EVALUATION OF THE GROWTH AND SURVIVAL OF MANGROVE SEEDLINGS UNDER DIFFERENT LIGHT INTENSITIES: SIMULATING THE EFFECT OF MANGROVE DEFORESTATION

– Environmental factors, especially light, temperature, and edaphic conditions are of great importance for the establishment of plant communities. In areas degraded by deforestation, these factors can vary greatly, which may aﬀ ect the recolonization of the typical populations in the altered area. This study evaluated the development of seedlings of pioneer mangrove species under diﬀ erent levels of shading in soil substrate degraded by deforestation, aiming to assess the eﬀ ect of deforestation on the recolonization of mangrove forests, which may be of help in the production of seedlings and recovery of deforested areas. The study was conducted in the municipality of Acaraú in the state of Ceará, Brazil. The species Avicennia schaueriana Stapf and Leechm. ex Moldenke (1939) and Laguncularia racemosa C. F. Gaertn (1807) were used in the study, and the substrate was collected from an area impacted by deforestation. The propagules and seedlings were exposed to full sun and 30, 50 and 70% shading. The results revealed that the treatment at full sun had the lowest germination rate of 86.66% for L. racemosa . Seedlings of both species showed a signiﬁ cant diﬀ erence and better quality between shading treatments and full sun. The height of the seedlings showed a correlation above -0.90 with ambient temperature. L. racemosa tested in full sun had a viable seedling reduction of 50% and A. schaueriana was superior. It is concluded that the natural regeneration of a mangrove area can be compromised under the conditions of total deforestation with high sun exposure and degraded soils. Human intervention in providing 50% shading is essential for the development of seedlings and regeneration of the area impacted by deforestation.


INTRODUCTION
Mangroves are ecosystems established in coastal areas with average temperatures above 16 °C, brackish water, unconsolidated soil, and characterized by interactions between soil, freshwater and seawater (Saenger, 2002;Tomlinson, 2016;Duke, 2017). The mangrove forest is a component of great importance to the mangrove ecosystem, providing essential resources for the survival of numerous species, including humans, besides promoting the control of hydrodynamics and erosion, protecting against storms, and stabilizing climatic conditions along the coastline (Alongi, 2008;Lee et al., 2014).
Mangrove forests in the state of Ceará are composed of four plant species, namely Avicennia germinans L. (Stearn 1958), A. schaueriana, Laguncularia racemosa, and Rhizophora mangle L. (1753), while another species, Conocarpus erectus L (1753), occupies higher ground and receives low tidal infl uence (Camargo Maia and Coutinho, 2012;Vale and Schaeff er-Novelli, 2018). Anthropogenic changes in the environment have a strong impact on this area, such as the deforestation of mangroves for fi rewood and charcoal, residential construction, making fi shing artifacts, and various developments (Maia et al., 2018).
The canopy of mangrove forests can present a wide variety of sunfl ecks, depending on the plant species, arrangement of canopies, and conservation of the mangrove ecosystem (Asaeda et al., 2016). Sunfl ecks are small spaces in the canopy structure of a forest where the light passes through. Therefore, they are essential for the germination and growth of propagules, seedlings, and juvenile plants in the soil of mangroves (Maciel et al., 2002). Intense changes in this environment, such as deforestation, will result in a great increase in light as reported by Querino et al. (2013), albedo elevation of 49%, fertility decline, and soil destruction and loss (Arruda, 2013).
These changes aff ect mangrove seedlings, infl uencing their establishment, early development and resilience (Clarke, 2004;Krauss et al., 2008;Balke et al., 2015). The adaptation of plant species to new environmental conditions, particularly light, is especially important in juvenile plants because it conditions morphogenetic and physiological alterations in their structure and function, determining the success of regeneration (Pérez et al., 2017).
In the case of mangrove forest species, studies addressing issues such as seedling development under diff erent shading levels and recolonization of clearings are scarce. Hence, this study evaluated the development of seedlings of pioneer mangrove species under diff erent levels of shading in soil substrate degraded by deforestation, aiming to assess the eff ect of deforestation on the recolonization of mangrove seedlings and generate subsidies for the recovery of deforested areas and production of seedlings.

Study area
The study was carried out at the Federal Institute of Ceará -Acaraú Campus (-2.889037 °S and -40.113054 °W) located in the Municipality of Acaraú in the State of Ceará, Northeastern Brazil. According to the aridity index (AI) from several rainfall stations in the state of Ceará (Funceme, 2017), the municipality shows a sub-humid type of regional climate (65 ≤ AI <100). Data from the National Institute of Meteorology (INMET) (Lima Junior et al., 2016) showed a mean annual temperature of 27.6 °C, relative humidity (RH) of 80.73%, and wind speed of 4.2 m/s in the period of 1976 to 2011.

Choice of species and procedures with propagules, substrate and irrigation
The species A. schaueriana and L. racemosa were chosen because they are pioneer mangrove species (Costa et al., 2014). Propagules were collected near the Curral Velho village in the municipality of Acaraú -CE (-2.889037°S and -40.077642°W), according to Silva and Maia (2018). The propagules were collected by manual harvesting, found on the soil, and they were viable and non-germinated. Propagules were visually sorted for apparent health and size in the laboratory to select propagules with similar characteristics. Propagules were subsequently washed in running water, immersed in 1% sodium hypochlorite (NaClO) for 5 min for disinfection, and washed again in running water according to Silva and Maia (2018).
The substrate for seedling production was collected between zero and 20 cm deep in the fi rst layer of soil and 15 m from the riverbank in an area impacted by deforestation (-2.876921°S and -40.126909°W), because most environmental factors undergo modifi cations in deforested areas (Arruda, 2013). This collection method was chosen to avoid the soils of the riverbank where nutrient concentration is maintained in the system, mainly due to the mixing of fresh water with salt water (Baran and Hambrey 1998;Alves, 2001).
The substrate was sun-dried, fragmented into small-sized clumps and homogenized, ensuring greater uniformity. Afterwards, the propagules were sown and irrigated using the public water supply. The seedlings were irrigated with 40 ml/plant on average in each treatment, twice a day, at 8 am and 5 pm.

Time of experiment, design, evaluation, and data collection
The study was carried out from May to July of 2017 (60 days), ending when the development of the seedlings was considered completed. The beginning of the juvenile phase, on average 60 days after sowing, is recognized by the loss of the cotyledons (embryonic leaves) in L. racemosa and A. racemosa and by the formation of adult-like leaves representing the end of the supply of compounds from the embryonic (Oliveira, 2017).
We used a 2 x 4 x 3 factorial completely randomized design (two species, four levels of shading, and three replications). Each treatment contained ten seedlings of each species, totaling 240 seedlings (10 seedlings x 2 species x 4 treatments x 3 replicates). The seedlings were exposed to full sunshine (0% shading) and levels of 30, 50 and 70% shading using polyethylene screens. The seedlings were monitored daily, and data on plant structure and germination percentage were collected every eight days after the stems were completely erect and showed the fi rst pair of leaves in A. schaueriana and L. racemosa.
The structural data collected were height, rootcollar diameter (RCD), and phenological conditions of seedlings (coloration and number of leaves and branches). A graduated ruler was used to measure height, and an analog caliper with an accuracy of 0.005 mm was used to measure the RCD.
The Dickson quality index (DQI) (Dickson et al., 1960) was used to obtain structural height (H) and RCD data using nine randomly collected seedlings of each species (three seedlings per replicate per treatment). Subsequently, the aerial part of the root system was separated and dehydrated at 60 °C for 20 h in a drying oven to obtain root, shoot, and total dry matter. A precision balance (SHIMADZU AY220), with a precision of 0.0001 g, was used to weigh fresh matter and root and shoot dry matter. The following formula was used to calculate the quality index: DQI = (total dry matter)/(((H+RCD+aerial dry matter)/ (root dry matter))).
For the analysis of total dry matter and water content of the seedling (total fresh matter minus total dry matter) the data obtained for DQI were used. The temperature of the experimental units was measured using a mercury thermometer placed for seven minutes in each experimental unit during the hottest period of the day, between 2 and 4 pm; these data were collected twice a month.

Statistical analysis
Descriptive analysis was performed on the following parameters: number of viable plants, percentage of germination, and physiological and phenological characteristics. Two-way ANOVA was used (species and treatment) to evaluate the height, RCD, dry matter and water content data. One-way ANOVA was performed to analyze the variation in RCD during the experimental time. Tukey's test was used to compare results when diff erences between treatments were observed at a 95% confi dence interval. Linear correlation analysis was conducted to verify the correlation between shading percent and germination rate, height, and temperature in the experimental units. Pearson's correlation was used in three sample pairs with 12,000 simulations at a 95% confi dence interval, resulting in the r correlation.

RESULTS
The results showed that germination rates were good for both species tested (Table 1). A. schaueriana had rates above 90% and the treatments with 50 and 0% shading reached 100% germination; however, 0% shading did not maintain 100% of seedlings during the experiment time. We also found that the higher the level of shading the better the germination rate in the fi rst week; thus, better germination rates were achieved with greater shading , with a positive Pearson correlation of r = 0.9852. On the other hand, the propagules of L. racemosa in the treatment with 50% shading had a 100% germination rate and the treatment in full sun (0% shading) produced the lowest germination rate.
All treatments with shading showed a signifi cant diff erence in height compared to 0% shading for L. racemosa (F 3,99 = 23.984, p = 0.00001) (Figure 1). The seedlings with 50 and 70% shading diff ered signifi cantly from those with 30% shading, but there was no diff erence between the two higher shading levels. A. schaueriana seedlings in the 50 and 70% shading treatments diff ered signifi cantly in height from those produced at 0% shading, but did not diff er from those with 30% shading (F 3,109 = 6.3428, p = 0.00053).
The L. racemosa species was signifi cantly diff erent in the 30 and 50% shade treatments, with the largest RCD (F 3, 99 = 9.0339, p = 0.00002) ( Figure  2). A. schaueriana showed no signifi cant diff erence between shading treatments, the mean RCD was 3.4296 mm for full sun, 3.6893 mm for 30% shading, 3.6800 mm for 50% shading and 3.4000 mm for 70% shading, proving to be a species with good adaptability variation in light.
A. schaueriana seedlings showed no signifi cant diff erence in total dry matter gain; however, seedlings grown under 50% shading showed the highest mean of 0.9897 g for dry matter gain, followed by 30% shading with 0.9554 g, 70% shading with 0.8372 g, and full sun exposure with 0.7674 g. A. schaueriana seedlings also showed no signifi cant diff erence in water content between treatments. The seedlings showed the following amounts of water content: full sun exposure, 2.9767 g; 30% shading, 3.8967 g; 50% shading, 3.8589 g; and 70% shading, 3.3800 g.
In relation to the nursery temperature and seedling height, linear correlation analysis showed that the lower the ambient temperature, the greater growth was for both study species with -0.981 for L. racemosa and -0.984 for A. schaueriana. The best DQI was obtained in seedlings in the 50% shading treatment for both species. The lowest indices were obtained in full sun and 70% shading treatments (Table 2).
Through the monitoring of the structure of the seedlings it was possible to observe branching development. L. racemosa had a higher percentage of branches at the 30% shading level. A. schaueriana showed a higher percentage of branches at the 50% shading level ( Figure 3A). Finally, the quality of seedlings grown at diff erent levels of shading showed some percentage of chlorinated or nongerminated seedlings, probably due to the fact that the substrate comes from degraded area soil. However, degraded soils and in full sun impacted the seedlings negatively. Therefore, both species studied had a higher percentage of chlorinated, non-germinated and dead seedlings with 0% shading treatment, where A. schaueriana was more aff ected ( Figure 3B).

DISCUSSION
Germination rates were good in A schaueriana for all treatments, showing rates above 90%, which may be related to nutritional reserves and the water content level in the propagules, making germination less sensitive to treatments. L. racemosa had high germination potential, except with full sun treatment (0% shading), which showed the lowest germination rate of 86.66%. This demonstrated that in deforested areas, the germination potential of L. racemosa can be reduced. Even so, these results corroborate those of other studies where germination tests with L. racemosa propagules resulted in a high germination rate (Silva et al., 2017;Silva and Maia, 2018).
The growth rates of A. schaueriana and L. racemosa were very similar in the treatments evaluated, with greater growth in the 70% shading treatment and lower growth in full sun exposure. Pereira et al. (2009) produced seedlings in in full sun and irrigation with fresh water, which are conditions similar to those used in the present study in the 0% shading treatment. These data reinforce the eff ect of direct sunlight on mangrove seedlings, reducing their size.
The seedlings of L. racemosa and A. schaueriana of the 50 and 70% shading treatments had the best heights, but the did not diff er signifi cantly. These results corroborate those reported by Clara and Giordano (2014) in the state of São Paulo when monitoring the development of L. racemosa seedlings in full sun exposure and 35, 50, and 80% shading, with the best result obtained with 50% shading. Similar results have been obtained in forest species in other ecosystems and agroecosystems (Paiva et al., 2003;Oliveira and Perez, 2012;Rezende et al., 2017). Figure 2 -Total dry matter and Water contained in the seedlings, average + standard error in Laguncularia racemosa seedlings according to diff erent shading levels (0%, 30%, 50%, and 70%). Diff erent letters indicate signifi cant diff erences according to the Tukey's Test. Figura 2 -Total de matéria seca e água contida nas mudas, média + erro padrão nas mudas de Laguncularia racemosa de acordo com diferentes níveis de sombreamento (0%, 30%, 50% e 70%). Letras diferentes indicam diferenças signifi cativas de acordo com o teste de Tukey.    The 30 and 50% shading treatments provided greater environmental comfort for L. racemose, with a consequent increase in photosynthesis and photoassimilates in plant tissue, which signifi cantly aff ected the RCD. In general, diff erences in light conditions may lead to variations in chlorophyll content (Brant et al., 2011) and thus refl ect on biomass. The RCD of seedlings exposed to 0% shading was probably restricted by the strong sunlight and substrate drying up, which compromise photosynthesis (Pereira et al., 2015).
Dry matter analysis showed that the best growth was obtained with 50% shading. According to Paiva et al. (2003), the total amount of dry matter accumulated by a plant, as a parameter of growth evaluation, is a direct refl ection of the net photosynthetic production added to the amount of absorbed mineral nutrients, reinforcing the results of Clara and Giordano (2014).
It is believed that 50% shading has a positive eff ect on seedling transpiration, and this may explain the higher amount of water in L. racemosa and A. schaueriana seedlings produced under this level of shading. Dutra et al. (2012) evaluated physiological parameters in copaíba (Copaifera langsdorffi i Desf.) under the same four levels of shading used in this study and found that the 50% shading treatment provided the lowest values of daily transpiration and at time points measured throughout the day in the plants. This result may explain the greater amount of water in L. racemosa and A. schaueriana seedlings produced under 50% shading.
The results also showed a signifi cant linear correlation between the height of seedlings and ambient temperatures provided by the shading levels, so the higher the temperature (full sun) the smaller the seedling. Dalastra et al. (2012) and Costa et al. (2011) also observed greater heights in cultivars of arugula and Mimosa lettuce with increasing shading levels (0, 30, 40, and 50%). Shading may reduce vapor pressure defi cit, air and soil temperature, and wind speed, which were signifi cantly lower in the shaded system in coff ee crops compared to full sun (Lunz, 2006).
The best DQI was obtained in seedlings with the 50% shading treatment in both species. The 50% shading treatment provides seedlings with the greatest robustness and biomass balance. Although seedlings in the 70% shading treatment showed higher mean height values than those in the 30% shading for A. schaueriana and 50% for L. racemosa, they displayed lower quality indices after 0% shading treatment. This was caused by cellular stretching, which contributes to greater heights under shade environments (King, 1994;Carvalho et al., 2006).
The development of leaves was not aff ected by any of the treatments in either species. A. schaueriana had three pairs of leaves and L. racemosa four pairs of leaves at the end of the experiment. Clara and Giordano (2014) followed the development of L. racemosa seedlings for 130 days and did not observe a signifi cant diff erence in leaf gain between diff erent shading levels. Morandi et al. (2009) performed a study with Calophyllum brasiliensis Camb. under full sun exposure and 30, 50, 70, and 90 shading and did not observe a diff erence in the number of leaves during the study.
Regarding lateral development, both species developed branches in all treatments. L. racemosa showed more branched individuals in the 30% shading treatment, representing 50% of the branched seedlings. A. schaueriana showed more branched individuals in the 50 and 30% shading treatments, representing respectively 53.85% and 30.77% of the branched seedlings. Thus, increased branching is not attributed to greater light intensity and solar radiation in mangrove seedlings because mangrove forest altered by deforestation showed more branched individuals of L. racemosa (Santos et al., 2012).
We found that the higher the shading, the better the seedling performance in soils degraded by deforestation. L. racemosa responded well to the 50% shading treatment with a high rate of viable seedlings and low rate of non-germinated and dead seedlings. A. schaueriana seedlings exhibited a better performance in the 50 and 70% shading treatments and appeared to be more sensitive to areas degraded by deforestation, showing a low rate of viable seedlings and high rate of chlorinated, non-germinated, and dead seedlings. In summary, soils degraded by deforestation and exposed to full sunshine show decreased seedling survival rate and high rates of chlorinated individuals that lead to mortality and non-germinated and dead seedlings.

CONCLUSION
We conclude that deforestation of mangroves interferes with the recolonization of the mangrove by reducing the germination rate of L. racemosa species and mortality of seedlings of both species studied. L. racemosa seedlings exhibited higher physiological resistance in soil conditions of deforested areas under full sun. Deforested areas with high light intensity require intervention with shading, planting and replanting due to the high mortality of seedlings. 50% shade treatment provided higher quality and stability of seedlings in deforested environments.

6.ACKNOWLEDGEMENTS
The authors thank the fi nancial, structural and logistical support of the following organizations: the Pro-rectory of Research, Postgraduate and Innovation (PRPI) of the Federal Institute of Education, Science and Technology of Ceará (IFCE) by the fi nancing through the announcement PROAPP / Postgraduate. To the National Council for Scientifi c and Technological Development (CNPq), for the granting of scholarship. To the IFCE and the Mangrove Ecology Laboratory (ECOMANGUE) for providing space, equipment and collaborators in the execution of this research, especially to technicians and interns / scholars.