Abstract From an ecological perspective, urban green roofs can be viewed as green islands surrounded by an urban matrix. Island biogeography theory suggests that species richness on an island is the outcome of a dynamic equilibrium between immigration and extinction. Immigration is affected by the distance of an island from a colonizing source. Building height may act as a limiting factor for many species on green roofs. Here, we used green roofs to consider two distance components – vertical (building height), and horizontal (building distance from open areas). We first review the green roof literature on these two spatial aspects. We then focus on a new study conducted in Haifa, Israel. In our review, we found only a small number of studies addressing the effects of distance from city edge and building height on insects’ occurrence on roofs, and usually, these studies considered a small number of roofs. In the Haifa study, we compared arthropod abundance and richness by visiting a set of 16 pots (four flowering plant species) placed on 18 rooftops and on adjacent ground-level yards. Yards had significantly more species and there was significant community dissimilarity between roofs and yards. This dissimilarity and the large proportion of unique species appearing exclusively on roofs suggest that green roofs add to the city’s biodiversity. Species richness at the city edge was significantly larger than in the city core. These results suggest that the distance from open areas and building height can be limiting factors for insect colonization on roofs.

Introduction: Salamandra infraimmaculata is considered as a threatened species in Israel. However, sincewhen we started studying Salamandra, we have found many more breeding sites in the past 10 years than were known previously. In addition, O. Rybak (unpublished results) has found that Salamandra does well in urban habitats on Mount Carmel. Salamandra infraimmaculata has strong pond site fidelity that likely results in differential genetic structure among breeding sites. Strong pond site fidelity has been found by multiple researchers. An exception is that on Mount Carmel, S Bar David found dispersal as far as 1280 m, which is likely to result in overlapping genetic structuring. During our first genetics study, in which-nine breeding sites in Mount Carmel and 11 sites in the Galilee were surveyed and tissue samples were taken from 475 adults. Microsatellites were used for observing differences in genetic structure in the study which was performed in Prof Juha Merila’s Lab (University of Helsinki, Finland). In this study, we found that there were substantial genetic differences between Mount Carmel and the Galilee. Allelic richness was much higher in the Galilee than in Mount Carmel; 40 unique alleles were found in the Galilee and 0 in Mount Carmel. Unrooted neighbour joining tree diagrams resulted in pure separations between Mount Carmel and the Galilee. Structure analysis showed strong differences between the Galilee and Mount Carmel. A second microsatellite study was conducted in Finland in Juha Merila’s Lab. The goal was to consider genetic diversity in peripheral populations compared to populations closer to the core. We collected 692 tissue samples from adult and juvenile fire salamanders from 33 breeding sites (13 from upper Galilee, ten from lower Galilee, nine from Mount Carmel and one from Tel Dan). This study also considered vegetation types and meteorological aspects such as elevation, average temperature and precipitation. Maximum entropy analysis was also used to score major regions. The lower Galilee had the lowest stability values of the three regions. Allelic richness increased with maximum entropy scores in the Upper Galilee. Allelic richness also increased with latitude. A Bayesian analysis of population structure (BAPS) also demonstrated that Mt Carmel and the upper Galilee were homogeneous genetically while the lower Galilee contained genetically differentiated populations. Lastly, we performed transcriptomics/gene expression studies on fire salamander larvae. We insisted that Salamandra tailfins (which do not cause damage to the Salamandra larvae) can demonstrate gene expression as opposed to using the whole body. We found that Salamandra larvae turn darker when exposed to ultra violet radiation and they turn darker with increased density. We are currently conducting gene expression studies on Salamandra larval development, oxygen ranges, colour change and temperature change.

תקציר בעברית: למרות הגידול בהיקף הקמתם של גגות ירוקים בעולם, בישראל מספר הגגות הירוקים עדיין מצומצם, והתהליך להתאמת הפרקטיקה לתנאים המקומיים עדיין בהתהוות. במחקר זה נבדקה במשך ארבע שנים השפעתם של מאפיינים רלוונטיים לישראל – רמות שונות של עושר מינים של צמחים חד-שנתיים, סוג המצע (טוף ופרלייט) ונוכחות או היעדר של יריעת ניקוז – על מגוון מדדים (כיסוי צמחי ירוק, מינים זרים שנוספו, לחות וטמפרטורה של המצע) המשקפים את תפקוד הגג הירוק. הממצאים הראו שעושר מינים התחלתי גבוה הגביר את הכיסוי הצמחי הירוק בכל השנים, אף על פי שעושר המינים בפועל בשני הטיפולים (גבוה ונמוך) התכנס למספר מינים דומה לקראת סוף הניסוי. בחלקות עם עושר התחלתי גבוה מספר המינים הזרים שנוספו במהלך השנים היה נמוך יותר. סוג המצע השפיע על הכיסוי הירוק בחלקות באופן שונה על פני השנים – בשנה הראשונה הכיסוי היה גבוה יותר בחלקות הטוף, ואילו בשאר השנים הכיסוי היה גבוה יותר בחלקות הפרלייט. בכל ארבע השנים הלחות במצע הייתה גבוהה יותר והטמפרטורה נמוכה יותר בחלקות הפרלייט, דבר המעיד על תרומת טיפול זה לבידוד הגג, לאגירת משקעים וליצירת בית גידול טוב יותר לצמחים. יריעת הניקוז היא רכיב יקר בגג הירוק, ולמרות זאת לא נמצאה להימצאותה השפעה כלשהי. ממצאים אלה מספקים את הבסיס להקמת גגות ירוקים אקסטנסיביים אפקטיביים בעלות מופחתת.

Green walls can improve indoor air-quality by reducing concentrations of carbon dioxide (CO2) and other air pollutants. Our study focused on the spider plant, Chlorophytum comosum, and devil's ivy, Epipremnum aureum, both common greenwall plants that have been found to be efficient CO2 absorbers. Both species have multiple variants with varying degrees of leaf green-white segmentation. Since photosynthesis depends on the concentration of leaf chlorophylls, we hypothesized that green variants are more efficient carbon absorbers than green-white variants. In addition, we tested the hypothesis that the photosynthetic rate of plants is affected by pot volume, as suggested by previous studies. We used a portable gas exchange system to determine the rate of photosynthesis of the study plants. No evidence was found for better photosynthetic performance in the green vs. green-white variants of each species. In fact, our results suggest the opposite. It was observed that a spider plants assimilated carbon more efficiently when grown in a larger pot volume. In conclusion, our study shows that in terms of carbon assimilation, green-white variants of spider plants are the better choice for indoor green walls. Their efficiency can be improved dramatically by increasing pot volume.

Green roofs are expected to contribute to higher biodiversity in urban surroundings. Typically, green roofs have been designed with low plant diversity. However, plant diversity can be enhanced by controlling resource availability and creating distinct niches. Here we hypothesize that by using different drainage heights during the short plant-growing season in a semi-arid green roof system we can create distinct niches and plant communities. Our experiment took place at the University of Haifa, north Israel. We tested three different heights of drainage outlet: 10 cm under the surface of the substrate (Low), 1 cm under the surface of the substrate (Medium) and 3 cm above the surface of the substrate (High) on plant species-composition in green-roof gardens. Grasses cover was higher in High and Medium drainages while forbs cover was higher in Low drainage. Species richness was the highest in Low drainage while diversity indices showed the opposite trend. We conclude that by changing the height of the drainage we can create different niches and change species composition in a short time period of one growing season. This way we can create more diverse green roof communities and enhance biodiversity in urban areas, particularly in semi-arid regions.

The Near Eastern fire salamander (Salamandra infraimmaculata) reaches its southern distribution range in Israel. Although the population structure has been analysed in central Israel and at the southern distribution limit, we lack knowledge on populations in the northern area, such as along Mount Hermon. S. infraimmaculata occurs at Mt. Hermon along an altitudinal gradient and appears to be fragmented by urban and agricultural landscape. We studied the genetic structure of four populations based on microsatellite loci and the mitochondrial D-loop to determine the genetic diversity and connectivity between populations. We observed moderate gene flow at lower parts, i.e. from Tel Dan and Nimrod Castle to Banias indicating extant but limited connectivity. Genetic diversity and gene flow declined along the altitudinal gradient at Mt. Hermon, reaching rock-bottom levels in the highest located population of Nimrod Pool. The observed isolation-by-elevation gradient might induce a higher extinction risk for the highland populations of S. infraimmaculata.

Fire retardants are commonly used for fighting wildfires. Retardant chemicals washed via runoff into aquatic systems may be concentrated, thus exposing aquatic species to high levels of ammonium, phosphate, and iron. These chemicals directly affect individual species, which can also cascade to the rest of the aquatic community. We investigated the effects of a prevalent fire retardant, FR CROS 134 (FR), at various concentrations (0 to 11.4 × 102 mg l−1) on larval fire salamander (Salamandra infraimmaculata) and aquatic community structure (bacteria, algae and invertebrates) using mesocosms. We show a negative effect of the presence of FR on the survivorship and time to metamorphosis of salamander larvae (94–17% and 48–64 days, respectively). Moreover, increasing FR concentrations were found to inhibit the predation rate of salamander larvae on mosquito larvae (89–14%). FR decreased the invertebrate community diversity with changes in composition and shifts in evenness and dominance (e.g., cladoceran species abundance increased, whereas calanoid copepod species declined). Finally, cyanobacteria and eukaryotic algal abundance increased, while heterotrophic bacterial abundance decreased with FR addition. The effects of fire retardants on biodiversity found here and the growing use of fire retardants worldwide call for additional evidence-based assessment of their impact, especially in aquatic ecosystems.

1. An observed community structure is often shaped by a combination of bottom‐top and top‐bottom processes, affected by biotic and abiotic factors. These factors affect community structure either directly (each species separately) or indirectly, via other species. Few studies have observed the combined effects of biotic and abiotic factors on primary producers and consumer communities composed of numerous species.2. This study investigated the combined effects of a predatory hemipteran and water salinity on abundance and species richness of the periphyton and arthropod communities in an artificial pool experiment.3. The abundance and species richness of both periphyton and arthropod communities were affected by the combination of salinity and predator.4. It is suggested that the observed consumer community is composed of arthropod species that vary in salinity tolerance and response to the predator. In addition, the abundance of euryhaline species or species that are not consumed by this particular predator increases indirectly in specific treatment combinations that eventually result in reduced interspecific competition. Periphyton variables were strongly correlated to densities of larval Ochlerotatus caspius (Diptera: Culicidae), suggesting that this species may be largely responsible for structuring the producer community.5. It is suggested that O. caspius distribution is the result of female oviposition habitat selection that is manipulated by chemical signals. Therefore, chemical signals that inform about habitat suitability also play an important role in shaping both consumer and primary producer communities.

1. Fire retardants are composed of fertilising salts that are commonly used for fighting wildfires. These chemicals have various effects on individual arthropod species and aquatic communities.2. This study investigated the effects of four treatments of a prevalent fire retardant [FR CROS 134 (FR), applied at concentrations of 0, 3.8 × 102, 7.6 × 102 and 11.4 × 102 mg litre−1] on oviposition habitat selection (OHS) by Culiseta longiareolata and Culex laticinctus mosquitoes in an outdoor mesocosm. Additionally, larval development of C. longiareolata was examined in response to this fire retardant.3. The results demonstrate that OHS increased in both species, whereas survivorship (0–86.7%) and time to metamorphosis (5.4–23.1 days) of C. longiareolata were not affected by any of the treatments. However, cyanobacterial biomass increased and heterotrophic bacteria decreased in response to the addition of FR.4. It was found that FR applications can have unintended consequences which can increase OHS by female mosquitoes. The growing use of fire retardants worldwide, and especially in Israel, justify evidence‐based assessment and environmental management of their use. This study is a first critical step in filling the knowledge gap with respect to the impact of FR on aquatic ecosystems.

Effective amphibian conservation requires knowledge of both the aquatic and terrestrial phases of life. As extinction probabilities are a function of population size, it is crucial not only to understand the habitat requirement of the species but also to estimate its population size. In this work, we studied the endangered fire salamander, Salamandra infraimmaculata, and analyzed the population size at a total of 14 sites-eight temporary and six permanent. For identifying the local and landscape scales factors predicting S. infraimmaculata's breeding sites we monitored 54 aquatic sites. We found that permanent sites support larger populations of adult salamanders. The breeding site characteristics analyses revealed that at the local scale water depth and shade were the most important factors and two regional variables were found to be important: proximity to another breeding site and elevation. This work provides two clear conservation implications permanent breeding sites will support much larger populations compared to temporary sites, particularly if close to other potential breeding sites, and both terrestrial and aquatic features are important for a site to be suitable for breeding.

A number of factors contribute to the process of amphibian decline, among them population fragmentation and road-kills. In this work we evaluated the effects of a road that separates the hibernation area of the Fire Salamander population from their breeding site on the demography and movement behavior. For that we monitored the population for four years using transects along the road and along a river that runs in parallel to the road and serve as the breeding site. We found that the estimated percent of the population killed on the road steadily increased minimally from 2.56 % to 10.78 % over a four year period of increasing vehicular activity. Interestingly, only a small number of individuals were documented on both sides of the road, suggesting there is a potential for population fragmentation due to the road. Additionally, we used geostatistics to reveal that the spatial distribution pattern of the population and road-kills along the road is not random. Thus, conservation efforts that focus on preventing salamanders from accessing the road surface should focus on the areas where the population and road-kills aggregate.

1. Hydroperiod duration has been identified as the main factor determining the faunal composition and structure of aquatic communities in temporary habitats. It is hypothesised that desiccation will positively affect mosquito oviposition habitat selection during the post‐drought period due to the lack of antagonists.2. An experiment was carried out in outdoor mesocosms to assess whether desiccation events have post‐drought effects on community richness and mosquito oviposition.Three different treatments were randomly assigned to the mesocosms: (i) eight mesocosms were left dried for a week and then reflooded; (ii) eight mesocosms stayed dry for 4 weeks before they were reflooded; (iii) eight mesocosms were maintained at a constant water volume of 30 litres during the entire experiment as controls. Mosquito oviposition and invertebrate community richness were monitored in every mesocosm, along with environmental parameters [water temperature, pH, conductivity, dissolved oxygen, total suspended solids (TSS) and chlorophyll a concentration].3. Post‐drought mosquito oviposition and larval abundance were higher in the short‐drought and long‐drought pools than in the control. Desiccation negatively affected the biomass of the filter feeder invertebrates in both desiccation treatments. Chlorophyll a concentrations were higher in the long‐drought pools than in controls. The negative impact of desiccation on zooplankton led to a post‐drought increase in algae, associated with an increase in mosquito oviposition.4. Despite immediate negative effect on mosquitoes, pulsed disturbances can benefit mosquitoes as they favour oviposition during the post‐disturbance recovery period due to a lower abundance of mosquito antagonists and higher food resources for their offspring.

Green roofs, which are roofs with growing substrate and vegetation, can provide habitat for arthropods in cities. Maintaining a diversity of arthropods in an urban environment can enhance the functions they fill, such as pest control and soil development. Theory suggests that the creation of a heterogeneous environment on green roofs would enhance arthropod diversity. Several studies have examined how arthropod diversity can be enhanced on green roofs, and particularly whether substrate properties affect the arthropod community, but a gap remains in identifying the effect of substrate heterogeneity within a green roof on the arthropod community. In this paper, it is hypothesized that creating heterogeneity in the substrate would directly affect the diversity and abundance of some arthropod taxa, and indirectly increase arthropod diversity through increased plant diversity. These hypotheses were tested using green roof plots in four treatments of substrate heterogeneity: (1) homogeneous dispersion; (2) mineral heterogeneity-with increased tuff concentration in subplots; (3) organic heterogeneity-with decreased compost concentrations in subplots; (4) both mineral and organic heterogeneity. Each of the four treatments was replicated twice on each of three roofs (six replicates per treatment) in a Mediterranean region. There was no effect of substrate heterogeneity on arthropod diversity, abundance, or community composition, but there were differences in arthropod communities among roofs. This suggests that the location of a green roof, which can differ in local climatic conditions, can have a strong effect on the composition of the arthropod community. Thus, arthropod diversity may be promoted by building green roofs in a variety of locations throughout a city, even if the roof construction is similar on all roofs.

Genetic studies on core versus peripheral populations have yielded many patterns. This diversity in genetic patterns may reflect diversity in the meaning of “peripheral populations” as defined by geography, gene flow patterns, historical effects, and ecological conditions. Populations at the lower latitude periphery of a species’ range are of particular concern because they may be at increased risk for extinction due to global climate change. In this work we aim to understand the impact of landscape and ecological factors on different geographical types of peripheral populations with respect to levels of genetic diversity and patterns of local population differentiation. We examined three geographical types of peripheral populations of the endangered salamander, Salamandra infraimmaculata, in Northern Israel, in the southernmost periphery of the genus Salamandra, by analyzing the variability in 15 microsatellite loci from 32 sites. Our results showed that: (1) genetic diversity decreases towards the geographical periphery of the species’ range; (2) genetic diversity in geographically disjunct peripheral areas is low compared to the core or peripheral populations that are contiguous to the core and most likely affected by a founder effect; (3) ecologically marginal conditions enhance population subdivision. The patterns we found lead to the conclusion that genetic diversity is influenced by a combination of geographical, historical, and ecological factors. These complex patterns should be addressed when prioritizing areas for conservation.

A key environmental factor that varies both spatially and temporally in surface waters is dissolved oxygen (DO). In stagnant ephemeral freshwater ponds, DO can fluctuate diurnally and seasonally, while the constant mixing of water in streams typically maintain DO levels close to saturation with only minor fluctuations. Larvae of the Near Eastern fire salamander (Salamandra infraimmaculata) develop in a range of waterbodies that vary in flow and permanence. To study inter-population variation in larval response to environmental change, we translocated larvae between stream and pond habitats and exposed larvae sampled from different habitat types to hypoxic and normoxic conditions in the laboratory. Larvae transferred from stream to pond retain gill size, while larvae transferred from pond to stream show a reduction in gill size. Larvae that were caged within their native habitat, either stream or pond, display a decrease in gill size similar to larvae transferred from pond to stream. When exposed to experimentally manipulated levels of DO in the laboratory larvae, respectively, increase and decrease gill size under hypoxic and normoxic conditions. Habitat-type origin had a significant effect on the degree of change in gill size with larvae from permanent streams demonstrating the lowest absolute variation in gill size. There was no interaction between DO level (hypoxic/normoxic) and the larvae habitat-type origin. These results suggest that S. infraimmaculata larvae are locally adapted to their aquatic breeding habitat through the plastic ability to respond to the prevailing respiratory conditions by rapidly decreasing or increasing gill size.

For a vector species, understanding their egg raft predation (consumption) or destruction is essential for both ecological and human health reasons since it directly influences its fitness. In a mesocosm experiment, we assessed differences in Culiseta longiareolata egg raft predation/destruction by three aquatic predators Notonecta maculata (backswimmers), Sympetrum fonscolombii (dragonflies) and Ommatotriton vittatus (newts), both in the presence and absence of an alternate prey (Culex larve). Egg raft predation and destruction significantly differed between predators types, and strongly influenced by the presence of alternate prey. Backswimmers attacked and destroyed (broke down) all egg rafts until they disintegrated and sank in water regardless of whether an alternative prey was present. Egg raft predation by dragonflies was common in the absence of alternative prey, but rare when alternative prey was present. Predation by newts was rare regardless of whether there was an alternative prey. The number of alternate prey consumed also significantly differed between predators (P<0.001) with backswimmers being the most effective predator. Relatively few studies have tested for egg raft predation/destruction. Hence it is crucial that we conduct similar trials in other landscapes since such predators can prove to be key agents for the biological control of mosquitoes.

Green roofs are often used to cool and insulate the buildings below them, but there is a lack of information about optimum design of green roofs to achieve this function in Mediterranean climates. In such climates, the cooling effect of green roofs can also be important on warm days during the winter. In this study, we compared temperatures of non-green roofs, green roofs planted with Sedum sediforme (Jacq.) Pau, and green roofs planted with the same density of Sedum together with annual plants. These plants were chosen because annual plant communities are an important part of semi-arid environments, which are analogous to green roofs. Temperatures were measured below green roof plots and in the substrate. We found that temperatures below green roofs were around 7 degrees C cooler than below non-green roofs, and that temperatures below Sedum roofs were, in general, 1.5 degrees C cooler than below annuals + Sedum roofs. The difference in temperatures between green roofs types was seen in summer in the substrate and in winter below the roof. Substrate moisture was higher in Sedum roofs in the summer. Vegetation cover was higher in annuals + Sedum roofs in winter and in Sedum roofs in summer, as Sedum grown alone grew more than Sedum grown with annuals. Overall, Sedum roofs appear to be more effective in providing cooling benefits in this climate, due to higher substrate moisture and vegetation cover in the summer. Thus, to achieve the cooling benefits of an extensive green roof in a Mediterranean climate, it may be prudent to use succulent species alone, without annuals that would compete with Sedum in winter, while not providing a cooling effect in summer.

The mosquito Aedes phoeniciae is a potential disease vector that inhabits the coastal rock-pools of the Southeastern Mediterranean Sea. Our year-long study examined the abundance and distribution of Ae. phoeniciae in 49 rock-pools along HaBonim Beach Nature Reserve (Israeli coast) on a monthly basis (September 2016 to August 2017). Additionally, the physical, chemical, and biological characteristics of the rock-pools were measured. Our results showed a correlation between the abundance of Ae. phoeniciae and abiotic (salinity, pool volume, and pH) and biotic (bacterial, micro-phytoplankton, and chironomid abundance) characteristics. A complementary experiment was conducted to examine the role of bacteria and phytoplankton on Ae. phoeniciae larval performance by rearing larvae in seawater (SW) or seawater without microbes (FSW, 0.2-µm). Ae. phoeniciae grown in SW exhibited a high survivorship rate (~ 77%), while lower survivorship rate was measured in the FSW treatments (~ 45%). Furthermore, a higher number of adult females were found in the SW compared to FSW treatments (35 and 11, respectively), while the number of male adults remained similar. Our results suggest that Ae. phoeniciae larvae rely on the water characteristics and especially on the microbial communities that habitat the rock-pools. These results may enable improved mosquito control of Ae. phoeniciae along the Southeastern Mediterranean Sea.