Advanced

Igenväxande hagmarkers förekomst och tillstånd i västra Götaland : en utveckling av två arbetsmetoder

Setterby, Ylva (2004) In Lunds universitets Naturgeografiska institution - Seminarieuppsatser
Dept of Physical Geography and Ecosystem Science
Abstract (Swedish)
Detta examensarbete är gjort på uppdrag av skogsvårdsstyrelsen i Västra Götaland. Det övergripande syftet är att finna en metod som visar förekomsten av igenväxande hagmarker i Västra Götaland, samt ett sätt att beskriva tillståndet på dessa. Tillståndet skall beskrivas ur naturvårdssynpunkt och produktionssynpunkt. Tyngdpunkten ligger på naturvård. Även rådgivning om de praktiska åtgärderna behandlas.
Abstract
The rural landscape of Sweden has changed rapidly during the 20:th century. Measurements from 1927 compared to those from 1990 show that the area of non managed pasture in Sweden has declined by 86.7 %. Since a lot of threatened and rare species are dependent of this kind of land, these figures are alarming. During the last 30-40 years non managed pasture has declined due to abandonment, tree planting and preparation with fertilization and weed control. Today most cattle and grazing animals are found on sown and fertilized fields. This is largely due to the economical financing and the management advice that the Government provided until the 1980’s. Today tree production and environmental issues are described as equally important in the... (More)
The rural landscape of Sweden has changed rapidly during the 20:th century. Measurements from 1927 compared to those from 1990 show that the area of non managed pasture in Sweden has declined by 86.7 %. Since a lot of threatened and rare species are dependent of this kind of land, these figures are alarming. During the last 30-40 years non managed pasture has declined due to abandonment, tree planting and preparation with fertilization and weed control. Today most cattle and grazing animals are found on sown and fertilized fields. This is largely due to the economical financing and the management advice that the Government provided until the 1980’s. Today tree production and environmental issues are described as equally important in the forestry act. One of the main tasks of the National Board of Forestry is to work towards the goals of that act.
This thesis is based on an assignment from the National Board of Forestry. The aim is to find an effective and reliable method to examine the occurrence and the state of overgrowing pasture in Västra Götaland in Sweden. The state of these areas is supposed to be examined both in terms of nature restoration and tree production. Furthermore there will be a discussion over suitable counter-measures for the chosen areas. The areas of interest are non managed pasture, which stopped being grazed 10-40 years ago. The work area is located 10 km northeast of Borås and had to fulfill three criteria: plentiful of overgrowing pasture, high values of nature interest and that the area was covered by the new satellite images at the National Board of Forestry.
The state of these areas is examined by fieldwork, so the field protocol is an important part of the work. The field protocol was created in cooperation with the National Board of Forestry and is very extensive as a lot of criteria were to be examined. Different kinds of values of nature interest, like different kinds of production conditions, demand different kinds of treatment. The highest values of nature interest were mostly found among the big older trees. When finding single objects of high natural value like these big older trees it is enough to keep the vegetation low around the tree, so that it stands freely and much light comes in. If there are more values of nature interest left, rotation pasture could be one solution. Some animal species are benefit from higher vegetation but herbs and mushrooms often are benefit from lower and some insects demand a mix of vegetation in its immediate surroundings. If there are no such values and the tree cover has developed quite far this could be used for production, if the quality is satisfying.
A factor that would have been of interest in the field protocol that is now missing is canopy density. This would be of interest both from a tree production point of view and to draw conclusions about the spectral signature. This can easily be measured in field and would also give a more certain measure of dominating tree species.
It is of great importance to take several aspects in account when deciding how to take care of an area. There is a big risk that some natural values are lost if only herbs and species richness among those get attention. It is also of great importance that the decision maker is open for the possibility that some natural values could have come after the non managed pasture area was abandoned. This state is unstable and would not continue to evolve freely. However intensive pasture is not recommended either.
Studies of the possibility to use digital satellite images for the identification of areas with overgrowing pasture were made on two SPOT5 scenes acquired 2003-06-03. Two different methods were tested for classification of satellite data: supervised and unsupervised. The collected field data was used as reference material in the classifications and evaluation of the results. The goal was that the resulting data file should cover as many visited field sites with overgrowing pasture as possible and at the same time cover as small an area as possible, so that the data file wouldn’t cover other kinds of lands. The best result was achieved with an unsupervised maximum likelihood standard classification with 40 classes, which had the highest amount used. When evaluating the accuracy it became clear that the resulting file covered the overgrowing pastures in the working area with few exceptions. The problem was that other kinds of overgrowing land (as fields) and boundaries between open land and forest were also covered. Since the mapping results were improved as the amount of classes increased, it seams like what a larger amount of classes, for example 45 or 50, would give a better result.
To get an acceptable and reliable result when mapping overgrowing pasture with classification of satellite images, both the spectral and the spatial resolution must be high, in particular the spectral resolution. (Less)
Abstract (Swedish)
Populärvetenskaplig sammanfattning: Naturbetesmarkerna minskade med 86,7% från 1927 till 1990. Denna minskning är alarmerande eftersom många idag hotade och sällsynta arter är beroende av dessa. En stor del av anledningen till detta är att staten ända in på 1980-talet i sin rådgivning danade för att dessa marker antingen skulle bearbetas med konstgödsel och ogräsbekämpning eller planteras med träd (företrädesvis granskog). Alternativet var att lämna dem orörda. Numera finner man betesdjur framför allt på sådd, gödslad åker. I dagens skogspolitik är produktion och miljö lika viktiga.

Detta examensarbete är gjort på uppdrag av skogsvårdsstyrelsen i Västra Götaland. Det övergripande syftet är att finna en metod som visar förekomsten av... (More)
Populärvetenskaplig sammanfattning: Naturbetesmarkerna minskade med 86,7% från 1927 till 1990. Denna minskning är alarmerande eftersom många idag hotade och sällsynta arter är beroende av dessa. En stor del av anledningen till detta är att staten ända in på 1980-talet i sin rådgivning danade för att dessa marker antingen skulle bearbetas med konstgödsel och ogräsbekämpning eller planteras med träd (företrädesvis granskog). Alternativet var att lämna dem orörda. Numera finner man betesdjur framför allt på sådd, gödslad åker. I dagens skogspolitik är produktion och miljö lika viktiga.

Detta examensarbete är gjort på uppdrag av skogsvårdsstyrelsen i Västra Götaland. Det övergripande syftet är att finna en metod som visar förekomsten av igenväxande hagmarker i Västra Götaland, samt ett sätt att beskriva tillståndet på dessa. Tillståndet skall beskrivas ur naturvårdssynpunkt och produktionssynpunkt. Tyngdpunkten ligger på naturvård. Även rådgivning om de praktiska åtgärderna behandlas.

Metoden som skall visa förekomsten har skapats med hjälp av flyg- och satellitbilder och data insamlad i fält. Resultatet visar igenväxande hagmarker med få undantag, men även andra igenväxande marker, som igenväxande äng, och kantzoner mellan öppen mark och skog. Det bästa resultatet uppnåddes med den klassificering av satellitbilder som gjordes med flest klasser, vilket var 40 stycken. Antagligen skulle fler klasser, till exempel 45 eller 50 stycken, ge ett ännu mer precist resultat. Dock är det nog svårt att helt undvika igenväxande åker. Om möjligheten finns att subtrahera de två felkällorna från resultatet så skulle detta vara det bästa. Användningen av denna metod bör kombineras med digitala ortofoton (flygbilder) och man bör vara medveten om felkällorna. En förutsättning för att få ett rättvisande resultat är också att satellitbilderna har en hög upplösning både rumsligt och spektralt.

Det fältprotokoll som tagits fram är mycket omfattande eftersom många kriterier skall undersökas. Olika naturvärden, liksom olika produktionsförutsättningar kräver olika åtgärder och därför måste så många som möjligt av dessa beaktas vid inventering. En faktor som hade bidragit till slutsatser om såväl trädproduktionen som metoden om förekomst baserad på fjärranalys, men som nu saknas, är ett täthetsmått på det uppväxande trädskiktet. Detta hade gått att mäta antingen med relaskop eller dataklave och måttband. Båda dessa mätningar är enkla att genomföra och kräver relativt lite tid i anspråk, framför allt mätning med relaskop. I övrigt fungerar fältprotokollet relativt bra.

Det är viktigt att inventera ett område utifrån flera aspekter innan beslut om skötsel fattas. Risken är stor att vissa naturvärden går om intet om bara exempelvis gräs och örter och deras artrikedom beaktas. Exempel på förlorade naturvärden skulle vara insekter, svampar, mossor och lavar.

Man bör även vara öppen inför den möjligheten att ett högt naturvärde kan ha tillkommit efter att marken övergetts. Detta innebär dock inte att lokalen för den skull skall fortsätta att utvecklas fritt eftersom detta stadium i regel är instabilt. (Less)
Please use this url to cite or link to this publication:
author
Setterby, Ylva
supervisor
organization
alternative title
Overgrowing pasture in västra Götaland in Sweden : an evaluation of two working methods
year
type
H1 - Master's Degree (One Year)
subject
keywords
geomorphology, satellitbilder, pasture, inventering, nature restoration, flygbilder, digital satellite images, pedology, cartography, naturgeografi, climatology, physical geography, geomorfologi, marklära, kartografi, klimatologi
publication/series
Lunds universitets Naturgeografiska institution - Seminarieuppsatser
report number
102
language
Swedish
additional info
Patrik Olsson och Emma Roland, Skogsvårdsstyrelsen.
id
1332767
date added to LUP
2005-10-12
date last changed
2011-12-12 16:11:03
@misc{1332767,
  abstract     = {The rural landscape of Sweden has changed rapidly during the 20:th century. Measurements from 1927 compared to those from 1990 show that the area of non managed pasture in Sweden has declined by 86.7 %. Since a lot of threatened and rare species are dependent of this kind of land, these figures are alarming. During the last 30-40 years non managed pasture has declined due to abandonment, tree planting and preparation with fertilization and weed control. Today most cattle and grazing animals are found on sown and fertilized fields. This is largely due to the economical financing and the management advice that the Government provided until the 1980’s. Today tree production and environmental issues are described as equally important in the forestry act. One of the main tasks of the National Board of Forestry is to work towards the goals of that act.
This thesis is based on an assignment from the National Board of Forestry. The aim is to find an effective and reliable method to examine the occurrence and the state of overgrowing pasture in Västra Götaland in Sweden. The state of these areas is supposed to be examined both in terms of nature restoration and tree production. Furthermore there will be a discussion over suitable counter-measures for the chosen areas. The areas of interest are non managed pasture, which stopped being grazed 10-40 years ago. The work area is located 10 km northeast of Borås and had to fulfill three criteria: plentiful of overgrowing pasture, high values of nature interest and that the area was covered by the new satellite images at the National Board of Forestry.
The state of these areas is examined by fieldwork, so the field protocol is an important part of the work. The field protocol was created in cooperation with the National Board of Forestry and is very extensive as a lot of criteria were to be examined. Different kinds of values of nature interest, like different kinds of production conditions, demand different kinds of treatment. The highest values of nature interest were mostly found among the big older trees. When finding single objects of high natural value like these big older trees it is enough to keep the vegetation low around the tree, so that it stands freely and much light comes in. If there are more values of nature interest left, rotation pasture could be one solution. Some animal species are benefit from higher vegetation but herbs and mushrooms often are benefit from lower and some insects demand a mix of vegetation in its immediate surroundings. If there are no such values and the tree cover has developed quite far this could be used for production, if the quality is satisfying.
A factor that would have been of interest in the field protocol that is now missing is canopy density. This would be of interest both from a tree production point of view and to draw conclusions about the spectral signature. This can easily be measured in field and would also give a more certain measure of dominating tree species.
It is of great importance to take several aspects in account when deciding how to take care of an area. There is a big risk that some natural values are lost if only herbs and species richness among those get attention. It is also of great importance that the decision maker is open for the possibility that some natural values could have come after the non managed pasture area was abandoned. This state is unstable and would not continue to evolve freely. However intensive pasture is not recommended either.
Studies of the possibility to use digital satellite images for the identification of areas with overgrowing pasture were made on two SPOT5 scenes acquired 2003-06-03. Two different methods were tested for classification of satellite data: supervised and unsupervised. The collected field data was used as reference material in the classifications and evaluation of the results. The goal was that the resulting data file should cover as many visited field sites with overgrowing pasture as possible and at the same time cover as small an area as possible, so that the data file wouldn’t cover other kinds of lands. The best result was achieved with an unsupervised maximum likelihood standard classification with 40 classes, which had the highest amount used. When evaluating the accuracy it became clear that the resulting file covered the overgrowing pastures in the working area with few exceptions. The problem was that other kinds of overgrowing land (as fields) and boundaries between open land and forest were also covered. Since the mapping results were improved as the amount of classes increased, it seams like what a larger amount of classes, for example 45 or 50, would give a better result.
To get an acceptable and reliable result when mapping overgrowing pasture with classification of satellite images, both the spectral and the spatial resolution must be high, in particular the spectral resolution.},
  author       = {Setterby, Ylva},
  keyword      = {geomorphology,satellitbilder,pasture,inventering,nature restoration,flygbilder,digital satellite images,pedology,cartography,naturgeografi,climatology,physical geography,geomorfologi,marklära,kartografi,klimatologi},
  language     = {swe},
  note         = {Student Paper},
  series       = {Lunds universitets Naturgeografiska institution - Seminarieuppsatser},
  title        = {Igenväxande hagmarkers förekomst och tillstånd i västra Götaland : en utveckling av två arbetsmetoder},
  year         = {2004},
}