Dinâmica e relação solo-vegetação do componente arbustivo-arbóreo da regeneração natural em fragmento de floresta atlântica, Pernambuco

Abstract

The objective of this study was to characterize the dynamics of the natural regeneration of shrub-woody species, to verify the nutritional efficiency of the species with the highest absolute density and to relate the natural regeneration with the soil characteristics in a fragment of Lowland Rainforest in Sirinhaém, Pernambuco. To obtain the estimates of natural regeneration, four subplots were allocated within 40 sample units of 250 m2. The inclusion level adopted was circunference at breast hight < 15 cm and height (H) measurements, divided into three classes, whose class 1 included individuals with H ≥ 1.0 and H ≤ 2.0 m, class 2 with individuals H > 2.0 and H ≤ 3.0 m and class 3 with individuals with H > 3.0 m. Of the four subplots, three were allocated with differentiated sizes, taking into account the increase of height classes: class 1 - subplots of 1 m x 25 m; Class 2 - 2 m x 25 m and class 3 - 3 m x 25 m. For the sampling of the three height classes together, a plot of 5 m x 5 m was allocated on the opposite side. The phytosociological parameters, the regeneration dynamics and the Shannon indices (H') and the Pielou equability (J') were calculated. For the determination of leaf biomass, all leaves of three individuals per selected species were collected and weighed in the field and dried in an oven to determine the percentage of moisture. The content and efficiency of the biological utilization of the nutrients by the species were calculated. In order to study soil chemical attributes as indicators of the spatial distribution of the species with the highest absolute density, the characterization of the spatial variability of the species and of the soil attributes was performed using the geostatistical analysis. The richest families, common to both types of sampling, were kept with the largest number of species during the whole period evaluated (2014 to 2016), especially Myrtaceae and Fabaceae. The sampling method using different plot sizes proved to be more efficient, since it brought together a larger number of families and species. The species with the best performance against the natural regeneration index in all evaluations were Brosimum rubescens and Tovomita mangle. The Shannon index was, on average, 3.85 nats ind.-1, for sampling in plots of 5 m x 5 m and 4.05 nats ind.-1, for plots of different sizes and the equability of J' of 0.81 and 0.82, respectively. In the period evaluated for the sampling performed in plots of 5 m x 5 m there was a mortality record of 310 ind. ha-1 and entry of 900 ind. ha-1, and in plots of different sizes by height class a mortality of 526 ind. ha-1 and admission of 1122 ind. ha-1. In relation to the basal area, the percentages of gain were also higher than the losses. As for the floristic similarity between the adult shrub and arboreal component in regeneration, we found 90 common species, the Sorensen Similarity index was 81%. The species utilized nutrients more efficiently in the following order: P>Mg>K>Ca>N. In tropical soils of low natural fertility, mainly in P, the use of these species can be recommended in reforestation projects. The soils of the area presented high acidity, high saturation by aluminum and low saturation by bases, indicating low natural fertility. The species Tovomita mangle can be indicated for soils with different chemical characteristics and of diverse natural fertility; Brosimum rubescens and Inga capitata for acid soils and low natural fertility, but with moderate Al exchangeable contents; Talisia retusa should be recommended for albic and Caraipa densifolia soils for non-alic soils; and the species Anaxagorea dolichocarpa and Protium arachouchini for soils of high natural fertility.