Secondary Succession in the Lowland Forests of the Marlborough Sounds Maritime Park
This study documents aspects of the forest recovery process in secondary communities of the Marlborough sounds Maritime park. some 39 types of seral vegetation were recognised as being common in the lowland zone of the Marlborouqh Sounds. Data on vegetation structure and composition, forest flooor biomass and nutrient status and soil nutrient status were collected from 39 seral communities ranging in age from 2 to 84 years and were compared with data collected. from 6 adjacent undisturbed lowland forest coutuunities. Three seral pathways were represented at the study sites, the vegetation was developing toward either: - Weinmannia racemosa dominated forests - mixed broadleaved species forests or Dysoxylum spectabile coastal forests. A number of vegetation structural parameters (crown cover, specific vegetation space, basal area and stem density) showed rapid recovery to levels similar to those of undisturbed forests. However, other structural parameters (mean canopy height, total stand volume, total stand foliage) had not recovered to levels typical of adjacent undisturbed forests some 70 to 80 years after disturbance. The pattern of vertical development of the forest foliage during succession was from short concentrated foliage distributions in young stands to tall, relatively evenly spread distributions throughout the height profile in older stands. Foliage in stands older than 25 years tended to be concentrated in the height class near, or just below, the mean canopy height of the stand. shrub understorey development conulenced some 45 years after disturbance. Few trends in species diversity were evident during succession. Species richness (d) was very high and at a maximum in the youngest study stand. species richness was relatively constant after two years and changes in species diversity during succession were due mainly to decreases in species evenness (J'). Duration of seral vegetation in the Marlborough Sounds Maritime park was related to the life spans of the dominant pioneering species at a particular site. Gahnia pauciflora dominates a site for only 1 year, Cassinia leptophylla can dominate for 20 years, Leptospermum scoparium for some 45-55 years and L. ericoides can dominate the vegetation for between 70 and 85 years. Linear relationships were derived between age and d.b.h. measurements for three seral tree species (weinmannia racemosa, Leptospermum scoparium and L. ericoides). Leptospermum scoparium and L. ericoides have similar average diameter growth rates of 0.6 cm/yr for most of their growing period. The average diameter growth rate of Weinmannia racemosa was O.29 cm/yr. The variation in weight of the forest floor at any one study site was very high and the sample size employed, in this study was inadeguate for estimating the weight of the forest floor within reasonable limits. However broad trends in the recovery of the forest floor during succession were noted. There were two main patterns of forest floor accumulation. Sites dominated by Pteridium esculentum showed rapid recovery of forest floor biomass to very high levels. sites dominated by Leptospermum and Cassinia species showed much slower rates of forest floor recovery and never approached the weights attained by Pteridium-dominated stands. Two types of forest floor (laminated mors and moroid-mulls) developed under seral vegetation but were not present under undisturbed forests. There was a trend toward. a decline of the forest floor under Leptospermum-dominated stands after 40 years and this was coincident with a decrease in the contribution of L. scoparium to a stand's basal area and crown cover and an increase in dominance in a stand, by L .ericoides and Pseudopanax arboreus. Seral stands located on slopes and ridges never attained forest floor weights similar to those under virgin forests on comparable sites. Rates of nitrogen, organic carbon, total cation and total phosphorus accumulation in the forest floor layers were directly related to the rate of forest floor biomass accumulation. No large accumulation of nitrogen was noted in the forest floors during the early stages of succession. A high proportion of the available nutrient capital of many of the seral stands and the undisturbed forests was located in the forest floor and the top 1 cm of mineral soil. Nitrogen was the one element that appeared to be in ample supply in the mineral soil to a depth of 30 cm. No relationships between rates of regeneration and measured soil parameters were evident in this study. It did appear, however, that the broad seral patterns of vegetation recovery were related to the soil and topographical pattern of the Marlborough sounds. soils of valley floors and terraces and steep colluvial slopes support seral communities that are developing toward a mixed broadleaved species forest or to a Dysoxylum coastal forest. The pioneering shrub Leptospermum scoparium was not important in seral vegetation at such sites. soils of the gentler slopes and ridges support communities that are tending toward either a weinmannia racemosa-dominated, forest or a mixed broadleaved species forest. Leptospermum scoparium is usually an important component of the seral vegetation at these sites. All sites that supported vegetation which was developing toward. a weinmannia racemosa forest possessed low or very low exchangeable calcitrm levels in the A horizon soils. seasonal fluxes of nutrients in litterfall were monitored in two adjacent younq seral stands. one stand was dominated by a Leptospermum ericoides - L. scoparium low forest and the other stand by a Coriaria arborea - Aristotelia serrata low forest. Quantities of litter returned to the forest floor under both stands were higher than those reported from undisturbed forests in New Zealand. The annual litterfall under the Leptospenrum-dominated stand. was 7.80 tonne/ha and Under the coriaria-Aristotelia stand it was lO.07 tonne/ha. High quantities of nutrients were returned to the forest floor under both stands. Leaf-litter decomposition in the Coriaria-Aristotelia stand was rapid. After 6 months on the forest floor the leaf litter reached such a degree of disintegration that it could not be separated from the mineral soil. In constrast the decomposition of Leptospermum leaf litter was slower Approximately 40% of the litter weight of the Leptospermum material was lost after 13 months on the forest floor. The nutrient flux behaviour of the litter differed at the two seral stands and this reinforced the soil nutrient differences of the two sites. The value of young regenerating vegetation to the integrity of the Marlborough Sounds lowland forest ecosystem has not been appreciated by authors of recent land use studies of the Marlborough Sounds. A critique of, some of the recommendations made in the Land use studies has been undertaken. Guidelines to assist with the management of, scenic reserves in the Marlborouqh Sounds Maritime Park have been drawn up and are discussed within the context of, modern ideas of successional theory.