Tarnava Mare 2017 Biodiversity Survey Summary Report Report editor & lead scientist: Dr Bruce Carlisle – Geography & Environmental Sciences, Northumbria University. Science team: Zuni Askins, Silvia Cojocaru, Graham Forbes, Sian Green, Paul Leafe, Chris Mackin, Cecilia Montauban, James O’Neill, Huma Pearce, Sophie Perry, Peter Thomas. Project leader: Toby Farman. Assisted by: Bogdan Ciortan, Paul Hangan, Mihaela Hojbota, Dragos Luntraru, Valentin-Ioan Marcos, Bogdan-Mihai Mehedin, Alin-Marius Nicula, Silviu Simula, Ovidiu Tanasa, Daniela Vasilache. With thanks to all the staff at Fundatia ADEPT, all the dissertation students and volunteers.
65
Embed
Tarnava Mare 2017 Biodiversity Survey Summary Report · June to 8 August 2017, in eight villages within the Tarnava Mare Natura 2000 site. In total, 48 days fieldwork were undertaken,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Tarnava Mare 2017 Biodiversity Survey Summary Report
Report editor & lead scientist: Dr Bruce Carlisle – Geography & Environmental Sciences, Northumbria University.
Science team: Zuni Askins, Silvia Cojocaru, Graham Forbes, Sian Green, Paul Leafe, Chris Mackin, Cecilia Montauban, James O’Neill, Huma Pearce, Sophie Perry, Peter Thomas. Project leader: Toby Farman. Assisted by: Bogdan Ciortan, Paul Hangan, Mihaela Hojbota, Dragos Luntraru, Valentin-Ioan Marcos, Bogdan-Mihai Mehedin, Alin-Marius Nicula, Silviu Simula, Ovidiu Tanasa, Daniela Vasilache. With thanks to all the staff at Fundatia ADEPT, all the dissertation students and volunteers.
2.2 Land use ........................................................................................................................................ 5
11.0 Site Trends ..................................................................................................................................... 49
Much of the survey work is carried out along “the transects” which are 3 linear routes per village.
Each route was selected with the aim of traversing land covers and land uses that are representative
of the village’s surroundings. The routes are constrained by accessibility. The “central transect” is
approximately 4km long and runs along the valley floor, upstream and downstream of the village.
This transect runs through the village, usually alongside a road, near to the stream, and through
more intensely farmed land. “West” and “east” transects are approximately 6km long and each takes
a roughly semi-circular route from the valley floor up the valley sides, usually into less intensely
farmed land, meadow grassland, pasture and woodland. There have been no significant changes to
the transect locations over the five years.
Page 5
There are seven main survey teams covering farmer interviews, grassland plants and land use
mapping, grassland butterflies, birds, small mammals, large mammals and bats. An Orthoptera
survey was also undertaken this year, after a trial run in 2016. Further details of the methods of each
team, and any notable alteration of methods, are given in the following sections.
2.1 Farmer interviews An extensive set of farm interviews were carried out in 2017, for the second year (the first being
2015). In 2017 a total of 137 interviews were completed, with between 6 to 22 interviews at each
village. Very few interviews took place in 2016 due to staff injury. In 2015 a total of 153 interviews
were completed, with between 9 to 29 interviews at each village. 41 and 48 interviews were
completed in 2013 and 2014 respectively. The number of farmers interviewed varied amongst the
villages, depending on the presence and effectiveness of a local person to make contacts, the
willingness of farmers to participate, and how busy the farmers were. There was no strategy to
selecting farmers – the participants were whoever was willing and available to be interviewed. The
number of interviews in 2015 and 2017 is noticeably higher than other years. This is primarily due to
the time and persistent effort put into arranging and carrying out the interviews. The years with
small sample sizes mean year-on-year farm statistics derived from the interviews are unreliable.
However, data from the 2015 and 2017 interviews will be much more representative of each village’s
farm characteristics.
The farmer interviews involved asking a fixed set of questions covering topics such as farm
characteristics (size, age etc.), crops grown, livestock, hay cutting dates and so on. The questions
asked in 2013 and 2014 have been repeated in all subsequent years. Additional questions were
added from 2015 onwards, to investigate mowing technique, use of communal grazing and future
plans. These additional questions were actually first trialled during the second half of the 2014
season.
2.2 Land use A land use survey was first undertaken in 2013, and then repeated in 2016 at the first four villages,
and 2017 (first 5 villages). The transect routes were walked and notes were taken of the land use
types adjacent to the walked route, marking the transition points with GPS, using a fixed set of land
use types. A high resolution satellite image (Worldview 2 imagery) was also consulted and annotated
to help determine the extent of each land use and to survey areas not adjacent to the route if their
land use type was discernible from a distance.
The GPS points are being loaded into a GIS (Geographic Information System), displayed over the high
resolution satellite image, and used as a template for digitising polygons depicting the land use
around each transect. There is ongoing work to produce a land use map of the whole Tarnava Mare
Natura 2000 site, using image processing of satellite imagery. Although a preliminary map has been
produced and provided to Fundatia ADEPT, the mapping technique continues to need refinement. It
is proving challenging to distinguish between several of the land use types from satellite imagery.
Also, it can be difficult to ascertain some land uses in the field. For example, it can be difficult to tell
whether an area of grassland continues to be mown, is used for occasional low intensity grazing, or is
abandoned. Results of this analysis will be reported independently.
Page 6
2.3 Grassland plants The plant team re-surveyed the sites from previous years, using the same methods. Apold, Crit and
Daia have now been surveyed over four years, while the other 5 villages have 5 years of surveys. To
decide on locations of sites in 2014 and 2013, grassland was visually partitioned into high, medium
and low nature value (HNV, MNV, LNV) categories based on indicators such as the presence of farm
weed species, evidence of current use, shrub encroachment, and abundance and variety of
wildflowers. On each transect a minimum of six plot locations were identified with the target of 2
HNV, 2 MNV and 2 LNV plots. This was not always achieved due to the prevalence or absence of
grassland categories.
Each grassland plant plot is 50m by 5m. The surveyors walk the length of the plot counting the
number of individuals of 30 species defined as indicators of HNV dry grassland in Fundatia ADEPT’s
guide “Indicator Plants of the High Nature Value Dry Grasslands of Transylvania” (Akeroyd &
Bădărău, 2012). Betony was also counted as, although it is an indicator for damp grasslands, it is
relatively abundant and widespread on the surveyed grasslands.
The species in flower change as the fieldwork season progresses. Surveying a plot on a different date
is likely to give different results. This is of particular relevance when comparing data from different
years to assess change. Also, as the season progresses, the number of mown fields increases and the
number of fields available for survey, with standing wild flowers, decreases. This could affect the
representativeness of a village’s plant surveys, and could also affect comparisons between years if
the survey date is not similar. In 2017 there were different surveyors for the first 5 and last 3 villages.
The 2017 data for Daia is currently missing from this report as it is hiding in an unknown box in
Saschiz village.
2.4 Grassland butterflies The grassland plant plots are also used for the butterfly surveys, although they are extended to 50m
by 10m. All butterflies seen in a 5 minute walk along the length of the plot are counted. Butterfly
counts take place between 10am and 4pm, to avoid the cooler parts of the day. Butterfly counts do
not take place if it is raining. However, there still remains wide variation in the abundance of
butterflies due to weather conditions and time of day. The team aims to repeat the survey of each
site two or three times (dependant on suitable weather conditions) to reduce the impact of weather
conditions on the data. The number of times plots were surveyed is summarised in Table 2.2. Nearly
all plots were surveyed two or three times. Weather caused 7 sites to be surveyed just once. The
“Not surveyed” sites are now considered as a reserve set of sites. There is a growing set of nearby
and similar alternative plots to allow surveys even if the main site has been mowed. Each year the
butterfly survey leader has changed, although the same leader has been used in 2015 and 2017.
Table 2.2. Summary of how many times plots were surveyed at each village.
Village N sites Not surveyed Once 2 times 3 times N surveys
Apold 12 - - 12 - 24
Crit 18 2 3 12 1 30
Daia 11 - - 12 - 24
Malancrav 12 1 - 9 2 24
Mesendorf 15 3 - 7 5 29
Nou Sasesc 12 - - 3 9 33
Richis 12 - - 12 - 24
Viscri 13 1 4 4 4 24
Page 7
2.5 Birds Standing point counts were undertaken at 500m intervals along each of the three transects for each
village, giving a target of 13 point counts per east and west transect, and 9 point counts per central
transect. The 2017 point count locations were very similar to those of 2015 and 2016. Some point
counts from 2014 and 2013 were removed in 2015 due to proximity to a point on another transect.
Each point count lasted 10 minutes and all individuals seen or heard were counted. The surveys
began soon after dawn, between 0545 and 0615, and were usually completed before midday.
The time of year and amount of mown grass will affect the numbers and species of birds being
recorded. Also as the morning progresses, there is a very noticeable decrease in the amount of bird
song and activity. So, points further along a transect tend to have fewer birds. Most surveys were
repeated, walking the transect in the opposite direction to compensate for the time of day effect.
Five Apold West points, two Daia Central points, the Daia East transect, and Mesendorf South
transect were only surveyed once due to heavy rain. A few points at Malancrav and Viscri could not
be surveyed at all, or only once, due to presence of shepherd dogs. The 2017 survey leader was the
same as in 2015 and 2016. There were different surveyors in 2014 and 2013.
In addition to the point counts, the mist netting and ringing survey was continued in 2017. Three nets
were set up from dawn until about 1100 in scrub areas adjacent to farmland, across bird movement
corridors. In 2017 the mist netting and ringing took place at 6 villages, compared to all 8 villages in
2016 and 2015, and 5 villages in 2014.
A night-time corn crake survey was also continued at the first three villages. The approximate
distance and direction of corn crake calls was recorded on linear walks through potential corn crake
habitat. The survey was relatively late in the breeding season and so very few records were obtained,
which does not give an accurate estimate of corn crake abundance. Data from the corn crake survey
is not presented in this summary report.
2.6 Small mammals The small mammal survey methods were re-designed for 2014 and continued in 2015, following
limited trapping success in 2013. Cheaper plastic traps were used instead of folding Sherman traps.
The lower cost meant more traps could be bought, and replaced when stolen. Grids of 4 by 5 or
single lines of 20 traps were laid out in different habitat types (low and high nature value (LNV and
HNV) grassland, and scrub/woodland edge), dependent on characteristic and shape of the habitat
type. In 2016 and 2017, more expensive traps were used – but not as expensive as the 2013 Sherman
traps – as they are better for animal welfare and hopefully more effective at trapping small
mammals. The same basic trap grid layout was used as in 2014 and 2015, but the locations of some
trap grids were adjusted to reduce chances of trap damage or theft, and due to habitat changes from
mowing and grazing. Traps were set each evening and checked the following morning. The trap lines
/ grids were in place for at least 4 nights. Surveyors have changed each year.
2.7 Large mammals The large mammal surveys commenced in 2014 have continued in each subsequent year. Two survey
techniques are used: camera traps and observation of signs such as scat and tracks.
Page 8
Camera traps were set up in woodland locations. At Richis, 10 cameras were set up for 3 days, then
another 4 for 2 days. This first week revealed some malfunctioning cameras. In the other villages, 9
to 11 cameras were set up in two sets of locations for 4 or 5 days. The number of cameras varied due
to malfuntioning and one incident of attempted theft. The cameras were placed in strategically
chosen woodland locations that seemed likely to experience frequent large mammal activity. Unlike
previous years, no cameras were actually stolen, partly due to attaching cameras with padlocked
cables. Batteries and SD card were stolen from one camera on one occasion.
The survey of large mammal signs involved walking the east and west transects, recording sightings,
scat, tracks, digging and any other signs of large mammal presence, and GPS coordinates of their
location. The same technique and routes have been used every year from 2014 to 2017. The large
mammal survey team leader has changed every year.
2.8 Bats An extensive and systematic survey of bats has been repeated very year from 2014 to 2017. Various
methods were used at each village, including roost surveys, bat activity transect surveys, static
detector surveys and mist netting.
The bat survey results will be reported separately.
2.9 Orthoptera Surveying of Orthoptera – grasshoppers and bush crickets –was trialled in 2016. The trial ran at Nou
Sasesc and Mesendorf. The team tried different techniques to assess the diversity and abundance.
Then in 2017 Orthoptera surveys were undertaken at 5 villages (Mesendorf to Malancrav). The
technique involves arranging 5 to 8 people in a line spaced approximately 5 metres apart. Each
person then walks forward 5 steps, or 10 steps, 15 steps and so on, so that they form a diagonal line.
At that spot each person then identifies the first few Orthoptera they spot, or uses nets and pots to
capture species for later identification. After 3 minutes the surveyors re-position so that the end
result is a large X formation. The possible survey techniques are heavily constrained by the need to
minimise trampling of the hay meadow plants, so for example a sweep netting technique cannot be
used. The technique has questionable rigour and repeatability and a better approach is needed to
produce more thorough and representative data on Orthoptera communities.
Page 9
3.0 Vital statistics
This section presents selected summary information to give a concise overview of the data.
Remember that various factors can influence the data including natural fluctuations in wildlife
populations, natural variation from year to year due to changing vegetation phenology, timing of the
survey relative to the day of the year and time of day, surveyor knowledge and experience, and
sample size. The methods described above have been designed to limit these issues, while allowing a
relatively rapid biodiversity assessment across the Tarnava Mare.
Figures 3.1 to 3.3 summarise the farm interview data. Figure 3.1 shows the mean farm size for each
village’s interview respondents for each year. There is variation between years, particularly for
Mesendorf. The 2015 and 2017 data are based on a lot more interviews and can be considered more
representative. The graph illustrates that the 2014 and 2013 interviews did not give an accurate
representation of each village. Consequently, only comparisons between 2015 and 2017 are made in
this report. Richis and Viscri appear to have smaller farms than other villages.
Figure 3.1. Farm size, showing the mean total farm area for 2013 to 2017 interviewees. Village
abbreviations: AP – Apold, CR – Crit, DA – Daia, MA – Malancrav, ME – Mesendorf, NS – Nou Sasesc,
RI – Richis, VI – Viscri.
Figure 3.2 shows the mean extent of cultivation, hay meadows and other agricultural land use at
each village in 2015 and 2017. The large difference between the 2015 and 2017 data for Nou Sasesc
is probably due to a relatively small sample of 6 farmers for this village in 2017. Five villages have
larger total farmed area in 2017 than 2015, possibly a sign that farm sizes are increasing. There has
been greater increase in “Other” than hay or cultivation. This other category includes pasture used
Page 10
for livestock grazing. The “Other” category increases at all villages except Richis. Changes to the hay
and cultivation categories vary between villages. Despite the greater sample size in 2015 and 2017, it
is still felt that this may not give an accurate picture of the extent of different farming types across
the villages. This is partly due to the still limited sample size, and also the potential inaccuracy of
farmer responses. But these differences in the 2015 and 2017 data are signs that should be watched
over the coming years.
Figure 3.2. Farm land use, showing the 2015 and 2017 mean area of cultivation, hay, and other use
as shaded stacked columns. Village abbreviations: AP – Apold, CR – Crit, DA – Daia, MA – Malancrav,
ME – Mesendorf, NS – Nou Sasesc, RI – Richis, VI – Viscri.
Page 11
Figure 3.3 shows the mean number of milk cattle, ewes and lambs at each village in 2015 and 2017.
There are notable differences between villages and between years. Crit and Daia have a large
number of sheep on average. The number of sheep is much greater than the number of milk cattle at
all villages except Nou Sasesc. Nou Sasesc seems to have fewer livestock than the other villages.
Again, despite the greater sample size in 2015 and 2017, it is still felt that this may not give an
accurate picture of the number of livestock across the villages. The very large difference at Viscri
between 2015 and 2017 is probably at least partly due to sampling biases. There is large variation
amongst farms. Small traditional farms may have one or two cows and a few sheep or goats. More
specialised farms have large flocks of sheep. The results shown depend heavily on how many of these
different types of farm were included in the survey. However, all villages apart from Malancrav and
Richis have greater numbers of lambs in 2017 than 2015. This difference needs to be monitoried in
the coming years.
Figure 3.3. Farm livestock, showing mean number of lambs, ewes and milk cattle in 2015 and 2017.
Village abbreviations: AP – Apold, CR – Crit, DA – Daia, MA – Malancrav, ME – Mesendorf, NS – Nou
Sasesc, RI – Richis, VI – Viscri.
The village farming summaries listed below have been produced by compiling all of the farmer
interview responses (see section 4 for details). The previously described caveats due to limited
sampling apply here too. There are a number of signs that farming is changing, with more livestock
grazing seeming to be the most common type of change.
Apold increased intensification - due to less hay production and more livestock
low change potential
Crit reduced intensification – due to less cultivation, fewer livestock
increased change potential – favouring more silage and cultivation
Page 12
Daia slightly lower intensification – fewer livestock, more communal grazing
reduced change potential – all becoming more stable
Malancrav reduced intensification – due to reduction in all farming aspects, i.e. less farming overall
reduced change potential – all becoming more stable
Mesendorf increased intensification – due to less communal grazing, less hay production
increased change potential – favouring more silage, crops, livestock
Nou Sasesc slightly increased intensification – more livestock, less communal grazing, more hay, but less hand-mowing
increased change potential – favouring less hay, more silage
Richis slightly increased intensification – less hand-mown hay
low change potential
Viscri increased intensification – due to more livestock, more hay production
low change potential
Figures 3.4 and 3.5 summarise the grassland plant data. For each survey site, a “3-way diversity”
score has been calculated (see section 4 for details) and is summarised for each village in figure 3.4.
All villages have a wide range of “3-way diversity” scores, although this is less so for Apold. No village
has scores that are noticeably greater than other villages. At Crit the median score is tending to
increase year on year. At Daia and Nou Sasesc the median score is tending to decrease.
Figure 3.4. Site-level grassland plant survey “3-way diversity” scores, summarised for each village, for
each year. Higher scores indicate higher diversity of indicator species. In each boxplot: the horizontal
line represents the median value; the height of the box represents the inter-quartile range (IQR); the
length of the whiskers represents whichever is shorter of the maximum/minimum value or 1.5 times
the IQR; circles represent outliers (data points beyond the whisker range).
Page 13
Figure 3.5 shows the total abundance of indicator plants across the years at each village, and all
villages combined. For all villages combined the abundance decreased each year to 2016 which was a
potential cause for concern. However, in 2017 that trend was reversed. No individual village has a
consistent trend in indicator plant abundance over the 5 years. Abundance at Richis and Nou Sasesc
does show signs of a decreasing trend. This may be because the surveys at these villages are at the
start of the season, and this has become earlier over the years. Crit and Viscri seem to generally
increase in abundance. Again this may only be due to the changing dates of the surveys. Apold has
notably fewer indicator plants than other villages. Crit has notably higher numbers – this is primarily
caused by a very high amount of Betony at a few Crit sites.
Figure 3.5. Total indicator plant abundance per ha for each village, and all-village average, for each year.
Page 14
Figure 3.6 summarises the grassland butterfly diversity. There is a wide range of butterfly diversity levels at all of the villages, but less so at Viscri, Apold and Daia. All villages seemed more diverse in 2016 compared to previous years, with the exception of Crit (not surveyed). This trend seems to have continued in 2017 at Apold, Mesendorf, Nou Sasesc and Viscri, while the other villages have stayed releatively constant over the last 2 years. In 2016 Malancrav had notably higher diversity indices than other villages, but this has become more “normal” in 2017. Viscri has a consistently lower median value every year. There are no clear signs of a reduction in butterfly diversity at any of the villages.
Figure 3.6. Plot-level butterfly diversity data, summarised per village, for each year. See Figure 3.4 for explanation of the box plot elements.
Page 15
Red-backed shrike abundance is summarised in figure 3.7. Daia consistently has noticeably higher
numbers of red- backed shrike than the other villages. In all villages fewer red-backed shrike were
seen in 2015 compared to 2014. In 2016 numbers declined further in 3 villages, but in 2017 those
trends were reversed. Nou Sasesc is the only village that now has signs of a possible consistent
decline in red-backed shrike numbers. This needs to be monitored closely in future years to check
whether this is a more long-term change.
Figure 3.7. Number of red-backed shrike per point count, per village, for 2013 to 2017.
Page 16
Figure 3.8 indicates that small mammal abundance has fluctuated markedly at all villages. 2017 was
the most abundant small mammal year at all villages except Richis and Viscri. The population crash in
2015 has been followed by recovery in numbers at all surveyed villages. High fluctuation in
abundance seems to be a normal pattern, as can often be the case with small mammals.
Figure 3.8. Small mammal abundance per trap night, per village, for 2013 to 2017.
Page 17
The large mammal signs of presence data summarised in Figure 3.9 show that all villages had less
frequent signs in 2017 than 2016. This may be due to the generally drier conditions in 2017 giving
hard ground and fewer prints. Mesendorf has consistently had more signs than other villages, but
not in 2017. Nou Sasesc, Richis and Viscri consistently have fewer signs than other villages.
Figure 3.9. Signs of large mammal presence per kilometre, per village, 2014 to 2017.
Page 18
Figure 3.10 summarises the number of species of Orthopteran at each village. There are no clear
differences between the 5 villages, although Crit and Malancrav have wider variation between sites
than other villages. Mesendorf has a consistently high Orthopteran species richness.
Figure 3.10. Plot-level orthoptera species richness in 2017, summarised for each village. See Figure 3.4 for explanation of the box plot elements.
Page 19
4.0 Farmer interviews
The data collected during the farm interviews over the 4 years (2016 is excluded due to very small
sample numbers) is presented in table 4.1. Note that in both 2013 and 2014 the number of
interviews that could be conducted was low. This reduces the reliability of the data, both in terms of
comparing villages and considering changes from year to year. A lot more interviews were
conducted in 2015 and 2017, and the results differ notably from the previous years (see figure 3.1
and table 4.1). It is assumed that the 2015 and 2017 data are more representative and reliable. Only
the 2015 and 2017 data are compared in this report.
In Table 4.1, a greater than 50% difference between 2015 and 2017 is highlighted in green or red, for
an increase or decrease respectively. These highlighted cells reveal some potentially interesting
differences between villages. The changes at each village can be summarised as:
Apold: more cultivation, more other, more milk cattle, more lambs
Crit: less cultivation, less ewes, more lambs
Daia: less beef cattle, more lambs
Malancrav: more other, less beef cattle, less lambs
Mesendorf: more other, less milk cattle, less beef cattle
Nou Sasesc: more hay, more other, less beef cattle, more lambs
Richis: less other, less beef cattle, less lambs
Viscri: more hay, more other, less beef cattle, more ewes, more lambs
Also wolf and bear attacks are reported to have increased in 5 villages and overall – but decrease in
Nou Sasesc and Richis. This may be a change in the awareness and recollection of attacks amongst
the interviewees. Or this may be a real increase in wolf and bear attacks, perhaps as a result of an
increase in the number of livestock.
In 2015 and 2017 additional questions on mowing technique, use of communal grazing and future
plans were included. The farm interviews capture a wide range of information. Two index values
have been calculated to summarise this range of information and to try to pick out key differences
between villages. The data used to calculate the indices and the indices are shown in tables 4.2 and
4.3.
The intensification index is the average of the following 5 scores:
Livestock score: mean number of livestock per interviewee divided by 150 (uses the sum of
all the types of livestock recorded). More intense farming can involve larger herds/flocks.
Communal grazing score: 1 minus the proportion of interviewees who use the communal
grazing. Intensification can involve abandoning the communal grazing system and grazing
your own animals on private pasture.
Hand mown score: 1 minus the proportion of the hay area that is mown by hand. More
intense farming involves using hay cutting machinery instead of hand mowing.
Hay score: 1 minus the proportion of the total farm area that is used for hay. More intense
farming is associated with abandonment of hay meadows.
Cultivation score: the proportion of the total farm area that is used for cultivation. More
intense farming is associated with more crop cultivation.
Page 20
Table 4.1. Part 1. Farm interview results for 2013 to 2017. Green – 2017 data 50% or more greater than 2015. Red – 2017 data 50% or less than 2015.
Interviews Years Farm
area (ha) Cultivation
(ha) Hay (ha)
First hay cut
Other (ha)
Milk cattle
Beef cattle
Ewes Lambs Goats Pigs Horses & donkeys
Buffalo Wolf and
bear attacks
Ap
old
2014 7
24.7 (6 to 40)
25 (0.75 to 90)
6.2 (0 to 15)
9 (0.75 to 20)
10 Jul (01 Jul to 08 Aug)
9.9 (0 to 60)
11.9 (0 to 45)
1.4 (0 to 4)
33.6 (0 to 120)
12.4 (0 to 80)
3.3 (0 to 18)
6 (1 to 15)
1 (0 to 2)
-- 0
2015 13
17.2 (4 to 37)
14.9 (0 to 54)
3.5 (0 to 14)
10.1 (0 to 49)
26 Jun (15 Jun to 01 Aug)
1.5 (0 to 15)
2.7 (0 to 20)
0 (0 to 0)
65.2 (6 to 294)
5.1 (0 to 20)
4.2 (0 to 37)
5.5 (0 to 15)
0.8 (0 to 4)
-- 2
2017 17
23.9 (1 to 50)
31.2 (0 to 180)
7.7 (0 to 34)
8.9 (0 to 75)
08 Jul (30 May to 01 Aug)
14.5 (0 to 71)
9.9 (0 to 107)
0 (0 to 0)
71.6 (0 to 400)
13.6 (0 to 80)
3.6 (0 to 50)
1.9 (0 to 7)
0.6 (0 to 2)
0.3 (0 to 5)
17
Cri
t
2013 11
19.9 (8 to 40)
69.3 (3.5 to 200)
12.8 (0 to 65)
24 (0 to 115)
22 May (01 Jun to 20 Jul)
32.5 (0 to 140)
19.2 (0 to 87)
9.1 (0 to 75)
308.1 (0 to 2000)
101.4 (0 to 850)
56.5 (0 to 300)
5.6 (0 to 40)
0.6 (0 to 2)
-- 6
2014 5
24 (15 to 40)
32.8 (3 to 120)
12.9 (0 to 60)
19.6 (2 to 60)
24 Jun (30 May to 01 Jul)
0.3 (0 to 1.5)
10.4 (3 to 30)
2.4 (0 to 11)
56.8 (0 to 250)
47.4 (0 to 230)
4.6 (0 to 10)
1.6 (0 to 4)
0.8 (0 to 2)
-- 0
2015 29
22.8 (1 to 95)
21.4 (0 to 100)
10.5 (0 to 60)
12.5 (1 to 50)
28 Jun (01 Jun to 01 Aug)
3.4 (0 to 42)
14.8 (0 to 100)
0.3 (0 to 3)
92.8 (0 to 1600)
1.8 (0 to 20)
13 (0 to 150)
6.9 (0 to 100)
0.7 (0 to 4)
-- 4
2017 21
23.9 (1 to 50)
15.9 (0 to 76)
2.7 (0 to 20)
9.1 (0 to 40)
19 Jun (01 May to 15 Jul)
4.2 (0 to 35)
12 (0 to 88)
0.3 (0 to 4)
38.5 (0 to 300)
5.7 (0 to 80)
5.7 (0 to 77)
2.7 (0 to 12)
0.4 (0 to 2)
0 (0 to 0)
25
Dai
a
2014 4
23.8 (8 to 42)
27 (7 to 60)
5.8 (2 to 10)
10 (5 to 20)
09 Jul (01 Jul to 20 Jul)
11.3 (0 to 45)
18.8 (1 to 45)
0.3 (0 to 1)
302.5 (0 to 1200)
150.5 (0 to 600)
26.8 (0 to 107)
9.3 (0 to 24)
0.5 (0 to 1)
-- 3
2015 24
20.9 (3 to 50)
21.8 (3 to 80)
4.9 (0 to 18)
8.9 (1 to 60)
27 Jun (15 May to 01 Aug)
8.3 (0 to 70)
14.8 (0 to 41)
6.1 (0 to 25)
92.1 (0 to 1200)
6.3 (0 to 100)
2.5 (0 to 51)
3.9 (0 to 15)
1 (0 to 3)
-- 2
2017 21
22 (2 to 50)
26.5 (2 to 100)
5.7 (0 to 20)
10.4 (1 to 97)
20 Jun (01 May to 15 Jul)
10.5 (0 to 50)
13.1 (0 to 50)
0 (0 to 0)
51 (0 to 1000)
23.9 (0 to 500)
1 (0 to 15)
3.2 (0 to 13)
1.1 (0 to 4)
0 (0 to 0)
8
Mal
ancr
av
2013 9
28.3 (2 to 80)
26.8 (3 to 50)
7.7 (1.5 to 25)
6.5 (1.5 to 20)
02 Jul (01 Jul to 10 Jul)
12.6 (0 to 40)
14 (5 to 30)
1.2 (0 to 5)
91.2 (0 to 260)
30.8 (0 to 80)
1 (0 to 4)
5.8 (0 to 26)
1 (0 to 2)
-- 6
2014 10
14.3 (2 to 30)
8.7 (0.5 to 40)
4.1 (0.5 to 10)
1.8 (0 to 5)
25 Jul (01 Jul to 15 Aug)
2.9 (0 to 25)
6.7 (1 to 40)
1.4 (0 to 10)
25.5 (0 to 170)
5.6 (0 to 35)
1.2 (0 to 9)
3.6 (0 to 20)
0.3 (0 to 1)
-- 4
2015 20
15.4 (3 to 40)
13.5 (0 to 53)
5.5 (1 to 25)
5.5 (0 to 25)
29 Jun (15 May to 01 Aug)
3.5 (0 to 50)
8.3 (0 to 31)
1.5 (0 to 10)
49.3 (0 to 500)
11.3 (0 to 80)
6.6 (0 to 93)
5.9 (0 to 32)
0.8 (0 to 3)
-- 8
2017 19
19.8 (3 to 50)
16.1 (1 to 50)
5.5 (1 to 25)
5.1 (0 to 25)
26 Jun (15 May to 30 Jul)
5.6 (0 to 32)
6.8 (0 to 25)
0.2 (0 to 3)
38.2 (0 to 300)
3.5 (0 to 30)
1.4 (0 to 25)
4.3 (0 to 30)
0.4 (0 to 2)
0.5 (0 to 5)
10
Mes
end
orf
2013 6
29.2 (6 to 100)
197.1 (0.03 to 1000)
53.7 (0 to 300)
47.5 (0 to 200)
30 Jun (30 Jun to 01 Jul)
95.9 (0 to 500)
103.5 (0 to 560)
7 (0 to 30)
54.2 (0 to 250)
46 (0 to 250)
14.3 (0 to 70)
3 (0 to 6)
1.8 (0 to 10)
-- 13
2014 6
15.3 (9 to 20)
172.3 (7 to 680)
11.5 (0 to 40)
75.8 (5 to 300)
22 Jun (01 May to 07 Jul)
85 (0 to 380)
124.3 (2 to 650)
21.5 (0 to 64)
105.8 (0 to 600)
34.2 (0 to 200)
13.7 (0 to 70)
4.8 (0 to 20)
2.8 (0 to 15)
-- 6
2015 29
17.3 (2 to 35)
16.3 (0 to 100)
5.8 (0 to 40)
10.6 (1 to 60)
25 Jun (15 May to 15 Jul)
2.8 (0 to 30)
16.1 (0 to 200)
6.8 (0 to 100)
31.1 (0 to 450)
12.6 (0 to 185)
42.1 (0 to 500)
3.2 (0 to 14)
1.5 (0 to 7)
-- 2
2017 22
27.7 (6 to 52)
32.3 (0 to 312)
5.5 (0 to 61)
8.2 (0 to 50)
22 Jun (01 Jun to 15 Jul)
18.7 (0 to 236.59)
6.8 (0 to 70)
0.1 (0 to 2)
42.6 (0 to 500)
16.6 (0 to 200)
29.7 (0 to 300)
2.4 (0 to 20)
1.5 (0 to 8)
20.7 (0 to 439)
16
Page 21
Table 4.1. Part 2.
Interviews Years Farm
area (ha) Cultivation
(ha) Hay (ha)
First hay cut
Other (ha)
Milk cattle
Beef cattle
Ewes Lambs Goats Pigs Horses & donkeys
Buffalo Wolf and
bear attacks
No
u S
ases
c
2013 4
15.5 (10 to 29)
29 (4.8 to 53)
3 (0 to 6)
4.9 (0 to 15)
01 Jul (01 Jul to 01 Jul)
21.1 (0 to 53)
5.5 (0 to 18)
2.8 (0 to 10)
14.3 (0 to 35)
6 (0 to 17)
0 (0 to 0)
2.3 (0 to 3)
0.5 (0 to 2)
-- 0
2014 3
15.7 (10 to 23)
50.3 (5 to 100)
14.3 (2 to 30)
27.3 (3 to 70)
28 May (20 May to 10 Jun)
8.7 (0 to 26)
10 (2 to 24)
4 (0 to 12)
23.3 (5 to 35)
8.3 (0 to 14)
0 (0 to 0)
2.7 (0 to 4)
0.3 (0 to 1)
-- 0
2015 11
17.9 (5 to 24)
24 (4 to 60)
10.4 (3 to 29)
9.8 (1 to 30)
30 May (15 May to 01 Jul)
3.8 (0 to 25)
14.1 (0 to 65)
7.1 (0 to 24)
10.8 (0 to 40)
6.1 (0 to 25)
0 (0 to 0)
4.2 (0 to 15)
0.7 (0 to 3)
-- 8
2017 6
19.2 (2 to 60)
49.8 (12 to 120)
8.2 (0 to 20)
19 (6 to 50)
01 Jun (01 Jun to 01 Jun)
22.7 (0 to 80)
20.2 (0 to 40)
0.2 (0 to 1)
13.7 (0 to 80)
22.2 (0 to 130)
0 (0 to 0)
1.3 (0 to 5)
0.3 (0 to 1)
1.7 (0 to 10)
0
Ric
his
2013 5
20.2 (3 to 45)
8.6 (1.5 to 16)
3.2 (0.5 to 5)
3.6 (0 to 10)
04 Jul (01 Jul to 15 Jul)
1.8 (0 to 7.5)
3.4 (1 to 6)
2 (0 to 7)
30.8 (0 to 150)
10.2 (0 to 50)
2.6 (0 to 13)
5.4 (2 to 9)
1 (0 to 2)
-- 0
2014 7
19 (6 to 44)
5.6 (2.5 to 12)
2.1 (1 to 4)
3.5 (1 to 10)
22 May (01 May to 10 Jun)
0 (0 to 0)
2.9 (0 to 10)
0.9 (0 to 4)
43.9 (0 to 300)
10.1 (0 to 70)
0 (0 to 0)
3.7 (1 to 7)
1.6 (1 to 2)
-- 0
2015 18
22.4 (1 to 50)
12.3 (0 to 70)
4.2 (0 to 15)
3.5 (0 to 13)
26 May (05 May to 01 Jun)
5 (0 to 56)
3.8 (0 to 18)
1.3 (0 to 8)
54.9 (0 to 300)
10.1 (0 to 58)
0.2 (0 to 3)
5.4 (0 to 14)
1.3 (0 to 4)
-- 2
2017 11
21 (5 to 27)
10.8 (1 to 40)
4 (1 to 20)
4.6 (0 to 20)
11 Jun (01 Jun to 01 Jul)
2.2 (0 to 19)
4.4 (0 to 30)
0.1 (0 to 1)
49.7 (0 to 400)
1.9 (0 to 20)
0 (0 to 0)
4 (0 to 9)
0.7 (0 to 2)
0 (0 to 0)
0
Vis
cri
2013 6
18.2 (6 to 25)
14.3 (5 to 28)
1.8 (0 to 3.5)
8.4 (2.5 to 25)
08 Jul (01 Jul to 30 Jul)
4.1 (0 to 14.75)
7.2 (0 to 29)
0.7 (0 to 3)
28 (0 to 60)
12.7 (0 to 40)
0 (0 to 0)
2.7 (0 to 5)
0.2 (0 to 1)
-- 0
2014 6
20.3 (2 to 50)
9.25 (5 to 23)
2.6 (0 to 10)
5.65 (2.5 to 7.4)
01 Jul (01 Jul to 01 Jul)
1 (0 to 6)
4.33 (0 to 10)
1.83 (0 to 4)
20 (0 to 76)
9.17 (0 to 30)
0 (0 to 0)
4.17 (0 to 15)
0.5 (0 to 1)
-- 0
2015 9
20.3 (14 to 25)
6.9 (1 to 16)
1.6 (1 to 3)
4.6 (1 to 7)
30 Jun (15 Jun to 07 Jul)
1 (0 to 7)
6 (0 to 10)
0.4 (0 to 3)
19.7 (0 to 55)
1.1 (0 to 4)
0 (0 to 0)
3 (0 to 8)
0.3 (0 to 1)
-- 0
2017 20
25.8 (0 to 60)
16.7 (0 to 90)
0.9 (0 to 12)
9.3 (0 to 60)
29 Jun (15 Jun to 01 Jul)
6.2 (0 to 25)
7.7 (0 to 30)
0.2 (0 to 3)
82.7 (0 to 600)
46.3 (0 to 600)
0.2 (0 to 3)
3 (0 to 20)
1.4 (0 to 8)
0 (0 to 0)
49
All
2013 41
22.5 (2 to 100)
59.3 (0 to 1000)
13.9 (0 to 300)
17.0 (0 to 200)
22 Jun (1 Jun to 30 Jul)
28.4 (0 to 500)
25.4 (0 to 560)
4.3 (0 to 75)
119.9 (0 to 2000)
44.4 (0 to 850)
17.8 (0 to 300)
4.5 (0 to 40)
0.9 (0 to 10)
-- 25
2014 48
19.2 (2 to 50)
37.8 (0.5 to 680)
6.5 (0 to 60)
17.0 (o to 300)
27 Jun (1 May to 15 Aug)
14.3 (0 to 380)
22.9 (0 to 650)
4.1 (0 to 64)
64.9 (0 to 1200)
27.9 (0 to 600)
5.1 (0 to 107)
4.4 (0 to 24)
1 (0 to 15)
-- 13
2015 153
19.5 (1 to 95)
17.1 (0 to 100)
6 (0 to 60)
8.8 (0 to 60)
21 Jun (05 May to 01 Aug)
4 (0 to 70)
11.5 (0 to 200)
3.3 (0 to 100)
58.1 (0 to 1600)
8.2 (0 to 185)
12.4 (0 to 500)
4.9 (0 to 100)
1 (0 to 7)
-- 28
2017 137
17 (2 to 40)
23.3 (0 to 312)
4.7 (0 to 61)
8.7 (0 to 97)
23 Jun (01 May to 01 Aug)
9.9 (0 to 236.59)
9.5 (0 to 107)
0.1 (0 to 4)
51.2 (0 to 1000)
17.1 (0 to 600)
6.5 (0 to 300)
2.9 (0 to 30)
0.9 (0 to 8)
3.6 (0 to 439)
125
Page 22
The change index is intended to capture how much the farming system is likely to change in the near
future towards greater intensification. The index uses questions about whether interviewees are
likely to increase or decrease various aspects of their farming, such as numbers of sheep, or area of
cultivation, or amount of hay mown by tractor for example. An “increase” response scores +1, while
a decrease response scores -1. No response or “no change” scores 0. These scores can be summed
for each village to give a village-level measure of likelihood of further intensification. If every
interviewee responded “increase” the score would be the number of interviewees. Or if everyone
responded “decrease” the score would be minus the number of interviewees. The change index is
the average of the following 4 scores:
Hay change score: based on adding together the response sums for more/less hay mown by
hand, mower and tractor. The score is re-scaled to range from 0 to 1 where 0 would
represent all interviewees saying “increase” to all types of hay cutting, and 1 would
represent all saying “decrease”.
Silage change score: based on the response sum for more/less silage production. The score is
re-scaled to range from 0 to 1 where 0 would represent all interviewees saying “decrease”,
and 1 would represent all saying “increase”.
Crop change score: same method as silage change score but using more/less crops question.
Livestock change score: based on adding together the response sums for more/less milk cattle, beef cattle and sheep. The score is re-scaled to range from 0 to 1 where 0 would represent all interviewees saying “decrease” to all the types of livestock, and 1 would represent all saying “increase”. Table 4.2. Interview data used in the calculation of village intensification and change indices.
AP CR DA MA ME NS RI VI All
Total Farm area (ha) 2015 193.2 598.39 502.45 257.1 471.72 263.95 220.56 61.99 2569.36
The intensification and change indices are visualised in figure 4.1. The two indices show some differences between the villages. Lower left areas on the diagram represent more extensive, and less changing farming practices. Upper right areas represent more intensive, and likely-to-change farming. The thicker, black horizontal and vertical lines show that several of the arrows lie at least partly in the upper right area, as do 4 villages’ arrow ends (i.e. the 2017 values).
Fig. 4.1. The intensification and change indices for each village. Village abbreviations: ALL – all villages, AP – Apold, CR – Crit, DA – Daia, MA – Malncrav, ME – Mesendorf, NS – Nou Sasesc, RI – Richis, VI – Viscri. Synthesising information from tables 4.1 to 4.3 and figure 4.1, each village can be summarised as follows (this is the same material as in Section 3 – Vital Statistics):
APAPCR
CR
DA
DA
MA
MAME
MENS
NS
RI RI
VI
VI
ALL ALL
0.45
0.5
0.55
0.6
0.65
0.3 0.4 0.5 0.6 0.7
Change I
ndex
Intensification Index
AP
CR
DA
MA
ME
NS
RI
VI
ALL
Page 24
Apold increased intensification - due to less hay production and more livestock
low change potential
Crit reduced intensification – due to less cultivation, fewer livestock
increased change potential – favouring more silage and cultivation
Daia slightly lower intensification – fewer livestock, more communal grazing
reduced change potential – all becoming more stable
Malancrav reduced intensification – due to reduction in all farming aspects, i.e. less farming overall
reduced change potential – all becoming more stable
Mesendorf increased intensification – due to less communal grazing, less hay production
increased change potential – favouring more silage, crops, livestock
Nou Sasesc slightly increased intensification – more livestock, less communal grazing, more hay, but less hand-mowing
increased change potential – favouring less hay, more silage
Richis slightly increased intensification – less hand-mown hay
low change potential
Viscri increased intensification – due to more livestock, more hay production
low change potential The calculation of the intensification and change indices is experimental. The choice of data, and calculation method may not be appropriate. The interview data may not be representative of a village as a whole due to the limited sample size. Nonetheless this data is included in this report to promote thought and discussion.
It is important to keep collecting this farm interview data in future years to be able to more reliably
confirm whether these are genuine changes in the farming practices, or due to the sampling
differences of 2015 and 2017. However, there are a number of signs that farming is changing, with
more livestock grazing seeming to be the most common type of change.
Page 25
5.0 Grassland plants
The indicator plant data for each site have been converted to three measures to characterise the
indicator species diversity and abundance. These three measures have been combined into a single
“3-way diversity” score, which is presented in Figure 3.4 of the vital statistics. The three measures
are:
A. Richness: Species richness, the number of indicator species
B. Evenness: 1 – Berger Parker dominance index
C. Abundance: Total number of individuals of each indicator species
The “3-way diversity” score is calculated as: A + 10B + C/100. This re-scales the three measures to
similar ranges of values, and then adds them together.
Figures 5.1, 5.2 and 5.3 show the richness, evenness and abundance measures for each village, and
for all 5 survey years. There is a wide range in the values of the measures across the sites at each
village. There is some variation between years. This may be due partly to variation in the date of
survey. There will also be natural fluctuation. Annual plant species change their location from year to
year, and can change from lying within a 50m by 5m plot to outside from year to year. Year on year
changes must be interpreted with caution, and longer term trends over several years will be more
reliable. The only potentially consistent trends revealed in figures 5.1 to 5.3 are Richis richness
decreasing, Nou Sasesc and Richis evenness decreasing and Viscri evenness increasing.
Figure 5.1. Site-level plant indicator species richness measure, summarised for each village, for each
year. In each boxplot: the horizontal line represents the median value; the height of the box
represents the inter-quartile range (IQR); the length of the whiskers represents whichever is shorter
of the maximum/minimum value or 1.5 times the IQR; circles represent outliers (data points beyond
the whisker range).
Page 26
Figure 5.2. Site-level plant indicator species evenness measure, summarised for each village, for each
year. See figure 4.1 for boxplot specifications.
Figure 5.3. Site-level plant indicator species abundance measure, summarised for each village, for
each year. See figure 5.1 for boxplot specifications.
Table 5.1 presents data on the three measures and the “3-way diversity” score for each site, for all 5
years. Sites with a consistent change in the 3-way score have their name highlighted in the ‘Site’
column. A consistent change in 3-way score is deemed to be present if there is a birdlife. There is a
lot of fluctuation from one year to the next. 14 sites have consistent change, with 9 decreasing and 5
increasing – but 8 of these 14 sites do not have a full 5 years’ worth of data. The TOTAL 3-way score
Page 27
consistently decreased over the first four years, but has increased to its highest level in 2017. For
many sites, 2017 had good indicator plant diversity, particularly at Apold, Mesendorf and Viscri, but
most sites at Malancrav had notably lower values than 2016. There is a lot of variability amongst the
sites, and various factors could cause changes, including weather conditions, scheduling of the
surveys, and surveyors. However, there is evidence that the botanic biodiversity at some sites of the
Tarnava Mare may be declining, and this needs to be monitored closely.
Table 5.1. Indicator plant diversity and abundance measures for each site of each village. Dark green:
Table 5.3 shows the abundance of the 10 most common indicator species that were surveyed,
totalled for each village (the equivalent data for all indicator plants is in Appendix 1). This could
potentially mask the within-site natural fluctuations in abundance and reveal more systematic
trends. However, the differences in survey date remains an influencing factor. Table 5.3 contains a
real mixture of colours, indicating variation between years, between species and between villages.
Overall, there are 133 dark and light green cells compared to 140 red and orange cells – suggesting a
balance of increasing and decreasing abundances. The comparable figures for just 2017 are 41
increases and 37 decreases. Species that experienced a consistent decline or increase over the years
are listed in Table 5.2. These are species with a significant Spearman’s rank correlation (Prho <= 0.05)
between abundance and year. Some trends identified previously have not been maintained into
2017, while some new ones have been added. The number of decline incidences has stayed the same
from 2016 to 2017, while the number of increases in abundance has increased by 2. In terms of total
abundance across all indicator species (the righthand column of Table 5.3), no village has a
statistically significant consistent trend across all years. The 2016 report identified a possible overall
decline in indicator plant abundance. The 2017 data does not support this trend. Monitoring will
continue, and with each year there can be greater certainty as to whether these are genuine trends
in wildflower abundance, or natural variation, or due to surveying artefacts such as change in survey
date or surveying staff.
Table 5.2. Species with consistent change over five years at a village or all villages combined. Bold
indicates an additional trend added since the 2016 report. The lower half of the table lists species
where consistent change had been identified in the 2016 report, but 2017 data do not continue that
trend. Underlined species are in the top 10 in terms of average annual abundance.
Species showing consistent decline Species showing consistent increase
Jurinea – Malancrav, Nou Sasesc
Large speedwell - All
Sainfoin – Apold, Richis
Lady’s bedstraw – Daia, All
Yellow scabious –Richis
Greater selfheal – Daia
Dorycnium –Daia
Sword-leaved fleabane – Crit
Deptford pink – Daia
Betony – Richis
Greater milkwort – Mesendorf, All
White dwarf broom – Nou Sasesc,
Richis, All
Sainfoin - Viscri
Charterhouse pink – Daia, Nou Sasesc
Squinancywort – Daia
Lady’s bedstraw – Richis
Dorycnium – Mesendorf
Deptford pink – Crit
Betony – Crit, Viscri
Species no longer showing consistent decline Species no longer showing consistent increase
Lady’s bedstraw –All
Crown vetch – Apold
Dorycnium – Crit
TOTAL – Apold, All
Large speedwell – Crit, Nou Sasesc
Greater milkwort – Nou Sasesc
Siberian bellflower – Malancrav
Squinancywort –Mesendorf
Yellow scabious –Viscri
Sword-leaved fleabane – Nou Sasesc,
Viscri
Page 30
Table 5.3. Abundance of the 10 commonest indicator species at each village. Grey: no record for two consecutive years. Dark green: >= 50% increase. Light green: >= 20% increase. Yellow: <= 20% decrease. Red: <= 50% decrease.
This section reports on the 25 most abundant butterfly species, with an annual average abundance
greater than 10, as these show more reliable trends than species with few individuals observed.
Unidentified species of blue butterfly and all species of blue combined are also shown here. Data on
the full set of species are given in Appendix 2. Table 6.2 shows the abundance of each observed
butterfly species summed per village. Notable changes between years have been highlighted. These
should be interpreted with caution due to natural variability, the influence of weather during the
survey period, and changes in surveying staff.
The total number of butterflies recorded at each village should be less influenced by surveyor bias. In
2014 three villages showed notable decreases in abundance. In 2015 butterfly abundance recovered
by more than 20% in two of those villages – Richis and Viscri. But Nou Sasesc showed a further
decline. In 2016, there was a notable increase in total butterfly numbers at all villages, except
Mesendorf which experienced a 22% decrease. And 2017 was another good year for butterflies at
most villages, although Richis and Viscri had declines of greater than 20%.
Two species were recorded on the surveys for the first time in 2017. These are Scarce large blue
(Maculinea telejus) and Turquoise blue (Plebicula dorylas).
Table 6.1 shows the species that have consistently decreased or increased over the 5 years (4 years
for Apold, Crit and Daia). These are species with a significant Spearman’s rank correlation (Prho <=
0.05) between abundance and year. 19 species show a total of 50 incidents of consistent increase,
while 3 species show a total of 3 incidents of consistent decrease. So there are many more incidents
of increase rather than decline.
Butterfly biodiversity in all the surveyed villages appears to be in good health. The survey data gives
no causes for concern.
Page 32
Table 6.1. Species with consistent change over five years at a village or all villages combined. Bold
indicates an additional trend added since the 2016 report. The lower half of the table lists species
where consistent change had been identified in the 2016 report, but 2017 data do not continue that
trend. Species in red are used in the European Butterfly Indicator for Grassland Species (Van Swaay
et al., 2016)
SPECIES SHOWING CONSISTENT DECLINE
Marbled white – AP
Silver washed fritillary – NS
Essex skipper - DA
SPECIES SHOWING CONSISTENT INCREASE
High brown fritillary – RI
Weaver’s fritillary – AP, ME, NS, RI, All
Nickerls fritillary – ME, NS, RI, All
Small skipper – NS
Essex skipper – RI, All
Dingy skipper – AP, CR, VI
Small white – DA, ME
Wood white – DA, MA, NS, VI, All
Small heath – AP, DA, All
Chestnut heath –RI, All
Pale clouded yellow – DA
Ringlet – CR, DA, VI
All blues – DA
Silver studded blue – VI, All
Common blue – CR, DA, ME, RI
Short tailed blue – AP, DA, MA, ME, All
Osiris blue – DA, VI, All
Scarce swallowtail – DA, All
Map - All
Species no longer showing consistent decline
Marbled white – NS
Meadow brown – AP
Silver-studded blue – AP
Species no longer showing consistent increase
Marbled white – CR
Meadow brown – CR
Weaver’s fritillary – VI
Wood white – RI
Pale clouded yellow – CR, NS
Ringlet – AP
Common blue – VI
Blue sp. – AP, VI, All
Scarce swallowtail - AP
Page 33
Table 6.2. Grassland butterfly abundance (numbers per hectare) at each village. Grey: no sighting two years running. Dark green: >= 50% increase. Light green: >= 20% increase. Yellow: <= 20% decrease. Red: <= 50% decrease. The most abundant species with mean annual abundance > 10 shown here. Full species list in Appendix 2.
This section reports on the bird species that are listed by Birdlife International (2018) as being
associated with grassland habitats, and which were observed on average at least twice per year. Data
on the full set of species are given in Appendix 3. Table 7.2 shows the abundance of each grassland
bird species per point count at each village. The abundance as a percentage of the total number of
birds throughout the season is also used to help determine if a significant change has occurred. This
percentage partly compensates for differences due to change of surveyor each year. Overall, after a
relatively low total number of birds per point count in 2015, many more birds were recorded in 2016
and 2017 (right hand column of Table 7.2). All villages apart from Apold and Richis had a higher
number of birds per point count than in 2016. There was a substantial decline in abundance of many
species at Richis in 2017 – something to monitor next year. Three species were added that had not
previously been recorded during the point counts: common redstart (Phoenicurus phoenicurus),
lapwing (Vanellus vanellus) and purple heron (Ardea Purpurea).
The number of highlighted cells illustrates the fluctuations in species numbers between 2013 and
2017. This will partly be natural variation, but also change in surveying staff. For example, in 2014
there was a fall in the number of house sparrows and tree sparrows, but an increase in sparrow sp.,
with these trends reversed in 2015. This is very probably an artefact of the different surveyors.
Likewise there is a fall in the number of middle spotted woodpecker in 2014, but rises in great
spotted woodpecker, spotted woodpecker sp. and woodpecker sp. with the trends reversed in 2015.
The same person led the point surveys in 2015, 2016 and 2017. So these effects should be reduced
for the last three years.
The species showing a consistent trend over the 5 years in certain villages or overall are shown in
Table 7.1. These are species with a significant Spearman’s rank correlation (Prho <= 0.05) between
abundance and year. There are many more instances of a grassland species showing a consistent
increase (19) than a consistent decrease (8) at particular villages. Most of the declining trends
identified in the 2016 report, using 3 or 4 years of data, have not continued over the 5 year period.
No village stands out as having more prevalent bird population changes. All villages have both
increasing and declining species. The abundance of the declining species should be monitored closely
in the villages where they have declined. However, overall the grassland bird populations appear in
good health.
Table 7.3 summarises the ringing surveys of 2014 to 2017. The mist netting and ringing only occurred
at 5 of the villages in 2014. All eight were surveyed in 2015. Seven were surveyed in 2016, six in 2017.
In 2016 and 2017, the total number of birds ringed was slightly lower than in 2015. Most of the
notable declines highlighted in red in Table 7.3 are for species with less than 10 individuals caught in
any year. Many factors can cause these numbers to fluctuate so not too much should be inferred
from any increases or declines.
Page 36
Table 7.1. Species with consistent change over five years at a village or overall. Species in red are
associated with grassland according to Birdlife International’s (2018) online species database. Bold
indicates a new entry since the previous annual report. Striked out indicates a trend that was
identified in last year’s report but no longer continues into this year.
SPECIES SHOWING CONSISTENT DECLINE
Barn swallow –DA, MA
Bee-eater – MA, RI
Black redstart - DA
Common whitethroat - DA
Cuckoo – CR, VI
Great grey shrike – DA
Hoopoe – ME, VI
Magpie - VI
Red-backed shrike – CR, NS
Whinchat – DA, ALL
White stork – ME
Willow warbler – AP
Woodlark – VI
Wryneck – RI
Yellow wagtail – AP, ME
SPECIES SHOWING CONSISTENT INCREASE
Barn swallow – AP
Bee-eater - CR
Black redstart – CR, MA, NS
Blackbird – ALL
Goldfinch – AP, MA, NS
Great tit - RI
Hoopoe – AP
House sparrow - RI
Little owl – AP
Magpie - AP
Marsh warbler – RI, VI
Raven - ME
River warbler – CR, NS, ALL
Skylark - DA
Stonechat - CR
Thrush nightingale – DA
Tree pipit – RI
White stork – AP, VI
White wagtail – VI
Woodlark – AP, RI
Wryneck - NS
Yellowhammer – CR, NS
Page 37
Table 7.2. Bird abundance per point count for more common grassland species (species listed by Birdlife International (2018) as associated with grassland, and recorded on average more than twice per year). Dark green: >= 50% increase in both abundance per point count and % of season’s total. Light green: >= 20% increase. Yellow: <= 20% decrease. Red: <= 50% decrease.
Table 7.3. Number of individuals ringed for each species at each village and overall. Dark green: >= 50% increase in number of individuals. Light green: >= 20% increase.
Yellow: <= 20% decrease. Red: <= 50% decrease. Grey: none ringed two consecutive years.
Apold Crit Daia Malancrav Mesendorf Nou Sasesc Richis Viscri Species Total
Akeroyd, J., & Bădărău, S. (2012). Indicator plants of the High Nature Value dry grasslands of Transylvania. Fundatia ADEPT Transylvania. Retrieved from http://www.fundatia-adept.org/bin/file/Wildflowers_ENG(2).pdf
Birdlife International. (2018). Data Zone - Species Search. Retrieved April 12, 2018, from http://datazone.birdlife.org/species/search
Van Swaay, C. A. M., Van Strien, A. J., Aghababyan, K., Åström, S., Botham, M., Brereton, T., … Warren, M. S. (2016). The European Butterfly Indicator for Grassland species: 1990-2015. Wageningen. Retrieved from http://www.vlindernet.nl/doc/vs2016-019_european_butterfly_indicator_1990-2015_v3.pdf
Page 51
Appendix 1 Table A1. Abundance of each indicator species at each village. Grey: no record for two consecutive years. Dark green: >= 50% increase. Light green: >= 20% increase. Yellow: <= 20% decrease. Red: <= 50% decrease. Note: Six indicator species (Adonis vernalis, Viola hirta, Orchis militaris, Dictamnus albus, Echium maculatum, Gentianopis ciliate) were not present in any site in any year, and so are not included in this table. The 10 most abundant species are underlined.
Appendix 2 Table A2, part 1. Grassland butterfly abundance (numbers per hectare) at each village. Grey: no sighting two years running. Dark green: >= 50% increase. Light green: >= 20% increase. Yellow: <= 20% decrease. Red: <= 50% decrease.