Publikasjoner
NIBIOs ansatte publiserer flere hundre vitenskapelige artikler og forskningsrapporter hvert år. Her finner du referanser og lenker til publikasjoner og andre forsknings- og formidlingsaktiviteter. Samlingen oppdateres løpende med både nytt og historisk materiale. For mer informasjon om NIBIOs publikasjoner, besøk NIBIOs bibliotek.
1994
Forfattere
Knut Solbraa Jon Olav BrunvatneSammendrag
The project started in 1976/77 with a series of experiments in five burned areas. These were:Fyresdal, 80\" E and 591\" N, 610 m.a.s.l. on shallow moraine soilHeddal, 95\" E and 594\" N, 380 m.a.s.l. on shallow moraine soilElverum, 114\" E and 605\" N, 225 - 265 m.a.s.l., sites on shallow moraine soil and sandy, sedimental soil respectivelyDeset,\"113\" E and 612\" N, 300 m.a.s.l. on sandy, sedimental soilkrestrmmen, 112\" E and 614\" N, 300 m.a.s.l. in a rocky, former river bed. The site at Deset burned in 1959, krestrmmen in 1980 and the others in 1976. The experiments covered more than 30 000 plants, mainly of Scots pine (P. sylvestris), but also a few lodgepole pine (P. contorta). Direct seeding after scarification was tried as well, but in a small scale only. Mainly the last years results are fully reported here, because the others are published earlier.The following conclusions are drawn from the complete project.1. High soil temperatures in addition to fresh cut stumps and dead trees permitted the accumulation of pathogens after the fire. These cover Hylobius abietis and other insects feeding on bark or needles of the young conifer plants. Their attacks culminated the year after the fire, and were neglectible from the third year. Treating the plants with pyretroides before planting significantly reduced bark damages. Treatment against needle eaters should be repeated annually, but is not permitted in Norway.2. The fungus Rhizina undulata killed a great share of the transplants near fresh cut stumps in Elverum. These plants died within a few weeks after planting. Also this attack continued during the first two years only. The picture of attacks in Elverum was confirmed after a prescribed burning, close to the wild-fire area, in 1984. A suspension of systemic fungicides gave the plants a good protection against Rhizina undulata. The fungus was also found in high densities at krestrmmen and with a few specimen in Fyresdal and Heddal. It seems however, to be no relations between the presence of the fungus and mortality of plants these places.3. Direct seeding after ground scarification showed good results in areas not exposed to erosion. These small pine seedlings were not attacked to the same extent as older plants. Potential seedtrees should therefore be left in order to initiate natural regeneration.4. The combination of low temperatures during the growth season and the attack of the fungus Gremmeniella abietina was the main reason for total mortality of the Scots pine plants at Deset.5. If chemical protection against insects and fungi is unwanted, planting on sandy soil should be done in the spring the third year after the fire. Treatment with insecticides and fungicides may allow earlier planting.6. On sandy soil, the annual height increment culminated after 9 years, and decreased rapidly from about 35 to 10 centimeters. Planting later than three years after fire reduced growth compared to earlier planting. Application of nitrogen fertilizers initiated continued high growth rate. On sandy soil, fertilization with approximately 130 kg nitrogen should start 8 to 10 years after the fire. The nitrogen status of plants may be checked by needle analysis. A concentration of up to 1,6 - 1,8 % N is probably suitable at a tree age of 10 years. A mulch layer of 5 cm of crude sewage sludge gave the best growth of all treatments. On moraine, growth reductions were considerable less than on sandy soil.7. Application of nitrogen fertilizers increased damage both by Gremmeniella abietina and moose. Heavy fertilization may also increase the risk of climatic damages and reduces the final lumber quality. Therefore, fertilization should be limited to stands on deep soil layers, favourable climatic conditions and areas that are not exposed to heavy winter browsing by moose.
Forfattere
Petter NilsenSammendrag
Det er ikke registrert sammendrag
Forfattere
Christian Nellemann Mette Goul ThomsenSammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag
Forfattere
Halvor SolheimSammendrag
Det er ikke registrert sammendrag
Forfattere
Erik NæssetSammendrag
Det er ikke registrert sammendrag
Forfattere
Øystein Johnsen Gunnar Haug Ola Gram Dæhlen Borgny Sveen Grønstad Anne Tove RognstadSammendrag
Det er ikke registrert sammendrag
Forfattere
Petter NilsenSammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag
Forfattere
Tron EidSammendrag
Information about the number of trees per ha in stands in development class III-V is useful for several reasons; for yield forecasts on forest level with economic calculations, for planning of thinning regimes and for cost calculations with respect to logging.The aim of this work has primarily been to map variations within stands for the number of trees per ha in order to predict how many sample plots to be distributed in a stand with certain requirements for standard error. Different sample plot sizes have been considered in this context. Several questions have also been discussed in order to settle inventory instructions for sampling number of trees per ha in practical forest management planning.The data materials have been collected from 230 stands with 4836 sample plots. The number of trees has been sampled on 100 m2, 200 m2, 300 m2 and 400 m2 sample plots, with the majority on 200 m2 plots (Table 1). The time consumption, when the sampling of the number of trees is carried out by an inventory crew of one person, has been studied for 17 stands.Table 2 shows that the mean standard deviation and the mean coefficient of variation between plots within stands are 256 trees/ha and 39.5% for 200 m2 sample plots. There are quite large differences between sites. In general the standard deviation increases and the coefficient of variation decreases when the number of trees per ha increases.Regression functions have been developed in order to predict standard deviation and coefficient of variation between plots within stands for the number of trees per ha. Different stand attributes are used as independent variables, and the functions are based on 200 m2 sample plots (Table 3). R2 are generally low. It is therefore quite likely that the predicted values become too large in some cases, and too small in others. Table 4 shows how this works for the single sites included in the data material.Table 5 shows standard deviations and coefficients of variation for stands where the number of trees is recorded with different plot sizes in the same stand, and with identical plot centers.Table 6 shows additions and deductions when the standard deviation and the coefficient of variation for other plot sizes than 200 m2 are predicted. These figures have to be used together with the regression function in Table 3. It should be emphasized that the additions and deductions are based on data from relatively few stands.Table 7 shows the mean time consumption per plot for measurements on different plot sizes. Table 8 shows the estimated number of sample plots to be distributed in a stand according to different requirements for standard error.Fig. 1 shows the estimated time consumption per stand according to different requirements for standard error. In development class III the estimated time consumption is lowest with 100 m2 plots, while the differences between the plot sizes in development class IV-V are very small.Fig. 2 shows how a regression function (S2) might be used to predict the standard deviation between plots within stands for number of trees per ha. Fig. 2 also shows how many plots which have to be distributed in a stand according to different requirements for standard errors. Sampling number of trees in practical planning is discussed in chapter 3.3.In a relascope survey it is recommended to sample the number of trees directly instead of indirectly by means of the tariff number. A direct method generally provides for the most accurate results. A direct sampling of number of trees also provides for lower time consumption than indirect sampling, if the requirements for accuracy are the same.In a relascope survey the most efficient strategy is to distribute the same number of sample plots in each stand both for sampling the number of trees and for sampling the basal area. Sample plot sizes of 100 m2 in development class III and 200 m2 in development class IV-V usually provide for a satisfactory accuracy.It is recommended to use 200 m2 sample plots in development class III and 400 m2 sample plots in development class IV-V if the requirements for accuracy are high. Also if the estimation of volume in each stand is carried out by means of aerial photographs, a direct sampling of number of trees through field work will be the most accurate method. A direct sampling in the field, however, will be more expensive than a sampling by means of interpretation on aerial photographs.A more precise comparison of these two methods, both with respect to accuracy and time consumption, should be carried out. If a systematic sample plot survey for large areas is carried out, a direct sampling of number of trees in each stand might be carried out if the economical part of the prognosis is important.