Better together: Combining expert and citizen science data improves our understanding of occurrence patterns of lynx and wolves in Lower Saxony, Germany

doi:10.21203/rs.3.rs-5038065/v1

The return of the two large carnivores wolf and lynx to the federal state of Lower Saxony, Germany, is a conservation success story. However, their return is accompanied by conflicts, which have to be resolved by comprehensive management strategies. Basis for such management is rigorous monitoring of spatial and temporal occurrence patterns of both species. Currently, there are two different monitoring approaches executed in Lower Saxony: the official one, established with the species’ return, is based on reporting opportunistic findings by the general public complemented with systematic camera trap surveys and scat searches. The other approach was implemented in 2014 as part of the “Wildlife Survey Lower Saxony”, an annual questionnaire sent out to owners and tenants of hunting districts, with the goal to obtain state-wide information on huntable wildlife. In this study, we therefore aimed to compare both monitoring approaches in terms of general, spatial and temporal congruence using an internal classification scheme. We showed that the different monitoring approaches provide similar information on the general development of lynx and wolf occurrence across Lower Saxony. Spatial differences were mainly found at the edges of known distributional ranges. In terms of temporal dynamics, the wildlife survey data seemed to be slightly ahead of the official monitoring programmes. We also found species-related differences, which may be related to different attitudes towards the two species. Overall, our findings indicate that the different approaches complement each other and inferences on species occurrence should be made in conjunction of the two data sets.

Annual survey data

Canis lupus

citizen science

complementary monitoring approaches

Lynx lynx

SCALP

validation

Due to human persecution, habitat loss and fragmentation, large carnivores in Europe have been severely pushed back to remote areas and were on the brink of extinction for many decades. However, intense conservation efforts, protective legislation, measures encouraging coexistence within human-dominated landscapes and also improved public opinions favoured the expansion of large carnivores to their former ranges where they have re-established stable or even increasing populations (Chapron et al. 2014; Cimatti et al. 2021; Ledger et al. 2022). Despite their potential favourable impact on entire ecosystems in terms of more stable population structures of their prey, controlling mesopredators and top-down regulation of wildlife diseases (Ripple et al. 2014; Tanner et al. 2019), the return of lynx and wolves is accompanied by a variety of challenges for human-carnivore coexistence in densely populated landscapes (Bautista et al. 2019; König et al. 2020; Rode et al. 2021; Kiffner et al. 2022). Human-dominated landscapes are severely dissected, and collision with traffic, in addition to illegal killings, is a major mortality factor, possibly resulting in constrained species dispersal, lack of genetic exchange and, therefore, impaired resilience to changing environmental conditions (Andrén et al. 2006; Musto et al. 2021; Ripari et al. 2022). Large carnivores also frequently come into conflict with humans, as they predate on livestock and compete for game species (Khorozyan and Heurich 2023a, b; Singer et al. 2023). To balance the different interests, management of possibly conflictual carnivores is often advised to minimise negative outcomes on both sides (Linnell et al. 2001; Grossmann et al. 2020). The basis for such wildlife management strategies is rigorously collected and scientifically validated data for species occurrence and occupancy as well as population dynamics, on which legally compliant and widely accepted decisions can be made (Fryxell et al. 2014; Sells et al. 2018). However, collecting such data is very costly and logistically challenging (Nichols and Williams 2006; Jones 2011), especially for monitoring across broad spatial scales and long periods of time. Citizen science is increasingly considered to overcome these challenges and complement biodiversity monitoring efforts (Dickinson et al. 2010; Theobald et al. 2015; McKinley et al. 2017). Despite their positive additional outcomes, data quality of citizen science studies is often doubted because of concerns related to skills, habits and attitudes of participants involved (Fraisl et al. 2022).

There are currently two large carnivores present in Germany with different success stories regarding their re-establishment. The Eurasian lynx (Lynx lynx) was re-introduced starting in the 1970s with several repopulation programmes, resulting in three distinct populations across Germany (Wölfl et al. 2020; Port et al. 2021), with the first repopulation programme decided by a federal government in the Harz Mountains since 2000. The lack of interconnectivity between these three populations, the restrained dispersal of individual lynx and also the isolation from other lynx populations in Central Europe motivated ongoing projects (e.g., www.luchs-thueringen.de/en) to overcome these challenges and sufficiently reinforce habitats suitable as stepping stones for the European metapopulation (Port et al. 2021). In contrast, the grey wolf (Canis lupus), re-established itself in Germany without direct human assistance via migration mainly from the east (Jarausch et al. 2021), with the first reproducing pack confirmed in 2000 on a military training ground in Lusatia, Saxony (Reinhardt et al. 2019). Since then, owing to its strong adaptability within human-dominated landscapes, wolves are swiftly and successfully dispersing mainly westwards across northern Germany, with only few individuals migrating to the south, leading to at least 169 reproducing packs in the monitoring year 2022/23 (DBBW 2024).

In Lower Saxony, a federal-state in Northern Germany, there are currently two different approaches to gain insights into the occurrence of both large carnivores. (1) The official monitoring was established with the onset of the species’ return to the federal state: it is a mixed approach, combining reports of opportunistic findings by the general public, verified and classified by experts according to national monitoring standards, with systematic monitoring measures to answer concrete scientific questions (e.g., concerning reproduction, dispersal). (2) Starting a few years after the carnivores’ return, questions on their occurrences were implemented within the already established wildlife survey programme of Lower Saxony (Wildtiererfassung Niedersachsen). Here, an annual questionnaire is sent out to hunting district owners and tenants (hereafter: participants) asking, besides other wildlife and hunting related questions, for the regularity of wolf and lynx signs within their hunting district. Thus, the main differences between both approaches is that the first one heavily relies on finding signs of occurrence, but also on reporting them to the authorities afterwards (opportunistic sampling), while the second approach is actively reaching out to gain area wide information (annual inquiry), which is on the contrary not verified by experts. It is therefore important to compare both approaches, and identify their individual strengths and weaknesses.

A first evaluation of the data on grey wolves obtained by the wildlife survey with the corresponding (official) opportunistic monitoring data showed a high spatial congruence between both datasets (Ronnenberg et al. 2017). However, Ronnenberg et al. (2017) focused only on the first two years (2014/15) of questioning participants on wolf occurrences within their hunting districts. Furthermore, no such validation was performed for the Eurasian lynx. Aside from describing the two different approaches we also contrast their information on species' occurrences in terms of (1) general agreement, (2) spatial patterns and (3) temporal dynamics since the implementation of the carnivore occurrence question within the wildlife survey programme. We expect that both approaches show the same general agreement of species occurrence and their range expansion. We also hypothesise that most spatial differences are located at the periphery of already confirmed areas, i.e., areas of potential or ongoing range expansion. Regarding their temporal dynamics, we were particular interested in how long it would take for the two approaches to converge and how long this convergence is maintained. We assume that information on species occurrences from the wildlife survey is slightly ahead of the official monitoring. Moreover, we expect species-related differences due to their different abundances in Lower Saxony, their varying biology but also because of different public awareness of, or concerns towards, lynx vs. wolves.

2.1 Study area of Lower Saxony

Lower Saxony is a federal state in north-west Germany with a total area of about 47.710 km² and is characterized by an exceptionally diverse landscape, ranging from marine areas to extensive heathland and wooded low mountain ranges. Almost 60% of the federal state is in agricultural use, nearly 22% is forested and the remaining area is primarily comprised of settlements, industry and traffic related infrastructure (Federal Statistical Office 2023). Many municipalities in Lower Saxony have merged to handle their administrative tasks, resulting in a total of 430 local administrative units (LAU). Based on common natural features (e.g., climate, geology, biosphere) the natural regions of Lower Saxony were firstly introduced in the early 1980s and then revised in 2010 (v. Drachenfels 2010). For evaluations within the wildlife survey programme this official classification was adopted and grouped into 12 different regions based on their importance for different wildlife species (Figure 1, see www.wildtiermanagement.com/naturraeume). The (1) Ems-Weser-Marsh is characterized by salt vegetation of the mudflats, the sandy dune vegetation of the islands as well as the meadows and pastured of the diked marshes. The dyke hinterland of (2) the Elbe Lowlands is predominantly used as grassland. The areas of the glacial valley have a slightly higher proportion of arable land, with beech-oak forests and pedunculate oak-hornbeam forests. Dunes in this area are overgrown with pine trees. The (3) East Frisian-Oldenburg Geest is characterized by its richly structured and diverse landscape. Its original vegetation of raised bogs and woodland has largely been replaced by arable land and grassland. Meadows and pastures of the (4) Staader Geest are mainly used for dairy farming and cattle breeding, while the arable land is oftentimes cultivated with cereals. Former heaths and moors used to characterise the landscape of the (5) Ems-Hunte Geest, but have been converted into grassland, arable land and pine forests. This also characterises the (6) Duemmer Geest Lowlands, which in contrast to the former region has a higher diversity of landscape features. The (7) Lueneburg Heath and Altmark are both known for their vast heath areas, which have continued to decline due to cultivation (i.e., arable land, cattle pastures and reforestation with pine) since the mid of the 19^th century. Glacial valleys dominate the (8) Weser-Aller-Lowlands with different landscapes: near-natural raised bogs, arable and grassland, but also pine forests on the sandy soils and deciduous forests on the better soils. The (9) Lower-Saxony Boerden are characterised by fertile loess soils with extensive arable land, but also small-scale waterlogged sites and elevations with near-natural deciduous forests. Small-scale mosaic of forests, settlements and agricultural land characterises the (10) Lower Weser Uplands, while its typical for the (11) Weser Leine Uplands that loess-covered, arable basins are oftentimes alternating with steeply rising, wooded mountain ranges. The (12) Harz is the highest mountain range of Lower Saxony (971 m above sea level). Typical for this region are beech and spruce forests, numerous rock formations and near-natural raised bogs, but also mountain meadows, old dammed ponds and other evidences of a long history of mining.

The right to hunt in Germany is tied to land ownership, thus hunters are either landowners or tenants who lease the hunting rights in a specific area. However, for reasons of wildlife biology, legislation requires a minimum size (contiguous, huntable area of 0.75 km²) for hunting districts, so that many hunting districts exceed property boundaries. Furthermore, boundaries can change more or less frequently due to changes in land ownership or consolidations. In Lower Saxony, hunting is allowed across approximately 84.4% of the federal state (40,274 km², with pacified areas included), with a varying amount of hunting districts with an average size of 4.7 km².

2.2 Official monitoring programmes and study species

For both species examined in this study, the reported evidence is verified and confirmed by experts using criteria based on the project “Status and Conservation of the Alpine Lynx Population” (SCALP). The SCALP criteria (Molinari-Jobin et al. 2012) were adopted in a modified form as national monitoring standards for large carnivores in Germany (Reinhardt et al. 2015): This scheme covers confirmed hard facts (C1, e.g., georeferenced pictures, captured or dead found animals, genetics), records confirmed by a trained expert (C2, e.g., prey remains, tracks) and unconfirmed records (C3, e.g., too old or badly documented signs and direct observations without proper documentation). Despite potentially containing false-positive observations, the C3 records are of particular value to indicate range expansion or further monitoring needs (Molinari-Jobin et al. 2021). This opportunistic sampling design is complemented by systematic surveys (camera trap studies, scat transects) to better understand local species dynamics if needed. However, opportunistic surveys are prone to false-negative sampling errors, thus, we cannot infer the species’ absence if no signs were actively reported in a given area. Therefore, areas without reported signs are referred as areas with a ‘pseudo’-absence.

2.2.1 Eurasian lynx

Between 2000 and 2006, 24 (9 male, 15 female) zoo born lynx have been released into the Harz Mountains (Figure 1) in central Germany (Anders and Middelhoff 2021). The Ministries for Agriculture and Conservation of Lower Saxony accompanied by the Hunting Association of Lower Saxony were executors of the reintroduction project. The practical work was carried out by the Harz National Park, who is also responsible for the monitoring of lynx in Lower Saxony and Saxony Anhalt since 2000. The majority of the lynx observations come from hunters, foresters and the general public who report chance sightings, possible lynx tracks, scats and prey or images from camera traps. Since 2020 people can report information via an online platform, which also provides a cartographic overview of the reports from all monitoring years.

Active monitoring with camera traps takes place primarily in areas with lynx reproduction. Lynx can be distinguished individually based on their fur pattern (Weingarth et al. 2012). Camera traps (systematically on forest roads or opportunistically, for example on prey remains) as well as genetic monitoring (saliva, scats or hair samples) are used for the identification of individual lynx. Besides that, the genetic monitoring is an important measure of the genetic diversity of the population (Mueller et al. 2022).

In the monitoring year 2010/11 for the first time five out of 25 grid cells of the 100 km² grid cell of the EU-reference grid 10 km were outside the Harz Mountains (Anders and Middelhoff 2021). Until the monitoring year 2022/23, the number of cells of the EU monitoring grid occupied by the Harz lynx population has increased to 88. 58 (66%) of them do not touch the Harz Mountains. Most of the latter are located west and south of this area. Inside the Harz Mountains, the first evidence of lynx reproduction has been detected in 2002 (Anders and Middelhoff 2021). In each of the following years, lynx offspring were recorded. Since 2010/2011 the first reproduction outside the Harz Mountains was documented. In the meantime, reproduction has also taken place in four more areas outside the Harz in distances of 30 to 70 km from the population centre although the Harz Mountains are surrounded by major roads and landscapes with low forest cover percentages.

2.2.2 Grey wolf

Since December 2011, the Lower Saxony Hunting Association (Landesjägerschaft Niedersachsen e.V.) has been responsible for the official monitoring of wild wolves in Lower Saxony, as part of a cooperation agreement with the federal state government. This monitoring is largely passive, relying heavily on camera trap images captured by hunters, where wolves appear as bycatch. Online entry of reports is possible since 2017. This passive monitoring is supplemented by active measures in areas with high likelihood of wolf presence. Wolves often utilize the anthropogenic network of forest roads for energy-efficient movement, particularly in wooded regions (Bojarska et al. 2020). Targeted use of camera traps and genetic monitoring through scat collection on these forest roads provide crucial information about the territorial status of wolf populations.

Conflict-prone livestock depredations also yield valuable insights. As livestock owners usually tend to their animals daily, any instances of predation are quickly discovered. These incidents, through genetic sampling of saliva found on preyed livestock, contribute significant information about the local wolf occurrences. Although wildlife kills can also provide valuable data, their detection is less frequent due to their remote locations and the advanced state of decomposition often observed upon discovery, making them less suitable for genetic analysis.

Starting with 138 wolf reports in the first monitoring year of 2011/12 (May 1, 2011 - April 30, 2012), the number of wolf reports has steadily increased, reaching 6,356 reports in the 2021/22 monitoring year. Over this period, a total of 29,230 datapoints have been collected. These data have been documented and evaluated according to the SCALP criteria: 41.12% as C1-confirmations, 2.84% as C2-indications, and 50.73% as C3-indications. The remaining 5.31% were either false reports or could not be evaluated due to insufficient information.

The growth in reporting corresponds to the increase in confirmed wolf territories. In the 2011/12 monitoring year, the first resident wolf pair was confirmed in the central Lueneburg Heath (Figure 1). Initially, wolves predominantly settled in nature-rich areas with high habitat suitability (Planillo et al. 2024). Subsequently, they expanded into less suitable habitats in adjacent areas and intensively used cultural landscapes farther from the core areas. By the end of the 2022/23 monitoring year, the total number of territories in Lower Saxony had reached 54, consisting of 39 reproducing packs, 14 pairs, and 1 territorial single female wolf.

2.3 Wildlife survey Lower Saxony

In 1991, the Lower Saxony Hunting Association, supported by funds of the Lower Saxony Ministry of Food, Agriculture and Consumer Protection, established the "Lower Saxony Wildlife Survey" to gain obtain information on selected, huntable wild animals, providing area wide information throughout the whole federal state. As part of the wildlife survey, participants have since been asked to answer a wide range of questions related to hunting and wildlife in their hunting districts every year. Questions on the occurrence of wolves have been incorporated into the wildlife survey starting with the hunting year 2013/14, and for the lynx three years later, in 2016/17 (Table 1), with the following four reply categories: “yes, regularly”, “yes, sporadically”, “no”, “not specified”. Overall participation of hunting districts (total number ranging between 9,366 to 9,474, depending on merging and de-merging processes of districts) is relatively stable over the years, with about 80.9% on average during the past decade. However, also considering the amount of “not specified”-replies or the ones that entirely ignored the aforementioned question (“ignored”), the number of hunting districts providing explicit information on species occurrences of the two carnivores is slightly lower, averaging 68.8% for wolves since 2013/14 and 63.6% for lynx since 2016/17.

Table 1: Data structure comparison of the different monitoring approaches in Lower Saxony

	Wildlife survey Lower Saxony		Official Monitoring
	C. lupus	L. lynx	C. lupus	L. lynx
Starting Year	2013/14	2016/17	2011	2000
Monitoring Year	01.04. – 31.03.		01.05. – 30.04.
Temporal resolution	annual		exact date
Reporting type	questionnaire on species occurrence (regular, sporadic, no)		opportunistic reports, classified according to SCALP-criteria
Spatial resolution	hunting district		point-data
Data type	presence/absence		presence only
Spatial extent	Lower Saxony		Lower Saxony

2.3 Data analyses

Data cleaning and analyses were performed with R version 4.3.2 (R Core Team 2023). To compare the different monitoring approaches, the spatial resolution had to be generalized to the highest common spatial resolution available. To protect the privacy of the hunting district owners and tenants, this is the level of the LAUs. Thus, individual responses participants evaluating the occurrence of lynx and wolves in their hunting districts were aggregated by LAU. Specifically, each of the reply categories is expressed as percentage of the hunting districts providing explicit information on species occurrences (regular, sporadic or no) within the specific LAU. Point data from the official monitoring programmes of wolf and lynx were allocated to the LAU they were reported in, using the R-package “sf” (Pebesma et al. 2023). Although the monitoring year for both carnivores starts at May 1 and lasts until April 30 of the following year, points were assigned to the hunting year (April 1 to March 31 of the following year) to match the temporal resolution of the wildlife survey. The number of records was then also aggregated on this level and is expressed as absolute counts per LAU as well as density value (number of the three different records per 1 km²), accounting for the different LAU sizes. The ratio between hard/confirmed facts (C1+C2) and unconfirmed records (C3) was calculated by dividing the sum of the C1 and C2 facts by the amount of C3 records, to see if there is a proportional shift of the different categories.

General agreement of the official monitoring with the wildlife survey was first assessed in an exploratory way to identify potential cut-off values for further classification (Table 2) and investigation. For both species, we checked the number of reported records (hard/confirmed facts (C1+C2), unconfirmed records (C3) and the sum of all three) from a LAU coinciding with a given reply from participants within that LAU. Non-replies or no specific information was categorized as “not available”, so that four reply categories were looked upon. Cut-off values to classify reported signs (total signs per LAU) into one of the three occurrence categories (regular, sporadic, no) were derived with the median values as a measure of central tendency of the amount of reported signs corresponding to the wildlife survey replies denoting regular and sporadic occurrence. To map spatial patterns of agreement, the wildlife survey data was also classified (Table 2), so that each LAU also had a single occurrence state based on the relative shares of replies within its boundaries. For a regular occurrence, more than 50% of the annual replies from participants providing explicit information on predator occurrence needed to report that they regularly spot signs of the respective carnivore in their hunting district. On the contrary, 80% (median) of the annual replies had to state “no occurrence” to be counted as such, while everything else is accounted as “sporadic occurrence”. Analogous to this, a simplified classification scheme was also applied, in which only presence/absence was categorized based on the median of total reports corresponding to the related wildlife survey reply category (Table 2). The consistency of this classification schemes was assessed with two interrater reliability metrics: (1) Overall accuracy, as a measure of the proportion of corresponding categories, and (2) Cohen’s Kappa, which adjusts the accuracy by accounting for the possibility of a correct classification match by random chance, provided by the R-package “caret”’ (Kuhn et al. 2023). For both metrics, a value of zero equals no agreement, while a value of one indicates a perfect match. Kappa values within 0.3 and 0.5 indicate for example a reasonable agreement (Kuhn and Johnson 2013).

Table 2: Cut-off values for the three-class and two-class classification approach for both monitoring approaches.

			Lynx (Lynx lynx)		Wolves (Canis lupus)
	Score	State	wildlife survey	off.-monitoring	wildlife survey	off.-monitoring
three-class scheme	3	Regular occurrence	> 50 % replies	>= 3 signs	> 50 % replies	>= 26 signs
	2	Sporadic occurrence	everything between	>= 1 < 3 signs	everything between	>= 6 < 26 signs
	1	No occurrence	> 80 % replies	< 1	> 80 % replies	< 6
two-class scheme	2	Occurrence	else	>= 1 signs	else	>= 11 signs
two-class scheme	1	No Occurrence	> 80 % replies	< 1 signs	> 80 % replies	< 11 signs
	0,01	Not available or overall response equals zero	NA	-	NA	-

Additionally, the ratio of agreement between the wildlife survey data and the official monitoring programmes was calculated by dividing the occurrence score of the wildlife survey with the occurrence score calculated for the official monitoring programmes and taking its logarithm of ten, so that zero indicates an agreement, a negative value indicates a higher occurrence score according to the official monitoring programme and a positive value is indicative for a higher occurrence score according to the wildlife survey data. In this context, the mean absolute deviation can be interpreted as the agreement variability between the two approaches and a mean absolute deviation lower than 0.18 (which equals a difference of one class), is interpreted as indicative for a good concordance.

Temporal agreement was assessed for each LAU with the two-class scoring scheme (presence/absence) by simply counting the number of years of agreement and the amount of years of disagreement until the next agreement was reached.

3.1 Species occurrences according to the different monitoring approaches

3.1.1 Official monitoring programmes

The number of LAU in which signs of lynx were reported and classified slightly increased from 44 (10.2%) in 2016/17 to 58 (13.5%) in 2021/22 with a peak of 66 (15.3%) in 2018/19 (Table 3). However, only within the natural regions of the Weser-Leine-Uplands and the Harz, signs were reported continuously over the whole study period in more than 50% of the associated LAU (Suppl.-Table 1). The sum of all reported signs increased from 248 in 2016/17 to 415 in 2021/22, with the majority of signs being classified as either C1 or C3. The ratio shifts slightly in favour of the unconfirmed records, from 1.3 to 1.1. Most signs were reported from the Weser-Leine-Uplands (50.4%) and the Harz (36.2%), but also from the Lower Saxony Boerden (10.6%). The average number of reported signs per 1 km² in Lower Saxony increased from 0.01 to 0.03 signs per 1 km² (Table 3), the Harz Mountains being the natural region with the highest reporting density of about 0.16 signs per 1 km² and year on average (Suppl.-Table 1).

Table 3: Temporal development of reported hints and evidences (signs classified into C1, C2 and C3) for lynx and wolves across Lower Saxony (430 local administrative units (LAU)) according to the official monitoring. See Suppl.-Table 1 for the same summary broken down to natural regions. Maximum values are highlighted in bold-italic.

		sum of reported signs				(C1+C2) C3	average reported signs per 1 km²				[%] of LAU with …
		C1	C2	C3	all	(C1+C2) C3	C1	C2	C3	all	C1	C2	C3	any
Lynx (Lynx lynx)	2016/17	71	70	107	248	1.3	0	0	0	0.01	5.3	*4.7*	8.4	10.2
	2017/18	139	26	106	271	1.6	0	0	0	0.02	7.0	2.1	8.4	11.2
	2018/19	147	23	147	317	1.2	0	0	0	0.01	9.3	3.0	12.1	*15.3*
	2019/20	128	11	157	296	0.9	0	0	0	0.02	8.6	1.9	11.6	13.7
	2020/21	*205*	13	*217*	*435*	1.0	0	0	0	0.02	*9.8*	1.2	*12.6*	14.7
	2021/22	200	15	200	415	1.1	0.02	0	0	0.03	8.6	2.3	11.9	13.5
Canis lupus	2013/14	162	0	329	522	0.5	0.26	0.05	0.56	0.88	7.4	2.1	18.8	20.2
	2014/15	275	0	552	885	0.5	0.52	0.09	1.07	1.68	11.9	1.2	35.1	36.0
	2015/16	490	0	962	1,489	0.5	0.88	0.06	1.79	2.73	16.0	2.3	47.4	48.8
	2016/17	912	9	1,291	2,212	0.7	1.60	0.01	2.36	3.98	21.9	0.7	46.5	48.8
	2017/18	1,116	84	1,672	2,872	0.7	2.14	0.18	3.21	5.53	28.8	4.9	58.8	61.9
	2018/19	1,170	224	1,828	3,222	0.8	2.08	0.40	3.42	5.90	29.5	8.6	64.9	67.4
	2019/20	1,603	*300*	2,081	3,984	0.9	2.97	0.52	3.99	7.49	35.6	9.1	66.7	69.3
	2020/21	2,340	29	2,588	4,957	0.9	4.26	0.05	5.06	9.37	39.5	3.3	68.8	70.5
	2021/22	*3,272*	53	*2,724*	*6,049*	1.2	5.82	0.10	5.43	11.36	*44.7*	5.8	*70.7*	*73.7*

For wolves, the number of LAU from which signs were reported increased considerably from 87 (20.2%) in 2013/14 to 317 (73.7%) in 2021/22 (Table 3) and this development is evident nearly in every natural region (Suppl.-Table 1). The sum of all of reported signs increased from 522 in 2013/14 to 6,049 in 2021/22. At the beginning of the monitoring programme, there were more unconfirmed signs, however, the ratio shifts continuously from 0.5 to 1.2. Most signs were reported from the Lueneburg Heath and Altmark (51.8%), the Stader Geest (18.1%) and the Weser-Aller-Lowlands (14.1%). Mean reported signs per 1 km² increased from 0.88 to 11.4, with the Lueneburg Heath and Altmark having the highest reporting density of about 27.4 signs per 1 km² and year on average (Suppl.-Table 1).

3.1.2 Wildlife survey Lower Saxony

The total number of participants providing explicit information on lynx occurrence in their hunting districts decreased slightly from 6,436 to 5,834, while the proportion of replies stating a sporadic or regular lynx occurrence was relatively constant, ranging from 9.1% in 2016/17 to 11.4% in 2019/20 (Figure 2, left). Corresponding to that, the number of LAU also remained stable during this time (111 ± 4). The spatial distribution of replies revealed a clustered pattern of lynx distribution in Lower Saxony, with the majority being reported from southern parts of the federal state (Suppl.-Figure 1), namely the Harz Mountains and the Solling, which is a range of forested hills situated in the Weser Leine Uplands. Another, yet not so prominent cluster, can be seen in the Lueneburg Heath.

The response rate for wolf occurrences increased by 22% after the first year of the survey and remained more or less constant afterwards (Figure 2, right). In contrast to the lynx, proportion of replies stating a sporadic or regular occurrence of wolves increased considerably over the survey period, with a mean annual increase of 150%, resulting in 2,321 and 1,735 hunting districts, respectively. The number of LAU with an occurrence stated by the participants increased from 39 in 2013/14 to 363 in 2021/22, which means that from 84% of the LAU information on wolf presence was available in the course of the wildlife survey. Distributional patterns indicate a rapid longitudinal spread of wolves from the Lueneburg Heath and beyond (Suppl.-Figure 2).

3.2 General agreement

If the two approaches are plotted against each other (Figure 3), it can be seen that the number of reported signs of the official monitoring programmes correspond to the replies of the hunting districts owners and tenants provided in the wildlife survey. Thus, the higher the amount of reported signs within a LAU, more participants of the wildlife survey are reporting signs of occurrences of the respective carnivore within their hunting district. This pattern can be seen with the hard and confirmed facts (C1 and C2) and also with the unconfirmed signs (C3), but seems to be more pronounced for wolves as for lynx.

Despite this general agreement, there were also some differences, highlighted by the amount of participants reporting a none-occurrence within their hunting districts although there were signs reported to the official monitoring programmes within the respective LAU. Interestingly, by taking a closer look at the none-occurrence outliers, the majority of them are above the median value of regularly occurrence replies. Also, the ‘not available’ category covers a wide range of the number of the reported signs from the official monitoring programmes.

Within the study period from 2016/17 to 2021/22, there were the same ten LAUs every year, where none of the participants answered the lynx related question in the wildlife survey. However, in eight of these LAUs, there were single to few signs reported to the official lynx monitoring. From these fourteen records in total, 50% were classified as C1 and the other 50% as C3. There were no cases where more than 90% of the hunting districts owners or tenants of a LAU (with more than 50% participation rate) said they haven’t spotted signs of lynx occurrence in their hunting district, despite signs reported to the official lynx monitoring programme. On the other hand, there are 21 LAUs in which more than 50% of participants within a LAU said that they have spotted signs of lynx occurrence in their hunting district, but no signs were reported to the official lynx monitoring. From the 183 potential WTE replies from this 21 LAUs, 81 stated they have regular signs of lynx occurrence, while 40 said they have only a sporadic occurrence.

In 90 cases (from 3,870, 430 LAUs per studied year for wolves (9)) throughout the whole study period, information on wolf occurrences were only available from the official monitoring. From the majority of the aforementioned 90 cases, no signs were reported to the official wolves monitoring. However, in 20 of these, numerous signs were reported to the official monitoring programme, with two of them being hot spots of wolf occurrence (257 and 114 total records in 2022 alone) within the natural region of Lueneburg Heath and Altmark. In 37 cases, more than 90% of the hunting district owners and tenants of a LAU stated that there were no signs of wolf occurrence, but at least one sign was reported to the official wolf monitoring. However, the maximum count of reported signs were four and all of them were only unconfirmed (C3) records. On the other hand, also in 37 cases, more than 50% of the participants of a LAU said that they have signs of wolves within their hunting district, but no sign was reported to the official wolf monitoring. However, only 8.5% (24 of 284) of these participants stated that they spotted signs of wolves on a regular basis.

3.3 Comparison of classification schemes

Overall accuracy was highest for the lynx with the two-class classification (0.9) and relatively stable over the past years (Table 4, left). Cohen’s Kappa averaged at 0.5 (three-class) and 0.6 (two-class). Both performance parameters seemed to relate positively (but non-significant) with the participation rate across Lower Saxony (Table 4, left). For wolves, the accuracy was lower, averaging at 0.6 for both classification approaches. Cohen’s Kappa was also considerable lower, with 0.3 for the three-class, and 0.2 for the two-class classification. Between 2013/14 and 2021/23, overall accuracy decreased considerably from 0.9 to 0.4 for both classification schemes. Kappa also decreased from reasonable fit to nearly zero during this time span. Contrary to the lynx, both performance parameters seemed to negatively relate (but non-significant) to the participation rate in the wildlife survey across Lower Saxony (Table 4, right). For both species, both classification schemes behave similarly, with a slight tendency in favour of the two-class scheme for lynx, and the three-class scheme for wolves.

Table 4: Temporal development of the performance parameters (overall accuracy and Cohen’s Kappa) of the different classification schemes for the two monitoring approaches and the participation rate (pR) of the wildlife survey data for the carnivore related questions.

Lynx (Lynx lynx)						Wolf (Canis lupus)
three-class		two-class					three-class		two-class
Acc.	Kappa	Acc.	Kappa	pR	Year	pR	Acc.	Kappa	Acc.	Kappa
-	-	-	-	-	2013/14	56.6	0.9	0.3	0.9	0.4
-	-	-	-	-	2014/15	70.0	0.9	0.4	0.9	0.5
-	-	-	-	-	2015/16	69.5	0.7	0.3	0.7	0.2
0.9	0.5	0.9	0.6	70.6	2016/17	71.2	0.6	0.3	0.6	0.2
0.8	0.5	0.9	0.6	65.5	2017/18	68.8	0.5	0.2	0.5	0.2
0.8	0.5	0.9	0.6	68.0	2018/19	69.9	0.5	0.2	0.5	0.2
0.8	0.4	0.9	0.6	57.3	2019/20	64.4	0.4	0.2	0.4	0.1
0.8	0.5	0.9	0.6	64.1	2020/21	68.1	0.5	0.2	0.5	0.2
0.8	0.5	0.9	0.6	63.7	2021/22	67.4	0.4	0.2	0.4	0.1
0.8	0.5	0.9	0.6	64.9	overall	67.3	0.6	0.3	0.6	0.2
0.7	0.7	0.7	0.6		r_P		-0.3	-0.1	-0.4	-0.3
0.4	0.5	0.4	0.4		R²		0.1	0.0	0.1	0.1
> 0.1					p		> 0.3
Acc.: accuracy; Kappa: Cohen’s Kappa, pR: WTE participation rate for the carnivore related questions. r_P is the pearson correlation coefficient between the participation rate (pR) and the respective performance parameters (Acc. or Kappa).

3.4 Spatial patterns and temporal dynamics

The average ratio of agreement for lynx between the different monitoring approaches with the two-class classification was -0.02 (-0.03 for the three-class-classification) in Lower Saxony over the complete study period (Suppl.-Figure 3: A), with a mean absolute deviation of 0.04 (two-class) and 0.05 (three-class). Mean absolute deviation for the three-class classification scheme was highest in the natural region of the Weser-Leine-Uplands (0.26), following by the Ems-Weser-Marsh and the Lueneburg Heath and Altmark (both 0.13) and the Weser-Aller-Lowlands (0.10). With the two-class classification scheme, Ems-Weser-Marsh (0.13), Lueneburg Health and Altmark (0.13) and Weser-Aller-Lowlands (0.10) had the highest values.

For wolves, the ratio of agreement with the two-class classification averaged at 0.05 (0.03 for the three-class-classification) over the study period (Suppl.-Figure 3: B), with a mean absolute deviation of 0.08 (two-class) and 0.05 (three-class). Natural regions with the highest mean absolute deviation were Elbe Lowlands, Weser-Aller-Lowlands and the Ems-Hunte Geest (0.24, 0.24 and 0.23 respectively with the three-class classification scheme). For the two-class classification scheme, it was also these three natural regions, but variability was highest within the Weser-Aller-Lowlands (0.38), followed by Ems-Hunte-Geest (0.22) and Elbe Lowlands (0.21).

Spatial patterns of disagreement and agreement (presence/absence) can be seen in Figure 4 for the lynx and in Figure 5 for wolves. For both species there were more areas, where occurrence can be presumed according only to the wildlife survey data, than areas with occurrences only based on officially reported signs. Yet, for the lynx, there are more LAUs were the official monitoring programme is more sensitive than the wildlife survey data. In most cases, the areas of disagreement are in close proximity to areas with an agreement of both monitoring approaches.

The average amount of years of agreement for lynx was 5.3 years (of 6 years studied), with 1.8 years of disagreement on average until agreement was reached. Years of agreement before the schemes disagreed averaged at 2.3 years. For wolves, the average amount of years of agreement were 5.5 years (of 9 years studied). It took about 2.1 years before an agreement was reached, which, on averaged, lasted about 3.6 years.

Since the return of lynx and wolves to Lower Saxony, Germany, much effort has been undertaken to monitor the species’ occurrence throughout the federal state. The official, mainly passive, monitoring programmes started subsequently with the species’ return, whereas the carnivore-related questions were implemented in the annual wildlife survey a few years later, i.e., 16 years after the first lynx was released into the Harz Mountains in 2000 and two years after the first wolf pack was confirmed in the Lueneburg Heath in 2011. Besides this temporal gap at the initial phase of the species’ establishment, we show that the different monitoring approaches provide similar information on the development of lynx and wolf occurrences across Lower Saxony (i.e., in terms of general trends of reported signs and within the wildlife survey). Hence, we confirm the findings of Ronnenberg et al. (2017), stating a high congruence of the official monitoring programme and the wildlife survey for wolf occurrences, and also expand this findings for lynx. Despite the similarities between the two monitoring approaches, we were able to identify differences in terms of general agreement, spatial patterns and temporal dynamics, indicating that inferences regarding species occurrence should be made by combining both data sets.

As expected, the higher the amount of signs reported to the official monitoring programmes within a LAU, the more participants of the wildlife survey reported occurrence, with more signs also corresponding to a higher occurrence reply category (sporadic vs. regular). This relationship was similar for hard and confirmed facts (C1+C2) and also for unconfirmed signs (C3). Differences are emphasized by the amount of participants reporting a none-occurrence within their hunting districts, although there were signs reported to the official monitoring programmes within the respective LAU. However, because we had to aggregate the number of reported signs per LAU to link them with the wildlife survey replies, it is possible that even if an LAU has official occurrence, the species might not occur in all (or the majority) of the hunting districts within that LAU. The “not available” category also covers the whole range of the amount of reported signs from the official monitoring programmes, with a slight bulk at the lower end of the range. Together, these findings highlight the importance of participation in the annual inquiry of the wildlife survey. Discriminative power of the number of reported signs seemed to be higher for distinguishing between the reply categories “sporadically” and “regularly” for wolves, as for the lynx. This may be due do the different magnitude of reported signs but could also be indicative of different awareness of, or concerns towards, the respective species (Arbieu et al. 2019; Rode et al. 2021; Fraisl et al. 2022; Ostermann-Miyashita et al. 2023; Whiley and Tzanopoulos 2024). Lynx in Northern Germany is far less abundant and more elusive than the wolf, probably resulting in a lower detection probability of the former (Hofmeester et al. 2021). Additionally, media coverage differs greatly between the two species and is much more extensive and emotionally charged in the case of wolf instead of lynx (Zscheischler and Friedrich 2022; Whiley and Tzanopoulos 2024).

Cut-off values for the classification of every LAU based on occurrence were much lower for lynx than for wolves. However, this was expected due to the much higher abundance (and therefore reported signs) of wolves in Lower Saxony and the regionally restricted occurrence of lynx. Overall accuracy of both classification schemes was high for lynx across Lower Saxony and in each year studied. Cohen’s Kappa also indicated a good agreement between the official monitoring programme and the wildlife survey. However, for wolves, we found a considerable decrease in these two performance parameters over the years studied. One reason for this could be the unbalanced distribution of the different classes, meaning that none-occurrence has a much higher proportion for lynx and at least for wolves in the beginning of their establishment process, leading to a higher probability of both monitoring approaches to hit the same score just by chance. Another explanation may, again, be connected to the different detection probabilities of and attitudes towards the species.

As already expected by the results of the classification evaluation, the average ratio of agreement for both species is about zero, indicating again a good overall fit of both classification schemes for the two different monitoring approaches. Species-related differences can be seen by the unlike signs, where information on lynx occurrence is slightly in favour of the official monitoring programme, while for wolves, the wildlife survey seems to provide more information on their occurrence. This implicitly highlights the difference between the two species in terms of their biology, dispersal behaviour and sign detection probability (Blanc et al. 2013; Ausband et al. 2014; Molinari-Jobin et al. 2018) and again in terms of stakeholder concerns and media coverage. This could also result in the higher agreement variability expressed by the higher mean absolute deviation for wolves. On the other hand, this phenomenon could also be a result of some kind of reporting fatigue as seen in wildlife health surveillance (Heiderich et al. 2024) or a shift in reporting attitudes of the participants in both monitoring approaches (Locke et al. 2019; Larson et al. 2020; Maund et al. 2020). This may also be supported by the clear shift in the ratio of confirmed to unconfirmed signs in favour towards the confirmed signs for wolves, while this ratio is relatively stable for lynx.

At the beginning of the lynx related questions in the wildlife survey two occurrence areas with an agreement between both approaches can be identified in the south of Lower Saxony, with the Harz Mountains and its adjacent forelands to the east, and the deciduous forests of the Weser-Leine-Bergland to the west. These areas are separated by the motorway A7 and the Leine River. However, information derived from the wildlife survey participants suggests connectivity between those areas despite the potential barriers in the early years of the survey. In subsequent years, both areas merge into one area of agreement on lynx occurrence. Interestingly, there is an area in the centre of the natural region of the Lueneburg Heath and Altmark, where lynx are initially present only according to the official monitoring programme. As this area is also a hot-spot of wolf presence in Lower Saxony, we speculate that this could be due to different reasons: (1) Overall reporting activity to the official monitoring programmes is higher as the public may “hunt” for wolf signs in this area and lynx photos are only reported as a “bycatch”. (2) Participants of the wildlife survey are too focussed on wolf occurrence and how they are personally affected of the rising numbers in their hunting districts, that they may oversee signs of lynx, which are naturally more elusive und way lower in numbers. With the integration of the wolf related questions in the wildlife survey there is a cluster of agreement in the already confirmed Lueneburg Heath. In subsequent years, this area is widening, oftentimes following the assessment of the wildlife survey participants. In general, it can be stated that these differences are mainly recognised, as expected, in the peripheral areas of already confirmed carnivore occurrences. Additionally, participants reported the presence of the carnivores on average two years before the official monitoring programmes confirmed this presence. Hence, the wildlife survey provides meaningful insights into potential dispersal pathways and indicates further monitoring requirements.

Both monitoring approaches benefit from the integration of hunters as an affected stakeholder group, as they are actively involved in wildlife management, know their hunting districts and are likely more interested in reporting valuable information (Lüchtrath and Schraml 2015; Reding and Leschinski 2016; Cretois et al. 2020). However, bias or lack of neutrality are among the common concerns regarding citizen science data quality (Fraisl et al. 2022). While the opportunistic monitoring is reimbursed by quality assurance through national monitoring standards (Hočevar et al. 2020), this is not the case for the wildlife survey of Lower Saxony, which is therefore susceptible to false positives. Nevertheless, this and many other problems can be addressed with careful design, study implementation, training and management/evaluation of data derived from citizen science projects (Miller et al. 2013; Aceves-Bueno et al. 2015; Fraisl et al. 2022). It is essential for both approaches to keep the motivation of participants high and establish a feedback system to enhance their performance (Zhou et al. 2020), but also to gain insights into their reporting behaviour. Participants of both monitoring approaches should be actively informed on the outcome of their valuable and deliberately delivered data, so that they feel engaged in the course of decision-making (Newman et al. 2017; Suškevičs et al. 2021). In case of the wildlife survey programme, we suggest to provide participants either with clear guidelines on separating the two response categories on the frequency of sign occurrences or simply ask for information on presence or absence of the respective species. Our findings also stress the need to at least have centroids of hunting districts available, so that spatial inferences are more accurate while protecting the privacy of their owners or tenants.

We conclude that inferences based solely on one of the two data sets might be incomplete, i.e., spatially inaccurate and/or time-delayed. Given our findings, we therefore call for the continuation of both approaches, the official monitoring programmes and the carnivore-related questions within the wildlife survey of Lower Saxony.

CrediT

Conceptualization: CL, ES, NB, RG; Data curation: CL; Formal Analysis: CL; Funding acquisition: ES, NB, RG; Visualization: CL; Writing - original draft: CL, ES, LM, NB, OA, RG, RR; Writing - review & editing: CL, IK, ES, LM, NB, OA, RG, RR.

Funding: This work was supported by the Ministry for Food, Agriculture and Consumer Protection of Lower Saxony (Niedersächsisches Ministerium für Ernährung, Landwirtschaft und Verbraucherschutz) through the “Jagdabgabemittel des Landes Niedersachsen” (Grant-No.: 406-04032-13014/2022).

Interests: The authors have no competing interests to declare that are relevant to the content of this article.

Ethics: Not applicable

Data availability: The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.

Acknowledgements: We thank all the volunteers who reported signs on carnivore occurrence and the hunting district owners and tenants for participating in the wildlife survey of Lower Saxony.

Aceves-Bueno E, Adeleye AS, Bradley D, et al (2015) Citizen science as an approach for overcoming insufficient monitoring and inadequate stakeholder buy-in in adaptive management: criteria and evidence. Ecosystems 18:493–506. https://doi.org/10.1007/s10021-015-9842-4
Anders O, Middelhoff TL (2021) The development of the Harz lynx population. Cat News Spec Issue 14:24–28
Andrén H, Linnell JDC, Liberg O, et al (2006) Survival rates and causes of mortality in Eurasian lynx (Lynx lynx) in multi-use landscapes. Biol Conserv 131:23–32. https://doi.org/10.1016/j.biocon.2006.01.025
Arbieu U, Mehring M, Bunnefeld N, et al (2019) Attitudes towards returning wolves (Canis lupus) in Germany: Exposure, information sources and trust matter. Biol Conserv 234:202–210. https://doi.org/10.1016/j.biocon.2019.03.027
Ausband DE, Rich LN, Glenn EM, et al (2014) Monitoring gray wolf populations using multiple survey methods. J Wildl Manag 78:335–346. https://doi.org/10.1002/jwmg.654
Bautista C, Revilla E, Naves J, et al (2019) Large carnivore damage in Europe: Analysis of compensation and prevention programs. Biol Conserv 235:308–316. https://doi.org/10.1016/j.biocon.2019.04.019
Blanc L, Marboutin E, Gatti S, Gimenez O (2013) Abundance of rare and elusive species: Empirical investigation of closed versus spatially explicit capture–recapture models with lynx as a case study. J Wildl Manag 77:372–378. https://doi.org/10.1002/jwmg.453
Bojarska K, Sulich J, Bachmann S, et al (2020) Opportunity and peril: How wolves use a dense network of forest roads. Mamm Biol 100:203–211. https://doi.org/10.1007/s42991-020-00014-0
Chapron G, Kaczensky P, Linnell JDC, et al (2014) Recovery of large carnivores in Europe’s modern human-dominated landscapes. Science 346:1517–1519. https://doi.org/10.1126/science.1257553
Cimatti M, Ranc N, Benítez‐López A, et al (2021) Large carnivore expansion in Europe is associated with human population density and land cover changes. Divers Distrib 27:602–617. https://doi.org/10.1111/ddi.13219
Cretois B, Linnell JDC, Grainger M, et al (2020) Hunters as citizen scientists: Contributions to biodiversity monitoring in Europe. Glob Ecol Conserv 23:e01077. https://doi.org/10.1016/j.gecco.2020.e01077
DBBW (2024) Map of wolf territories in Germany. https://www.dbb-wolf.de/wolf-occurrence/confirmed-territories/map-of-territories. Accessed 21 Mar 2024
Dickinson JL, Zuckerberg B, Bonter DN (2010) Citizen science as an ecological research tool: challenges and benefits. Annu Rev Ecol Evol Syst 41:149–172. https://doi.org/10.1146/annurev-ecolsys-102209-144636
Federal Statistical Office (2023) Floor area according to types of use and federal state. In: Fed. Stat. Off. https://www.destatis.de/EN/Themes/Economic-Sectors-Enterprises/Agriculture-Forestry-Fisheries/Land-Use/Tables/areas-laender.html. Accessed 29 Dec 2023
Fraisl D, Hager G, Bedessem B, et al (2022) Citizen science in environmental and ecological sciences. Nat Rev Methods Primer 2:1–20. https://doi.org/10.1038/s43586-022-00144-4
Fryxell JM, Sinclair ARE, Caughley G (2014) Wildlife ecology, conservation, and management, Third edition. John Wiley & Sons, Chichester, West Sussex
Grossmann CM, Patkó L, Ortseifen D, et al (2020) Human-large carnivores co-existence in Europe – A comparative stakeholder network analysis. Front Ecol Evol 8:. https://doi.org/10.3389/fevo.2020.00266
Heiderich E, Keller S, Pewsner M, et al (2024) Analysis of a European general wildlife health surveillance program: Chances, challenges and recommendations. PLOS ONE 19:e0301438. https://doi.org/10.1371/journal.pone.0301438
Hočevar L, Fležar U, Krofel M (2020) Overview of good practices in Eurasian lynx monitoring and conservation. University of Ljubljana, Biotechnical Faculty, Ljubljana
Hofmeester TR, Thorsen NH, Cromsigt JPGM, et al (2021) Effects of camera-trap placement and number on detection of members of a mammalian assemblage. Ecosphere 12:e03662. https://doi.org/10.1002/ecs2.3662
Jarausch A, Harms V, Kluth G, et al (2021) How the west was won: Genetic reconstruction of rapid wolf recolonization into Germany’s anthropogenic landscapes. Heredity 127:92–106. https://doi.org/10.1038/s41437-021-00429-6
Jones JPG (2011) Monitoring species abundance and distribution at the landscape scale. J Appl Ecol 48:9–13. https://doi.org/10.1111/j.1365-2664.2010.01917.x
Khorozyan I, Heurich M (2023a) Patterns of predation by the Eurasian lynx (Lynx lynx) throughout its range: Ecological and conservation implications. Mammal Rev 53:177–188. https://doi.org/10.1111/mam.12317
Khorozyan I, Heurich M (2023b) Where, why and how carnivores kill domestic animals in different parts of their ranges: An example of the Eurasian lynx. Glob Ecol Conserv 46:e02585. https://doi.org/10.1016/j.gecco.2023.e02585
Kiffner C, Uthes S, Ostermann-Miyashita E-F, et al (2022) Patterns of livestock loss associated with a recolonizing wolf population in Germany. Front Conserv Sci 3:
König HJ, Kiffner C, Kramer-Schadt S, et al (2020) Human–wildlife coexistence in a changing world. Conserv Biol 34:786–794. https://doi.org/10.1111/cobi.13513
Kuhn M, Johnson K (2013) Applied predictive modeling. Springer New York, New York, NY
Kuhn M, Wing J, Weston S, et al (2023) caret: Classification and Regression Training
Larson LR, Cooper CB, Futch S, et al (2020) The diverse motivations of citizen scientists: Does conservation emphasis grow as volunteer participation progresses? Biol Conserv 242:108428. https://doi.org/10.1016/j.biocon.2020.108428
Ledger S, Rutherford CA, Benham C, et al (2022) Wildlife Comeback in Europe - Opportunities and challenges for species recovery. Final report to Rewilding Europe by the Zoological Society of London, BirdLife International and the European Bird Census Council. London
Linnell JDC, Swenson JE, Anderson R (2001) Predators and people: Conservation of large carnivores is possible at high human densities if management policy is favourable. Anim Conserv Forum 4:345–349. https://doi.org/10.1017/S1367943001001408
Locke CM, Anhalt-Depies CM, Frett S, et al (2019) Managing a large citizen science project to monitor wildlife. Wildl Soc Bull - Rec Set Error 43:4–10. https://doi.org/10.1002/wsb.943
Lüchtrath A, Schraml U (2015) The missing lynx — Understanding hunters’ opposition to large carnivores. Wildl Biol 21:wlb.00855. https://doi.org/10.2981/wlb.00068
Maund PR, Irvine KN, Lawson B, et al (2020) What motivates the masses: Understanding why people contribute to conservation citizen science projects. Biol Conserv 246:108587. https://doi.org/10.1016/j.biocon.2020.108587
McKinley DC, Miller-Rushing AJ, Ballard HL, et al (2017) Citizen science can improve conservation science, natural resource management, and environmental protection. Biol Conserv 208:15–28. https://doi.org/10.1016/j.biocon.2016.05.015
Miller DAW, Nichols JD, Gude JA, et al (2013) Determining occurrence dynamics when false positives occur: Estimating the range dynamics of wolves from public survey data. PLOS ONE 8:e65808. https://doi.org/10.1371/journal.pone.0065808
Molinari-Jobin A, Breitenmoser U, Breitenmoser C, et al (2021) SCALP: Monitoring the Eurasian lynx in the Alps and beyond. SI:50–52
Molinari-Jobin A, Kéry M, Marboutin E, et al (2012) Monitoring in the presence of species misidentification: the case of the Eurasian lynx in the Alps. Anim Conserv 15:266–273. https://doi.org/10.1111/j.1469-1795.2011.00511.x
Molinari-Jobin A, Kéry M, Marboutin E, et al (2018) Mapping range dynamics from opportunistic data: spatiotemporal modelling of the lynx distribution in the Alps over 21 years. Anim Conserv 21:168–180. https://doi.org/10.1111/acv.12369
Mueller SA, Prost S, Anders O, et al (2022) Genome-wide diversity loss in reintroduced Eurasian lynx populations urges immediate conservation management. Biol Conserv 266:109442. https://doi.org/10.1016/j.biocon.2021.109442
Musto C, Cerri J, Galaverni M, et al (2021) Men and wolves: Anthropogenic causes are an important driver of wolf mortality in human-dominated landscapes in Italy. Glob Ecol Conserv 32:e01892. https://doi.org/10.1016/j.gecco.2021.e01892
Newman G, Chandler M, Clyde M, et al (2017) Leveraging the power of place in citizen science for effective conservation decision making. Biol Conserv 208:55–64. https://doi.org/10.1016/j.biocon.2016.07.019
Nichols J, Williams B (2006) Monitoring for conservation. Trends Ecol Evol 21:668–673. https://doi.org/10.1016/j.tree.2006.08.007
Ostermann-Miyashita E-F, Pernat N, König HJ, et al (2023) Attitudes of wildlife park visitors towards returning wildlife species: An analysis of patterns and correlates. Biol Conserv 278:109878. https://doi.org/10.1016/j.biocon.2022.109878
Pebesma E, Bivand R, Racine E, et al (2023) sf: Simple Features for R
Planillo A, Wenzler-Meya M, Reinhardt I, et al (2024) Understanding habitat selection of range-expanding populations of large carnivores: 20 years of grey wolves (Canis lupus) recolonizing Germany. Divers Distrib 30:71–86. https://doi.org/10.1111/ddi.13789
Port M, Henkelmann A, Schröder F, et al (2021) Rise and fall of a Eurasian lynx (Lynx lynx) stepping-stone population in central Germany. Mammal Res 66:45–55. https://doi.org/10.1007/s13364-020-00527-6
R Core Team (2023) R: A language and environment for statistical computing
Reding R, Leschinski L (2016) Involving an affected interest group in the monitoring of a conflict species. In: Wildbiologische Forschungsberichte. Vereinigung der Wildbiologen und Jagdwissenschaftler Deutschlands, Trippstadt
Reinhardt I, Kaczensky P, Knauer F (2015) Monitoring von Wolf, Luchs und Bär in Deutschland. Bundesamt für Naturschutz, Bonn- Bad Godesberg
Reinhardt I, Kluth G, Nowak C, et al (2019) Military training areas facilitate the recolonization of wolves in Germany. Conserv Lett 12:. https://doi.org/10.1111/conl.12635
Ripari L, Premier J, Belotti E, et al (2022) Human disturbance is the most limiting factor driving habitat selection of a large carnivore throughout Continental Europe. Biol Conserv 266:109446. https://doi.org/10.1016/j.biocon.2021.109446
Ripple WJ, Estes JA, Beschta RL, et al (2014) Status and ecological effects of the world’s largest carnivores. Science 343:1241484. https://doi.org/10.1126/science.1241484
Rode J, Flinzberger L, Karutz R, et al (2021) Why so negative? Exploring the socio-economic impacts of large carnivores from a European perspective. Biol Conserv 255:108918. https://doi.org/10.1016/j.biocon.2020.108918
Ronnenberg K, Habbe B, Gräber R, et al (2017) Coexistence of wolves and humans in a densely populated region (Lower Saxony, Germany). Basic Appl Ecol 25:1–14. https://doi.org/10.1016/j.baae.2017.08.006
Sells SN, Bassing SB, Barker KJ, et al (2018) Increased scientific rigor will improve reliability of research and effectiveness of management. J Wildl Manag 82:485–494. https://doi.org/10.1002/jwmg.21413
Singer L, Wietlisbach X, Hickisch R, et al (2023) The spatial distribution and temporal trends of livestock damages caused by wolves in Europe. Biol Conserv 282:110039. https://doi.org/10.1016/j.biocon.2023.110039
Suškevičs M, Raadom T, Vanem B, et al (2021) Challenges and opportunities of engaging biodiversity-related citizen science data in environmental decision-making: Practitioners’ perceptions and a database analysis from Estonia. J Nat Conserv 64:126068. https://doi.org/10.1016/j.jnc.2021.126068
Tanner E, White A, Acevedo P, et al (2019) Wolves contribute to disease control in a multi-host system. Sci Rep 9:7940. https://doi.org/10.1038/s41598-019-44148-9
Theobald EJ, Ettinger AK, Burgess HK, et al (2015) Global change and local solutions: Tapping the unrealized potential of citizen science for biodiversity research. Biol Conserv 181:236–244. https://doi.org/10.1016/j.biocon.2014.10.021
v. Drachenfels O (2010) Überarbeitung der Naturräumlichen Regionen Niedersachsens. Informationsdienst Naturschutz Niedersachs 30:249–252
Weingarth K, Heibl C, Knauer F, et al (2012) First estimation of Eurasian lynx (Lynx lynx) abundance and density using digital cameras and capture–recapture techniques in a German national park. Anim Biodivers Conserv 35:197–207
Whiley FL, Tzanopoulos J (2024) Public acceptance of Eurasian lynx (Lynx lynx) in Germany. J Nat Conserv 77:126535. https://doi.org/10.1016/j.jnc.2023.126535
Wölfl S, Anders O, Middelhoff TL, et al (2020) Status des Luchses in Deutschland. 0028-0615 96:2–10. https://doi.org/10.17433/1.2021.50153865.2-10
Zhou X, Tang J, Zhao Y (Chris), Wang T (2020) Effects of feedback design and dispositional goal orientations on volunteer performance in citizen science projects. Comput Hum Behav 107:106266. https://doi.org/10.1016/j.chb.2020.106266
Zscheischler J, Friedrich J (2022) The wolf (Canis lupus) as a symbol of an urban–rural divide? Results from a media discourse analysis on the human–wolf conflict in Germany. Environ Manage 70:1051–1065. https://doi.org/10.1007/s00267-022-01719-3

No competing interests reported.

Supplementary.docx

Better together: Combining expert and citizen science data improves our understanding of occurrence patterns of lynx and wolves in Lower Saxony, Germany

Status:

Version 1

Abstract

Figures

1. Introduction

2. Material and Methods