tissue production (i.e. high LAR, and high SLA) allow plants to effectively capture and use the limited sunlight resource. The ability of seedlings to respond to variation in the light environment through adjustments in these morphological traits is critical during the early developmental stages, when seedlings and saplings are easily shaded by neighboring plants due to their small size, even when they are growing in relatively open conditions or without canopy. These responses are common among plants of different ecosystems, including terrestrial and wetland environments (Menges and Waller 1983; Hall and Harcombe 1998; Lucas et al. 2013; Kitajima et al. 2013; Baird et al. 2017; Cisneros-Silva et al. 2017). Additionally, when there is high sunlight availability, biomass allocation to leaf tissues follows a different pattern which generally results in reduced water loss by transpiration. A general strategy in this respect, which Wetlands Ecol Manage Fig. 2 Values of leaf area ratio (LAR), leaf weight ratio (LWR) and specific leaf area (SLA) for A. glabra and P. aquatica in different conditions of soil flooding, sunlight, and herbivory. The columns are mean values and the vertical lines are one standard error coincides with what we observed in this study, is a proportional reduction in foliar area (i.e. LAR, SLA; a relative decrement in transpiration area). Although we did not monitor leaf temperature and transpiration, some studies have demonstrated that smaller leaves have a reduced boundary layer, which promotes a high cooling rate and decreases transpiration (Boardman 1977; Packham and Willis 1982; Cunningham et al. 1999; Cornelissen et al. 2003; De la Barrera and Smith 2009). As previously discussed, flooded soils limit oxygen uptake reducing photosynthate production for tissue building (Pezeshki 1994; Pezeshki and DeLaune 2012). Coupled with limited sunlight (i.e. closed canopy conditions) and the presence of herbivores, soil flooding may have intensified the foliar biomass reduction observed in A. glabra, as has been reported in other species, such as Eugenia uniflora, Lindera melissifolia, Mimosa pigra, and Paspalum dilatatum (Striker et al. 2008; Mielke and Schaffer 2010; NurZhafarina and Asyraf 2017; Lockhart et al. 2018). Herbivory had important effects on the growth rates and foliar traits of A. glabra and P. aquatica; both response variables showed higher values when herbivores were excluded. However, the variance explained by herbivory was higher in P. aquatica than in A. glabra. This difference between species may have been influenced by their contrasting phenological patterns, as well as by the herbivores’ feeding preferences. During the time of our field experiment, the adults of P. aquatica at our study site were leafless (Infante 2004). Therefore, presumably the herbivores of this species consumed the biomass of the available—experimental—seedlings. Contrastingly, the adults of A. glabra did have leaves (Infante 2004), which represented an important food source for herbivores. Perhaps this was why the impact of 123 Wetlands Ecol Manage Fig. 3 Effects of flooding on root biomass and root:shoot ratio for a A. glabra and b P. aquatica. The black line indicates the flooding experimental factor (its levels are described in the text) and the ellipse represents the error (p B 0.05). Lines that extend outside the error ellipse indicate a significant effect of flooding on root biomass or root:shoot ratio. The plotted values of root biomass and root:shoot ratio are log values herbivores on the studied seedlings of A. glabra was not as relevant as for P. aquatica. Another growth pattern associated with shaded environments is the occurrence of relatively slender stems, as we predicted and observed in our results. Both species displayed higher height:diameter ratio values under closed compared to no canopy conditions, i.e. plants growing under low sunlight levels elongated at a faster rate. This is an architectural feature that allows plants to quickly reach the higher canopy and increase the acquisition of sunlight (Niinemets 2010b). This response, which was more evident in P. aquatica seedlings, may allow them to maintain their leaves above the water level reached during the rainy season, and thus form seedling banks. 123 In general, it is expected that flooding will promote high root biomass and root:shoot ratios (Tiner 2016), but our results pointed in the opposite direction. Apparently, oxygen limitation resulting from flooded conditions, restricted root development. Similar results in relation to a reduction in root biomass in flooded soils have been reported in other studies (Lopez and Kursar 1999; Pisicchio et al. 2010; Pezeshki and DeLaune 2012; Oliveira et al. 2015; Steven and Gaddis 2017). Additionally, Vartapetian et al. (2003) have suggested that this reduction in root growth may allow the plant to conserve energy and maintain an optimal metabolism under conditions of oxygen limitation. In connection with this, it is worth mentioning that the seedlings of both study species developed stem lenticels, and P. aquatica also grew Wetlands Ecol Manage adventitious roots. These traits are known to improve oxygen diffusion throughout the root system and also occur in other wetland species. These traits are generally interpreted as adaptations that allow plant survival under high levels of waterlogging (Sena and Kozlowski 1980; Moss 2010; Oliveira et al. 2015). Conclusions Soil flooding, sunlight, and herbivory played important roles on the growth of A. glabra and P. aquatica seedlings, but these factors did not affect their survival. Overall, the most suitable environment for the growth of A. glabra and P. aquatica seedlings was the no canopy plot with mesic soils and herbivore exclusion. As a response to limited sunlight, seedlings of A. glabra and P. aquatica developed broader and thinner leaves, as well as more slender and taller stems. Under oxygen limitation in the flooded soils, both species restricted their root biomass. The results of this work suggest that further studies on the environmental factors that affect the growth and establishment of tropical swamp tree species, and therefore community composition and dynamics, will be a fruitful field of research that will offer valuable insights into the functioning of tropical wetland ecosystems. Further studies with other species and a wider array of environmental factors are required for a better understanding of wetland ecosystems. Acknowledgements The authors thank Guillermo Angeles and Marı́a Luisa Martı́nez for useful comments on the manuscript. We would like to acknowledge the support of C. Madero, V. del Castillo and M. Arias during fieldwork. This study was funded by SEMARNAT-2002-C01-0190, the Canadian International Development Agency-University of Waterloo S-061870, and the Instituto de Ecologı́a, A.C. (90217). We are also grateful to CONACYT (#164467) for support awarded to the Dulce Infante-Mata. References Agriculture and Resource Management Council of Australia and New Zealand, Australian and New Zealand Environment and Conservation Council and Forestry Ministers (2001) Weeds of national significance pond apple (Annona glabra) strategic plan. National Weeds Strategy Executive Committee, Launceston Anderson JT, Landi AA, Marks PL (2009) Limited flooding tolerance of juveniles restricts the distribution of adults in an understory shrub (Itea virginica; Iteaceae). Am J Bot 96:1603–1611. https://doi.org/10.3732/ajb.0800366 Baird SA, Anderegg LDL, Lacely M, HilleRisLambers J, van Volkenburgh E (2017) Comparative leaf growth strategies in response to low-water and low-light availability: variation in leaf physiology underlies variation in leaf mass per area in Populus tremuloides. Tree Physiol 37(9):1140–1150 Barton KE, Hanley ME (2013) Seedling-herbivore interactions: insights into plant defense and regeneration patterns. Ann Bot 112(4):643–650 Battaglia LL, Foré SA, Sharitz RR (2000) Seedling emergence, survival and size in relation to light and water availability in two bottomland hardwood species. J Ecol 88:1041–1050 Batzer DP, Cooper R, Wissinger SA (2007) Wetland animal ecology. In: Batzer D (ed) Ecology of freshwater and estuarine wetlands. University of California Press, Berkeley, pp 242–284 Boardman NK (1977) Comparative photosynthesis of sun and shade plants. Annu Rev Plant Physiol 28:355–377 Canelo T, Gaytán Á, González-Bornay G, Bonal R (2018) Seed loss before seed predation: experimental evidence of the negative effects of leaf feeding insects on acorn production. Integr Zool 13:238–250. https://doi.org/10.1111/ 1749-4877.12292 Castillo-Campos G, Medina AME (2002) Árboles y arbustos de la Reserva Natural de La Mancha, Veracruz. Instituto de Ecologı́a, A. C, México Cisneros-Silva A, Castillo G, Chávez-Pesqueira M, Bello-Bedoy R, Camargo DI, Nuñez-Farfán J (2017) Light limitation reduces tolerance to leaf damage in Datura stramonium. Evol Ecol Res 18:351–362 Cornelissen JHC, Lavorel S, Garnier E, Dı́az S, Buchmann N, Gurvich DE, Reich PB, ter Steege H, Morgan HD, Van Der Heijden GA, Pausas JG, Poorter H (2003) A handbook of protocols for standardized and easy measurement of plant functional traits worldwide. Aust J Bot 51:335–380 Costa AN, Vasconcelos HL, Bruna EM (2017) Biotic drivers of seedling establishment in Neotropical savannas: selective granivory and seedling herbivory by leaf-cutter ants as an ecological filter. J Ecol 105:132–141. https://doi.org/10. 1111/1365-2745.12656 Cunningham AS, Summerhayes B, Westoby M (1999) Evolutionary divergences in leaf structure and chemistry, comparing rainfall and soil nutrient gradients. Ecol Monogr 69:569–588 Day RH, Doyle TW, Draugelis-Dale RO (2006) Interactive effects of substrate, hydroperiod, and nutrients on seedling growth of Salix nigra and Taxodium distichum. Environ Exp Bot 55:163–174 De la Barrera E, Smith WK (2009) Perspectives in biophysical plant ecophysiology: a tribute to Park S. Nobel. Universidad Nacional Autónoma de México, México De Oliveira VC, Joly CA (2010) Flooding tolerance of Calophyllum brasiliense Camb. (Clusiaceae): morphological, physiological and growth responses. Trees 24:185–193 Diamond JM (1975) Assembly of species communities. In: Diamond JM, Cody ML (eds) Ecology and evolution of communities. Harvard University Press, Boston, pp 342–344 123 Wetlands Ecol Manage Gattringer JP, Ludewig K, Harvolk-Schöning S et al (2018) Interaction between depth and duration matters: flooding tolerance of 12 floodplain meadow species. Plant Ecol 219:973–984. https://doi.org/10.1007/s11258-018-0850-2 Gómez-Pompa A, Lot A, Vázquez Yanes C, Soto M, Diego N (1972) Estudio preliminar de la vegetación y la flora de la región de Laguna Verde. Instituto de Biologı́a, Universidad Nacional Autónoma de México, México, Veracruz Grime JP (2007) The scale-precision trade-off in spatial resource foraging by plants: restoring perspective. Ann Bot 99:1017–1021. https://doi.org/10.1093/aob/mcm026 Hall RBW, Harcombe PA (1998) Flooding alters apparent position of floodplain seedlings on a light gradient. Ecology 79(3):847–855 Infante MD (2004) Germinación y establecimiento de Annona glabra (Annonaceae) y Pachira aquatica (Bombacaeae) en Humedales, La Mancha, Actopan, Ver. M.Sc. Dissertation. Instituto de Ecologı́a A.C., Xalapa, México Infante MD (2011) Estructura y dinámica de las selvas inundables de la planicie costera central del Golfo de México. Ph.D. Dissertation. Instituto de Ecologı́a A.C., Xalapa, México Infante MD, Moreno-Casasola P (2005) Effect of in situ storage, light, and moisture on the germination of two wetland tropical trees. Aquat Bot 83:206–218 Infante-Mata D, Moreno-Casasola P, Madero-Vega C (2014) >Pachira aquatica, un indicador del lı́mite del manglar? Rev Mex Biodivers 85:143–160. https://doi.org/10.7550/ rmb.32656 Keddy PA (2000) Wetland ecology: principles and conservation. Cambridge University Press, New York Kitajima K, Cordero RA, Wright SJ (2013) Leaf life span spectrum of tropical woody seedlings: effects of light and ontogeny and consequences for survival. Ann Bot 112(4):685–699 Kotowski W, Beauchard O, Opdekamp W et al (2010) Waterlogging and canopy interact to control species recruitment in floodplains. Funct Ecol 24:918–926. https://doi.org/10. 1111/j.1365-2435.2009.01682.x Kozlowski TT (1997) Responses of woody plants to flooding and salinity. Tree Physiol 17:490. https://doi.org/10.1093/ treephys/17.7.490 Lenssen JPM, Menting FBJ, der Putten WHV (2003) Plant responses to simultaneous stress of waterlogging and shade: amplified or hierarchical effects? New Phytol 157:281–290. https://doi.org/10.1046/j.1469-8137.2003. 00666.x Li F, Li Y, Qin H, Xie Y (2011) Plant distribution can be reflected by the different growth and morphological responses to water level and shade in two emergent macrophyte seedlings in the Sanjiang Plain. Aquat Ecol 45:89–97. https://doi.org/10.1007/s10452-010-9334-8 Lockhart BR, Gardiner ES, Leininger TD et al (2018) Biomass accumulation in the endangered shrub Lindera melissifolia as affected by gradients of light availability and soil flooding. For Sci 64:631–640. https://doi.org/10.1093/ forsci/fxy024 Lopez OR, Kursar TA (1999) Flood tolerance of four tropical tree species. Tree Physiol 19:925–932 López RH, Tolome RJ (2009) Medición del potencial redox del suelo y construcción de electrodos de platino. In: Moreno- 123 Casasola BP, Warner BG (eds) Breviario para describir, observar y manejar humedales. Serie Costa Sustentable No. 1. RAMSAR, Instituto de Ecologı́a, A.C., CONANP, US Fish and Wildlife Service. US State Department, Press, Xalapa, pp 131–138 Lot HA (1991) Vegetación y flora vascular acuática del estado de Veracruz. Ph.D. Dissertation, Facultad de Ciencias, Universidad Autónoma de México, México, D.F., México Lot HA, Novelo A (1990) Forested wetlands of Mexico. In: Lugo AE, Brinson M, Brown S (eds) Forest wetlands, encyclopedia ecosystems of the world, vol 15. Elsevier Press, Amsterdam, pp 287–298 Lucas CM, Bruna ME, Nascimento NMC (2013) Seedling cotolerance of multiple stressors in a disturbed tropical floodplain forest. Ecosphere 4:1–20 Menéndez LF (1976) Los manglares de la Laguna de Sontecomapan, Los Tuxtlas, Veracruz, Estudio florı́stico-ecológico. B. Sc. Dissertation, Facultad de Ciencias, Universidad Autónoma de México, Mexico, D.F., México Menges SE, Waller MD (1983) Plant strategies in relation to elevation and light in floodplain herbs. Am Nat 122(4):454–473 Mielke MS, Schaffer B (2010) Photosynthetic and growth responses of Eugenia uniflora L. seedlings to soil flooding and light intensity. Environ Exp Bot 68:113–121. https:// doi.org/10.1016/j.envexpbot.2009.11.007 Moreno-Casasola P, Infante MD, López-Rosas H (2012a) Tropical freshwater swamps and marshes. In: Batzer DP, Baldwin AH (eds) Wetland habitats of North America: ecology and conservation concerns. University of California Press, Berkeley, pp 267–282 Moreno-Casasola P, Rosas HL, Rodrı́guez-Medina K (2012b) From tropical wetlands to pastures on the coast of the gulf of Mexico. Pastos 42:185–217 Moss B (2010) Ecology of freshwaters. A view for the twentyfirst century. Wiley-Blackwell, London Niinemets Ü (2010a) Responses of forest trees to single and multiple environmental stresses from seedlings to mature plants: past stress history, stress interactions, tolerance and acclimation. For Ecol Manag 260:1623–1639. https://doi. org/10.1016/j.foreco.2010.07.054 Niinemets Ü (2010b) A review of light interception in plant stands from leaf to canopy in different plant functional types and in species with varying shade tolerance. Ecol Res 25(4):693–714 Novelo RA (1978) La vegetación de la Estación Biológica el Morro de la Mancha, Veracruz. Biotica 3:9–23 NurZhafarina A, Asyraf M (2017) Effects of biotic and abiotic environmental stimuli on the morphology and biomass allocation of Mimosa pigra L. Sains Malays 46(8):1241–1248 Oliveira AS, Ferreira CS, Graciano-Ribeiro D, Franco AC (2015) Anatomical and morphological modifications in response to flooding by six cerrado tree species. Acta Bot Bras 29(4):478–488 Orozco SA, Lot HA (1976) La vegetación de las zonas inundables del sureste de Veracruz. Biotica 1:1–44 Packham JR, Willis AJ (1982) The influence of shading and of soil type on the growth of Galeobdolon luteum. J Ecol 70:491–512 Wetlands Ecol Manage Pezeshki SR (1994) Plant response to flooding. In: Wilkinson RE (ed) Plant-environment interactions. Marcel Dekker, New York, pp 289–321 Pezeshki SR, DeLaune RD (2012) Soil oxidation-reduction in wetlands and its impact on plant functioning. Biology 1:196–221 Pisicchio CM, Bianchini E, Pimenta JA, Sert MA, DavasoFabro VM, Medri ME (2010) Heliocarpus popayensis Kunth (Malvaceae) tolera a hipoxia do substrato? Acta Sci 32(2):201–209. https://doi.org/10.4025/actascibiolsci. v32i2.3566 Poorter H, Nagel O (2000) The role of biomass allocation in the growth response of plants to different levels of light, CO2, nutrients and water: a quantitative review. Funct Plant Biol 27:1191. https://doi.org/10.1071/pp99173_co Rico-Gray V (1982) Estudio de la vegetación de la zona costera inundable del noroeste del estado de Campeche, México: Los Petenes. Biotica 7:171–190 Sánchez LOI (2018) Evaluación de los cambios en la estructura y funcionamiento de la selva inundable de la ANP Ciénaga del Fuerte, Tecolutla, Veracruz, durante las acciones de restauración. M.Sc Dissertation. Instituto de Ecologı́a A.C., Xalapa, México Sena GAR, Kozlowski TT (1980) Growth responses and adaptations of Fraxinus pennsylvanica seedlings to flooding. Plant Physiol 66:267–271 Steven JC, Gaddis AL (2017) Drought and simulated deer herbivory reduce growth in Atlantic white cedar seedlings. Botany 95:531–538. https://doi.org/10.1139/cjb-20160207 Striker GG, Insausti P, Grimoldi AA (2008) Flooding effects on plants recovering from defoliation in Paspalum dilatatum and Lotus tenuis. Ann Bot 102:247–254. https://doi.org/10. 1093/aob/mcn083 Sultan SE (2000) Phenotypic plasticity for plant development, function and life history. Trends Plant Sci 5:537–542. https://doi.org/10.1016/S1360-1385(00)01797-0 Tiner RW (2016) Plant indicators of wetlands and their characteristic. In: Tiner RW (ed) Wetland indicators, a guide to wetland formation, identification, delineation, classification, and mapping. CRC Press, Boca Raton, pp 157–240 Tsindi MF, Kupika OL, Moses M, Simbarashe M (2016) Seasonal variation in population structure and status of selected herbivores in the Mana Pools National Park flood plain, Zimbabwe. Poult Fish Wildl Sci. https://doi.org/10.4172/ 2375-446x.1000154 Vartapetian BB, Andreeva IN, Generozova IP, Polyakova LI, Maslova IP, Dolgikh YI, Stepanova AY (2003) Functional electron microscopy in studies of plant response and adaptation to anaerobic stress. Ann Bot 91(2):155–172 Vázquez G, Legaria-Moreno L (2006) Las algas. In: MorenoCasasola P (ed) Entornos veracruzanos: la costa de La Mancha. Instituto de Ecologı́a, A.C., Veracruz, Mexico, pp 247–260 Wantzen KM, Marchese MR, Marques MI, Battirola LD (2016) Invertebrates in neotropical floodplains. In: Batzer D, Boix D (eds) Invertebrates in freshwater wetlands: an international perspective on their ecology. Springer International Publishing, Cham, pp 493–524. https://doi.org/10.1007/ 978-3-319-24978-0_14 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 123 View publication stats