Os manguezais de Bertioga localizam-se próximos ao limite sul de distribuição dessa comunidade vegetal na costa brasileira e no hemisfério sul. Esses bosques são compostos por três espécies: Rhizophora mangle L., Avicennia schaueriana Stapf & Leechman e Laguncularia racemosa Gaertn f. e apresentam estrutura pouco desenvolvida. O manguezal do Rio Iriri possui um diâmetro médio de 8,48 cm, altura média de 6,55 m, densidade de 216 ind/0,1 ha e área basal de 1,29 m2m2/0,1 ha. Já no Rio ltapanhaú, .o diâmetro médio do bosque é de 10,41 em, com altura média de 6,83 m, densidade de 173 ind/0,1 ha e área basal de 1 ,69 m2/0,1 ha. Estimativas de biomassa nos dois locais de estudo resultaram em valores de 42,3 t/ha para o Rio iriri e 59,7 t/ha para o Rio Itapanhaú e a produção anual de madeira foi de 3,6 e 1,9 t/ha respectivamente. Esse baixo desenvolvimento estrutural é provavelmente devido às baixas temperaturas alcançadas no inverno e aos diversos estresses aos quais estão submetidos esses bosques em função das atividades humanas desenvolvidas na região. A produção de serapilheira desses bosques reflete seus baixos índices estruturais, embora trate-se de bosques de franja inundados duas vezes ao dia por marés de mais de 1 m de amplitude. A taxa média anual de produção de serapilheira foi de 5,6 t/ha no Rio lriri e 4,6 no Rio Itapanhaú. Padrões sazonais foram evidentes para os diversos componentes da serapilheira assim como uma sequência temporal das diversas fases fenológicas. As folhas apresentaram maior produção durante o verão. As flores mostraram picos no final dessa estação seguidas de picos de propágulos no outono. A queda de madeira apresentou maiores taxas principalmente nos meses de inverno. Os fatores que controlam essa sazonalidade estão provavelmente associados à temperatura e pluviosidade. Nesses manguezais a renovação desse material é alta, apresentando baixas quantidades de estoque de serapilheira no sedimento, cujos valores variam de 50 a 164 gPS/m2. Essa renovação se dá através da decomposição, que é bastante rápida, com taxas diárias médias de 0,006 para Rhizophora, 0,011 para Laguncularia e 0,015 para Avicennia. Além da diferença entre as espécies, foi observada uma diferença entre as estações e os sítios. As taxas do verão foram significativamente mais altas que as do inverno. Além disso, o Rio Iriri apresentou taxas médias mais elevadas que o Itapanhaú. Embora o processo de decomposição seja intenso nesses manguezais pela existência de condições favoróveis, a maior parte da renovação se dá através da exportação da serapilheira pelo movimento das marés. A partir da elaboração de um modelo ecológico da dinâmica da serapilheira incluindo sua produção, estoque e decomposição, foi possível estimar essa exportação cujos valores estão em torno de 0,7 gPS/m2 por dia com uma exportação acumulada anual de cerca de 2,5 tPS/ha, o que significa que estes manguezais exportam aproximadamente 50 por cento da serapilhelra que produzem. Essa dinâmica da serapilheira influencia também a ciclagem de nutrientes nesse ecossistema. Para compensar a exportação de materiais esses manguezais apresentam retranslocação de nutrientes como o nitrogênio e o fósforo, mecanismo no qual esses elementos são reabsorvidos da folha antes de sua queda. A imobilização, outro mecanismo para conservar nutrientes durante a decomposição, não foi significativa nesse estudo, ocorrendo apenas com nitrogênio em folhas de Laguncularia. Apesar da retranslocação registrada nesses bosques, tanto para o nitrogênio quanto para o fósforo, esses manguezais não se mostraram eficientes no uso de nutrientes, provavelmente, porque estes estejam disponíveis no sedimento em quantidades adequadas. A variação anual da concentração de nutrientes nas folhas da serapilheira mostrou um padrão sazonal com maiores valores no inverno e menores no verão. Já as folhas verdes não apresentam essa sazonalidade tão nítida:o que ocorre, portanto, é uma maior retranslocação durante a época de maior produção de folhas, mantendo os teores desses nutrientes na copa.

The mangroves in Bertioga estuary are mixed forests composed by three species: Rhizophora mangle L., Avicennia schaueriana Stapf & Leechman and Laguncularia racemosa. Gaertn f. Both study sites are frequently flooded by tide waters and can be classified as fringe forests, according to the physiographic types. The Rio Iriri site had a DBH of 8,48 cm, and a mean height of 6,55 m, density in this site was 216 trees/0,1 ha and the basal area was 1,29 m2/0,1 ha. In Rio Itapanhaú, mean DBh was 10,41 cm, with mean height of 6,83 m, density of 173 trees/0,1 ha and basal area of 1,69 m2/0,1 ha. Based on the structural parameters it is possible to see that DBH and tree height of these forests are equivalent to other fringe sites, but density and basal area are much lower. The DBH increment was similar to measurements presented in other studies. However wood production was much lower, because of low density. Biomass estimate of each site forest were: for rio Iriri the biomass was 42.3 t .ha-1 while in rio Itapanhaú it was 59.7 t. ha-1. Mean annual littertall rates were 5.6 t.ha-1 .yr -1 in Rio Iriri and 4.6 t.ha-l.yr-1 in Rio Itapanhaú. However, these forests are fringe mangroves which normally have higher production rates. In this study litter fall production was seasonal with a trend of alternate peaks for the various components which appears to be a sequence within the year. Leaf fall is higher in the summer which is the rainy season (from November through February) followed by the miscellaneous peak in the autumn (from March through May) while the period of higher wood production is from June to October. This pattern suggests a coordination among phenological phases. The factors controlling those variations are still not clearly established. Environmental conditions related to rainfall and temperature are the most probable responsibles for this influence. The decomposition potential of the mangrove sites in Bertioga region (23º51'5) was high and this might be due to high temperature associated to high rainfall and frequent flooding that create good conditions (hot and humid environment) to accelerate the organic matter degradation. Species showed different decomposition rates. The highest value of k were observed for A. schauerianna (0.015 d-1) followed by L. racemosa (0.011 d-1 ). The lowest value of k occurred in R. mangle (0.006 d-1 ). Brazilian studies of litter decomposition with the same species showed the same gradient of species specific rates. Those differences can be explained by differences in the resource quality. Chemical composition of leaf material is very important in determining its decomposition rate. There was also a significant difference in decomposition rates among species and between sites and seasons. Iriri had higher decomposition rates than Itapanhaú. Summer season had higher decay rates than winter. The processes of decomposition are regulated by the combined effects of the resource quality and the physico-chemical environment on the community of decomposer organisms. According to these main controlling factors, differences observed among species should be attributed to differences in substrate quality, while differences among sites or seasons would be related to different environmental conditions. Seasons differences would be more related to climate features while site differences in this study would be more determined by edaphic features. Although there was an increase in nutrient percentage during decomposition, there was not an increase in their absolute concentration or net accumulation. In Bertioga mangrove forests immobilization was low or absent perhaps because the studied sites are fertiles. In addition, it is only observed in Laguncularia leaves because it is the species with lowest nitrogen concentration what would be a limiting factor. The high content of nitrogen in Rhizophora leaves could be the reason why this nutrient had no net immobilization in this study. Apparently, the absence of phosphorus immobilization could be its availability in both ways, plant tissue and sediment. Mean values of litter standing stock ranged from 50 to 164 g.m-2 throughout the year which are low values, but considering that both are fringe sites it would be expected. Seasonal patterns of litter standing crop can be related to seasonal patterns of litter production, decomposition or hydrological variations. Since in Bertioga decomposition plays a small role because most of litter is removed and hydrology does not show a marked seasonal pattern, litter export and stock reflect the seasonality of litter production and may be a little modified by variations in tidal height. This is not valid for wood component that showed a different behaviour because it is not very influenced by export. Litter turnover rates of this study ( 5.1 and 4.7 yr-1) are higher than other values found in other mangrove forests. The values obtained for leaf fall (14.92 to 24.22 yr-1) in this study are very high even for fringe forest. Analysis of turnover rate for each month indicates that these rates vary throughout the year. especially for leaf fall. The period of higher turnover rates seems to be related to higher litterfall rates, but it may be also an interaction with the variation of tidal amplitude and decomposition. It is possible to see that there is a coincidence of the months of higher turnover rates with the months of higher tidal amplitude. Since measurement of litter export is very difficult because water movements are very complex in a mangrove site and due to the importance of litter dynamics to the tunction of a mangrove ecosystem. the development of a litter dynamics model was useful to understand those processes and organic matter fluxes and to the estimate of export rates. The litter model includes littertall as the input of organic matter, litter standing crop as the storage of organic matter and decomposition and export as outputs of organic matter from the system. Based on these figures it is possible to roughly estimate an annual export of 2,5 t.ha.yr -1. The conclusion is that in these mangrove forests at least 50 per cent of its production is exported by tides. Nutrient content in canopy of mangroves in Bertioga is high and although retranslocation takes place, nutrient concentrations in litterfall are still high. C:N of leaf litter in these forests are low and decomposition should not be nutrient limited. Then litter decay occurs fast with low or none immobilization. These are characteristics of sites with high availability of nutrients. In these forests, although nutrient use efficiency is not very high, nutrient recycling is occurring via retranslocation and not via immobilization during decomposition probably because they are fringe forests. The high retranslocation rates could be a mechanism to prevent loss of nutrients by export. Another important characteristic of these mangroves was the seasonal pattern of nutrient cycling. Nutrient content of leaf fall varies throughout the year with high concentration during winter. Retranslocation also showed a seasonal pattern with higher rates in the summer, resulting in lower nutrient levels in leaf litter and the consequence is a higher nutrient use efficiency in this period of leaf production.