Establishment of marram grass Ammophila arenaria from culms, seeds and rhizomes on the island of Voorne, Zuid, Holland, the Netherlands
van Der Putten W.H. (1990) Establishment of Ammophila arenaria (marram grass) from culms, seeds and rhizomes. Journal of Applied Ecology, 27, 188-199.
BackgroundBecause of its natural ability to colonize and stabilize sand dunes, marram grass Ammophila arenaria is often used to control coastal erosion. In this Dutch study three different methods of establishing marram were assessed: planting bundles of culms (traditional method), sowing seeds, and disc-harrowing rhizomes. The effects of applying a slow-release fertilizer on growth and establishment were also examined.
ActionStudy site: The field experiment was undertaken on the coastal dunes on the island of Voorne, the Netherlands (51.5ºN, 4.05ºE).
Experimental design: In March 1987, marram grass was established in 5 x 5 m plots from culms (the traditional method), seeds and rhizomes.
Culm stock used was cut 10 cm below the sand surface from a vigorous wild stand. Bundles of 13 culms were planted in holes 15-cm deep in a 50 x 75 cm grid. Most (95%) of tillers had at least two apparently viable buds.
Seeds were sown (mixed with sand to improve their distribution) at 200, 400 and 600 seeds/m² (7.5, 15 and 22.5 kg/ha, respectively). The sand surface was lightly disc-harrowed to bury the seeds 3-5 cm deep.
Rhizomes were collected from the top 2 m soil layer in a vigorous marram stand and planted at 20, 40 and 60, 15-cm pieces/m² at about 10-15 cm depth (15-cm lengths had 2-5 viable buds).
Prior to planting/sowing, a slow-release fertilizer (slow-release 'Osmocote', active for 12-14 months at a soil temperature of 21ºC) was broadcast and disc-harrowed to a depth of 10-1 5 cm at three levels: 0, 80-20-20 and 160-40-40 kg NPK/ha.
All treatments (planting method, density and fertilizer level) were replicated four times.
After applying treatments (except culm planting) the sand surface of all plots was stabilized, to a degree, by disc-harrowing to 5 cm depth of 0.5 kg/m² of straw. Finally, culms were planted with fertilizer added directly to planting holes (instead of by broadcasting) at 80-20-20 and 120-30-30 kg NPK/ha in some additional plots.
Sampling and evaluation: In July 1987, emerged seedlings, primary shoots (i.e. emerged buds of the rhizomes), and tillers (from culms and rhizomes) were counted. The sampled area, or numbers of plants counted, depended on the variability within the treatment. In every plot with seedlings, five subplots of 1 m² were examined; in every plot with rhizomes, eight subplots of 0.18 m; and 25 bundles were examined in culm plots.
In October 1987, a 3 x 1 m strip was cut at ground level in all seedlings or rhizome plots. A subsample was weighed fresh and the numbers of tillers counted before drying to determine dry weight. Numbers of tillers of the bundles of culms were assessed and dry weight was estimated by linear regression. Plant samples were dried, ground and analysed for N, P and K in order to assess effectiveness of the slow-release fertilizer. Fertilizer grains were collected in April 1988 and analysed.
Sowing seeds at 200, 400 and 600/m² and rhizomes (15 cm long fragments each with at least two buds) at 20, 40, and 60 fragments/m², showed that higher rates resulted in higher numbers of seedlings and primary shoots (as one might expect) in July 1987. However, although the intermediate sowing rate resulted in significantly more seedlings than the low rate, the high rate gave no further increase.
In July 1987, the rhizomes produced more tillers as fertilizer levels increased. Fertilization also increased tiller production of the bundles. There were no significant differences between broadcast and planting-hole fertilizer application.
In October 1987, after one growing season, above-ground dry matter production, number of tillers, and the weight per tiller were no longer affected by sowing rate or rhizome planting density. Production of biomass and tillers was only increased in fertilized plots.
Although recovery in above-ground plant parts of N and P was fairly low, slow-release fertilizer (80-20-20 kg/ha) increased biomass significantly. Fertilizing rhizomes with 160-40-40 kg NPK/ha produced significantly more dry matter than with 80-20-20 kg, but this had no effect on sown or planted marram.
Rhizomes produced more tillers and biomass than bundles of culms and seedlings when fertilizer was applied. Without fertilizer, culms gave most biomass production, but seedling growth was very poor.
Conclusions: This study suggests that good marram establishment can be expected from disc-harrowing rhizomes in combination with sand stabilization by straw and the application of slow-release fertilizer. However, as there are other considerations, e.g. costs and long-term vegetation development, all three methods of marram establishment trialed are being applied experimentally on a large scale in the Netherlands.
Note: If using or referring to this published study, please read and quote the original paper, this can be viewed at: