Habitat Preferences for the Corn Crake (Crex crex) in Sweden
Erik Andersson
Degree project in biology, Master of science (2 years), 2009 Examensarbete i biologi 30 hp till masterexamen, 2009
Biology Education Centre and Biology Education Centre, Department of Population Biology, Uppsala
University
Table of contents
Abstract ... 2
Sammanfattning ... 3
Introduction ... 4
History ... 4
Biology and Ecology ... 4
Population status ... 5
Method ... 6
Results ... 7
Vegetation type ... 7
Moist condition ... 8
Vegetation height ... 8
Distance to water ... 9
Multiple logistic regression ... 9
Discussion ... 10
Vegetation type ... 10
Moist condition ... 11
Distance to water ... 11
Conservation actions ... 11
Acknowledgments ... 12
References ... 12
Appendix ... 14
Abstract
The Corn crake (Crex crex) is a bird within the rail family that have during the second half of the 20
thcentury suffered a major decline in numbers, not only in Sweden but in Europe as well and is classified vulnerable (VU) by the Swedish red list and near threatened (NT) by the IUNC red list 2008. From being a common bird in Sweden during the 18
thand the 19
thcentury with an estimated population size of 60 000 pairs the population today is believed to only consist of 250 ‐1000 individuals. The most important factor behind the decline is the agricultural revolution that took place between 1910‐1940 across most parts of Europe.
Instead of as previously using hand held tools on small meadow patches, machines were now introduced and the meadows were drained and combined which led to big habitat losses all over Europe. This forced the Corn crake out onto more dangerous areas such as hay and silage meadows where the bird easily became victim to the fast going grass mowing machines. To learn more of which kind of habitat the Corn crake prefers in a present day agricultural landscape in central Sweden an inventory was done where 27 locations with Corn crakes were identified as well as 27 places without. Analyses of the various factors that were collected at each location indicate that the most important feature when it comes to habitat choice is the vegetation height followed by the vegetation type. Whether the area is wet or dry seemed to be of importance as well but there were no significance with distance to water.
Sammanfattning
Kornknarren (Crex crex) är en fågel inom rallfamiljen som under den senare delen av 1900‐
talet genomgått en drastisk minskning i antal individer, inte bara i Sverige utan även i stora delar av Europa. Den är idag klassad som hotad (VU) enligt den svenska rödlistan och missgynnad (NT) enligt IUNC’s rödlista 2008. Från att ha varit en vanlig fågel under 1700‐
1800 talet med en beräknad populationsstorlek upp mot 60 000 par så har antalet sjunkit till dagens 250‐1000 individer. Den största anledningen bakom minskningen tros vara den revolution inom lantbruket som skedde mellan 1910‐1940 över större delen av Europa.
Istället för handredskap och små ängar så började man nu använda sig av maskiner samt dränera de tidigare små ängarna för att få större, vilket innebar habitatförluster för kornknarren. Arten blev nu tvungen att söka sig till farligare platser såsom hö och ensilageängar där fågeln lät föll offer för de snabbgående slåtterkrossarna. För att undersöka vilken typ av habitat som kornknarren föredrar i ett modernt jordbrukslandskap i centrala Sverige så utfördes en inventering där 27 platser med kornknarr identifierades samt 27 platser där den saknades. Analyser av en rad olika faktorer som noterades vid varje plats tyder på att vegetationshöjden följt av vegetationstyp är de viktigaste faktorerna när det kommer till att välja habitat. Mindre viktigt var det om området var fuktigt eller ej samt om det var nära till vatten.
Introduction History
During the latter half of the 20
thcentury the agricultural landscape in Sweden went through a drastic change. The old standard, where most of the work was done by man and animal on a small scale disappeared. Instead machines were introduced which led to more efficient farming methods and the small fields were drained and enlarged (Berg & Pettersson 2007).
This revolution occurred not only in Sweden but in most parts of Western Europe as well and it had a strong negative impact on the birdlife connected to the agricultural environment.
According to Fuller et al. (1995), who did a comparison between two atlases from 1970 and 1990 in England, 24 of 28 farmland bird species have had their distribution range reduced and 15 of 18 species had suffered a population decline. No such survey has ever been done in Sweden but 46 % of the animals on the Swedish Red List are connected to farmland areas and 15 % of these are farmland birds which give an indication that there has been a similar downfall in Sweden.
The Corn crake (Crex crex) is a farmland bird that often could be heard on the fields and meadows in the beginning of the 20
thcentury (Pettersson 2007). Due to the change in agricultural management many of these areas disappeared and this has forced the Corn crake to more dangerous areas such as; grasslands that are managed for hay and silage and crop fields (Green et al. 1997) These are mainly agricultural areas with a lot of disturbances and a high risk of being killed by machines during the hay season. According to Crockford et al. (1998) there has been a 20 ‐ 50 % decline in Corn crake numbers during the last 20 years across Europe. Since the Corn crake is well known among farmers and the general public as an indicator of a healthy farmland landscape its value is highly regarded and much research is done to save it.
Biology and Ecology
The Corn crake belongs to the Rallidae family and can reach a length of 27‐30 cm and weigh between 140‐200 g. It is smaller in body size compared to the more common water rail (Rallus aquaticus) and has light brown wings and red‐brown legs (Ulfstrand 2003). The male has a very distinct song (sounding like “errp‐errp”) which can be heard from a distance of up to one kilometer and can continue for several hours during the night (Pettersson 2007). The song can also be heard during other hours but then in a less frequent manner (Green et. al 1997). The female on the other hand only makes a few sporadic, almost silent calls; instead it spends its time hiding in tall vegetation (Tyler & Green 1996).
In Sweden, predators are not considered to affect the populations particularly much.
However, in very local areas a Corn crake can be eaten by a mink or a cat but it is not considered common (Pettersson 2007). The Corn crake is an omnivorous bird and their diet varies. During the breeding season the birds feeds off of many different invertebrates that are found on the ground or on plants (Ulfstrand 2003) but it has been recorded to eat small fish and amphibians as well (Green et al 2007). The diet also seems to differ geographically.
In Poland the Corn crake mainly lives of insects whereas in Scotland and Ireland it prefers
mollusks and earthworms. The prey of Corn crakes can be found in high abundances in other
areas than the chosen habitat. Researchers have therefore drawn the conclusion that the
Figure 1a) The distribution range of the Corn crake with its breeding grounds shown in grey and its wintering areas in black (Wikipedia 2008). b)Swedish Corn crake locations based on data from 2008 (Nilsson & Södercrantz 2008).
birds are more specified in accordance to the habitat they live in rather than the diet (Green 1997).
In contradiction to earlier beliefs the Corn crake male is polygynous, meaning that a male mates with several females during breeding season (Pettersson 2007). The female is attracted to the male by his singing and after the mating the male leaves to try to find another mate and the female is left to build a nest and to raise the offspring. The nest is usually found close to the location where the male sung. The male’s song does not only attract females but other males are also drawn close to the area and thus it is not unusual to hear two or more birds at one location. When the male mates again he often does so in another territory (Pettersson 2007).
The bird can be found in a large belt stretching from the western part of central Europe in to the western part of Siberia (Figure 1a). The largest population is found in Russia where numbers can be as high as 100 000 – 1000 000 individuals. This is mainly because the modernization of the agriculture has not been as drastic as it has been in Western Europe. In Sweden the total numbers of Corn crake is estimated to be somewhere between 250‐1000 (Crockford et al. 1998; Green et al. 1997, Schäffer & Green, 2001). In the end of the 19
thcentury the Corn crake were much more abundant, researchers believe that there could be up to 60 000 breeding pairs in Sweden (Ulfstrand, 2003). The Corn crake arrives to the breeding grounds in the end of May, beginning of June, and returns to its wintering areas in southern Africa in the end of August (Brink et al. 1960). In Sweden the bird can normally be found in the southern part of Sweden with extra strong populations in the county of Uppland and on the islands Öland and Gotland (Pettersson 2007) (Figure 1b).
Population status
After the drastic decrease in Corn crake numbers during the 20
thcentury the Corn crake’s
status is now considered to be vulnerable (VU) according to the Swedish red list (Gärdenfors
2005) and near threatened (NT) according to the IUCN red list (IUCN 2008). The number of
observed Corn crake’s has increased since the middle of the 1980’s but the figures varies between years (Berg & Pettersson 2006).
The population in Sweden is believed to be too small to be able to sustain itself; instead it is believed that birds are added from areas surrounding Sweden. This is supported by the fact that the increase in Sweden is also evident in nearby areas within the Corn crake’s distribution range such as the Baltic countries (Pettersson 2007). A recent study also indicates that the genetic variation is similar to some degree in the different populations (Kempe 2008).
The aim of this study was to make an inventory of the Corn crake in Uppsala County and try to identify possible Corn crake habitats preferences and see if there is any type of habitat that is more preferred compared to others.
Method
The inventory took place between the 3
rdand the 25
thof June 2008. The method applied was partially derived from Åke Berg’s inventory done in 2006 in Uppsala and Västmanland’s County (Berg 2006). The main differences were that there were no pre‐chosen areas in this census. The areas were instead decided by the help of the Svalan database, a bird reporting website on the internet (Nilsson & Södercrantz 2008), oral communication with ornithologists Mats Edholm and Ulrik Löthberg as well as mail correspondence with members of the public.
The sites primarily chosen were well known areas with continuous yearly visits by Corn crakes as well as some areas with less known activity. Each of the areas was first inventoried by car to get an overview and to separate the areas with corn crake from those without. The Corn crake can be heard from distances of up to 1 km and it is therefore possible to do this by car (Berg 2006). In each area stops were made at a 500m‐intervall and took approximately five minutes, at locations that were considered Corn crake suitable habitats.
Since the bird can easiest be heard during night hours, each site was visited once between 11 p.m. to 4 a.m., on nights when weather conditions were favorable (in this case no precipitation and a clear sky) (Berg 2006).
For each location where a bird was found a randomized spot 100 meters from the original position was chosen as a comparative location for areas without Corn crake. This was done by randomly drawing a direction from a compass and then walking for 100 meters in that direction. The sites were re‐visited the day after to measure and collect different variables.
In total 28 Corn crakes were found on 11 different locations (Table 1) (Appendix 1).
The different variables were; vegetation type, moist condition, vegetation height and distance to water. The vegetation type was divided into six different categories (settlement, pasture, undisturbed, forest, agricultural land and fallow). Settlement includes all kinds of areas where interference made by man, such as houses etc. etc., could be found.
Undisturbed areas are areas where the vegetation has been left untouched and started go give way for brushes and small shrubberies. The forest category includes all types of forests and the fallow category includes grasslands, hay fields and fallows.
The moist condition was estimated by using a scale between 1 and 5 with 1 being dry and 5 very wet. Vegetation height was measured by using the sward stick technique (Stewart et al.
2001) and the distance to water by using the GPS‐coordinates for each location and comparing it with the closest stretch of water from a map.
Table 1. The eleven areas that were visited during the inventory.
Name Coordinates
Alstasjö (Hamra) ! Lat: N 59º 44' 30.03"
! Long: E 17º 14' 7.00"
Alstasjö (Skjutfält) ! Lat: N 59º 45' 15.81"
! Long: E 17º 16' 1.04"
Bergby (Vendel) ! Lat: N 60º 10' 26.72"
! Long: E 17º 35' 53.84"
Bälinge mosse ! Lat: N 60º 1' 50.47"
! Long: E 17º 26' 36.04"
Exarby ! Lat: N 60º 14' 13.27"
! Long: E 17º 37' 9.78"
Gunnarsbo/Månkarbov. ! Lat: N 60º 13' 3.19"
! Long: E 17º 28' 54.56"
Gunnarsbo ! Lat: N 60º 13' 31.35"
! Long: E 17º 28' 33.69"
Kyrksjön ! Lat: N 60º 14' 49.90"
! Long: E 17º 42' 48.31"
Marma ! Lat: N 60º 30' 23.51"
! Long: E 17º 28' 57.47"
Revelsta ! Lat: N 59º 48' 51.43"
! Long: E 16º 56' 41.48"
Salsta ! Lat: N 60º 2' 47.27"
! Long: E 17º 44' 20.97"
Results
The results show a significant difference between sites occupied by Corn crakes and random sites for vegetation type, moist condition and vegetation height but not for distance to water (see below). A multiple logistic regression (Chatfield & Collins 1980) showed that vegetation height was more strongly related with Corn crake presence than the vegetation type while neither moist condition nor distance to water were significance at all (table 2).
Vegetation type
Analysis of the vegetation type were done with a !2‐test and showed a difference between
areas with Corn crake and areas without (!"#"$%&'()*+#,()-#<0.001). The largest differences
were found in forest, where no Corn crake’s were found, and in fallow areas, where most
Corn crake’s were found (Figure 2).
Figure 2. The number of Corn crake’s found at the various vegetation types.
Moist condition
The result for moist conditions indicates that the Corn crake has a preference for wetter areas compared to more drier o./0)12#3%&4,()*+#,&%3$()-#0.0034) (Figure 3).
Figure 3.Moist preferences for Corn crake, where 1 is very dry and 5 very wet. Vertical bars show standard deviation.
Vegetation height
According to the results the Corn crake prefers 2566/7) 8/9/252:;.) 12#,%'"() *+#$4() -<0.0001)
with a median height of 39.6 centimeters.
Figure 3. Grass height preferences for Corn crake.
Distance to water
There were no significant difference (2#<%44() *+#,3%=3() -#0.082) between areas with Corn crake and distance to water compared to areas without Corn crake (Figure 4).
Figure 4. The distance to water in areas with or without Corn crake.
Multiple logistic regression
A multiple logistic regression showed a significance for vegetation height (p#0.000768) and 8/9/252:;.)2>-/)1-#0.025) b?2).;2)+;7)@;:02)A;.*:2:;.)1-#&%<,B);7)*:025.A/)2;)C52/7)1-#&%"DB) (Table 2).
Table 2. Deviance table showing the result from the multiple logistic regression.
Df Deviance AIC LRT Pr(Chi)
Vegetation type 1 47.01 55.01 5.05 0.025*
Vegetation height 1 57.6 65.6 15.64 0.000768***
Moist condition 1 44.05 52.05 2.09 0.15
Distance to water 1 43.13 51.13 1.16 0.28