Sten Rundgren

Food- and density-dependent dispersal: evidence from a soil collembolan

1. Dispersal in a fungivore species of Collembola, Onychiurus armatus, in a homo- geneous soil was theoretically approached as a deterministic compartmental process, in which the movement of animals between discrete patches could be described by a series of differential equations or by the discrete geometric probability distribution, assuming a constant probability to move from one patch to the next. 2. Experiments were designed for a release of collembolans in two types of soil in a physical arrangement of distinct patches represented by vials connected by tubings to cover a distance of 40 cm. Dispersal distances of individuals were determined, and theoretical and observed distributions compared. 3. Dispersal rates, estimated from transfer rate constants, ranged from 0 020 to 1 42 day-1, suggesting that an average 0. armatus moved less than 10 cm day -1. The probability to leave a patch varied between 0 10 and 0 50. 4. Dispersal was dependent on population density, soil type and length of fungal mycelium; it was almost twice as high at a high than at a low density (90 000 and 30000 individuals m-2) in a mor soil and four times as high in a sandy compared with a mor soil. Dispersal rate decreased as the mycelial length increased, especially in a sandy soil. 5. Collembolans in a feeding phase had a higher tendency to disperse than those that were moulting. 6. Enriching the soil patch at 40 cm distance from the release point with a favoured food item, the fungal species Mortierella isabellina, increased dispersal rate by more than four times in a mor soil, suggesting that fungal odour can attract collembolans from a large distance and enhance their rate of movement more than a three-fold increase of their population density does. (Less)

Evolutionary response of earthworms to long-term metal exposure

The occurrence of a number of soil intertebrate species in severely polluted soils raises the question whether they have evolved resistance, e.g. by adaptation or acclimatization. This was examined in experiments on the earthworm Dendrobaena octaedra (Sav.) using specimens from sites polluted by a brass industry. One site, exposed since the 17th century, was situated close to an old mill, another was near a modern mill that has been operating for the last two decades, and the third was a non-polluted reference site. Adults and juveniles of D. octaedra from each site were reared in each of the three different soils. Growth (increase of body weight), survival, and cocoon production of the populations were followed over 225 d

Relative abundance and resistance traits of two Collembola species under metal stress

Population differentiation was studied in Onychiurus armatus (Tullb.) and Isotoma notabilis Schaffer from a reference site and two highly polluted (Cu, Zn) sites (long-term and short-term exposed, respectively) in the Gusum area, SE Sweden. Growth, survival and reproduction were followed in a F1 generation of O. armatus reared on a diet of the fungus Verticillium bullosum (W. Gams and Malla) grown on a series of Cu+Zn amended agar plates. Survival and reproduction were observed in a P generation of I. notabilis incubated in soils with enhanced concentrations of Cu and Zn. Metals affected growth rate but not mean maximum length of O. armatatus. The growth rates of both populations decreased significantly with increasing metal concentration

Growth Changes Caused by Metal Uptake in a Population of Onychiurus armatus (Collembola) Feeding on Metal Polluted Fungi

The effects of uptake of copper and lead on the growth of Onychiurus armatus (Tullb.) (Collembola) were studied in laboratory experiments. Specimens were reared in petri dishes, supplied with a soil borne fungus Verticillium bulbillosum, (W. Gams and Malla), which was grown on a substrate contaminated with different amounts of metals. Growth of specimens of 0. armatus was measured in the F1 and F2 generations. Growth data were fitted to a growth model and the best estimates of growth rate and average maximum length were calculated and compared for different metal contaminations. The fungus accumulated metals efficiently, reaching levels of 1000 to 3500 ppm on substrates containing 15 to 300 ppm metals. 0. armatus concentrated metals during the first two weeks of its life cycle, and then excreted metals towards a steady state level. Copper concentration in the body was higher than lead concentration, and the metals behaved antagonistically when mixed. Neither metal was accumulated by Collembola, and equilibrium concentrations of copper varied from 150 to 600 ppm. Growth rate and average maximum lengths were higher at a moderate metal pollution, both in F1 and F2, and significantly reduced on substrate containing 90 ppm metal or more. The lowest concentration of copper and lead in specimens with significantly reduced growth rate was about 200 ppm. This concentration was also found in a field population in the vicinity of a brass mill. The consequences of reduced growth rate on survival and reproduction in a population are discussed.

Effect of Bird Predation on Lumbricid Populations

The effects of predation by golden plover Pluvialis apricaria (L.) on populations of Allolobophora caliginosa (Sav.) and Lumbricus rubellus Hoffm. in early spring on a hayfield in Iceland were estimated by two independent methods. Field experiments (lasting for 22 days) with nets preventing bird predation showed that the mean abundance of lumbricids was 238 ind m-2 on protected sites and 107 ind m-2 on exposed sites (P < 0.001). Observations of feeding plovers showed that about 4.5 lumbricids were taken per m2 and day; i.e. about 100 ind m-2 in the experimental period. This agrees closely with the result obtained by the first method. No selective predation could be confirmed but certain differences between the two species of lumbricids are discussed.

Modelling Dispersal Distances in a Soil Gradient: The Influence of Metal Resistance, Competition, and Experience

A discrete probability model was developed to describe the distributions of dispersal distances of two soil-living collembolan species, Folsomia fimetarioides and Isotomiella minor, in an experimental gradient with metal polluted soil. The model assumes that animals move stepwise between adjacent habitats with a probability that is influenced by soil metal concentration, metal induced mortality and resistance, interspecific competition and the capability of individuals to perceive concentration differences. Theoretical and observed distributions were compared to determine the contribution of each factor to the dispersal pattern. Mor soils from a contaminated area were used to construct gradients consisting of three connected 20-cm3 PVC rings filled with soil of different zinc and copper concentrations. Adult collembolans were added to each ring, and survivors and newly hatched juveniles were counted after 2 and 10 weeks of incubation. The probability that individuals remained in the soil where they were released was generally < 0.5. They responded to increasing metal concentrations by an exponentially decreasing probability to settle. The model suggested that dispersal distances of F fimetarioides were largely affected by resistance to metals, which was supported by independent observations of enhanced reproduction by individuals collected in the polluted field soil compared with individuals from an unpolluted soil in the same area. Both species preferred to settle in the unpolluted part of the gradient, but E fimetarioides was slower due to a poorer dispersal ability or a slower response to high metal concentrations. Significantly more F fimetarioides remained in the most polluted soil during the first 2 weeks of incubation when L minor was also present in the gradient, but the contribution of competition to the model description of dispersal distances was weak. The closest fit to the experimental data was attained when individuals were assumed to perceive differences in metal concentrations as they moved from one soil to another.

Methods of Sampling Lumbricids

The study was carried out on a permanent pasture and in a beech wood in southern Sweden in April and August 1967. Times and localities were chosen to obtain differences in soil physical properties and in associations, abundance and size classes of earthworms. Hand sorting was compared with extraction methods (formaldehyde and potassium permanganate solutions). The extraction methods were tested with and without added detergent. The formaldehyde method was as efficient as hand sorting when estimating populations of Lumbricus terrestris L. during periods of activity. Hand sorting was the best method for studying Allolobophora spp., whereas Dendrobaena spp. and L. rubellus Hoffm. can be studied using all the tested methods. There was a tendency for hand sorting to under-estimate small specimens. Detergent did not increase the efficiency of the extraction methods.

Selective predation on lumbricids by golden plover Pluvialis apricaria.

Predation by golden plovers Pluvialis apricaria (L.) on lumbricids was studied on a hayfield in Iceland in early June 1977 by means of stomach analyses and formalin sampling. Lumbricids, Allolobophora caliginosa (Sav.) and Lumbricus rubellus Hoffm., constituted about 90% (fresh wt) of food remains in stomachs. The plovers had taken relatively more A. caliginosa than L. rubellus and a relatively larger portion of medium-sized (by biovolume) than smaller and larger lumbricids. Thus the predation provided a selective advantage to L. rubellus over A. caliginosa and to the smallest and largest over the medium-sized individuals. The results are discussed in terms of degree of pigmentation (cryptic colouration), burrowing behaviour, activity, and density of the lumbricids. It is suggested that the selective effects of predation may vary considerably from one year to another.

Short-Term Colonization Success of Lumbricid Founder Populations

Small founder units (two or four individuals) of adult Allolobophora caliginosa (Sav.) and Lumbricus rubellus Hoffm. kept in netbags were introduced into a hayfield in NE Iceland. After 14 months 82% of the A. caliginosa units and 35% of the L. rubellus units still contained live adults, and the rates of individual survival were of the same magnitude. Differences in survival between units of A. caliginosa were presumably associated with position of the netbags in the field (exposure, etc.) and initial density of founders; thus, survival seemed to decrease with increasing density. A. caliginosa produced significantly more cocoons and newly hatched individuals than L. rubellus. Colonization success is discussed in relation to the experimental conditions and the difference between the two species in terms of differences in their ecology. It is suggested that A. caliginosa achieves a relatively high post-dispersal colonization success by having a relatively long life-span as adult rather than having a high birth rate.