Individual study: Responses of plant and the microbial communities to burning, mowing and turf translocation in the restoration of a former boreal hayfield at Tylldalen, Hedmark, Norway
Antonsen H. & Olsson P.A. (2005) Relative importance of burning, mowing and species translocation in the restoration of a former boreal hayfield: responses of plant diversity and the microbial community. Journal of Applied Ecology, 42, 337-347
The extensive loss of species-rich grasslands in Europe due to land-use changes has led to many attempts to recreate biodiverse grassland. A field experiment was performed in a boreal former hayfield in Norway to test the efficacy of turf inoculation, mowing and spring burning to reinstate a species-rich sward. Changes in plant diversity, composition, productivity and the soil community were recorded over a 3-year period.
Study site: The experiments were conducted in Tylldalen, a subalpine mountain area in Hedmark County, eastern Norway. The site experiences a continental climate. Annual precipitation in 2000, 2001 and 2002 was 555 mm, 712 mm and 497 mm, respectively, and mean annual temperatures 2.3 °C, 0.6 °C and 1.3 °C, respectively. The growing season (days with average temperature ≥ 5 °C) ranges 130 to 140 days.
The receiver site, a species-poor hayfield, was situated at c. 600 m above sea level. Soil was loamy sand (pH 5.7). Over the last 50 years field management involved mowing once each summer and autumn grazing. Crop rotation (3 years with potatoes and 7 under grass) had reduced the need and use of pesticides. The field was last ploughed in 1996 and then sown with a seed mix of 60% timothy Phleum pratense ssp. pratense, 15% meadow fescue Festuca pratensis, 10% cock's-foot Dactylis glomerata and 15% clover Trifolium spp. Once a year, until 1999, the field had been sparsely top-fertilized with commercial fertilizer (18:3:15 NPK, 200 kg/ ha).
The donor site was a relatively species-rich old field situated at c. 900 m above sea level. Traditional management comprised haymaking, cattle and sheep grazing, and occasional ploughing. Ploughing and haymaking ended in 1986 and 1990, respectively, since when the field had been only grazed at low-intensity, by sheep. In 1997, 43 plant species were recorded, including species normally found in traditionally managed meadows, such as viviparous bistort Bistorta vivipara, moonwort Botrychium lunaria, eyebright Euphrasia stricta, thread rush Juncus filiformis, autumn hawkbit Leontodon autumnalis, highland cudweed Omalotheca norvegica and alpine cinquefoil Potentilla crantzii. The soil was a brunisol gleisol (pH 5.4). The level of available soil P was 80% lower than at the receiver site.
Experimental design: The experiment was a randomized block design with seven replicate blocks. Within each block, there were six 3 m × 3 m plots, separated from each other by a 1 m pathway. In May 2000, turfs (20 × 20 × 20 cm³) were excavated at the donor site and transplanted in the centre of each of 21 of the 42 experimental 1 x 1 m plots at the receiver site. The management treatments i.e. mowing, burning and set-aside (control) were randomly applied within the transplant and non-transplant plots.
Mowed plots were cut to 2–3 cm with a motorized bar mower in mid-June and late August–early September each year. Cuttings were raked off. Burning occurred in mid-May after snowmelt. The litter was raised up by raking prior to ignition to promote a more efficient burn. Burning removed the litter and produced a ground cover of ash, but because of the short duration and low fire intensity, direct effects on the soil community were not expected to occur. As the hayfield had been cut in 1999, it was not feasible to burn in the first year (2000) because of low amounts of litter.
Vegetation surveys: In mid-July to early August in 2000 and 2002, plant species composition in two permanent 0.5 × 0.5 m subplots within each plot was recorded. In plots that had turf transplants, the monitoring squares were placed halfway between them. A grid, consisting of 16 squares (12.5 × 12.5 cm), was placed over the subplots and the number of squares in which each species was present was counted.
Net primary productivity (NPP) was determined each year by cutting two 0.5 × 0.5 m squares per plot at the end of the growth season. In mown plots, cuttings were also sampled prior to the first mowing and added to the autumn cuttings to estimate NPP. Samples were dried and weighed.
Soil temperature, moisture, respiration and nutrients: Soil temperature was measured at three depths (0, 2 and 5 cm) in burned and control plots (at six randomly chosen locations per plot), 1 week and 1 month after burning in 2001, using a digital thermometer. Soil moisture in burned and control plots was measured by taking four soil cores (0–5 cm depth, core diameter 2 cm) randomly within each plot. The cores were divided into two (0–2 cm and 2–5 cm) and pooled for each plot. These samples were weighed immediately and after drying to estimate water content. Soil respiration (CO2 flux) was measured in situ using a portable infrared gas analyser. Three measurements were taken per plot between 10:00 and 14:00 h (to minimize any diurnal fluctuation) in July 2000 and 2002.
Six additional soil cores (0–10 cm depth, core diameter 2 cm) were taken randomly within each plot on five occasions: June 2000, September 2000, June 2002, July 2002 and August 2002. Soil for subsequent fatty acid analysis was frozen. The rest was air-dried and analysed for pH, total carbon (C), N, P and plant-available P, K, magnesium (Mg) and calcium (Ca).
Fatty acid analysis: Lipids were extracted from soil samples (see original paper for methods). The fatty acid residues of the neutral lipids and phospholipids were transformed into free fatty acid methyl esters, which were identified and quantified using gas chromatography. Bacterial and saphrophytic fungi biomassess were subsequently estimated (see original paper for methods used).
Few changes in the vegetation and soil communities could be related to the transplanted turfs. Mowing increased plant species richness and diversity, mainly by enhancing the number and frequency of forb species. Small forb species were promoted by mowing, whereas tall leafy grasses declined. Mowing increased soil respiration and stimulated arbuscular mycorrhizal fungi. Thus whilst mowing caused significant changes in both the vegetation and the soil microbial community, spring burning caused only few changes relative to the control treatment.
Overall, the presence of transplanted turfs was poorly related to the observed changes in vegetation and below-ground community.
Plant diversity and production: The number of vascular plant species increased from 28 (six grasses, 22 forbs) to 32 (eight grasses, 24 forbs) during the 3-year experimental period. The most dominant grasses were smooth meadow-grass Poa pratensis, couch Elymus repens, Phleum pratense and Festuca pratensis, the most dominant forbs were creeping buttercup Ranunculus repens and dandelion Taraxacum spp.
The effect of treatments on species richness varied over time. In 2000, the average number of species did not differ between treatments (10-11 spp.), whereas in 2002 more species were found in mown plots (c.14 spp.) than the other plots (c.11 spp.). A similar effect on the species diversity index was found (Shannon Weiner Index: c.1.9-2.0 all plots in 2000; mown – c.2.2, burn – c.1.9, control – c.1.8 in 2002) . The turnover of species was also higher, more new species were found and fewer species disappeared from mown plots than control and burned plots.
The average numbers of grass and forb species were influenced by treatment. In 2002, the numbers of both grass and forb species were significantly higher in mown than in control plots, but only for forbs did mown plots have a significantly higher number than burned plots. The relative frequency (i.e. mean abundance per species per plot) of plants also changed in response to management. Mown plots had a lower frequency of grasses than controls after the first cut and the same relative difference was seen in 2002. Burned plots had a frequency of grasses intermediate between mown and control plots.
At the species level, several distinct responses to management treatments were seen. By 2002, grasses such as P.pratense and E.repens were significantly more abundant in control and burned plots relative to mown plots, whereas grasses such as F.pratensis and D.glomerata were more abundant in mown plots. The strong negative effect of mowing on P.pratense was evident after the first cut in 2000. Among the herbs, white clover Trifolium repens, thyme-leaved speedwell Veronica serpyllifolia and the rosette species Taraxacum spp. were significantly more abundant in mown than in control and burned plots in 2002. In addition, heath speedwell Veronica officinalis was only recorded in mown plots.
The NPP of different years reflected the length of the growing season (121, 99 and 109 days for 2000, 2001 and 2002, respectively). The effect of management on NPP varied between years. There was no difference in NPP between treatments in 2000 and 2002, but there was a significantly lower NPP in mown than in control and burned plots in year 2001.
Soil temperature and moisture after spring burning: On both recording dates and at all three depths the soil temperature was significantly higher in burned plots compared to controls, but soil moisture was not significantly affected by burning.
Soil nutrients, respiration and microbial biomass: Total C, N and P, plant-available P, Ca and Mg and pH were not affected by the treatments. Plant-available K varied between management treatments and years. In 2002, plant-available K was significantly lower in mown than in burned plots, whereas no effect of treatments was found in 2000.
The effect of management on soil respiration varied between years. In 2002, soil respiration was significantly higher in mown (447 (µmol CO2/m²/h) than in burned (334) and control plots (347), while there was no detected differences in 2000. Soil respiration was higher in plots with turfs added than controls. Mowing stimulated of arbuscular mycorrhizal fungi.
Conclusions: The results demonstrate the importance of reducing sward height in order to promote plant species diversity in former boreal hayfields. Eliminating accumulated litter by spring burning has little influence on species composition if the sward is allowed to grow tall. Mowing was most efficient way of enhancing biodiversity. The results also show that mowing-mediated changes in above-ground plant communities may stimulate below-ground symbiotic micro-organisms, potentially resulting in a positive feedback on ecosystem development.
Note: If using or referring to this published study, please read and quote the original paper, this can be viewed at: http://blackwellpublishing.com/submit.asp?ref=0021-8901