The influence of grazing by Hypogastrura viatica (Insecta: Collembola) on microbial activity in decomDosing kelp on Spitsbergen A PER SVEUM Sveum, P. 1986: The influence of grazing by Hypogasfrura viatica (Insecta: Collembola) on microbial activity in decomposing kelp on Spitsbergen. Polar Research 5 n.s.. 71 - 77. The influence of three different grazing levels (5-20 ind.) by Hypogastrura viarica (Collembola) on micro- bial activity in a Spitsbergen kelp Laminaria has been evaluated during a 10 day laboratoly experiment under oxic conditions. Respiration has been measured by infra-red gas analysis in the form of C O ,-emissions. The total and metabolically active fungal and bacterial biomass has been estimated by direct microscopical counts and epifluorescence techniques. Despite the results of other similar experiments, a n increase in microbial numbers and biomass was recorded following the introduction of grazing. Grazing increases the mean respiration between 0.6 to 1.4 times. A corresponding increase in the microbial biomasses was also recorded. Although a trend towards decreased mean bacterial cell size was found, the fungi were more influenced by grazing than the bacteria. Bacteria constituted the predominant part of the microbial biomass in all the experimental vessels. The results indicate that Collembola plays an important functional role by regulating the microbial activity, probably by nutrient mobilization. Per Sveum. SINTEF, Division of Applied Chemistry. Applied Biology Group, N-7034 Trondheim-NTH. Norway; March I986 (revbedJanuary 1987). Meiofauna has normally a limited direct influen- ce on the turnover of organic matter in ecological systems (Anderson et al. 1981; Petersen & Lux- ton 1982). Only a small fraction of the total C 0 2 respired, and thus carbon mineralized, is meta- bolized directly by the faunal component of the ecosystem. In many systems, however, the meio- and microfauna are known to enhance the minerali- zatiodnutrient cycling and carbon mineraliza- tion through their consumption of microbial biomass, i.e. fungal hyphae and bacteria (e.g. Johannes 1965). As Anderson (1975) pointed out, the study of the interaction between micro- fauna and microflora is one of the most intract- able problems of ecology. There have been few studies dealing with the microbial ecology of decomposing kelp heaps to date. As Griffith & Stenton-Dozey (1981) noted, little is known about the fate of stranded kelp. The aim of this study was to evaluate the influ- ence of one type of heterotrophic interaction, i.e. grazing by Collembola o n the decomposing dynamics. At the same time more general effects on decomposition and microbial activity by micro- grazers can also be evaluated. The present ex- periments were carried out as a part of a study of the ecology of seaweed heaps on Spitsbergen. This substudy has been very much influenced by similar investigations where comparable experi- mental designs and background philosophies have been employed, (see Hanlon & Anderson (1979) and Andren & Schniirer (1985)). Stranded kelp is initially rich both in energy a n d nutrients, and microbial activity is supported very well. As the seaweed loses its initial structu- ral support through water loss and decomposi- tion, a more or less solid layer of variable depth is formed, which is associated with a very limited exchange of gas. This enhances the formation of anoxic conditions dominated by anaerobic microbial activity. 72 Koop et al. (l982b) estimated the total amount of carbon mineralized by microbes to be about 70%. However, less than 4% of the nitrogen wrecked in macrophytal biomass is recorded as leaching. This implies a very high degree of microbial immobilization of nitrogen although ammonia is the principal end product of micro- bial decomposition both in aquatic and terrestri- al habitats (cf. Fenchel & Blackburn 1979) and might be lost to the atmosphere under airy condi- tions. Sveum (in prep.) found very high fractions of the total microbial biomasses in kelp beds o n Spitsbergen sea shores to be metabolically in- active. Seaweed heaps on Spitsbergen differ from those found i n central and southern Norway because of their lower biological activity, and consequently they have a heterogeneous and more aerobic gas composition than the anaerobic temperate varieties (Sveum, in prep.). Only one species of Collembola, Hvpogasfrura viatica was found in any number in the seaweed heaps (Sveum, i n prep.), and this species was selected as the microbial grazer for the experiment. H j p o - gastrura has earlier been found to graze on fungal biomass (Addison 1977). Koop et al. (1982a) reported that kelp (i.e. Eklomia maxima) is primarily decomposed by bacterial activity rather than by fungi. The Spitsbergen kelp beds were composed of two more or less distinct layers; an upper one where aerobic conditions predominated and a lower more or less anoxic layer. The kelp studied and reported in this paper is of the first type. Here H . viaticadominated among the invertebra- tes. The protozoa have not yet been studied in these kelp beds. Material a n d methods Experimental design Decomposing seaweed (Laminaria) was collect- ed at Kvadehuken in the outer part of Kongs- fjorden on the western coast of Spitsbergen. The decomposing seaweed had been on the shoreline for approximately 12 months. The experimental period of 20 days was divi- ded into two periods of 10 days each. The de- composing samples of seaweed were incubated at 5 O C in gas tight, 100 ml glass vessels. Then the seaweed was extracted to remove the meiofauna and was later supplied with a n equal amount of demineralized water to compensate for the water lost by evaporation during extraction. The vessels were allowed to acclimatize for 48 hours at the incubation temperature before they were closed. Incubation was carried out in a thermo- incubator, in total darkness. A total of 40 experimental vessels was used. After 10 days, 3 different levels of grazing by the Collembola species Hvpogastrura viatica were supplied to 30 of the incubation vessels (i.e. 10 replicates of each grazing level). Analytical techniques The COZ content of the vessels was measured every second day. To avoid hypertension and anaerobiosis in the vessels, fresh outdoor air was supplied by complete aeration. The CO, content of the gas samples was measured with an infrared gas analyzer (Siemens Ultramat I ) , modified for septum injection and calibrated with gas mix- tures of known COz concentration. The total CO, emission for each vessel was calculated by the numerical integration of the cumulative CO, curve. As the respiration rate of Hypogastrura viatica was not measured in the present study, the rate given by Aunaas et al. (1983) for Onychyums groenlandicus from a nearby site on Spitsbergen, was applied. Conversion from oxygen consumption to car- bon dioxide production was done using R Q = 0.8 (Andrtn & Schniirer 1985), and from microlitre t o microgram carbon by the general gas law. The collembolan respiration for each vessel was calculated as equal to the integrated animal respiration, assuming a linear growth curve from the number added at day 10 to the number present at day 20. The fresh weight of the Collembola was 3.33 times the dry weight (Pers- son 1983). At the end of the experimental period, bacteri- al counts and fungal measurements by epi- fluorescence microscopy were carried out on the content of each vessel. The total bacterial number was determined by staining with acridine orange (AO) (Trolldenier 1972) and by staining the metabolically active bacteria with fluorescein diacetate (FDA) (Lundgren 1981). The total of 73 fungal hyphae was measured after staining with methylene blue (Jones & Mollison 1948), while the length of the metabolically active fraction was measured with FDA ( S o d e r s t r ~ m 1977). The intersection method described by Olsson (1950) was used in both cases. Results and discussion Extensive studies of the feeding habits of H . viu- tica have not yet been camed out. However, microscopical examination of a few guts of the species obtained from the kelp beds on Spits- bergen showed that fungal hyphae were part of its content. It has been shown earlier that Col- lembola consume hyphae (Christiansen 1964; Petersen & Luxton 1982). H . viatica is a typical surface-living species in this habitat. According to Bodvarson (1970) surface-living species have a greater tendency to be fungal feeding than the strictly soil-living species. H . viatica occurs in several very different habitats on Spitsbergen. It is not known how far the predominant diet varies according to local adaptations within various habitats. The total respiration of the vessel communities for the first 10 days, with and without animals (days 11-20) is given in Table 1. The estimated Collembola respiration is also given in the same table. There is an increase in the total community respiration in the grazed vessels. The increase is linear with the rise in grazing pressure (r = 0.97, p <0.01). The differences between all combina- tions of grazed treatments are found to be statisti- cally significant. No significance is found be- tween the treatment groups in the ungrazed ini- tial period. The estimated grazing by Collembola does not account for the increased respiration activity, as it only comprises 0.3% t o 0.7% of the total respiration. The total increase ranges from 64% to 145% compared to the control. Bacteria constitute the predominant part of the microbial biomass in all the vessels. This is con- sistent with the results of Koop et al. (1982a). The abundance and masses of biota in the incubation vessels are given in Tables 2,3 and 4. Although the total length of fungal hyphae does not increase linearly with grazing pressure (i.e. hyphae are longer in an ungrazed state than at a low grazing pressure), a linear increase is found in the metabolically active fraction (r = 0.97, p