The broad cultivation of non-local species in urban areas or their introduction in natural vegetation, regardless of their origin, may offer a potential risk to local species in many aspects, as affecting the niches and interaction networks    . They are commonly called invasive or alien species. In this context, the biology of alien species received attention in recent studies  . Some seed germination features have been related to the invasive ability in many alien species and were evaluated in some places in Europe  , but there are no similar studies, as far as we know, regarding Neotropical plants.
Studies in the last two decades about seed germination patterns of Cerrado species  may allow some initial comparisons. Some differences in seed ecology between alien or invasive and local or native species were identified focusing on germinability, different light conditions for germination, stratification requirements, dormancy strategies, seedling growth or vigour and other traits      .
Tropical Malvaceae-Bombacoideae and Melastomataceae trees are widely cultivated around the world and some species of these groups present an invasive potential. Melastomataceae is a highly diverse worldwide family, with ca. 4500 species according to Clausing & Renner  , and some species show a high invasion potential around the world. The most famous invasive species are Clidemia hirta D. Don  and Miconia calvescens DC.  , but the list is increasing with Heterotis rotundifolia (Sm.) Jacq.-Fél., which has been considered invasive in many countries  . Recently, species from the Atlantic Forest as Tibouchina granulosa (Desr.) Cogn. and T. pulchra Cogn.  are being widely cultivated in urban areas of the Cerrado domain and also show some invasive potential. Tibouchina granulosa and T. pulchra are not viewed as alien or invasive species, but their cultivation may represent threats to local species. They usually occur as native species in the Atlantic Forest, but their habit and exuberant flowering has stimulated their cultivation in Brazilian urban areas  .
Special attention should be given to Heterotis rotundifolia, originated in African savannas  , which occurs from forests to open areas, roadsides and disturbed areas, because it is a fast-growing species, and due to its decumbent herbaceous habit that forms dense mats, it could displace natural vegetation   . Although it has viable seeds, this species has vegetative growth and reproduction and, for this reason, it was included in the Global Compendium of Weeds  , and was listed as invasive and a weed species around the world     . H. rotundifolia in Brazil was introduced in urban areas for cultivation, mainly for its dense mats and attractive flowers. Its present distribution in Brazil is sub-estimated due to the low number of surveys in urban areas, especially for non-arboreal plants, but it occurs in the states of Minas Gerais and Amazonas  .
In Bombacoideae (Malvaceae), some species also have a strong potential for cultivation in urban areas, especially due to their flowering and human seed consumption. Two of these species are cosmopolitan trees and cultivated around the world, Bombax ceiba, originated from Asia, and Pachira glabra, which has an uncertain origin. Robyns  suggests that P. glabra has its origin in the Tropical American Forests, with occasional and non-natural occurrence in disturbed and human altered areas of Cerrado. Pachira glabra is an alien in Cerrado areas, but it is commonly cultivated in urban areas inside the Cerrado Biome  .
This paper wants to describe the seed germination or seedling emergence pattern of five species of Bombacoideae and Melastomataceae commonly cultivated in urban areas surrounded by Cerrado, comparing them to local species from the same family and local flora.
2. Material and Methods
2.1. Species Selected and Seed Collection
Seeds of Heterotis rotundifolia, Tibouchina granulosa, T. pulchra (Melastomataceae), Bombax ceiba and Pachira glabra (Bombacoideae-Malvaceae) were collected in urban areas of Uberlândia (48˚18'W and 18˚59'S), Minas Gerais State, Brazil as described in Table 1. The region is included in the Cerrado domain, the Neotropical savanna of Brazil  . The climate of this region is characterized as Aw Megathermic, according to the updated classification of Köppen-Geiger  , a tropical wet climate with a dry winter (April to September) and a wet summer (October to March)  . The voucher specimens were included in the Herbarium Uberlandense (HUFU).
Mature fruits were collected directly from the mother plants and mixed for the analysis. The capsular fruits were kept in paper bags until seed dispersion. All dried seeds were kept in paper bags and stored in plastic pots with silica gel at room temperature (± 25˚C) until sowing. Additionally, for Pachira glabra, we
Table 1. Seed collection details for species of bombacoideae and melastomataceae cultivated in an urban cerrado biome area from Brazil.
Legend: NIS: number of individuals sampled, DSC: date of seed collection, DS: days stored before sowing, RN: replicate number, SNR; seed number by replicate, TSNS: total seed number sampled.
collected seeds on the soil, since the species shows large seeds with barochory. These seeds were collected on the same dates as the seeds collected directly from the trees to test the hypothesis if both seed groups have differences in the germination. We suspected that soil seeds show low physiological quality. See details of each collection in Table 1.
We collected seeds from only one tree of T. pulchra and B. ceiba since the species occurs in low density in the city with a low number of fruiting individuals. Only five alien species, segregated in two families (Melastomataceae and Bombacoideae), were included because other alien species from these two families do not occur in the city. We included only Melastomataceae and Bombacoideae alien species since we had sufficient data for comparative studies. In parallel we conducted specific studies to describe the seed germination patterns in these families.
2.2. Seed Experiments
For seed germination experiments of Melastomataceae species, the seeds were sown over filter paper inside Emanueli chambers, a plastic germination container  ; see photos and details in Mendes-Rodrigues & Oliveira  filled with distilled water (ca. 20 mL). For Bombacoideae species, the seeds were sown on 50 - 60 mL of fine vermiculite (expansion volume of 0.1 m3), inside transparent germination boxes, moistened, as necessary, with distilled water. The replicates from all species were distributed randomly and kept in a seed germination chamber (Model MDG2000, Seedburo Company, USA) under continuous white fluorescent light at temperatures of 25˚C.
For the seedling emergence, seeds of Bombacoideae species were sown ca. 1 cm below surface in multicelled polystyrene germination trays filled with a commercial substrate (Plantmax®, Eucatex Mineral Ltda, São Paulo, Brazil). They were maintained in a greenhouse, in an experimental field in Uberlândia, and moistened when necessary.
The number of seeds germinated was observed daily and the protrusion of any part of the embryo was used as germination criterion. The number of seedlings emerged was also observed daily and the appearance of any part of the seedling above the soil surface was used as emergence criterion. All replicates of each species were treated as a sample to process the descriptive analysis. The original experiments were installed in a conventional experimental design, because they belonged to a larger study for each family germination pattern characterization. The germination and emergence were observed up to two months after the last event (germination or emergence) in all replicates as the criteria to stop the evaluation of the experiment.
For each sample, germinability or emergence, mean germination or emergence time, time of the first and last seed germination or seedling emergence, coefficient of variation of germination or emergence time, uncertainty and synchrony were calculated  . Further details on the applications of these germination measurements can be found in Ranal & Santana  and Ranal et al.  .
2.3. Cerrado Species Comparisons
For this purpose, the data obtained for the species here analysed were compared with other Cerrado species and see species list in Table 1 and Table 2, including Melastomataceae and Bombacoideae. The majority of species included in the comparison were studied by  -  . The data were not estimated but obtained directly from the authors, and the seed germination measurements were recalculated as a single sample (all replicates were summed), without considering the original experimental design. When studies involved seed preparation treatments, we used only the control treatment (seeds moistened only with water, without dormancy breaking or additional treatment) and when they involved different individuals, fruit characters or seed types from the same plant, these data were grouped in a single sample. That was necessary due to differences in populations, individuals or seed classification according to the physiological qualities involved in each study. Working with the original replicates could lead to an error of analysis, since when comparing the species among themselves we would be comparing replicates with different sources of variation (e.g. replicates originated from individuals versus replicates originated from different types of fruit).
Independently of the experimental design, all studies included here were based on the same germination measurements and without time censoring (no experiment was interrupted and all necessary time for seed germination was allowed), making it possible to compare the studies. Besides, we also included only daily evaluated experiments. This criterion of paper selection minimizes the effects of experimental conditions making it possible to compare the species. Some studies were conducted in Uberlândia, in the same or similar conditions that we used in our seed germination experiments (same temperature, light conditions and same chamber model) or in the seedling emergence experiments (same greenhouse and environmental conditions). For seed germination comparisons, 36 species of 18 families of the Cerrado Biome were included. For seedling emergence, comparisons were made with 47 species of 20 families as described in Table 2. Details for each species, sampling and experimental design can be observed in the original papers listed in Table 1 and Table 2. In the present paper, data were used only for descriptive and comparative purposes, without statistical inferences.
As exploratory analysis, we performed Principal Component Analyses (PCA) to compare and group the alien (or cultivated) species with the native Cerrado species. In this analysis, we used germinability, germination or emergence time of the first seed or seedling, mean germination or emergence time, germination or emergence time of the last seed or seedling, coefficient of variation of the germination or emergence time, uncertainty of germination or emergence and synchrony of germination or emergence. The PCA was performed separately for
Table 2. Seed germination measurements for some species of malvaceae and melastomataceae cultivated in an urban area inside Cerrado, the brazilian neotropical savanna, compared with native species from Cerrado.
Legend: G: Germinability, GTFS: germination time of the first seed, MGT: mean germination time, GTLS germination time of the last seed, CVGT coefficient of variation of the germination time:, UG uncertainty of germination:, ZG: synchronization index of germination. The code refers to first and/or second letter from species name, followed by population number if more than one.
germination and emergence. The analyses were performed in R using the vegan and graphics packages  . All other data were presented in a descriptive approach since the data show some limitations and any inference was proposed.
3.1. Alien Species
The time for the first germination among the seven alien Melastomataceae species ranged from 5 to 9 days, and the time for the last germination for the same species ranged from 16 to 219 days as Table 2 demonstre.
Among the Melastomataceae, seeds of Heterotis rotundifolia showed high germinability (95.4%), and the highest values for mean germination time (77.85 days), for the last seed germination (219 days) and uncertainty (6.36 bits), and the lowest value for synchrony (0.0148) in relation to the species compared, characterizing its germination process as slow and asynchronous as described in Table 2. Seed germination was spread from 8 to 219 days, with few peaks of low relative frequency per day (below 4.5%, Figure 1), evidencing strong spreading of germination through time, when compared with the other Melastomataceae. The seed germination of H. rotundifolia was concentrated between 60 and 100 days as illustrated in Figure 1. Seeds of this species showed similar germinability and coefficient of variation of germination time to seeds of Clidemia hirta, but with more spreading through time, higher uncertainty and lower synchrony as illustrated in Figure 2(a) and Figure 3(a). In relation to synchrony and when compared to other Cerrado species, seeds of H. rotundifolia were the second most asynchronous, only behind Guazuma ulmifolia seeds as illustrated in Figure 3(a).
Tibouchina granulosa and T. pulchra seeds showed low germinability (lower 14%) and similar germination time (initial, final and mean) in relation to seeds of other Melastomataceae Cerrado species as illustrated in Figure 2(a) and Figure 3(a), except Heterotis rotundifolia. Seeds of T. pulchra had the highest
Figure 1. Relative frequency of seed germination of some Melastomataceae species as a function of time. The seeds of Tibouchina granulosa and Tibouchina pulchra (Atlantic Forest species) and Heterotis rotundifolia species were collected in urban areas of Uberlândia surrounded by Cerrado, the Neotropical savanna of Brazil.
Figure 2. Principal Component Analyses (PCA) and clustering based on seed germination measurements for some cultivated Bombacoideae-Malvaceae and Melastomataceae in the Cerrado Biome compared with to local species flora: (a) PCA; (b) Clustering; The cultivated species are Bc: Bombax ceiba, Hr: Heterotis rotundifolia, Pa: Pachira aquatica, Tig: Tibouchina granulosa and Tip Tibouchina pulchra.. See species code name and measurements details in Table 2 & Table 3.
(a) (b) (c)
Figure 3. Principal Component Analyses (PCA) for some cultivated Bombacoideae-Malvaceae and Melastomataceae in the Cerrado Biome, compared with local species: (a) PCA based on seed germination measurements for Melastomataceae; (b) PCA based on seed germination measurements for Bombacoideae. (c) PCA based on seed emergence measurements for Bombacoideae. The cultivated species are Bc: Bombax ceiba, Hr: Heterotis rotundifolia, Pa: Pachira aquatica, Tig: Tibouchina granulosa and Tip Tibouchina pulchra. See species code name and measurements details in Table 2.
spreading of germination around the mean time (coefficient of variation of the germination time of 56.04%) in relation to the other Melastomataceae as described in Table 2.
Bombax ceiba and Pachira glabra showed high germinability and percentage of seedling emergence, and similar synchrony in relation to the other Bombacoideae species (Malvaceae) from Cerrado as illustrated in Figure 2, Figure 3(a) ad Figure 3(b). No clear difference in the germination or emergence processes were observed among alien or local species of this group or in relation of the local flora as illustrated in Figure 3(b), Figure 3(c); Figure 4(a) and Figure 4(b). Low physiological quality of Pachira glabra seeds collected from the soil was observed during the emergence process, with only 5% of seedling emergence, which was slower and more asynchronous when compared to seeds collected directly from the trees as described in Table 2 and Table 3.
Figure 4. Principal Component Analyses (PCA) and clustering based on seed emergence measurements for some cultivated Bombacoideae-Malvaceae and Melastomataceae in the Cerrado Biome compared to local species flora: (a) PCA; (b) Clustering; The cultivated species are Bc: Bombax ceiba, Hr: Heterotis rotundifolia, Pa: Pachira aquatica, Tig: Tibouchina granulosa and Tip Tibouchina pulchra. See species code name and measurements details in Table 2 & Table 3.
Table 3. Seedling emergence measurements for species of malvaceae and melastomataceae cultivated in an urban area inside cerrado, the brazilian neotropical savanna, compared with native species from cerrado.
Legend: E: Emergence rate, ETFS: emergence time of the first seed, MET: mean emergence time, ETLS emergence time of the last seed, CVET coefficient of variation of the emergence time: UE uncertainty of emergence:, ZE: synchronization index of emergence. The code refers to first and/or second letter from species name, followed by population number if more than one.
3.2. Cerrado Species
For the seed germination process of Cerrado species, germinability ranged from 5%, for Enterolobium contortisiliquum Kuntze―Population 2, to 100% for Tapirira guianensis Aubl. The germination time for the first seed ranged from 1 day in seven species to 147 days for Aegiphylla sellowianna Cham., while the germination time for the last seed ranged from 3 days for Tapirira guianensis to 512 days for Trema micrantha (L.) Blume. The mean germination time ranged from 2.03 days for Tapirira guianensis to 231.89 days for Trema micrantha, and the coefficient of variation of germination time ranged from 10.53% for Trema micrantha to 151.12% for Plathymenia reticulata Benth. The uncertainty of germination ranged from 0.30 bits for Tapirira guianensis to 6.58 bits for Guazuma ulmifolia Lam.―Population 1, and the synchrony of germination ranged from 0.0042 for Guazuma ulmifolia-Population 1 to 0.9099 for Tapirira guianensis as described in Table 2.
For seedling emergence percentage the values ranged from 3% for Guarea guidonia (L.) Sleumer to 100% in three species (Diteryx alata Vogel, Pouteria torta Radk., Tapirira guianensis Aubl.). The emergence time for the first seedling ranged from 3 days for Anadenanthera colubrina (Vell.) Brenan and A. peregrina Speg. to 100 days for Syagrus oleracea Becc., while the emergence time for the last seedling varied from 10 days for Anadenanthera peregrina to 284 days for Hymenaea courbaril L. The mean emergence time ranged from 4.06 days for Anadenanthera peregrina to 117.40 days for Syagrus oleracea, with the coefficient of variation of emergence time ranging from 7.78% for Alibertia sessilis (Vell.) K. Schum. to 126.33% for Hymenaea courbaril. The uncertainty of emergence ranged from 1.85 bits for Anadenanthera peregrina to 5.24 bits for Ormosia arborea Harms, and the synchrony of emergence ranged from 0.0152 for Enterolobium contortisiliquum to 0.3671 for Anadenanthera peregrina as described in Table 3.
3.3. Comparison between Alien and Native Cerrado Species
Comparing Melastomataceae with other Cerrado species, seeds of Heterotis rotundifolia were similar to Guazuma ulmifolia, Lithraea molleoides and Schefflera morototoni, which formed a clear group with a high mean germination time and seed germination spread through time as illustrated in Figure 2. When Heterotis rotundifolia was compared with other Melastomataceae, it was similar to Clidemia hirta with high germinability and asynchronous seed germination, but the latter species has shorter seed germination time as illustrated in Figure 3(a). Tibouchina granulosa and T. pulchra formed a group with the other Cerrado Melastomataceae, with similar seed germination measurements as illustrated in Figure 3(a).
Bombax ceiba and Pachira glabra were similar to other Cerrado Bombacoideae species (Ceiba speciosa, Eriotheca spp. and Pseudobombax spp.), but with a different pattern in relation to species of other tribes of Cerrado Malvaceae for the germination and emergence processes as illustrated in Figures 2-4. The non-Bombacoideae species of Malvaceae showed a lower germinability and percentage of emergence compared to Bombacoideae, with higher spread of germination in time in relation to Bombacoideae species, probably related to dormancy. When compared to Cerrado native species, the alien Bombacoideae did not show any marked difference as illustrated in Figures 2-4.
The alien Bombacoideae species showed similar seed germination and seedling emergence pattern in relation to native Cerrado species of the same family, with high germinability and percentage of emergence, low germination or emergence time and synchronous processes    . Low germinability and percentage of emergence, with high germination or emergence time and asynchronous germination observed for seeds (or seedlings) of other Malvaceae tribes are associated with seed dormancy   . It is very important to observe some attributes of seeds during the germination or emergence process to consider dormancy, that is, the quantity activated that is expressed by embryo protrusion or seedling emergence through time. This is the velocity of the process, which is measured by means of distribution over time. Dormancy can also be measured by synchrony. Dormant seeds, in general, show low velocity of embryo protrusion or seedling emergence and asynchrony, spreading the process through time. In reality, more important than the mean germination/emergency time is the amplitude between the first and last event, as was demonstrated for diaspores of Lithraea molleoides (Vell.) Eng.  . The total seeds germinated or seedlings emerged frequently are more associated with viability than with dormancy.
For Melastomataceae, we found some differences. The seed germination pattern of Heterotis rotundifolia was similar to Clidemia hirta. The latter is considered an invasive species  , with high germinability, slow and asynchronous seed germination  . Seeds of both species are probably dormant, but the type of dormancy is unknown. The dormancy types and metabolism in Melastomataceae are poorly known, although some studies have showed multiple dormancy  , physiological dormancy  and water impermeability in seed coat  . The spreading of the germination through time, typical of seeds of several Melastomataceae, permits the inclusion of these seeds in the category of relative dormancy as proposed by Labouriau  , but the type of dormancy is unclear because the spreading of the germination through time can be due to physical, morphological, physiological or combination of these causes. Dormancy is also present in other invasive species    , reinforcing the risk related to the cultivation of Heterotis rotundifolia near natural areas.
The germination process for Melastomataceae species usually starts in less than 30 days and is associated with dormancy in many species   because more important than this limit, as was proposed by Baskin and Baskin  to consider seeds non-dormant, is the behaviour of the germination process through time. The spreading of the germination through time, that is, the broad range between the first and last germination or emergence is the most important reference point to consider dormancy and it is valid for species of other families. For example, seeds of Luehea divaricata, Ormosia arborea and Virola sebifera with dormancy, had a mean germination time of 24.1, 25.7 and 24 days, respectively, but with a range of the germination process from 4 to 104, 8 to 86, and 16 to 44 days, respectively  . The same species had a mean emergence time of 62.8, 80.4 and 57.7 days and a range of the emergence process from 29 to 105, 57 to 145 and 46 to 66 days, respectively. During this time, they reached 48.1%, 96% and 96% of germinability and 53, 89 and 15.7 of emergence. Seedlings of Copaifera langsdorffii and Genipa americana reached 95% and 84.5% of emergence with mean a emergence time of 27 and 28.7 days, ranging from 18 to 56 and 17 to 71 days  . All of them had the mean germination or emergence time lower than 30 days and relative dormancy due to the spreading of germination or emergence through time. Seeds of Luehea divaricata have relative dormancy caused by physiological dormancy, of Ormosia arborea have physical and mechanical dormancy, of Virola sebifera have physiological dormancy, and those of Copaifera langsdorffii have physical, chemical and physiological dormancy (see the relative frequency of germination or emergence in Ranal et al.  and more information for C. langsdorffii in Pereira et al.  . It is also important to consider that the low mean germination time of some species is associated with time censoring of the experiment, a particular decision of the researcher but, in these cases, the time measurements are underestimated. Another point is that for some species with dormant seeds that show a broad range between the first and last germination and very low synchrony, as is the case of Enterolobium contortisiliquum, Guazuma ulmifolia, Hymenaea courbaril  , Lithraea molleoides  , Senna sylvestris and Trema micrantha  , the mean germination time does not have the same importance or utility as for those that germinate very closely, even if this occurs after a long time. The relative frequency of germination or emergence and the survival analysis seems to be more appropriate for these species as pointed out by Berger et al.  . The last point is that seeds of some species could take a long time for the first germination, but most of them could germinate synchronously a long time after they were sowed. The seeds were dormant and this condition could be broken at the same time, weeks or months after sowing.
Regardless of the question of criterion, seeds of the Melastomataceae species, for example, specifically for Cerrado, show mean germination time values below 30 days        , reinforcing the importance of the high mean germination time and germination time of the last seed values found here for H. rotundifolia seeds.
The spreading of germination through time is associated with dormancy and this character permits soil seed bank formation for Melastomataceae, as was observed for Clidemia hirta, Miconia chartaceae, Miconia ferruginata, Microlicia fasciculata       and Miconia ferruginata (MA Ranal, personal observation). Seeds of Clidemia hirta when unearthed, show a faster and synchronic germination process than seeds stored in paper bags in laboratory conditions  , and these results represent irrefutable evidence of its ability to survive in soil and form soil seed banks. Thus, this behaviour is an important signal that seeds of Heterotis rotundifolia can persist in seed banks, and in favourable situations can compete with the natural vegetation, although this character needs to be investigated. Seeds of Miconia albicans established in grasslands have slow and asynchronous germination, and this fact could increase seedling survival in grasslands, a more unpredictable environment compared to woodlands  .
Another important factor for the establishment of invasive species is that a larger introduction is more important to survival than small and frequent insertions  . Heterotis rotundifolia could incorporate the two processes with multiple insertions, one of them by means of the relative dormancy of its seeds that spread the germination through time, which represents the small and frequent insertions, and a second type represented by large introductions due to the high germinability of the seeds present in the soil seed banks, when exposed to ideal conditions for germination. It was observed for Leucaena leucocephala (Lam.) de Wit, another invasive species of the Cerrado region, that it also has its invasive capacity related to seed bank formation due to dormancy and high seed production  .
The low germinability of Tibouchina pulchra and T. granulosa seeds is comparable to seeds of Miconia ferruginata  and Microlicia fasciculata  . Miconia ferruginata presents high embryoless seed percentages  , a fact that could be present in the seeds of the two Tibouchina species. Studies have demonstrated that germinability of 27.84% and 23.92%, respectively for seeds of Tibouchina pulchra and T. granulosa  is related to the absence of embryos in their seeds, a common character in Melastomataceae  .
For Tibouchina pulchra and T. granulosa, low germinability could also represent reproductive inadequacies for local conditions, for example, soil quality. The urban area that was studied here has a high plant and animal diversity, and probably has no deficiency in pollination services. As the seed quality and germination pattern in some Cerrado Melastomataceae could be also related to aluminium presence   , the presence of this element in Cerrado soils could be a disadvantage to non-Cerrado species as Tibouchina pulchra and T. granulosa. Heterotis rotundifolia might not be affected by this element taking into account its origin in African savannas, with possibly similar soil conditions. The last important point for some invasive species is that a low germination percentage has been found and that is compensated with other features as seedling vigour  .
Heterotis rotundifolia represents high potential risks to local species, especially because of the high germinability and spreading of germination through time with a not yet known dormancy mechanism. Allied to these factors, this species shows fast growth and vegetative propagation   . Thus, special attention is necessary in monitoring the spread of this species and the possible impacts in the local species of Cerrado.
Other seed germination characters have also been related to the success of alien species. We have the examples of Clausena excavata Burm., which germinates in broad light conditions  , Impatiens glandulifera Royle, which requires a short period of time for seed stratification  , and alien Impatiens species, which have a more successful local adaptation rate compared to local Impatiens species  .
Besides this information, seed germination characters are not generally related to the capacity of invasion, except high germination rates  , probably because of the low number of comparative studies for other metrics related to seed germination. Another point is the absence of studies that compare seed germination patterns of species when they are in their native area and when they are alien. Unfortunately, we do not know the seed germination measurements for the alien species studied here in their natural habitat.
Our data could present some limitations due to the low number of trees evaluated, statistical inference and low number of species and families studied, but they represent the first data for some species, regardless of being in natural or alien areas. These descriptive data could improve information to plan future experimental investigations.
In conclusion, our data show two possible different patterns. Bombacoideae alien species studied here show the same seed pattern of local species, without a distinctive factor in seed germination or seedling emergence. On the other hand, Melastomataceae shows one species with a distinct pattern, Heterotis rotundifolia, which could offer risks to local flora. This species had a similar behaviour of other dormant species of the Cerrado flora.
Although we did not evaluate the entire process of the introduction of alien or cultivated species, the analyses of seed germination and seedling emergence measurements, as the spreading of germination in time, could offer information that is related to the success of invasive species.
We would like to thank the Universidade Federal de Uberlândia; all authors that released the data and Roger Hutchings for the English review. This work was funded by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, Conselho Nacional de Desenvolvimento Científico e Tecnológico and Fundação de Amparo à Pesquisa do Estado de Minas Gerais. Some data of this study are part of the Doctoral thesis of the first author in the Programa de Pós-graduação em Ecologia e Conservação de Recursos Naturais.
 Maruyama, P.K., Vizentin-Bugoni, J., Sonne, J., Martín González, A.M., Schleuning, M., Araujo, A.C., Baquero, A.C., Cardona, J., Cardona, P., Cotton, P.A., Kohler, G., Lara, C., Malucelli, T., Marín-Gómez, O.H., Ollerton, J., Rui, A.M., Timmermann, A., Varassin, I.G., Zanata, T.B., Rahbek, C., Sazima, M. and Dalsgaard, B. (2016) The Integration of Alien Plants in Mutualistic Plant-Hummingbird Networks across the Americas: The Importance of Species Traits and Insularity. Diversity Distribution, 22, 672-681. https://doi.org/10.1111/ddi.12434
 Shaltout, K.H., Hosni, H.A., El-Kady, H.F., El-Beheiry, M.A. and Shaltout, S.K. (2016) Composition and Pattern of Alien Species in the Egyptian Flora. Flora-Morphology, Distribution, Functional Ecology of Plants, 222, 104-110.
 Dawson, W., Moser, D., Kleunen, M.V., Kreft, H., Pergl, J., Pysek P., Weigelt, P., Winter, M., Lenzner, B., Blackburn, T.M., Dyer, E.E., Cassey, P., Scrivens, S.L., Economo, E.P., Guénard, B., Capinha, C., Seebens, H., García-Díaz, P., Nentwig, W., García-Berthou, E., Casal, C., Mandrak, N.E., Fuller, P., Meyer, C. and Essl, F. (2017) Global Hotspots and Correlates of Alien Species Richness across Taxonomic Groups. Nature Ecology & Evolution, 1, 0186.
 Chrobock, T., Kempel, A., Fischer, M. and van Kleunen, M. (2011) Introduction Bias: Cultivated Plant Species Germinate Faster and More Profusely than Native Species in Switzerland. Basic Applied Ecology, 12, 244 250.
 Ranal, M.A., Santana, D.G. and Schiavini, I. (2010) Are There Germination Patterns for Cerrado Species? In: Del-Claro, K., Oliveira, P.S., Rico-Gray, V., Barbosa, A.A.A., Bonet, A., Scarano, F.R.G., Sampaio, F.J.M., Morris, M.V., Ramirez, N.M.R., Mar?al Júnior, O., Macedo, R.H.F., Marquis, R.J., Coelho, L., Martins, R.P., Rodrigues, S.C. and Luttge, U., Eds., Encyclopedia of Life Support Systems. Tropical Biology and Conservation Management, Savanna Ecosystems, UNESCO/EOLSS. Encyclopedia of Life Support Systems, Oxford, 106-159.
 Perglová, I., Pergl, J., Skálová, H., Moravcová, L., Jarosik, V. and Pysek, P. (2009) Differences in Germination and Seedling Establishment of Alien and Native Impatiens Species. Preslia, 81, 357-375.
 Vieira, D.C.M., Socolowski, F. and Takaki, M. (2010) Seed Germination and Seedling Emergence of the Invasive Exotic Species, Clausena excavata. Brazilian Journal of Biology, 70, 1015-1020. https://doi.org/10.1590/S1519-69842010000500014
 Skálová, H., Moravcová, L. and Pysek, P. (2011) Germination Dynamics and Seedling Frost Resistance of Invasive and Native Impatiens Species Reflect Local Climatic Conditions. Perspectives in Plant Ecology Evolution and Systematics, 13, 173-180. https://doi.org/10.1016/j.ppees.2011.03.005
 Marques, A.R., Costa, C.F., Atman, A.P.F. and Garcia, Q.S. (2014) Germination Characteristics and Seedbank of the Alien Species Leucaena leucocephala (Fabaceae) in Brazilian Forest: Ecological Implications. Weed Research, 54, 576-583.
 Sinclair, J.S. and Arnott, S.E. (2016) Strength in Size Not Numbers: Propagule Size More Important than Number in Sexually Reproducing Populations. Biological Invasions, 18, 497-505. https://doi.org/10.1007/s10530-015-1022-0
 Clausing, G. and Renner, S.S. (2001) Moleular Phylogenetics of Melastomataceae and Memecylaceae: Implications for Character Evolution. American Journal of Botany, 88, 486-498. https://doi.org/10.2307/2657114
 Dewalt, S.J., Denslow, J.S. and Hamrick, J.L. (2004) Biomass Allocation, Growth, and Photosynthesis of Genotypes from Native and Introduced Ranges of the Tropical Shrub Clidemia hirta. Oecologia, 138, 521-531.
 Medeiros, A.C. and Loope, L.L. (1997) Status, Ecology, and Management of the Invasive Plant, Miconia calvescens DC (Melastomataceae) in the Hawaiian Islands. Bishop Museum Occasional Papers, 48, 23-36.
 Randall, R.P. (2012) A Global Compendium of Weeds. Perth, Australia: Department of Agriculture and Food Western Australia, 1124 p.
 Silva, M.F.O., Potsch, A.R.H. and Guimaraes, P.J.F. (2013) Melastomataceae no Parque Estadual da Pedra Branca, Rio de Janeiro, RJ, Brasil. Hoehnea, 40, 679-700.
 Brizola-Bonacina, A.K., Arruda, V.M., Alves-Junior, V.V., Chaud-Netto, J. and Polatto, L.P. (2012) Bee Visitors of Quaresmeira Flowers (Tibouchina granulosa Cogn.) in the Re-gion of Dourados (MS-Brasil). Sociobiology, 59, 1253-1268.
 PIER (2013) Pacific Islands Ecosystems at Risk. HEAR, University of Hawaii, Honolulu. http://www.hear.org/pier/index.html
 Chong, K.Y., Tan, H.T.W. and Corlett, R.T. (2009) A Checklist of the Total Vascular Plant Flora of Singapore: Native, Naturalized and Cultivated Species. Raffles Museum of Biodiversity Research, National University of Singapore, Singapore, 273 p. https://lkcnhm.nus.edu.sg/app/uploads/2017/04/flora_of_singapore_tc.pdf
 Michelangeli, F.A. (2015) Heterotis in lista de espécies da Flora do Brasil. Jardim Botanico do Rio de Janeiro.
 Robyns, A. (1960) Contributional etude monographique du genre Bombax s.l. I. B. glabrum (Pasq.) A. Robyns comb. nov. Bulletin du Jardin Botanique de l’état à Bruxelles, 30, 473-484. https://doi.org/10.2307/3667346
 Duarte, M.C. (2015) Pachira in lista de espécies da Flora do Brasil. Jardim Botanico do Rio de Janeiro. http://floradobrasil.jbrj.gov.br/jabot/floradobrasil/FB25756
 Ribeiro, J.F. and Walter, B.M.T. (2008) As principais fitofisionomias do Bioma Cerrado. In: Sano, S.M., Almeida, S.P. and Ribeiro, J.F., Eds., Cerrado: Ecologia e Flora, Vol. 1, Embrapa Cerrados, Planaltina-DF, 152-212.
 Kottek, M., Grieser, J., Beck, C., Rudolf, B. and Rubel, F. (2006) World Map of the Koppen-Geiger Climate Classification Updated. Meteorologische Zeitschrift, 15, 259-263. https://doi.org/10.1127/0941-2948/2006/0130
 Ranal, M.A. (2003) Soil Spore Bank of Ferns in a Gallery Forest of the Ecological Station of Panga, Uberlandia, MG, Brazil. American Fern Journal, 93, 97-115.
 Ranal, M.A. and Araújo, E.A.P. (2005) Camara úmida para germinacao de sementes e desenvolvimento inicial de plantulas (Camara de Emanueli). Patent Number PI0520543-3 A2, Instituto Nacional da Propriedade Industrial.
 Mendes-Rodrigues, C. and Oliveira, P.E. (2012) Polyembryony in Melastomataceae from Brazilian Cerrado: Multiple Embryos in a Small World. Plant Biology, 14, 845-853. https://doi.org/10.1111/j.1438-8677.2011.00551.x
 Ranal, M.A., Santana, D.G., Ferreira, W.R. and Mendes-Rodrigues, C. (2009) Calculating Germination Measurements and Organizing Spreadsheets. Revista Brasileira de Botanica, 32, 849-855.
 Carvalho, M.P., Santana, D.G. and Ranal, M.A. (2005) Emergência de plantulas de Anacardium humile A. St.-Hil. (Anacardiaceae) avaliada por meio de amostras pequenas. Revista Brasileira de Botanica, 28, 627-633.
 Anastacio, M.R. and Santana, D.G. (2010) Características germinativas de sementes de Ananas ananassoides (Baker) L. B. Sm. (Bromeliaceae) Acta Scientiarum. Biological Sciences, 32, 195-200. https://doi.org/10.4025/actascibiolsci.v32i2.1693
 Anastacio, M.R., Santana, D.G., Camargos, R.C., Babata, M.M. and Oliveira, C.A.A. (2010) Maturacao e qualidade física de frutos na germinacao de pirenios de Schefflera morototoni (Araliaceae). Ciência Floresta, 20, 431-439.
 Mendes-Rodrigues, C., Araújo, F.P., Souza, C.B., Souza, V.B., Ranal, M.A., Santana, D.G. and Oliveira, P.E. (2010) Multiple Dormancy and Maternal Effect on Miconia ferruginata (Melastomataceae) Seed Germination, Serra de Caldas Novas, Goiás, Brazil. Revista Braseira de Botanica, 33, 93-105.
 Mendes-Rodrigues, C., Oliveira, P.E. and Ranal, M.A. (2011) Seed Germination and Seedling Growth of Two Pseudobombax Species (Malvaceae) with Contrasting Habitats from Brazilian Cerrado. Revista de Biologia Tropical, 59, 1915-1925.
 Mendes-Rodrigues, C., Ranal, M.A. and Oliveira, P.E. (2011) Does Polyembryony Reduce Seed Germination and Seedling Development in Eriotheca pubescens (Malvaceae-Bombacoideae)? American Journal of Botany, 98, 1613-1622.
 Luz, J.M.Q., Carvalho, A.F., Santana, D.G. and Silva, M.A.D. (2010) Germina??o de sementes e emergência de plantulas de carapiá: Espécie primitiva e medicinal. Horticultura Brasileira, 28, 107-110.
 Ranal, M.A., Mendes-Rodrigues, C., Teixeira, W.F., Oliveira, A.P. de and Rosana, R. (2016) Seed Germination of Microlicia fasciculata, an Apomictic and Aluminium Accumulator Species: Unexpected Intraspecific Variability in a Restricted Neotropical Savanna Area. Flora, 220, 8-16. https://doi.org/10.1016/j.flora.2016.02.001
 R Development Core Team (2017) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna.
 Berger, A.P.A., Ranal, M.A. and Santana, D.G. (2014) Variabilidade na dormência relativa dos diásporos de Lithraea molleoides (Vell.) Eng. Ciência Florestal, 24, 1-13. https://doi.org/10.5902/1980509814570
 Mendes-Rodrigues, C.M., Ranal, M. and Oliveira, P.E. (2008) Could Seed Dormancy and Polyembryony Explain the Success of Clidemia hirta (L.) D. Don. (Melas- tomataceae)? In: IX Simpósio Nacional Cerrado, II Simpósio Internacional Savanas Tropicais, Embrapa Cerrados, Brasília.
 Silveira, F.A.O., Ribeiro, R.C., Oliveira, D.M.T., Fernandes, G.W. and Lemos-Filho, J.P. (2012) Evolution of Physiological Dormancy Multiple Times in Melastomataceae from Neotropical Montane Vegetation. Seed Science Research, 22, 37-44.
 Ribeiro, R.C., Oliveira, D.M.T. and Silveira, F.A.O. (2015) A New Seed Coat Water-Impermeability Mechanism in Chaetostoma armatum (Melastomataceae): Evolutionary and Biogeographical Implications of Physiophysical Dormancy. Seed Science Research, 25, 194-202. https://doi.org/10.1017/S0960258515000070
 Pereira, R.S., Ranal, M.A. and Santana, D.G. (2009) Emergência de plantulas oriundas de sementes recém-colhidas e armazenadas de Copaifera langsdorffii Desf. (Caesalpinoideae), Triangulo Mineiro, Brasil. Revista árvore, 33, 643-652.
 Carreira, R.S. and Zaidan, L.B.P. (2007) Germinacao de sementes de espécies de Melastomataceae de Cerrado sob condicoescontroladas de luz e temperatura. Hoehnea, 34, 261-269. https://doi.org/10.1590/S2236-89062007000300001
 Silveira, F.A.O., Mafia, P.O., Lemos-Filho, J.P. and Fernandes, G.W. (2012) Species-Specific Outcomes of Avian Gut Passage on Germination of Melasto- mataceae Seeds. Plant Ecology and Evolution, 145, 350-355.
 Sales, N.M., Pérez-García, F. and Silveira, F.A.O. (2013) Consistent Variation in Seed Germination across an Environmental Gradient in a Neotropical Savanna. South African Journal of Botany, 87, 129-133.
 Escobar, E.D.F. and Cardoso, V.J.M. (2015) Longevity of Seeds and Soil Seed Bank of the Cerrado Tree Miconia chartacea (Melastomataceae). Seed Science Research, 25, 386-394. https://doi.org/10.1017/S0960258515000173
 Zaia, J.E. and Takaki, M. (1998) Seed Germination of Pioneer Species of Tibouchina pulchra Cong. and Tibouchina granulosa Cong. (Melastomataceae). Acta Botanica Brasilica, 12, 221-229. https://doi.org/10.1590/S0102-33061998000300004
 Silveira, F.A.O., Fernandes, G.W. and Lemos-Filho, J.P. (2013) Seed and Seedling Ecophysiology of Neotropical Melastomataceae: Implications for Conservation and Restoration of Savan-nas and Rainforests. Annals of the Missouri Botanical Garden, 99, 82-99. https://doi.org/10.3417/2011054
 Foxcroft, L.C., Richardson, D.M., Rejmánek, M. and Py?ek, P. (2010) Alien Plant Invasions in Tropical and Sub-Tropical Savannas: Patterns, Processes and Prospects. Biological Invasions, 12, 3913-3933.