Species vary over space and time according to a myriad of processes, describing variation in communities is now embedded in the metacommunity framework [1
]. In aquatic ecosystems, local limnological variables, hydrological processes, and habitat connectivity are the major determinants of species variation [2
]. On the other hand, species variation can also be generated through stochasticity, for instance, in probabilistic and limited dispersal, which is well described in Hubbell’s Neutral Theory [5
It is now a consensus that the relative importance of mechanisms above mentioned depends on the spatial scale in which communities are studied [6
]. For instance, at large spatial scales, even microorganisms with large dispersal ability can present some dispersal limitation [9
]. At local scales, priority effects, micro-habitat differences, and biological interactions take place as relatively more important mechanisms to explain spatial and temporal differences in communities [11
Diatoms are microalgae commonly used as bioindicators of water quality and environmental wealth of aquatic bodies, given their quick response to environmental variation [22
]. Diatoms, reproduce primarily via asexual division with rare instances of sexual reproduction. Some species may divide once per day in optimal environments. Importantly, diatom valves are typically well preserved in the ponds’ sediments [23
]. It is also reported that dispersal related mechanisms have a role in determining differences of local diatoms communities [24
]. In this sense, describing scales of the spatial variation and main correlations of diatoms community variation is the core to reveal the most important ecosystem processes determining aquatic metacommunities.
In addition to the known temporal variation, there is the spatial variation, floodplains like the Pantanal, have a high environmental heterogeneity in aquatic ecosystems, ranging from microhabitats to large scales [28
], and it include the habitats colonized by species, such as diatoms. Following this line, species can vary within and between such habitats, identifying how variation occurs may reveal the main ecological processes determining diatoms.
Historically, ecologists debate whether it is true that everything is everywhere, but the environment selects [31
] and there is dispersion to isolated spaces (e.g., ponds), with the success of immigration and colonization, with different biogeographic and biodiversity levels, depending primarily on the organism’s ability to disperse [34
]. Immigration of new species varies between scales, ecosystems, and organisms; however, what types of place are occupied by which species? Is very likely that due to the richness and structure of the species both by biogeographical factors as well as by local characteristics and connectivity forming a conundrum of likely underlying causes [35
]. For instance, a watershed is a landscape unit connected through the flow of water among its sources all along its mouth, and can be assessed through their mosaic composition, whether composed by corridors, patches, matrices, or subunits within in watershed for planning, management, and analytical purposes [36
Many are the responses projected with focus on wetlands around the world. However, water resources will not be sustainable or sufficient unless other indirect and direct drivers of change are addressed and do not lead to a reduction in the services provided by wetlands. Wetland studies are extremely important, because wetland do more than act as water filters, providing flood and erosion control, they also sustain plants and animals in the watershed. Studies of the diatom community at local scales add to this importance in wetlands, since first we have to understand how the dynamics of this community is structured in a micro-habitat, and then it can be applied in larger spatial scales. Knowing the dynamics and structure of diatoms, is essential for conserving, maintaining, or rehabilitating wetland ecosystems, and the Pantanal is part La Plata river basin, which represents the second largest surface for water resources in South America and the Guarani Aquifer System, the biggest unified groundwater aquifer in the world [36
In this study we analyzed the structure of the diatom metacommunity in three permanent ponds of the Pantanal of Mato Grosso State in Brazil, to understand how species variation occurs at different scales. We aimed to better understand how biota in these aquatic environments depends on the geographic distribution of water bodies (structural connectivity) and effective movements between species and bodies of water (functional connectivity). Therefore, this study will contribute to answer to the following questions and the concepts in metacommunities that will guide the hypotheses if the premises are true:
Do communities differ more than expected at random levels between local Pantanal habitats (i.e., ponds), or the difference is just a mathematical artifact?
If differences exist, which species characterize these differences? Does the different species composition a consequence of the environmental characteristics of the ponds?
Is there a homogeneous or heterogeneous spatial composition in the ponds?
Which is the scale that most contributes to the total diversity of diatoms: the samples made at each site, the compositional variation within the pond, or the compositional variation between different ponds? Which of these contributions are the greater, and which are less than expected by a null model?
A total of 119 species of diatoms distributed in 31 genera were reported from the qualitative analysis of the sampling of three permanent ponds located in the floodplain of the “Pantanal Mato-Grosso”. In general, the most representative genus in abundance was Aulacoseira Thwaites. The most abundant species was Aulacoseira italica (Ehrenberg) Simonsen. Considering species richness, the genus Eunotia Ehrenberg was the most representative in species, with 39 taxa recorded.
According to the Mennhinick Index, which considers diversity according to the registered numbers, a higher index (2.626) was observed for P1, due to the fact that this pond has the highest number of taxa (83) in relation to P2 (72 taxa) and P3 (78 taxa). Along with Fisher’s index, which assumes that abundance follows the distribution of the log series, it also resulted in a higher index (21,500) for P1 with the highest number of valves per ml (763,060) and for P3 with the lowest number of valves per ml (706,322). The Berger–Parker dominance index, which uses the measure of the numerical importance of the most abundant as a parameter, also generated greater value for P1. However, according to the Simpson and Shannon index, which consider diversity in relation to uniformity, it revealed the greatest diversity for P2, even with the smallest number of species, but with the best distribution of individuals (Table 2
Goal 1: the analyzed metacommunity has coherence and a non-random structure [52
]. Accordingly, ponds differ from each other more than would be expected by a null model, according to PERMANOVA (Figure 2
). Only the classification of ponds explains 24% of the spatial changes in community composition (F = 8.94; R2 = 0.24; p
< 0.001). It is visible in a PCoA diagram that there is little compositional superposition, however, the pond directly associated with the Cuiabá River (P1) is the one that most differs in species composition. It is also the pond that has the greatest numerical richness (83) in relation to the other two ponds.
Goal 2: Ponds also differ among each other for the set of environmental variables, with an explanation power of 27% (PERMANOVA F = 10.51; R2 = 0.27; p
< 0.001). The difference between all ponds is also visible in a PCoA scatterplot, but with a continuous change between ponds 1, 2 and 3 compared to the differences in species composition (Figure 3
, see also Figure 2
The highest concentrations of total nitrogen were observed at P1. The most conspicuous difference in environmental variables occurred between the sampling sites of P3; however, P2 had a high variation in turbidity between the 10 sampling sites (Table 3
In a coherent way, there were several indicator species for each pond. The list of indicator species with a significant indicator value considering a permutation test for the three ponds was longer for P3 with 14 indicator species (Table 4
Goal 3: Ponds also differed in compositional variation (F = 8.78; p
= 0.002), with P2 being the most variable considering the average distance to the centroid of distribution in a PCoA scatterplot, followed by P3 and P1 (see the size of species distribution clouds in Figure 2
; Average distance to the centroid: P1 = 0.252; P2 = 0.430; P3 = 0.292). Environmental heterogeneity was also different among ponds (F = 3.16; p
= 0.042). However, the compositional variation was not completely coherent the environmental heterogeneity of ponds. Although P1 was also the pond with the lowest environmental heterogeneity (average distance to the centroid: 1.831) and compositional variation (see above), P3 had the highest environmental heterogeneity value (2.793) but not compositional variation (results above), followed by P2 (2.202), which had intermediate environmental heterogeneity (Figure 3
) and also had intermediate absolute values of environmental variables.
Goal 4: There were evidences for spatial structuring in the composition of the three ponds (Mantel’s r and p-values for Pond 1, 2 and 3, respectively: r = 0.366; p = 0.021; r = 0.364; p = 0.004; r = 0.475; p = 0.007). Sites within ponds also shown spatial autocorrelation in environmental variables in P1 (r = 0.505; p = 0.002); but not in P2 (r = −0.083; p = 0.725) and Pond 3 (r = 0.137; p = 0.180). The relationship of compositional dissimilarity with environmental dissimilarity, controlling for geographical distance, was not significant for P1 (r = 0.139; p = 0.253) and P2 (r = −0.291; p = 0.890), but was significant for P3 (r = 0.5603; p = 0.008).
Goal 5: The additive diversity partitioning indicated that, in absolute terms, the variation in diversity between sites within the same pond (beta 1) mostly contributed the gamma diversity. However, this portion was lower than expected by the null model, indicating that the variation within the pond is relatively low when compared to the coexistence of species at each site (alpha) and to the compositional change between ponds (beta 2). In fact, both alpha and beta 2 diversity portions were significantly higher than the expected by the null model (Table 5
The Pantanal lives in constant transformation, contemplates periodic floods, and supplies countless permanent and non-permanent ponds, as well as in physical scales, storage of water on, and below ground, because they are not physically independent, since groundwater interacts with surface water; for example, the Guarani Aquifer System, which is under increasing stress, and is being threatened with pollution and contamination [36
]. Therefore, in period of flood, ponds may have a greater connection than in dry periods (not analyzed). Whereas, spatial variations in the structure and composition of communities are largely influenced by the seasonal flood pulse [53
], as well as by other predictors who have significant relationships with diatoms, such as pH, longitude, annual temperature, and precipitation [25
]. However, they can vary between spatial systems and scales, when the degree to which the dispersion limitation overlaps with environmental filtering process [18
] and comes to depend on specific ecological characteristics of each species [54
For Ferradura Pond (P1), the continuous influence on alteration of the communities may be due to the mass effects of the dispersion, or to the effects of the environmental structure. Paes and Blinder [55
] report that places with a greater number of species may have characteristic environmental properties, such as greater diversity of habitats, bigger area, species with greater range of distribution, among others. Otherwise, Remmer et al. [26
] report that beta diversity is greater in small lakes and decreases with the raise of the lake area, and that deeper ponds are less influenced than shallower ponds and the dispersion limitation generally increases with the raise of spatial distance between locations [18
]. In addition, diatoms preferentially respond to trophic gradients [4
], and metacommunity, made up of ponds with biogeographic restriction, it is reported that its local communities are highly resistant to invasion, unless there are significant disturbances [56
group and also Aulacoseira italica
were prominence species, in P1. The Eunotia
group is cited as the most diversified genus in acid and wetland environments [57
], Aulacoseira italica
is not a common species [61
] and when found in greater abundance is in acid and oligotrophic environments [64
At Burro Pond (P2), it is likely that only the effects of mass dispersion explain the compositional variation. The pond is considered the meeting site of the other rivers, facilitating the sustained dispersion in the pond. Even with the heterogeneous environment, the population’s existence is supported by immigration, assuming that the dispersion is modest due to the changes resulting from the floods. Moreover, local species differ from other ponds, thus supporting a metacommunity [1
]. As noted, species present different competitive skills, but remain in sub-optimal or even unfavorable habitats [67
] rescued by recurrent immigration [68
]. These source-sink dynamic mass effects modify species diversity [67
] and consequently affect the structure and dynamics of the community [68
]. The highlight for this pond (P2) was the largest number of species of the genus Aulacoseira
. The ecological preference of the Aulacoseira
show that the trophic gradient is the main driver of species distribution [66
], and many species of the genus are typical of the mesotrophic for eutrophic environments [63
For the Caracará Pond (P3), it is likely that both the effects of dispersion and environmental selection explain in a complementary way the compositional variation. It is the most environmentally heterogeneous pond, with a large variation among its 10 sampling sites. Studies suggest that the environmental filtering process is the main structuring factor for diatom communities [11
], modeled particularly by the availability of nutrients [21
] and widely influenced by the seasonal flood pulse [53
], with dispersion being second structuring factor, highlighting the importance of also considering processes related to dispersion in the interpretation of diversity patterns. However, studies show that both processes, dispersion and filtering, can be of equal equivalence [76
]. Pond 3 was the one presenting the most indicator species. Local communities within a metacommunity may result, according to Leibold et al. [77
], in several spatial dynamics, altering the diversity of local species, both directly and indirectly, and can go back and change characteristics of the regional biota. Since that diatom community structures are also the result of filtering and spatial processes [54
The diversity of each site in a pond, as well as the variation between ponds, was significant. Assuming that the dispersion is modest, and ponds maintain a metacommunity and not a homogenized local community, and if the dispersion were extremely low, it would depend fundamentally on stochastic processes and local interactions (species/environment). Studies report that the increase in alpha diversity occurs with the increase in dispersion rates [1
], but if it is maintained eventually through the immigration of source populations, in this case there is an increase in alpha diversity and consequently there is a reduction in beta diversity [79
]. However, the variation in diversity between sites in the same pond (beta 1) is what contributes the most to gamma diversity. Dispersal rates can have significant effect to species diversity and ecosystem stability [79
], in which richness and stability are maximized at intermediate dispersal [80
On the other hand, if even higher dispersion rates can make a difference if this effect occurs simultaneously for many species of the metacommunity, then the composition of the local communities will be homogenized between the spots. In this case, the beta diversity decreases, and the alpha of the local communities will approach the gamma diversity of the metacommunity. Moreover, if the dispersion rates are extremely high, it can be considered that the spatial fragmentation of the habitat is irrelevant for the species, and in this case, what we call a metacommunity is effectively just a local community [79
As it becomes increasingly evident that human actions are exercising ever greater control over the conditions and processes that allow our existence, diatoms have proven to be extremely powerful indicators to explore and interpret many ecological problems. Thus, applied studies based on diatoms are tools that closely meet the general expectations of the environmental managers [81
], as well as seeking to introduce environment related topics into projects and joint actions along transboundary water systems [36
]. Studies of the diatom community at local scales are extremely important, because first we have to understand how the dynamics of this community is structured in a micro-habitat, and then it can be applied in larger spatial scales. Large wetlands may also be a union of several smaller wetland types and are found around world. In this case, we need to start to study the basics to understand the functioning of any wetland.
Our results reveal that the processes related to the environment and dispersion are in control of the structure and composition of the diatom community in wetland. According to the data analyzed, it was particularly evident that different species responded to different limitations or restrictive environmental factors, with abundance of species that are more adapted to the environment.
Our studies reveal that environments that are more degraded have less local diversity and, because the areas maintained by the rivers receive an eutrophication load, there may be a change in the structure of the diatoms. Ponds that are more eutrophic showed less diversity and the pH and oligotrophy were the main factors to maintain the greatest diversity of species of the genus Eunotia and the greatest abundance of Aulacoseira italica. However, moderate dispersion also maintains species diversification, due to the fact that these ponds have a connection to a different river that supplies it.
Summing up, the ponds have unique dynamics, there are differences in the compositional variation, in the mass effect processes, and in the different indicator species for each pond. Knowing the dynamics and structure of diatoms, which are at the beginning of the food chain, is essential for conserving, maintaining, or rehabilitating wetland ecosystems, such as the area of our study. The Pantanal is part of La Plata river basin, which represents the second largest surface for water resources in South America and the Guarani Aquifer System, the biggest unified groundwater aquifer in the world. These results show that even environments that are of the same type (wetland) are unique and have differences from each other, so these areas cannot be generalized. Considering the context of the environmental fire destruction that the Pantanal suffered (2020), the results obtained in this study are of utmost relevance to raise new research for the area. Catastrophes such as these, are unpredictable. We understand that wetlands are fragile places that suffer greatly from floods; they should be monitored, both in the short and long term, in relation to changes in the environment so that conserving, maintaining, or rehabilitating actions can occur faster when needed.