Effects of automotive diesel oil on germination of Avicennia germinans and Laguncularia racemosa mangrove propagules

ABSTRACT Mangrove ecosystems are sensitive to oil, as spills can impair developmental processes of mangrove vegetation. Since the 2010s, the Brazilian equatorial margin, more specifically the Pará-Maranhão Basin and the mouth of the Amazonas River, has been affected by oil runoff from urban activities and the increased risk from exploratory deepwater drilling for oil extraction. Dispersal of mangrove propagules occurs during the tidal cycles, when the presence of tensors in the water can affect germination. We analyzed the effects of diesel oil on the germination of propagules of the two most common mangrove species in the region, Laguncularia racemosa and Avicennia germinans, in six treatments of diesel oil in the water (0.5%, 1%, 1.5%, 2%, 3% and 4%) and a diesel-free control. The response variables were germinability (G%), mean germination time, mean germination speed and the germination speed index (GSI). G% and GSI in L. racemosa propagules differed significantly between the control and all treatments (G% and GSI < 10% in the 3% treatment). Propagules of A. germinans were more resistant, and the physiological variables did not differ significantly among treatments and control (G% > 90% in all treatments). Our results indicate that, at the germination stage, L. racemosa was more susceptible than A. germinans to contamination by automotive diesel oil.


INTRODUCTION
Mangroves are among the most productive ecosystems in the world, providing important ecosystem services and wide connectivity with other coastal ecosystems such as coral reefs, marine prairies, as well as a continental-ocean interface (Yokoya 1995;Mochel 2011;ICMbio 2018;Kristian and Oktorie 2018;Van Der Stocken 2019).In 2018, global mangrove coverage was estimated at 150,000 km 2 and nearly 25% was damaged by human actions, mainly deforestation, pollution, coastal zone occupation, shrimp cultivation and climate change (Ferreira and Lacerda 2016;Kristian and Oktorie 2018).Brazil has approximately 10,020 km 2 of mangrove forest, 77.6% of which are in the Amazon coastal zone of the Brazilian states of Maranhão, Pará and Amapá (Diniz et al. 2019).
Oil pollution is one of the most recurrent impacts on coastal and marine systems due to accidents with oil tankers, offshore exploratory activities, leaks of fuel or lubricant from vessels, all of which damages the adjacent ecosystems (Thompson and Thompson 2020) and human health (Pena et al. 2020).Oil pollution can also occur during port operations with anchored vessels, from leaks in pipelines (Matos et al. 2019), during vessel cleaning (Alcântara and Santos 2005), leaks from stored fuels in tanks at gas stations (Finotti et al. 2001), pipeline, road and rail transport (Martins and Mochel 2022).
The impact of oil on coastal environments depends on the amount of spillage, the type of oil and its toxicity, and environmental conditions such as the predominant type of sediments, tidal circulation and prevailing climatic conditions at the time of contact with the oil (Soares et al. 2006;Chequer et al. 2017).Acute and chronic effects result from impacts of different types of oil in mangrove ecosystems (Hoff and Michel 2014).Diesel oil is the most consumed fuel in Brazil (Petrobras 2021).
Mangroves on the Amazonian coast are largely pristine, although areas near large urban centers such as the capital cities of Belém (Pará state), Macapá (Amapá), and São Luis (Maranhão), all located within the delta complex of the Amazonas River, may be heavily affected by urban pollution, including oil runoff.There are records of a diesel oil spill in São Marcos Bay (Maranhão) after the collision of a ship with a shipwreck stranded on a sandbar in 1990, and of diesel oil leakage in a mangrove during the fueling process of miningtrain locomotives in 2000, among other occurrences (Martins and Mochel 2022).Since the 2010s, the Brazilian equatorial margin has increasingly been the target of development plans and exploratory deepwater drilling for oil and natural gas extraction, raising the risk for possible oil leaks from activities related to these operations in the fragile and lush areas of mangroves on the Amazon coast (Silva Junior and Magrini 2014).
Mangrove tree species have viviparous reproduction as an inherent characteristic and propagule dispersal occurs through the tides (Hutchings and Saenger 1987;Elmqvist and Cox 1996;Tomlinson 2016).Dispersion success depends on the buoyancy and durability of propagules, as well as on their hydrodynamics and mortality from predation and damage caused by crabs and insects (Kathiresan and Bingham 2001;Hoff and Michel 2014;Hogarth 2015).It is in this highly vulnerable phase of the plant that the presence of tensors in the water can affect germination.Oil contamination can harm the germination process, as well as the growth, development and establishment of mangrove seedlings (Duke et al. 1998;Chindah et al. 2011;Hoff and Michel 2014;Chequer et al. 2017).
Previous studies with mangrove species in contact with various types of petroleum derivatives and crude oil were developed with the propagules planted directly in the sediment or transferred at the mangrove seedling stage and subsequently exposed to the contaminant (Zhang et al. 2007;Naidoo et al. 2010;Nardes et al. 2013;Amin et al. 2017;Chequer et al. 2017;Guedes et al. 2018).In the present study, we evaluated the germination of propagules submitted to different percentages of contamination of estuarine water with diesel oil under laboratory conditions, aiming to analyze the effects of the contaminant on the first stage after propagules have left the trees.Among the mangrove species that occur on the Amazonian coast, we opted to work with the two most common and widely distributed ones, the black mangrove, Avicennia germinans (L.) L., locally known as siriba or siriúba, and the white mangrove or inkworm, Laguncularia racemosa C.F. Gaertn.
The collection was made during high tide.This site is quite pristine and regularly used as a research site by the authors, being far from the municipal headquarters, where a community of fishermen lives from shrimp and shellfish fishing.The propagules were identified according to the specifications of each plant of the species: A. germinanslight propagule, its length varies from 25 mm to 30 mm, L. racemosa -light propagule, its length varies from about ACTA AMAZONICA 11.0 mm to 19.5 mm.Propagule collection and selection criteria followed the recommendations of Goforth and Thomas (1980) and Mochel and Fonseca (2019), to select healthy propagules with greater possibility of developmental success.The experiment was carried out at the Mangrove Recovery Center (CERMANGUE) of the Department of Oceanography and Limnology at the Federal University of Maranhão (Universidade Federal do Maranhão, UFMA).

Germination
To observe the effect of the automotive diesel oil on propagule germination in both species, we simulated an acute contamination event with six treatments: 0.5% (T 1 ); 1% (T 2 ); 1.5% (T 3 ); 2% (T 4 ); 3% (T 5 ) and 4% (T 6 ) and a control (C) without oil, for each species, based on the study by Kim (2014) who used 0%, 6%, 12% and 18% and Chequer et al. (2017) who used 3%.The experiment was carried out in triplicate, each replicate on a different date.Each replicate consisted of one glass bowl containing 500 mL of estuarine water at 28 g kg -1 of salinity for each species and oil percentage.Salinity was measured with a portable refractometer (Quimis Q767-3, Shanghai Precision & Scientific Instrument Co., Shanghai, China).For each bowl, we used 36 propagules of A. germinans and 114 propagules of L. racemosa (Figure 2).ACTA AMAZONICA Ti), where G1 to Gi = number of germinated propagules each day, and T 1 to Ti = time in days of the experiment (each day) (Maguire 1962;Borghetti and Ferreira 2004).

Statistical analysis
The Levene test was used to evaluate the homogeneity of variance.Variables that met the assumption for parametric analysis were compared among treatments and the control with ANOVA.Significant results were compared by the post-hoc Tukey test for pairwise comparison of the means.Variables with non-parametric distribution were analyzed with a Kruskal-Wallis test followed by Mann-Whitney tests for pairwise comparison.To further evaluate the association of the germination variables with the oil treatments in each species, we used principal component analysis (PCA) based on a correlation matrix (Valentin 2012;Legendre and Legendre 2012).The statistical significance of the difference between the groups obtained by the ordination technique was assessed with a one-way PERMANOVA test (Anderson 2001).The statistical analyses were performed with the Past 4.03 software (Hammer et al. 2001) with a significance level of 0.05 for all tests.

RESULTS
Laguncularia racemosa showed a significant difference among treatments for G% (F = 4.15; df = 6/14, p = 0.013), with significantly higher G% in the control and no difference among replicates within treatments (rejoinder) (Table 1).

Diesel oil treatment G% MGT (days) MGS (days) GSI
Laguncularia racemosa C 80.8 ± 30.Avicennia germinans, showed no significant difference in G% among the treatments and the control (F = 0.42; df = 6,14, p = 0.85) (Table 1).GSI varied significantly in L. racemosa (F = 6.20; df = 6,13, p = 0.002), with all treatments presenting significantly lower values than the control (Table 1).No significant difference was observed for GSI in A. germinans (F = 0.25; df = 6/14, p = 0.95).No significant variation was observed among treatments and control for MGT and MGS in both species (Table 1).Axis 1 and 2 of the PCA explained 46% and 35% of the data variance, respectively.G%, GSI and MGS of L. racemosa were positively correlated with axis 1, being associated with the control.MGT was negatively correlated with axis 1, being associated with treatments with higher concentrations of diesel oil (T 4 , T 5 and T 6 ).MGS and GSI of A. germinans were positively correlated with axis 2, being associated with the control and T 1 , while MGT was negatively correlated with axis 2, with association with treatments T 2 , T 3 and T 4 , and G% was close to the origin of the graph, indicating low representativeness in the sample, and little association with the treatments (Figure 3).The PERMANOVA indicated significant differences (F = 99.09;p = 0.01) among the control (CG) and the treatments with low T 1 , T 2 and T 3 ) and high (T 4 , T 5 and T 6 ) oil concentrations.

DISCUSSION
Our results showed that A. germinans and L. racemosa responded differently when exposed to diesel oil.Avicennia germinans exposed to crude oil in concentrations of 15 ml and 120 ml showed retardation in seedling development (Chindah et al. 2011), and the development and survival of planted propagules of A. schaueriana, R. mangle and L. racemosa exposed to 3% of marine diesel oil was significantly affected (Chequer et al. 2017).
Under unfavorable conditions, the seed germination process can be hampered (Baskin and Baskin 2014), which seemed to have been the case for L. racemosa, indicating that the species is sensitive to contact with diesel oil in the germinative stage, reinforcing the evidence from other studies that the components of petroleum derivatives or crude oil can inhibit both the germination process and the development of several mangrove species (Zhang et al. 2007;Hoff and Michel 2014;Guedes et al. 2018).
There was a significant variation in the germination speed index in L.racemosa, with greatest germination capacity of the propagules in the control, and decreasing capacity with the increase of diesel oil in the treatments.This same effect of decreasing germination capacity was also observed for halophytes from coastal wetlands exposed to diesel oil (Kim 2014).
Avicennia germinans showed no negative effects of diesel oil on any of the germination variables analyzed, but the ACTA AMAZONICA absence of a significant effect does not allow us to speculate to what extent the species would be resistant to diesel oil percentages greater than 4%.The mangrove response to contact with oil depends on the amount and type of oil with which it is impacted (Chequer et al. 2017;Soares et al. 2006).It is also necessary to consider the developental stage of the mangrove when it is hit by the contaminant (stage of germination, seedlings or established tree).Studies conducted on A. germinans seedlings using different concentrations of lubricating oil and crude oil (15ml, 120ml) caused their death (Proffitt et al. 1995;Chindah et al. 2011), while in Avicennia marina (Forssk.)Vierh.and Bruguiera gymnorrhiza (L.) Lam., marine fuel oil (bunker) impaired the healthy development and caused seedling abnormalities (Naidoo et al. 2010).Although there is extensive literature about the negative effects of oil spills on mangrove ecosystems, reporting short and long-term effects, the action of oil on propagules in the germination stage, when they are dispersed in the water, is still little known.

CONCLUSIONS
The germination of Laguncularia racemosa was significantly lower than that of Avicennia germinans when exposed to diesel oil at percentagens between 0.5 and 4%.This indicates that this species is especially susceptible to diesel oil in the germination stage of the propagules, suggesting a greater degradation potential than other mangrove species in areas affected by diesel oil spills.Avicennia germinans proved to be resistant to the percentages applied in this study, and it is recommended to test this species with higher percentages of oil.Our results contribute to the knowledge on the sensitivity of propagules during the germination process to tensors and the general process of ecological recovery of mangroves degraded by diesel oil residues.

Figure 1 .
Figure 1.Location of the collection points of Laguncularia racemosa and Avicennia germinans propagules in the Mangue Seco mangrove, Ilha do Maranhão, Maranhão, Brazil.This figure is in color in the electronic version.

Figure 2 .
Figure 2. Distribution of Laguncularia racemosa and Avicennia germinans propagules in the experimental vials containing estuarine water (500 mL) and different percentages of diesel oil.This figure is in color in the electronic version.