Habitat selection of Eurasian otters (Lutra lutra) in urban rivers: the effects of urbanization, water quality, and river structure

doi:10.21203/rs.3.rs-3595561/v1

Download PDF

Research Article

Habitat selection of Eurasian otters (Lutra lutra) in urban rivers: the effects of urbanization, water quality, and river structure

https://doi.org/10.21203/rs.3.rs-3595561/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

This study was motivated by the increasing presence of Eurasian otters (Lutra lutra) in densely populated urban areas despite their well-known susceptibility to urbanization. We assessed the spraint marking intensity indicators along the Gwangju Stream in South Korea in 2020 and 2021. In this study, we aimed to determine otter habitat preferences in an urban river and examine the impact of urbanization, water quality, and river structure on otter habitat selection. We tested two hypotheses: (1) otters inhabiting urban rivers are more likely to favor suburban areas over highly urbanized areas, and (2) heterogeneous habitat conditions in urban rivers promote otter presence. We utilized generalized linear mixed models (GLMMs) and self-organizing maps (SOMs) with 21 variables (eight related to river structure, seven related to urbanization, and six related to water quality) to elucidate our hypotheses. Our results indicated that otter spraint marking intensity indicators were more prevalent in highly urbanized areas, with river structure serving as the primary factor influencing otter habitat selection in urban rivers. Our SOM results highlight the possibility of otters inhabiting highly urbanized areas in scenarios where artificial structures perform a protective function and extensive gravel pits serve as alternative habitats. This study suggests that coexistence between humans and otters is possible, and it may contribute to the development of sustainable urban planning strategies.

habitat use

urban wildlife

river complexity

generalized linear mixed model

self-organizing map

Understanding species-specific distribution patterns in highly urbanized environments is crucial for facilitating the harmonious coexistence of humans and wildlife (Gordon et al. 2009). Numerous species have developed various adaptive strategies in response to anthropogenically altered landscapes (Meyer and Turner 1992; Foley et al. 2005). These species can be classified into three distinct categories based on their respective responses to urbanization: avoiders, adapters, and exploiters. Adapters and exploiters exhibit the ability to adapt and even thrive in novel environments resulting from human activities, whereas avoiders such as the Eurasian otter (Lutra lutra) are negatively affected by human disturbances, resulting in their displacement from many areas (McKinney 2002; Dettori et al. 2021).

The otter is widely recognized as an emblematic species of healthy aquatic ecosystems because it prefers habitats characterized by abundant food resources, secure shelters, and minimal external disturbances (Kruuk 1995). Their adaptability to inhabit diverse aquatic habitats, including rivers, streams, marshes, wetlands, reservoirs, and coastal areas, makes them valuable model species for evaluating ecological health across a spectrum of ecosystems, ranging from natural to urban habitats (Clavero et al. 2010; Marcelli et al. 2012). However, otter populations have experienced a sharp decline over the past century due to habitat deterioration induced by rapid urbanization (Khoo and Lee 2020). At present, otters are classified as "Near Threatened" by the International Union for Conservation of Nature (IUCN) on a global scale. In addition, South Korea has designated otters as a natural monument animal (No. 330) and classified them as an endangered wildlife species (Class 1) (Roos et al. 2015).

Although otters often display a general preference for natural environments, there has been a growing number of otter sightings in urbanized areas. This suggests their potential to adapt and thrive in human-dominated landscapes as long as the quality of their habitats remains relatively intact (Jo et al. 2017). By examining successful instances of otter recolonization in many countries, including South Korea, it has become evident that enhancements in public awareness, legal protection, and sustainable urban development have positive effects on the recovery of otter populations (Marcelli and Fusillo 2009; Hong et al. 2017; Khoo and Lee 2020). In particular, riparian vegetation plays a key role in the selection of resting sites by otters in landscapes that are predominantly influenced by human activities (Weinberger et al. 2019). Additionally, the implementation of alternative habitats aimed at mitigating the adverse effects of frequent river construction, such as habitat destruction and fragmentation, can assist in facilitating the adaptation of otters to urban environments (Pedroso et al. 2014; Van Looy et al. 2014; Martin-Collado et al. 2020). The findings of this study highlight both the positive habitat potential of otters in landscapes greatly impacted by human activities and the crucial importance of maintaining secure habitats in order to ensure their persistence.

Because otters have a nocturnal and secretive lifestyle, many researchers have used their characteristic signs, such as footprints, spraints, and latrines, to identify their habitat characteristics (Kruuk 1992; Ruiz et al. 2001; Clavero et al. 2006; Mowry et al. 2011). Although it is debatable whether these signs can provide useful information regarding otter populations, spraint counts can be considered a reliable indicator of population trends (Calzada et al. 2010; Hong et al. 2022). Improved survey methods, such as repetitive surveys using different types of spraint-marking intensities and sampling only fresh spraints, should be applied to more accurately evaluate otter habitat conditions (Lanszki et al. 2008; Parry et al. 2013; Martin-Collado et al. 2020).

Fine-scale surveys based on non-invasive methods are necessary to understand the complex habitat characteristics of otters. The utilization of coarse data, such as land cover types, to clarify the complexities of river systems is inherently limited, and procuring detailed information regarding these specific river structures during a large-scale survey is a formidable challenge (Quiñónez et al. 2018). Although large-scale spraint counts can provide insights into otter population status, modeling procedures that consider the fine-scale linearity of riverine environments are required because of the complexity of otter habitats (Ottaviani et al. 2009; Van Looy et al. 2014; Hong et al. 2020a). Given the contemporary focus on human aesthetics and the establishment of urban ecological rivers for aquatic conservation, understanding otter habitats solely through land cover analysis is inadequate to account for the increased otter presence in urban rivers.

Gaining a more comprehensive understanding of the factors underlying increases in the occurrence of otters within urban areas can provide important guidance for promoting sustainable urban development. Urbanization deteriorates water quality and leads to simplified river structures (Alberti et al. 2007). However, according to the "urban environmental transition model," the impact of urban activities on ecosystems decreases during the post-industrial phase. The recent increases in the occurrence of otters in urban areas, previously confined to forested regions as a byproduct of rapid economic development, can be interpreted as a positive indication of the adaptability of this species to diverse habitats. This offers a novel perspective on restoration. To the best of our knowledge, this is the first study to elucidate the underlying factors that contribute to the frequent occurrence of otters in urban settings, including densely populated cities with populations exceeding one million. To achieve this, the study employs a model that incorporates detailed variables related to urbanization, water quality, and river structures rather than solely relying on coarser data (e.g., land-cover data). Therefore, we aimed to determine otter habitat preferences in urban environments and evaluate the impact of urbanization, water quality, and river structure on otter habitat selection using a systematic field survey. We tested two hypotheses: (1) otters inhabiting urban rivers tend to exhibit a preference for suburban areas over highly urbanized areas, and (2) heterogeneous habitat conditions in urban rivers promote otter presence.

2.1. Study area

The target area in our study was the Gwangju Stream, an urban river situated within the Gwangju metropolitan area in South Korea (Fig. 1). The Gwangju Stream spans from 35° 6'5.12" N to 35° 8'54.89" N in latitude and 126°57'53.17" E to 126°49'47.42" E in longitude, with a total river length of 21 km. The study area has four distinct seasons and is influenced by the summer monsoon (spring: March–May; summer: June–August; autumn: September; and winter: December–February). In summer, there are high temperatures and humidity levels, accompanied by abundant rainfall as a result of the influence of an oceanic climate. Conversely, in winter, the area is characterized by cold and arid conditions caused by a continental climate. The river originates from the west side of the Mudeung Mountain National Park, traverses the downtown urban area, and discharges into the Yeongsan River.

2.2 Field survey and spraint marking intensity indicators

We selected 15 survey sites that were at least 1 km apart from each other to ensure spatial independence (Fig. 1). In accordance with standard methods for field surveys, we traversed 600 m on foot at each survey site and recorded otter spraint coordinates using a global positioning system (GPS) device (GPSMAP 64S, Garmin, Korea) whenever otter spraints were found. Generally, otters produce excrement at noticeable locations to mark their territories. Otter spraints are frequently isolated or accumulated in a single location to form a latrine. Therefore, three spraint intensity indicators—the number of spraint sites, isolated spraint sites, and latrine sites—were used as response variables in this study in accordance with the methodology employed by Martin-Collado (2020). Field surveys were conducted once per season between January 2020 and October 2021: April (spring), July (summer), October (autumn), and January (winter). Otter spraints were removed during each survey to avoid sampling duplication in subsequent surveys. In order to maintain consistency across all seasonal surveys, all otter spraints were removed from 15 sites two months prior to the first survey (November 2019).

2.3. Environmental variables

We evaluated the impacts of 21 environmental variables across three categories: river structure, urbanization, and water quality (Table 1). These variables are considered key factors influencing otter habitat selection. Land-cover maps (1:5000 scale, 1 m x 1 m resolution, 2020) were obtained from the Environmental Geographic Information Service (EGIS). Based on these maps, seven land-cover areas (residential areas [m²], industrial areas [m²], commercial areas [m²], forest areas [m²], road areas [m²], water areas [m²], and riparian vegetation areas [m²]) were calculated from the 300 m buffers using ESRI ArcGIS Pro version 2.0. The field survey yielded three approximated values for each survey site: the percentage of gravel pits [%], water depth [m], and the number of artificial structures (e.g., bridges and culverts). We calculated the dimensions of the gravel pits using ArcGIS Pro 2, with monitoring data obtained from each field survey recorded within the Guru Maps Pro (version 5.4.4) application. The number of artificial structures (e.g., bridges and culverts) was recorded during the field surveys. We measured the depth of the water at several spots at each survey site using a measuring tape, and these values were averaged. With respect to water quality information, we downloaded monthly datasets for six water quality data parameters—biological oxygen demand (BOD, [mg/l]), total nitrogen (TN, [mg/l]), total phosphorus (TP, [mg/l]), dissolved oxygen (DO, [mg/l]), chemical oxygen demand (COD, [mg/l]), and suspended solids (SS, [mg/l]))—at seven spots in the Gwangju Stream from the Gwangju City Hall website. These parameters were determined in relation to the nearest survey sites. To assess the effects of human density, we extracted the total human population in each buffer based on data provided by the National Spatial Data Infrastructure Portal. To identify the impact of human activity, we extracted the floating population in each buffer based on data provided by the Gwangju City Hall website.

Table 1. Environmental variables used in our study. The letters in parentheses represent abbreviations, and the letters in brackets denote units.

Category	Environmental variables	Description and hypothesis based on the literature
River structure	Water [%]	Water areas within a 300-m radius buffer. Water is essential for otters (Remonti et al. 2009).
	Riparian vegetation (RV) [%]	Areas of riparian vegetation within a 300-m radius buffer. Otters prefer riparian vegetation belts as resting sites (Weinberger et al. 2019).
	Covered areas (CA) [㎡]	Areas covered by artificial structures within a 300-m radius buffer. Otters prefer to produce their sprainting sites in closed areas rather than open areas (Hon et al. 2010).
	Number of artificial structures (NAS)	Number of bridges and culverts within a 300-m radius buffer. Otter road accidents occur less frequently in areas with bridges or culverts (Jancke and Giere 2011).
	Gravel pits [%]	Areas of gravel pits within a 300-m radius buffer. Gravel pits in urban rivers can play an important role as substitute habitats for otters (Martin‐Collado et al. 2020).
	Depth [m]	The average water depth measured at several spots within the survey area. Otters prefer shallower rivers because deep waters may make fishing more difficult (Cho et al. 2009).
	Bank types	Average naturalness level of both sides of a river (0: Concrete, 1: Stonework, 2: Mix of stonework and artificial vegetation, 3: Mix of natural and artificial vegetation, 4: Natural vegetation) (Jo and Won 2019).
	River construction (RC)	Binary data. Large construction triggers habitat fragmentation for otters (Pedroso et al. 2014).
Urbanization	Human population [n]	The human population within a 300-m radius buffer. Human population negatively affects otter populations (Robitaille and Laurence 2002).
	Floating population [n]	Floating population within a 300-m radius buffer. Direct human activity negatively affects otter activities (Andrade et al. 2019).
	Residential areas [%]	Residential areas within a 300-m radius buffer. Buildings negatively affected otter occurrence (Romanowski et al. 2013).
	Industrial areas [%]	Industrial areas within a 300-m radius buffer. Pollution aggravates otter habitat conditions (Hong et al. 2018).
	Commercial areas [%]	Commercial areas within a 300-m radius buffer. Human activities negatively affect otter activities (Jo et al. 2017).
	Forest areas [%]	Forest areas within a 300-m radius buffer. Forests provide better habitat conditions to otters (Hong et al. 2018).
	Road areas [%]	Road areas within a 300-m radius buffer. The frequency of otter roadkill in cities is high (Fabrizio et al. 2019).
Water quality	Biological oxygen demand (BOD) [mg/l]	BOD data for each survey site. High BOD had negative impacts on otter occurrences (Acharya and Rajbhandari 2014).
	Total nitrogen (TN) [mg/l]	TN data for each survey site. TN is the major reason of eutrophication (Dodds et al. 1998).
	Total phosphorus (TP) [mg/l]	TP data for each survey site. High phosphate level is the main cause of eutrophication (Acharya and Rajbhandari 2014).
	Dissolved oxygen (DO) [mg/l]	DO data for each survey site. Low DO negatively affects otter prey items (Acharya and Rajbhandari 2014).
	Chemical oxygen demand (COD) [mg/l]	COD data for each survey site. Otter occurrence was negatively correlated with BOD level (Acharya and Rajbhandari 2014).
	Suspended solid (SS) [mg/l]	SS data for each survey site. High SS results in a low number of otters (Acharya and Rajbhandari 2014).

2.4. Statistical analysis

2.4.1. Urbanization score (US) and water quality score (WQS)

Data preprocessing of the 21 explanatory variables was performed via normalization to remove errors caused by variations in units of measurement (Pollesch and Dale 2016). In order to represent urbanization and water quality and account for multicollinearity among explanatory variables, we performed a principal component analysis (PCA). The scores extracted from the PCA results were expressed as US1, US2, WQS1, and WQS2 (Table 2). Subsequently, we conducted a correlation analysis of 12 variables (Table 3).

Table 2

Results of principal component analysis (PCA) to indicate the degree of human disturbance and water quality deterioration as a score.
Urbanization score (US)
Variables	US 1	US 2
Human population	0.7	0.3
Floating population	0.9	-0.2
Residential areas	0.9	-0.3
Industrial areas	-0.1	0.9
Commercial areas	0.7	-0.1
Forest areas	-0.7	-0.4
Road areas	0.8	0.4
Eigenvalues	3.9	1.2
Cumulative variance explained (%)	73.6	100.0
Water quality score (WQS)
Variables	WQS 1	WQS 2
Biological oxygen demand (BOD)	0.9	-0.1
Total nitrogen (TN)	0.7	-0.6
Total phosphorus (TP)	0.9	0.1
Dissolved oxygen (DO)	-0.3	-0.1
Chemical oxygen demand (COD)	0.9	0.4
Suspended solids (SS)	0.1	0.9
Eigenvalues	3.0	1.3
Cumulative variance explained (%)	69.5	100.0

Table 3

Results of correlation analysis with 12 variables. ^*: P-value < 0.05; ^**: P-value < 0.01
	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
Water areas (1)	0.31^**	0.51^**	0.00	-0.30^**	0.62^**	0.16	0.18^*	0.32^**	0.34^**	0.34^**	-0.19^*
RV (2)	1.00	0.16	0.07	-0.02	0.13	0.64^**	-0.08	-0.23^*	0.00	0.19^*	0.15
CA (3)		1.00	0.28^**	0.08	0.22^*	-0.07	-0.02	0.46^**	0.13	0.26^**	-0.22^*
NAS (4)			1.00	0.10	0.11	0.11	-0.03	0.45^**	-0.25^**	0.45^**	0.10
Gravel pits (5)				1.00	-0.21^*	0.23^*	-0.24^**	0.08	-0.02	0.05	0.07
Depth (6)					1.00	0.23^*	0.10	0.23^*	0.33^**	0.36^**	0.00
Bank types (7)						1.00	-0.13	0.07	0.12	0.37^**	0.22^*
RC (8)							1.00	-0.04	0.00	-0.07	-0.27^**
US1 (9)								1.00	0.15	0.48^**	-0.24^**
US2 (10)									1.00	0.22^*	-0.02
WQS1 (11)										1.00	0.03
WQS2 (12)											1.00

2.4.2. Identifying the effects of environmental variables using a generalized linear mixed model

We used a generalized linear mixed model (GLMM) to determine the effects of environmental variables on otters in the study area. The execution of this model may vary in accordance with the manner in which the dependent variables are distributed (Tuerlinckx et al. 2016). A negative binomial distribution was applied in our study because of the frequently encountered overdispersion in field data (Cox 1983). Four scores and eight river structure variables were used as fixed effects. We set season and year as random effects to eliminate the seasonal and annual effects. The GLMM with a negative binomial distribution was performed using the lme4 package in R software (version 4.2). Additionally, we ran a single dredge on the 12 variables to identify the most parsimonious model. Lastly, we averaged the models with a delta AICc of less than 2.

2.4.3. Characterizing the habitat preference of otters in the heterogeneous urban river using a self-organizing map

In order to characterize the preferred habitats of otters in the heterogeneous urban river, a self-organizing map (SOM) was employed that incorporated three distinct types of spraint intensities and identified significant variables via the use of the parsimonious models selected through the GLMM. A SOM is a type of unsupervised neural network that facilitates the alignment of high-dimensional data input with low-dimensional neurons, resulting in the formation of a map (Kohonen 2001). The SOM comprises input and output layers that are interconnected using calculated weights (i.e., connection intensities). The weights are trained to minimize the distance between the input and output layers. In the SOM process, a dataset with ten variables was applied to train the learning network. We set the form of the output layer to a hexagonal lattice with 42 (6 × 7) output neurons. This process was performed using the Kohonen package in R software (version 4.2). To conduct a comparative analysis of several clusters defined by the SOM, we used the Kruskal–Wallis test. In addition, we performed post-hoc comparisons using the Dunn test package of R software (version 4.2).

During the eight survey periods, 2,460 spraint sites (i.e., 116 latrines and 2,326 isolated spraint sites) were detected. Variability in the spraint intensity indicators was clearly observed for each survey site and period (Fig. 2). The survey site with the highest number of spraint sites was Site 7, with a mean value of 61.88 ± 14.59 SE, whereas the survey site with the lowest number of spraint sites was Site 4, with a mean value of 6.13 ± 2.36 SE. The survey site with the highest number of latrines was Site 7, with a mean value of 6.25 ± 2.00 SE, whereas the survey site with the lowest number of latrines was Site 3, with a mean value of 0.00 ± 0.00. The survey site with the highest number of isolated spraint sites was Site 7, with a mean value of 55.63 ± 13.10, whereas the survey site with the lowest number of isolated spraint sites was Site 4, with a mean value of 5.88 ± 2.22.

3.1. Relationship between environmental variables and otter spraint intensity indicators

Based on the analysis of 12 variables and the averaging of models with a delta AICc of less than 2, it was found that the number of spraint sites had significant relationships with six variables. In contrast, isolated spraint sites and latrines had significant relationships with five variables (Table 4). In the averaged model using the number of spraint sites, three variables (gravel pits, NAS, and the water) had positive effects on the number of spraint sites: 0.430 ± 0.077 for gravel pits, 0.373 ± 0.085 for the NAS, and 0.406 ± 0.084 for water areas. Three variables (RC, US2, and WQS1) had negative impacts on the number of spraint sites: -0.534 ± 0.253 for the RC, -0.249 ± 0.076 for the US2, and − 0.196 ± 0.089 for the WQS1. In the averaged model using the number of isolated spraint sites, three variables (gravel pits, NAS, and water) had positive impacts on the number of isolated spraint sites: 0.419 ± 0.075 for gravel pits, 0.365 ± 0.085 for the NAS, and 0.398 ± 0.086 for water areas. The two variables (US2 and WQS1) had negative impacts on the number of isolated spraint sites: -0.243 ± 0.076 for the US2 and − 0.214 ± 0.087 for the WQS1. In the averaged model using the number of latrines, four variables (water areas, CA, NAS, and gravel pits) had positive impacts on the number of latrines: 0.594 ± 0.237 for water areas, 0.408 ± 0.145 for the CA, 0.489 ± 0.192 for the NAS, and 0.452 ± 0.171 for the gravel pits. In addition, the WQS1 had a negative impact on the number of latrines: -0.544 ± 0.185 for the WQS1. Lastly, the construction was not significant for the number of isolated spraint sites and latrines. Averaged models showed that the effect of the construction was negative on the number of isolated spraint sites (estimate ± SE: -0.475 ± 0.247, P-value: 0.057) and latrines (estimate ± SE: -0.798 ± 0.6243, P-value: 0.207).

Table 4

Results of generalized linear mixed model (GLMM) analysis between spraint intensity indicators and 12 variables.
Variables	Spraint sites			Isolated spraints			Latrines
Variables	Estimate ± SE	Z value	P-value	Estimate ± SE	Z value	P-value	Estimate ± SE	Z value	P-value
(Intercept)	2.738 ± 0.335	8.071	< 0.001 ***	2.658 ± 0.365	7.196	< 0.001 ***	-0.484 ± 0.183	2.617	0.009 **
Water areas	0.406 ± 0.084	4.760	< 0.001 ***	0.398 ± 0.086	4.577	< 0.001 ***	0.594 ± 0.237	2.486	0.013 *
RV	-	-	-	-	-	-	-	-	-
CA	-	-	-	-0.063 ± 0.085	0.732	0.464	0.408 ± 0.145	2.791	0.005 **
NAS	0.373 ± 0.085	4.360	< 0.001 ***	0.365 ± 0.085	4.269	< 0.001 ***	0.489 ± 0.192	2.533	0.011 *
Gravel pits	0.430 ± 0.077	5.506	< 0.001 ***	0.419 ± 0.075	5.531	< 0.001 ***	0.452 ± 0.171	2.627	0.009 **
Depth	0.084 ± 0.096	0.869	0.384	0.103 ± 0.093	1.091	0.275	-0.401 ± 0.228	1.746	0.081
Bank types	-0.099 ± 0.075	1.308	0.190	-0.072 ± 0.073	0.987	0.324	-0.190 ± 0.153	1.224	0.221
RC	-0.534 ± 0.253	2.087	0.037 *	-0.475 ± 0.247	1.902	0.057	-0.797 ± 0.624	1.263	0.207
US1	0.123 ± 0.082	1.492	0.135	0.140 ± 0.080	1.738	0.082	-0.173 ± 0.182	0.941	0.347
US2	-0.249 ± 0.076	3.226	0.001 **	-0.243 ± 0.076	3.182	0.001 **	-	-	-
WQS1	-0.196 ± 0.089	2.180	0.029 *	-0.214 ± 0.087	2.420	0.016 *	-0.544 ± 0.185	2.905	0.004 **
WQS2	0.125 ± 0.110	1.122	0.261	0.132 ± 0.108	1.209	0.227	-0.143 ± 0.150	0.943	0.346

3.2. Comparison of otter spraint intensity and environmental variables between three clusters grouped by SOM

According to the SOM training results, all otter spraint intensity indicators exhibited the highest values in Cluster 1. Regarding environmental variables, Cluster 1 showed significantly higher values for US1, gravel pits, NAS, CA, and WQS1 than those of other variables. Conversely, Cluster 2 exhibited significantly higher values for US2, RV, and water areas than those of other variables, whereas Cluster 3 had the lowest value for WQS1 among all the clusters.

Although otter occurrence in urbanized areas is gradually increasing, identifying and assessing otter habitat characteristics by solely considering land cover, as in previous studies, remains controversial. Van Looy (2014) insisted that recent otter distribution models that solely consider broad-scale factors have disadvantages in terms of the accurate interpretation of land use in riparian environments. Therefore, we implemented a fine-scale habitat suitability model, which included river complexities, and tested two hypotheses to identify the environmental factors preferred by otters in human-dominated environments.

In contrast to the first hypothesis, our results imply that highly urbanized areas fail to impede otter movement and serve as primary habitats for otters. In our study, otter spraints were observed more frequently in highly urbanized areas than those in suburban areas. Furthermore, our parsimonious model suggested that US1 (scores associated with population) was not significantly correlated, whereas US2 (scores associated with industrial areas) was negatively correlated with otter spraint intensity. Our results are similar to those of other studies, suggesting that otters may adapt slightly to urban areas (Weinberger et al. 2016; Hong et al. 2017). In Cork City, 11 otters used the city as an important corridor connecting marine and freshwater habitats (White et al. 2013). Similarly, a project conducted in 2022 revealed the presence of 15 otters within Seoul, the largest metropolis in South Korea (SMG 2022). According to Weinberger (2016), otters have been observed to inhabit a diverse array of environments, spanning from pristine streams to extensively altered large rivers, and do not have a strong preference for pristine areas at a fine scale. Surprisingly, their flexible habitat use supports the potential for their coexistence with humans.

However, the tolerance of otters towards anthropogenic environments appears to be conditional. As discussed above, although there were several instances where urban environments failed to completely prevent otter inhabitation, the high density of industrial areas imposed high travel costs on otters (Hong et al. 2021). Moreover, dense industrial areas can cause severe water contamination, resulting in decreased food availability for otters (Bedford 2009; Acharya and Raijbhandari 2014). According to Hong et al. (2020b), high electrical conductivity (> 1,000 µs cm^− 1) resulting from saltwater intrusion or industrial zones can have a negative effect on spraint density. In our study, we confirmed that water quality deteriorates with an increase in the density of industrial areas. Furthermore, WQS1 had a negative correlation with all three types of spraint intensities. Therefore, the percentage of industrial areas may serve as an indirect factor that negatively impacts otter habitat conditions.

In addition to indirect factors that negatively affect otter habitats, frequent river construction can directly impact otter habitats. Our results indicated that the presence or absence of river construction had the greatest impact on all three types of spraint intensities. Large structures can lead to changes where otters can no longer inhabit certain habitats, reduce their activity range and breeding sites, or cause shifts in prey composition (Saavedra 2003; Chettri and Savage 2014; Pedroso et al. 2014; Ozkazanc et al. 2019; Bedmar et al. 2022). In our study area, which is smaller than the major rivers in Korea, construction can be more detrimental to otters in terms of feeding activity and habitat connectivity. In particular, the construction of small-scale urban rivers can completely overlook the possibility of movement to suitable habitats.

Although highly developed urbanization appears to have a negative impact on otter habitats, this notion is based on the assumption that higher densities in urban areas can inflict direct damage to otter habitats. Otter signs have been discovered in metropolises within South Korea, such as Daegu and Daejeon, where otter occurrence is greatly affected by environmental factors that directly damage otters, such as water quality and volume of traffic (Jo et al. 2017). In our study, US2 had a smaller effect on spraint intensity than that of other variables associated with the complexity of the river. Moreover, the statistical analysis did not show any significant correlation with latrine site intensity. These findings suggest that otter habitat selection was mostly affected by the availability of riverbank, which served as the main area for their resting and feeding activities.

The results of our analysis of the second hypothesis suggested that the complexity of urban river environments positively affects the occurrence of otters. Complex river structures can positively affect otter resource use. Otters are known to utilize naturally formed areas as shelters and exhibit more regular breeding in complex habitats (Mason and Macdonald 2009; Ruiz-Olmo and Jiménez 2009). According to our SOM results, areas with higher spraint intensities had higher human populations and lower riparian vegetation; however, these locations displayed a high number of artificial structures, a wide distribution of gravel pits, and numerous covered areas. Although the importance of riparian vegetation in promoting optimal otter habitats has been emphasized by numerous scholars, heterogeneous habitat compositions in urban rivers may enable flexible habitat use for otters, even in low-riparian vegetation areas (Remonti et al. 2011; Weinberger et al. 2019; Martin-Collado et al. 2020). In this context, our results implied that the frequent occurrence of otters in highly urbanized areas could likely be attributed to the suitability of river habitats in these environments.

Artificial structures provide otters with opportunities to inhabit urban rivers. In densely populated urban areas with heavy traffic and extensive road networks, artificial structures have been found to enhance aquatic mobility between different water bodies and reduce the risk of roadkill incidents (Jancke and Giere 2011; Serronha et al. 2013). Our models revealed a significant positive correlation between the latrine marking intensity and CA and NAS. This result suggests that enclosed spaces play an important role in providing security and protection in key areas, as opposed to areas that individuals solely traverse or explore. Breeding females prioritize habitat security over other factors for survival (Green 1984; Ruiz-Olmo et al. 2005). Several mammal species have adapted to or even prospered in urban areas because of the new resources that emerged from these novel environments (Rodriguez et al. 2021). For prey species, urban environments can sometimes serve as a refuge from predators, leading to greater prosperity for these species than in their natural habitats (McCleery 2010). Otters tend to inhabit the middle downstream regions of water bodies, where food resources are abundant, even in the face of potential exposure to human presence, as opposed to the more resource-scarce upper reaches. According to Wang (2021), otters utilize areas with high population densities during autumn because food resources migrate downstream. In addition, the areas under bridges or culverts are advantageous for otters in retaining their spraints, playing an important role in intra-specific communication (Kraševec et al. 2022). Consequently, artificial structures both provide excellent habitat security and offer superior territory-marking locations for otters (Crowley et al. 2012).

Artificial structures provide otters with a safe and suitable space to demarcate their territory, while the presence of natural objects inside these structures increases the cover available for resting and foraging activities. According to Marcelli and Fusillo (2009), the adverse impacts of human density and urban land use are less critical than those caused by natural factors on riverside land. Additionally, Serronha et al. (2013) highlighted the low confidence of otters when using culverts devoid of natural objects. In particular, artificially created gravel pits to alleviate river discontinuities that occur owing to frequent construction in anthropogenic landscapes can function as high-quality corridors connecting habitat patches. Similar findings have been reported regarding otters and other taxa (Šálek 2012; Vojar et al. 2016; Kerbiriou et al. 2018; Martin-Collado et al. 2020). Regarding Site 7, which consistently recorded the highest number of samples except for the final survey, none of the otter spraint sites were detected after the elimination of their gravel pits. Although additional research is required to determine the exact cause, it is plausible that objects present within artificial structures in urban rivers may assume the ecological function of riparian vegetation, thereby providing otters with a consistent resource base.

Asserting that our results would yield identical outcomes across all urban rivers is a challenging endeavor, primarily because of the limitations associated with a small sample size (Park et al. 2002). This discrepancy may arise from variations in scale (Quiñónez et al. 2018). Furthermore, our model omitted crucial fish abundance variables that significantly influence otter habitat use. Comprehending otter habitat use in urban rivers in relation to food resources may prove difficult. Nevertheless, our research encompassed environmental variables specific to urban rivers and considered alternative factors that can serve as indicators of fish abundance. We anticipate that future studies will contribute to the advancement of knowledge regarding otter habitat use in urban rivers via the incorporation of otter surveys conducted in other major cities and the inclusion of fish resources as variables.

In conclusion, our findings provide evidence that otters have the ability to thrive in environments that are heavily influenced by human activities, hence challenging prevailing assumptions. Furthermore, we determined that the recent increase in the occurrence of otters in urban areas can be attributed to the complexity of river structures. In urban environments with limited wetlands, large gravel pits and artificial structures were identified as significant factors that facilitate the successful adaptation and proliferation of others within cities. We anticipate two important implications from our findings. First, the adaptability of otter habitat selection indicates the potential for coexistence with human populations. Second, the significance of river complexity in improving otter habitat conditions must be emphasized in terms of sustainable urban planning.

Acknowledgements

We would like to thank Editage (www.editage.co.kr) for English language editing.

Funding: The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Competing Interests: All authors involved in this study declare that there are no potential conflicts of interest in relation to academic publishing.

Author Contributions: Research conception and design: Gun-Joo Lee, Hyun Jin Cho, Hak Young Lee; Data collection: Gun-Joo Lee, Cheon Seung Ju; Data analysis and interpretation: Gun-Joo Lee, Hyun Jin Cho, Hak Young Lee; First draft of the manuscript: Gun-Joo Lee; Critical revision of the manuscript: Gun-Joo Lee, Hak-young Lee; Final manuscript review: Gun-Joo Lee, Hak-young Lee.

Acharya PM, Rajbhandari SL (2014) Habitats of Lutrogale perspicillata in the Narayani River, Chitwan National Park, Nepal: assessment of water quality. J Indian Res 2(1):67-76.
Alberti M, Booth D, Hill K, Coburn B, Avolio C, Coe S, Spirandelli D (2007) The impact of urban patterns on aquatic ecosystems: An empirical analysis in Puget lowland sub-basins. Landsc Urban Plan 80(4):345-361. https://doi.org/10.1016/j.landurbplan.2006.08.001
Andrade AM, Arcoverde DL, Albernaz AL (2019) Relationship of Neotropical otter vestiges with environmental and anthropogenic factors. Acta Amazonica 49:183-192. http://dx.doi.org/10.1590/1809-4392201801122
Bedford SJ (2009) The effects of riparian habitat quality and biological water quality on the European Otter (Lutra lutra) in Devon. Biosci Horiz 2(2):125-133. https://doi.org/10.1093/biohorizons/hzp015
Bedmar S, Blanco‐Garrido F, Delibes M, Clavero M (2022) Temporal and spatial patterns in the shifting of otter diet to invasive prey after river damming. River Res Appl 38(8):1450-1459. https://doi.org/10.1002/rra.3961
Calzada J, Delibes-Mateos M, Clavero M, Delibes M (2010) If drink coffee at the coffee-shop is the answer, what is the question? Some comments on the use of the sprainting index to monitor otters. 10.1016/j.ecolind.2009.06.012
Chettri P, Savage M (2014) A distribution survey for otters along a river in Central Bhutan. IUCN Otter Spec Group Bull 31(2):65-74.
Cho HS, Choi KH, Lee SD, Park YS (2009) Characterizing habitat preference of Eurasian river otter (Lutra lutra) in streams using a self-organizing map. Limnology 10:203-213. https://doi.org/10.1007/s10201-009-0275-7
Clavero M, Hermoso V, Brotons L, Delibes M (2010) Natural, human and spatial constraints to expanding populations of otters in the Iberian Peninsula. J Biogeogr 37(12):2345-2357. https://doi.org/10.1111/j.1365-2699.2010.02377.x
Clavero M, Prenda J, Delibes M (2006) Seasonal use of coastal resources by otters: comparing sandy and rocky stretches. Estuar Coast Shelf Sci 66(3-4):387-394. https://doi.org/10.1016/j.ecss.2005.09.011
Cox DR (1983) Some remarks on overdispersion. Biometrika 70(1):269-274. https://doi.org/10.1093/biomet/70.1.269
Crowley S, Johnson CJ, Hodder D (2012) Spatial and behavioral scales of habitat selection and activity by river otters at latrine sites. J Mammal 93(1):170-182. https://doi.org/10.1644/10-MAMM-A-362.1
Dettori EE, Balestrieri A, Zapata-Perez VM, Bruno D, Rubio-Saura N, Robledano-Aymerich F (2021) Distribution and diet of recovering Eurasian otter (Lutra lutra) along the natural-to-urban habitat gradient (river Segura, SE Spain). Urban Ecosyst 24:1221-1230. https://doi.org/10.1007/s11252-021-01109-3
Dodds WK, Jones JR, Welch EB (1998) Suggested classification of stream trophic state: distributions of temperate stream types by chlorophyll, total nitrogen, and phosphorus. Water Res 32(5):1455-1462. https://doi.org/10.1016/S0043-1354(97)00370-9
Fabrizio M, Di Febbraro M, Loy A (2019) Where will it cross next? Optimal management of road collision risk for otters in Italy. J Environ Manage 251:109609. https://doi.org/10.1016/j.jenvman.2019.109609
Foley JA, DeFries R, Asner GP, Barford C, Bonan G, Carpenter SR, Snyder PK, et al (2005) Global consequences of land use. Science 309(5734):570-574. 10.1126/science.1111772
Gordon A, Simondson D, White M, Moilanen A, Bekessy SA (2009) Integrating conservation planning and landuse planning in urban landscapes. Landsc Urban Plan 91(4):183-194. https://doi.org/10.1016/j.landurbplan.2008.12.011
Green J (1984) A radio-tracking survey of otters Lutra lutra on a Perthshire river system. Lutra 271:85-145.
Hon N, Neak P, Khov V, Cheat V (2010) Food and habitat of Asian small-clawed otters in northeastern Cambodia. IUCN Otter Spec Group Bull 27(1):12-23.
Hong S, Chon TS, Joo GJ (2021) Spatial distribution patterns of Eurasian Otter (Lutra Lutra) in association with environmental factors unravelled by machine learning and diffusion kernel method. J Environ Inform 37:130-141. 10.3808/jei.202000443
Hong S, Di Febbraro M, Kim HG, Loy A (2022) Is scat marking a reliable tool for otter census and surveys at the landscape scale? J Environ Manage 315:115098. https://doi.org/10.1016/j.jenvman.2022.115098
Hong S, Di Febbraro M, Loy A, Cowan P, Joo GJ (2020a) Large scale faecal (spraint) counts indicate the population status of endangered Eurasian otters (Lutra lutra). Ecol Indic 109:105844. https://doi.org/10.1016/j.ecolind.2019.105844
Hong S, Do Y, Kim JY, Cowan P, Joo GJ (2017) Conservation activities for the Eurasian otter (Lutra lutra) in South Korea traced from newspapers during 1962–2010. Biol Conserv 210:157-162. https://doi.org/10.1016/j.biocon.2017.03.010
Hong S, Kim DK, Do Y, Kim JY, Kim YM, Cowan P, Joo GJ (2018) Stream health, topography, and land use influences on the distribution of the Eurasian otter Lutra lutra in the Nakdong River basin, South Korea. Ecol Indic 88:241-249. https://doi.org/10.1016/j.ecolind.2018.01.004
Hong S, Recknagel F, Kim HG, Chon TS, Joo GJ (2020b) Relationships of otter populations with fish, macroinvertebrates and water quality across three Korean rivers revealed by inferential modelling based on evolutionary computation. Ecol Inform 59:101108. https://doi.org/10.1016/j.ecoinf.2020.101108
Jancke S, Giere P (2011) Patterns of otter Lutra lutra road mortality in a landscape abundant in lakes. Eur J Wildl Res 57:373-381. https://doi.org/10.1007/s10344-010-0442-5
Jo YS, Won CM (2019) The sprainting behavior and habitat preference of the Eurasian Otter Lutra lutra along a montane stream in South Korea. Mammal Study 45(1):3-11. https://doi.org/10.3106/ms2018-0071
Jo YS, Won CM, Fritts SR, Wallace MC, Baccus JT (2017) Distribution and habitat models of the Eurasian otter, Lutra lutra, in South Korea. J Mammal 98(4):1105-1117. https://doi.org/10.1093/jmammal/gyx037
Kerbiriou C, Parisot-Laprun M, Julien JF (2018) Potential of restoration of gravel-sand pits for Bats. Ecol Eng 110:137-145. https://doi.org/10.1016/j.ecoleng.2017.11.013
Khoo MDY, Lee BH (2020) The urban smooth‐coated otters Lutrogale perspicillata of Singapore: A review of the reasons for success. International Zoo Yearbook, 54(1):60-71. https://doi.org/10.1111/izy.12262
Kohonen T (2001) Self-organizing maps, ser. Information Sciences. Berlin: Springer, 30.
Kraševec R, Trajbarič A, Fležar U (2022) Under the bridge: marking sites prove the presence of Eurasian otter Lutra lutra in the Poljanska Sora River valley, Slovenia. Natura Sloveniae 24(1):17-29. http://web.bf.uni-lj.si/bi/NATURA-SLOVENIAE/pdf/NatSlo_24_1_2.pdf
Kruuk H (1995) Wild otters: predation and populations. OUP Oxford.
Kruuk H (1992) Scent marking by otters (Lutra lutra): signaling the use of resources. Behav Ecol 3(2):133-140. https://doi.org/10.1093/beheco/3.2.133
Lanszki J, Hidas A, Szentes K, Révay T, Lehoczky I, Weiss S (2008) Relative spraint density and genetic structure of otter (Lutra lutra) along the Drava River in Hungary. Mamm Biol 73(1):40-47. https://doi.org/10.1016/j.mambio.2007.08.005
Marcelli M, Fusillo R (2009) Assessing range re-expansion and recolonization of human-impacted landscapes by threatened species: a case study of the otter (Lutra lutra) in Italy. Biodivers Conserv 18(11):2941-2959. https://doi.org/10.1007/s10531-009-9618-2
Marcelli M, Poledník L, Poledníková K, Fusillo R (2012) Land use drivers of species re‐expansion: inferring colonization dynamics in Eurasian otters. Divers Distrib 18(10):1001-1012. https://doi.org/10.1111/j.1472-4642.2012.00898.x
Martin‐Collado D, Jiménez MD, Rouco C, Ciuffoli L, de Torre R (2020) Potential of restored gravel pits to provide suitable habitats for Eurasian otters in anthropogenic landscapes. Restor Ecol 28(4):995-1005. https://doi.org/10.1111/rec.13129
Mason CF, Macdonald SM (2009) Otters: ecology and conservation. Cambridge University Press.
McCleery R (2010) Urban mammals. Urban Ecosyst Ecol 55:87-102. https://doi.org/10.2134/agronmonogr55.c5
McKinney ML (2002) Urbanization, Biodiversity, and Conservation: The impacts of urbanization on native species are poorly studied, but educating a highly urbanized human population about these impacts can greatly improve species conservation in all ecosystems. Bioscience 52(10):883-890. https://doi.org/10.1641/0006-3568(2002)052[0883:UBAC]2.0.CO;2
Meyer WB, Turner BL (1992) Human population growth and global land-use/cover change. Annu Rev Ecol Syst 23(1):39-61. https://doi.org/10.1146/annurev.es.23.110192.000351
Mowry RA, Gompper ME, Beringer J, Eggert LS (2011) River otter population size estimation using noninvasive latrine surveys. J Wildl 75(7):1625-1636. https://doi.org/10.1002/jwmg.193
Ottaviani D, Panzacchi M, Lasinio GJ, Genovesi P, Boitani L (2009) Modelling semi-aquatic vertebrates’ distribution at the drainage basin scale: The case of the otter Lutra lutra in Italy. Ecol Modell 220(2):111-121. https://doi.org/10.1016/j.ecolmodel.2008.09.014
Ozkazanc NK, Ozay E, Ozel HB, Cetin M, Sevik H (2019) The habitat, ecological life conditions, and usage characteristics of the otter (Lutra lutra L. 1758) in the Balikdami Wildlife Development Area. Environ Monit Assess 191:1-8. https://doi.org/10.1007/s10661-019-7833-1
Park C, Joo W, Seo C (2002) Eurasian otter (Lutra lutra) habitat suitability modeling using GIS; A case study on Soraksan National Park. Journal of Korea Spatial Information Society 10:501-513.
Parry GS, Bodger O, McDonald RA, Forman DW (2013) A systematic re-sampling approach to assess the probability of detecting otters Lutra lutra using spraint surveys on small lowland rivers. Ecol Inform 14:64-70. https://doi.org/10.1016/j.ecoinf.2012.11.002
Pedroso NM, Marques TA, Santos‐Reis M (2014) The response of otters to environmental changes imposed by the construction of large dams. Aquat Conserv 24(1):66-80. https://doi.org/10.1002/aqc.2379
Pollesch NL, Dale VH (2016) Normalization in sustainability assessment: Methods and implications. Ecological Economics, 130, 195-208. https://doi.org/10.1016/j.ecolecon.2016.06.018
Quiñónez AL, Fuller TK, Randhir TO (2018) A review of otter distribution modeling approach, scale and metrics. IUCN Otter Specialist Group Bulletin 35(3).
Remonti L, Balestrieri A, Prigioni C (2009) Altitudinal gradient of Eurasian otter (Lutra lutra) food niche in Mediterranean habitats. Can J Zool 87(4):285-291. https://doi.org/10.1139/Z09-011
Remonti L, Balestrieri A, Smiroldo G, Prigioni C (2011). Scent marking of key food sources in the Eurasian otter. In Annales Zoologici Fennici (Vol. 48, No. 5, pp. 287-294). Finnish Zoological and Botanical Publishing Board. https://doi.org/10.5735/086.048.0503
Robitaille JF, Laurence S (2002) Otter, Lutra lutra, occurrence in Europe and in France in relation to landscape characteristics. In Animal Conservation Forum (Vol. 5, No. 4, pp. 337-344). Cambridge University Press. https://doi.org/10.1017/S1367943002004109
Rodriguez JT, Lesmeister DB, Levi T (2021) Mesocarnivore landscape use along a gradient of urban, rural, and forest cover. PeerJ 9:e11083. https://doi.org/10.7717/peerj.11083
Romanowski J, Brzeziński M, Żmihorski M (2013) Habitat correlates of the Eurasian otter Lutra lutra recolonizing Central Poland. Acta Theriologica 58:149-155. https://doi.org/10.1007/s13364-012-0107-8
Roos A, Loy A, de Silva P, Hajkova P, Zemanová B (2015) Lutra lutra. The IUCN Red list of threatened species 2015: e. T12419A21935287. http://dx.doi.org/10.2305/IUCN.UK.2015-2.RLTS.T12419A21935287.en
Ruiz-Olmo J, Jiménez J (2009) Diet diversity and breeding of top predators are determined by habitat stability and structure: a case study with the Eurasian otter (Lutra lutra L.). Eur J Wildl Res 55(2):133-144. https://doi.org/10.1007/s10344-008-0226-3
Ruiz‐Olmo J, Margalida A, Batet A (2005) Use of small rich patches by Eurasian otter (Lutra lutra L.) females and cubs during the pre‐dispersal period. J Zool 265(4):339-346. https://doi.org/10.1017/S0952836905006424
Ruiz-Olmo J, Saavedra D, Jiménez J (2001) Testing the surveys and visual and track censuses of Eurasian otters (Lutra lutra). J Zool 253(3):359-369. https://doi.org/10.1017/S0952836901000334
Saavedra Bendito D (2003) Reintroduction of the Eurasian otter (Lutra lutra) in Muga and Fluvià basins (north-eastern Spain): viability, development, monitoring and trends of the new population. Universitat de Girona.
Šálek M (2012) Spontaneous succession on opencast mining sites: implications for bird biodiversity. J Appl Ecol 49(6):1417-1425. https://doi.org/10.1111/j.1365-2664.2012.02215.x
SMG (2022) Assessment and Optimal Management Approaches for Otter Habitats along the Han River. P97. Seoul Metapolitan Government, Seoul.
Tuerlinckx F, Rijmen F, Verbeke G, De Boeck P (2006) Statistical inference in generalized linear mixed models: A review. Br J Math Stat Psychol 59(2):225-255. https://doi.org/10.1348/000711005X79857
Van Looy K, Piffady J, Cavillon C, Tormos T, Landry P, Souchon Y (2014) Integrated modelling of functional and structural connectivity of river corridors for European otter recovery. Ecol Modell 273:228-235. https://doi.org/10.1016/j.ecolmodel.2013.11.010Get rights and content
Vojar J, Doležalová J, Solský M, Smolová D, Kopecký O, Kadlec T, Knapp M (2016) Spontaneous succession on spoil banks supports amphibian diversity and abundance. Ecol Eng 90:278-284. https://doi.org/10.1016/j.ecoleng.2016.01.028
Wang QY, Zheng KD, Han XS, He F, Zhao X, Fan PF, Zhang L (2021) Site-specific and seasonal variation in habitat use of Eurasian otters (Lutra lutra) in western China: implications for conservation. Zool Res 42(6):825. https://doi.org/10.24272/j.issn.2095-8137.2021.238
Weinberger IC, Muff S, de Jongh A, Kranz A, Bontadina F (2016) Flexible habitat selection paves the way for a recovery of otter populations in the European Alps. Biol Conserv 199:88-95. https://doi.org/10.1016/j.biocon.2016.04.017
Weinberger IC, Muff S, Kranz A, Bontadina F (2019) Riparian vegetation provides crucial shelter for resting otters in a human-dominated landscape. Mamm Biol 98:179-187. https://doi.org/10.1016/j.mambio.2019.09.001
White S, O’Neill D, O’Meara D, Shores C, O’Reilly C, Harrington A, Sleeman DA (2013) A Non-invasive genetic survey of otters (Lutra lutra). Urban Environ Pilot Study Citiz Sci IUCN Otter Spec Group Bull 30(2):103-11.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Habitat selection of Eurasian otters (Lutra lutra) in urban rivers: the effects of urbanization, water quality, and river structure

Status:

Version 1

Abstract

Figures

1. Introduction

2. Material & Methods

2.1. Study area

2.2 Field survey and spraint marking intensity indicators

2.3. Environmental variables

2.4. Statistical analysis

2.4.1. Urbanization score (US) and water quality score (WQS)

2.4.2. Identifying the effects of environmental variables using a generalized linear mixed model

2.4.3. Characterizing the habitat preference of otters in the heterogeneous urban river using a self-organizing map

3. Results

3.1. Relationship between environmental variables and otter spraint intensity indicators

3.2. Comparison of otter spraint intensity and environmental variables between three clusters grouped by SOM

4. Discussion

Declarations

References

Additional Declarations

Status:

Version 1