Anthropogenic changes in the environment have brought negative consequences that caused even the protected areas to be susceptible (Filla et al., 2008; Zappes et al., 2009). Triggered by human activities, these changes have been more intense in coastal environments, habitat of several cetaceans that are now vulnerable (Di Beneditto and Rosas, 2008; Crespo et al., 2010; Ross et al., 2011; Gusso-Choueri, 2015). Among the threats to the balance of marine environments, we can highlight: destruction of mangrove swamps, predatory fishing, land and ship pollution, climate change and bioinvasion (Pereira et al., 2014).

In the southern coast of São Paulo, southeast of Brazil, among many conflicts, the closure of Valo Grande canal—built in 1852—is frequently discussed by the locals, in protected areas management boards, or in judicial proceedings, once the estuary has been deeply influenced by the opening of this canal, originally 4.4 m wide, and widening rapidly ever since (Besnard, 1950; Saldanha, 2005; de Souza and de Oliveira, 2016; Prado et al., 2019). Because its margins could not stand the force of the water, they went through a process of erosion, overturning existing houses in the environs of Iguape (Valentin, 2006; de Souza and de Oliveira, 2016). In addition, the rapid siltation that prevented large crafts from navigating in the Valo took the region to economic decadence (Valentin, 2006; Saldanha, 2005).

Today, due to erosion, the canal is more than 300 m wide and 15 m deep at most, meaning that 70% of Ribeira river flow is channeled through it, greatly affecting the ecosystem as a whole, especially regarding the decrease in salinity (Besnard, 1950; Bonetti-Filho and Miranda, 1997; Saldanha, 2005). However, this shift in salinity allowed shoals of manjubas (Anchoviella lepidentostole) to enter the canal and go upstream (Carneiro, 2005; Saldanha, 2005; Souza, 2012). For this reason, this species has taken a greater proportion in comparison to other species, such as crustaceans and saltwater fish that no longer reach the northern part of the estuary (Carneiro, 2005; Saldanha, 2005).

In 1978, the State government decided to build a dam to close Valo Grande canal, causing, once again, changes in the ecosystem (Bonetti-Filho and Miranda, 1997). An improvement in the estuarine water quality was observed, with the rapid reappearance of shellfish, oysters and shrimp (Saldanha, 2005). This recovery, however, contrasted with the floods in the Ribeira river floodplains, which had already been intensively occupied by families and businesses, leading to conflicts among different perspectives on the multiple uses of water resources in the low basin (Souza, 2012).

In 1995, the dam broke, allowing the flow of fresh water, silt and dissolved substances into the Estuarine Lagunar Complex of Cananéia-Iguape, risking its environmental functions of regulation and biodiversity support for the third time (Bonetti-Filho and Miranda, 1997; Bernardes and Miranda, 2001; Souza, 2012).

This estuary is located in the largest continuous remnant of Atlantic Forest and represents one of the most well preserved ecosystems of the Brazilian coast (Souza, 2012). It is legally protected and has more than 40 Environmental Protection Areas at municipal, state and federal levels, which present different levels of restriction of occupation and use (ICMBio, 2015). These protection areas are called Lagamar Mosaic (ICMBio, 2015). In this estuary, there is a residing population of Sotalia guianensis which, according to the Official National List of Endangered Species of Fauna, published in 2014 by Ministry of the Environment, is listed as vulnerable (Monteiro-Filho, 1991; Santos and Rosso, 2008; Havukainen et al., 2011; MMA, 2014; Godoy et al., 2015). Among the threats mammals such as the studied species are subjected to, Begon et al. (2007) and Crespo et al. (2010) evaluate that degradation and habitat loss are the biggest.

In view of the changes described here, the aim of the present study is to assess the influence of the construction of Valo Grande canal on the occurrence and distribution of Guiana dolphins. Among the changes caused by the canal that had already been observed in the Estuarine Lagunar Complex of Cananéia, the decrease in salinity could lead to a probable loss of habitat with a decrease in the occurrence of Guiana dolphins (Sotalia guianensis), since these animals are more observed in areas with higher salinity (Eschrique et al., 2011; Godoy et al., 2015).

We conducted 131 surveys, accounting for 330 h of effective effort. During this period, we sighted 5,003 individuals, an average of 38 dolphins per survey (standard deviation = 2.47). The Valo Grande canal flow presented a daily average of 429 m3/s (standard deviation = 146.63) and a maximum of 1,087 m3/s.

The results of the best model 1 are shown in the table below according to the AIC value:

AIC (Poisson): 5259.2

AIC (Binomial Negative Binomial): 2530.2

Results show that the occurrence of dolphins throughout the study area was lower in spring and summer, compared to winter (intercept) (Table 1). The opposite was observed in the sectors where there was a significant increase in the occurrence of dolphins (sectors II and III). Also according to the data presented in Table 1, there was a negative significance of Valo Grande daily flow in relation to the number of dolphins in the region.

The statistical summaries of GLMs per sector (model 2) and season (model 3), with the flow of the Valo Grande as the explanatory variable, are presented in Tables 2, 3. Table 2 shows significant results correlating the occurrence of dolphins in sectors III and IV, while results in sectors I and II show that the occurrence of dolphins is not determined by the daily flow of Valo Grande. However, in models that take the four seasons into account (Table 3), summer is the only season in which the occurrence of dolphins does not correlate to Valo Grande flow.

Results of Generalized Linear Models pointed to a negative influence of Valo Grande canal daily flow on the number of dolphins in the study area. The predictive models showed that in sectors II, III, and IV there was an increase in the number of dolphins with the decrease of Valo Grande canal average flow. This increase was crescent from sectors II to IV, corresponding to the proximity of these sectors to Valo Grande canal. Sector I was the only one that had a non-significant decrease in number of dolphins with the increase of the canal flow, according to data presented in Appendix A,B. The results presented in the predictive model confirm that the decrease in flow would increase the dolphins' spatial distribution in the region, especially in sector IV, the closest to Valo Grande canal.

Ross et al. (2011) listed 10 principles that characterize the habitat needs for small cetaceans, contributing to delineate boundaries and priority habitats. One of the principles is that the priority habitat must take into account the external connections needed to ensure its integrity, since many anthropogenic actions in coastal waters are strongly impacting small cetaceans. Among these actions, the regulation of fluvial ecosystems charge and habitat loss have been threats to them (Begon et al., 2007). Examples of these threats are presented by Nabi et al. (2021) in their bibliographic review of habitat changes and habitat loss caused by the regulation of river flow on Indus River Dolphin populations (Platanista gangetica minor) and conservation strategies of Yangtze Finless Porpoise (Neophocaena asiaeorientalis).

In the southeast of Brazil, the opening of Valo Grande canal took place in 1852, and completely changed the fluvial charge of Ribeira river, leading to the construction of a dam, which was kept closed from 1978 to 1995, when it broke down (Souza, 2012; de Souza and de Oliveira, 2016; Prado et al., 2019). Nowadays, Valo Grande canal is responsible for 70% of the flow coming from the higher portion of Ribeira river (Alto do rio Ribeira). Many studies have been trying to assess the changes suffered by this environment, which is part of a region considered of extreme biological importance, identified as a priority area for conservation (MMA, 2007), with more than 40 Protection Areas.

According to Nascimento et al. (2008), the re-opening of Valo Grande canal caused an increase in the flow, which led to the appearance of sandy islands and the formation of sand spots at the bottom of the lagunar canal. In a study conducted by Mahiques et al. (2009), in the central part of the complex, there was a change in the depositional pattern: before the opening of Valo Grande canal there was a predominance of sands and nowadays, the deposited sediments present more than 65% of silt and 8% of clay. According to the same author, today, the presence of metal in the sediments within the Cananéia estuary comes from the Ribeira river hydrological system that used to be low when the canal was closed, but has increased substantially after the opening as a consequence of the discharge of heavy metals from ancient existing mining companies of Ribeira river toward the inner part of the estuary (Mahiques et al., 2009; Coelho, 2011; Saito and Oliveira, 2012; Salgado et al., 2021).

The flow of Valo Grande canal directly influences salinity and water transparency in the sectors assessed in this study (Bernardes and Miranda, 2001; Coelho, 2011). Salinity, on the other hand, is an important factor for the species distribution, especially in estuaries (Begon et al., 2007). This difference in salinity has brought consequences for the northern region of the estuary and in places closer to Valo Grande canal, observed from macrophytic banks found today within and around mangrove forests, suffering thus a reduction of areas covered by mangrove swamps (Cunha-Lignon et al., 2015; Sampaio et al., 2021). The APA CIP Management Plan (Environmental Protection Area of Cananéia-Iguape-Peruíbe) also points to a loss of mangrove swamps habitats in the northern part, resulting from changes caused by the construction of the canal, which has come to be considered one of the factors of greatest impact on mangrove and estuarine species of the region (ICMBio, 2016). Once acknowledged its importance for the food chains in estuaries and marine areas, the destruction of mangrove swamps could lead to the decline of coastal and estuarine fishing in the region (Pereira et al., 2014).

The southern part of the study area–Trapandé Bay, the area of highest concentration of Guiana dolphins, located >50 km from Valo Grande canal–is still influenced by the freshwater intake, which alters salinity (Geise et al., 1999; Santos and Rosso, 2007; Coelho, 2011). Studies with cetaceans also confirmed that salinity influences the spatial distribution of some species that live along the coast and estuaries (Dias et al., 2009; Stutz Reis, 2013; Godoy et al., 2015; Kanaji et al., 2016).

The region's ichthyofauna has also changed because of the freshwater intake in the Complex (Maciel, 2001). In his study of the region, using various methods for catching fish between 1996 and 1997, the freshwater species were more abundant in the region influenced by Valo Grande canal, especially in the rainy season, and the typically marine species were more abundant in areas nearer the adjacent ocean, where salinity was higher. However, Contente (2013) did not find any species, using data from before the opening of Valo Grande canal. According to Maciel (2001), the flow of Ribeira river is a tensor that modulates the structure of fish communities in the region

Studies on habitat use by cetaceans show that environmental characteristics effectively influence distribution and abundance of food resources, which, in turn, influence species distribution in a given area (Ballance, 1992; Davis et al., 1998; Hastie et al., 2004; Torres et al., 2008; Rupil et al., 2018). Therefore, the environmental variables used in most studies (depth, temperature, salinity, declivity, among others) once related to the abundance of preys, are indirectly linked to the presence of cetaceans, known to be sensible to these resources abundance (Baumgartner et al., 2001; Learmonth et al., 2006; MacLeod, 2009; Lambert et al., 2014; de Boer et al., 2014; Godoy et al., 2020). Most preys of Guiana dolphins are found in the estuary because they're from salt and brackish water, showing this region's importance for this species (Zanelatto, 2001; Oliveira et al., 2008; Lopes et al., 2012; Rupil et al., 2018). However, the decrease in salinity directly affects the occurrence of marine fish that enter the estuary to reproduce and feed, due to physiological limitations, since most species can only stand small variations of salinity (Maciel, 2001; Pascke and Lanzendorf, 2017).

Although the lower limit of salinity and the upper limit of exposure time for dolphins in low salinity environments are unknown, variations in salinity can alter the availability of prey and also cause physiological changes. In spite of the fact that cetaceans have a complex osmoregulation mechanism, some studies with Tursiops truncatus found physiological evidence of blood changes and skin lesions due to exposure to low salinity (Xu et al., 2013; Ewing et al., 2017; McClain et al., 2020; Takeshita et al., 2021).

To prevent the flow of freshwater in the estuary, the Public Prosecutor's Office of the state of São Paulo has determined that Valo Grande has to be closed again. The decision has been appealed and is currently under trial. If the final sentence is for its closure, a long term monitoring of the fauna and flora will be needed, in order to: (1) evaluate the recovery of areas of Manguezal; (2) carry out the survey of the estuary's ichthyofauna to confirm the decrease of freshwater species and consequent increase of marine species; (3) monitor dolphins in order to verify the increase in their occurrence in the areas closest to Valo Grande; (4) identify the exact cause of illness or death of Guiana dolphins using necropsy, virtopsy, pathology, histopathology, microbiology, and biochemistry techniques to assess the animal's health and understand how the various environmental factors can cause morbidity or mortality (as proposed by Nabi et al., 2021) and (5) monitor accumulation of heavy metals in the ecological system.

Although the closure of Valo Grande is expected to bring ecological benefits, it will possibly also have negative social impacts, since floods, which also came to occur with the open Valo, were more frequent when it was closed. These floods affected the local population living on the floodplains, leading to huge financial losses (Souza, 2012; Prado et al., 2019). Therefore, in case of closure, support to the communities in the floodplain areas of Ribeira river will be necessary. A community group may have to be formed with researchers, managers of Protection Areas, representatives of local communities, politicians and public prosecutors to discuss this social issue. To prevent the social impacts, the relocation of the population, dredging of the original course of the river or installation of a floodgate in the Valo could be proposed. In the case of installation of a floodgate, the biologists working in the area under recovery will need to be consulted to advise when it could be opened, since to re-enter of fresh water in the system can affect directly or indirectly the areas in recovery.

It is interesting to highlight that the collaboration between managers of Protection Areas and researchers makes the information resulting from research relevant to the management and directly applicable during the decision-making processes. Managers and researchers tend to agree on the importance of most themes and the need to respond to fundamental questions about the main threats to protected areas management (Cvitanovic et al., 2013).

Therefore, we can conclude through the predictive models that the flow coming from Valo Grande influences negatively the distribution of Guiana dolphins in the surroundings of Ilha de Cananéia, especially in the areas that are closer to the Valo Grande canal. In view of this, a thorough assessment of the pros and cons regarding the closing of Valo Grande is imminent, taking into account, however, that the loss of speccies of fauna and flora in certain regions of the estuary will have an inestimable value, often with no possibility of recovery.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethical review and approval was not required for the animal study because the data were obtained through observations of animals in their natural environment.

DF and AA contributed to the conception and design of the study. DF organized the database. HP performed the statistical analysis. DF and HP wrote sections of the manuscript. All authors contributed to the manuscript review, read, and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.