Case Study 1: Olive Fruit Fly (Bactrocera oleae)

The olive fruit fly Bactrocera oleae (Rossi) (Diptera, Tephritidae) (former name: Dacus oleae) is a phytophagous insect associated to olive trees (Olea europaea, Oleaceae).With its larvae feedingmonophagously onolive fruits, thefly is considered the most severe pest of olive cultivation causing tremendous economic losses. The current distribution of B. oleae encompasses the Mediterranean basin, Africa, the Canary Islands, theMiddle East, California andCentralAmerica (Daane and Johnson 2010; Nardi et al. 2005).

to pupation, the last instar leaves the fruit, drops to the soil and pupates at a depth of one to nine centimetres (Dimou et al. 2003;Sharaf 1980). When conditions are favourable, e.g. during summer, pupation may also take place inside the olive fruit (Dimou et al. 2003;Sharaf 1980). Finally, mature adults emerge from the exuviae of the pupa.
Over the course of a year, the olive fruit fly exhibits three to five overlapping generations with seasonal fluctuations of fly densities (Boccaccio and Petacchi 2009;Comins and Fletcher 1988;Kokkari et al. 2017;Pontikakos et al. 2010;Voulgaris et al. 2013). The population dynamics of a Mediterranean population and phenological characteristics of investigated B. oleae populations are summarized in Table 4.1. In the Mediterranean, B. oleae eggs first appear in August, with a peak of oviposition occurring in mid-October (Bento et al. 1999;Kapatos and Fletcher 1984;Petacchi et al. 2015). Larvae are observed from the end of August and pupation was recorded from the middle of September onwards (Bento et al. 1999). However, life stage-specific developmental times and thus B. oleae population dynamics may considerably vary between locations and are strongly driven by climatic conditions (Marchi et al. 2016;Ordano et al. 2015).
While B. oleae may be able to even lay eggs in fall and early winter in mild environments (Castrignanò et al. 2012;Petacchi et al. 2015), harsh winter conditions induce overwintering for populations in most Mediterranean regions. The fly overwinters in its adult-or, more commonly, in its pupa-stage in the soil (Kapatos and Fletcher 1984;Neuenschwander et al. 1981;Sharaf 1980). Low winter temperatures regularly cause extremely high mortalities (Table 4.2), with reductions of the adult population of up to 99.7% (Arambourg and Pralavorio 1970;Bigler and Delucchi 1981;Gonçalves et al. 2012). A maximum of two flies per tree was determined to be the overwintering population in a Greek olive orchard (Krimbas and Tsakas 1971). Pupae disappear in high quantities of up to 98.5% over the winter (Arambourg and Pralavorio 1970;Gonçalves et al. 2012). Consequently, B. oleae is assumed to frequently go through a bottleneck during winter (Ochando and Reyes 2000). Since all populations in the same area suffer the bottleneck more or less simultaneously, a recolonization by surrounding populations is unlikely (Ochando and Reyes 2000). The spring population consists of the surviving adults, insects arising from overwintering pupae or individuals produced from a spring infestation of unharvested fruits or wild occurring olive trees (Economopoulos et al. 1982;Marchini et al. 2017).

Population Characteristics
In order to assess the inheritance of genetic features, certain qualities of the regarded population need to be considered. Therefore, this section reviews two important aspects determining gene propagation in B. oleae populations: sex ratio and mating behaviour.

Sex Ratio
Trap catches constitute the major data source for population analyses of the olive fruit fly. The attractiveness of traps to olive fruit flies varies over the course of a year (Neuenschwander and Michelakis 1979a) with different effects of various trap systems (e.g. pheromone based traps, food odours) on the two sexes (Economopoulos and Stravropoulou-Delivoria 1984;Neuenschwander and Michelakis 1979a). These aspects need to be considered when looking at trap-determined sex ratios of wild populations. However, most studies report a sex ratio close to 1:1 (Ant et al. 2012;Moore 1962;Speranza et al. 2004). Katsoyannos and Kouloussis (2001) observed a slightly higher number of males (52.2% males) captured on coloured spheres over the course of two years and a total of 7518 flies. In contrast to these observations, in laboratory trials, adults emerged with a ratio of 1:2 (male:female) from reared larvae (Genç and Nation 2008b). Compared to trap catches, the knock-down method or so-called "sondage" (use of insecticides) is considered to be a relatively unbiased method for population analyses. With this technique, Neuenschwander and Michelakis (1979a) demonstrated that the sex ratio of wild populations stays closer to 1:1 than is suggested by trap catches.

Mating Behaviour
After emergence from the pupa, adults complete gonad maturation and reach sexual maturity within three to eight days Mazomenos 1984). After mating, females remain unreceptive to further mating for a few days or weeks and store the received semen until ovulation (Solinas and Nuzzaci 1984;Tzanakakis et al. 1968;Zervas 1982). Studies demonstrated that female B. oleae are oligogamous, mating only one to three times (Zervas 1982;Zouros and Krimbas 1970). Zouros and Krimbas (1970) demonstrated monogamy for a vast proportion of female flies and estimated a frequency of 17% for female polygamy (Zouros and Krimbas 1970). In contrast to these findings, males are polygamous, mating daily when receptive females are available (Zervas 1982). Strong intraspecific competition and sexual selection was observed for olive fruit flies (Benelli 2014;Benelli et al. 2012Benelli et al. , 2013Benelli et al. , 2015. Competition among male flies occurred for single leaf territories suitable for courtship display (wing vibration), whereas females competed for oviposition sites (Benelli 2014;Benelli et al. 2012). Aggressive behaviour was expressed by wing waving, fast running towards the opponent, pouncing and boxing on the head and thorax of the foe (Benelli 2014).

Environmental Tolerances
The life cycle of Bactrocera oleae is closely linked to environmental conditions, in particular to local climatic conditions (Fletcher et al. 1978). This section gives an overview on the two main climatic factors impacting the development of the olive fruit fly in Mediterranean regions: temperature and relative humidity.

Temperature
Local temperature is a major driver of olive fruit fly development and a determinant of the population dynamics (Marchi et al. 2016;Ordano et al. 2015). Consequently, many models on B. oleae phenology and population evolution are driven by local temperature data. For this, comprehensive knowledge on the effect of temperature on survival and development of different life stages of B. oleae is indispensable. A number of scientific studies have assessed temperature-dependent mortalities and developmental times of the olive fruit fly, e.g. in laboratory trials, the summarized data is displayed in Tables 4.3 and 4.4. In general, high mortalities of the fly are observed at high summer temperatures and at low temperatures over the winter months (Gonçalves et al. 2012). Thus, temperature conditions of the previous seasons were found to determine the rate of B. oleae infestation (Marchi et al. 2016). Mild winters are characteristically followed by high olive fruit fly infestations (Marchi et al. 2016).
Strong temperature effects have been observed on the reproduction of Bactrocera oleae (Tzanakakis and Koveos 1986). No ovarian development was observed at temperatures lower than 12°C (Fletcher and Kapatos 1983). The upper maturation threshold was found to be 29.3°C, with very few females maturing at 29°C (Fletcher et al. 1978;Fletcher and Kapatos 1983). In some Mediterranean regions, these temperature requirements annually cause two periods when no reproduction occurs: during the winter months and at the beginning of summer, a period characterized by high temperatures and low humidity (Economopoulos et al. 1982;Fletcher et al. 1978). In mild climates, however, B. oleae might even be actively reproducing in winter (Economopoulos et al. 1982).

Relative Humidity
The development of Bactrocera oleae, i.e. longevity, maturation and survival, is strongly correlated with relative air humidity (Broufas et al. 2009;Broumas et al. Genç and Nation (2008a) 2002; Fletcher et al. 1978). Especially long life spans of adult flies have been observed in high air humidity (Broufas et al. 2009). A summary of the effect of humidity on different life parameters of B. oleae is displayed in Table 4.5. Humidity may also impact the effectiveness of specific trapping systems for fly monitoring (Bueno and Jones 2002;Mazomenos et al. 2002).

Dispersal Dynamics
Scientists have utilized various techniques to determine the dispersal potential of adult olive fruit flies, ranging from traditional capture-recapture experiments and  performance analyses on flight mills to the use of molecular techniques. This section provides a literature review in order to provide data appropriate for dispersal models. Experimentally determined maximum and typical dispersal distances as well as observed impacts of laboratory treatments and genetic parameters are summarized for Mediterranean Bactrocera oleae populations.

Dispersal Distances
In suitable environmental conditions, non-dispersive movements are assumed for adult olive fruit flies (Remund et al. 1976), with typical distances of 180-190 m overcome within two weeks (Fletcher and Economopoulos 1976;Fletcher and Kapatos 1981). Radioactive labelling revealed that the major proportion of a wild B. oleae population did not disperse further than 1000 m within one month and no migration was observed (Pelekassis et al. 1963). However, extreme conditions, such as high population densities (overcrowding) or lack of oviposition sites, might induce long distance dispersal of olive fruit flies (Economopoulos et al. 1978;Fletcher and Kapatos 1981;Remund et al. 1976). A release experiment by Fletcher and Kapatos (1981) illustrates this situation for B. oleae populations on Corfu, Greece. After release to a grove with no new season crop, flies travelled more than 400 m during the first seven days, but only dispersed over 180 m when 30% of olives in the release grove bore fruit (Fletcher and Kapatos 1981). Maximum dispersal distances for B. oleae reported in the literature range from 4000 to 5000 m (Economopoulos et al. 1978;Pelekassis et al. 1963;Remund et al. 1976). A summary of dispersal distances recorded in scientific studies is displayed in Table 4.6.  Remund et al. (1976) The flight potential of B. oleae is likely to vary with fly characteristics, e.g. fly age and sex. Remund et al. (1976) determined flight performance by means of flight propensity, number of flights and total distance flown for laboratory-reared B. oleae on flight-mills. The authors observed a threefold increase of the distance overcome by 14-day-old adult flies compared to their 2-day-old conspecifics. At the same time, females displayed a higher flight capacity with regard to all parameters assessed than their male counterparts. This effect of fly sex, however, was not detected in a release experiment by Fletcher and Economopoulos (1976).
Some strategies to reduce population densities of B. oleae in the Mediterranean area depend on the release of laboratory-reared individuals that have been sterilised by irradiation treatments. As a consequence, experimenters investigated the impact of laboratory-rearing as well as different rearing conditions on the fly's dispersal potential, assessing both non-sterilised and sterilised specimens. Based on results of field experiments, Fletcher and Economopoulos (1976) suggested greater dispersal distances for wild-compared to laboratory-reared olive fruit flies. In their experiment, no additional negative impact of the irradiation procedure of laboratory-reared B. oleae was observed. Remund et al. (1976) found a significant impact of diet on flight performance of olive fruit flies. Compared to individuals reared on an artificial diet, flies feeding on olives flew 2-3 times greater distances (Remund et al. 1976).

Genetic Variability and Gene Flow
Genetic parameters of Bactrocera oleae indicate a high degree of genetic variability for Mediterranean populations, as displayed by polymorphism and the number of alleles per locus (Augustinos et al. 2005;Ochando and Reyes 2000;Segura et al. 2008). Compared to polyphagous fruit flies, a lower genetic variability was expected for B. oleae due to its high host specialization. However, actual genetic variability was comparable or even higher (Augustinos et al. 2005;Ochando and Reyes 2000). At the same time, analyses detected more variability within than among Mediterranean B. oleae populations, indicating the occurrence of gene flow among these populations (Segura et al. 2008). In the regarded studies, gene flow (Nm, with N as the effective population size and m as the net migration rate per generation) was estimated after Wright (1931). Despite the relatively short regular dispersal distances of B. oleae individuals observed in field experiments, molecular analyses indicate a high level of gene flow among Mediterranean populations (Augustinos et al. 2005;Ochando and Reyes 2000;Segura et al. 2008). Gene flow was as high as Nm = 8.9 between different populations from Spain that were separated by hundreds of kilometres of dry climate (Ochando and Reyes 2000). According to Wright (1931), an Nm equal to 1 would be sufficient to prevent differentiation among populations. In addition to ecological dispersal, social-ecological aspects, such as trade with olive cultivars, have to be considered as potential source of gene flow (Augustinos et al. 2005;Segura et al. 2008).
With regard to the fly's dependence on its olive host and its winter disappearance, a high potential for the occurrence of bottlenecks was acknowledged in different studies (Augustinos et al. 2005;Ochando and Reyes 2000). As bottlenecks might result in inbreeding, heterozygote deficiency detected in Spanish fly populations substantiated this assumption (Ochando and Reyes 2000). An interesting result was an inhomogeneity of genetic variability among loci, which indicates selective processes (Augustinos et al. 2005). Natural selection due to agricultural practices may be the factor responsible for the pattern of genetic variability (Ochando and Reyes 2000).

Eggs
Hosted underneath the olive's epidermis, B. oleae eggs are protected from most natural enemies (Bateman 1972;Daane and Johnson 2010). However, predation may occur by birds feeding on whole olives, such as thrushes (Bigler et al. 1986;Neuenschwander et al. 1983), or by the midge Prolasioptera berlesiana (Kalaitzaki et al. 2014). Larvae of the latter occur in olive punctures infested with the fungus Macrophoma dalmatica and were observed to destroy olive fly eggs (Solinas 1967). Recognized as a so-called biocomplex, relationships and potential dependencies between B. oleae, M. dalmatica and P. berlesiana have extensively been studied and tested for the use in biological pest management. Even though olive fly oviposition holes appear to be the main entrance for the fungus as well as for the midge (González et al. 2006;Neuenschwander et al. 1983), neither of the insects plays an important role as disseminating agent of M. dalmatica spores (Harpaz and Gerson 1966). According to Neuenschwander et al. (1983) P. berlesiana oviposition immediately followed B. oleae attacks. The midges may, to a minor extent, prey on olive fly eggs and young larvae, but rely on fungal diet for survival (Harpaz and Gerson 1966;Solinas 1967). Maximum B. oleae egg mortality induced by P. berlesiana was estimated at 30-50% for Cretan populations (Neuenschwander et al. 1983). Overall predation of eggs, however, is low, with 0-6% predation-induced egg mortality determined for traditional groves in Portugal (Gonçalves et al. 2012).

Larvae
After hatching, B. oleae larvae feed monophagous on the pulp of their olive hosts (Daane and Johnson 2010). Within the tissue of an unripe fruit, olive fruit flies depend on the presence of symbiotic microorganisms for growth and survival (Bateman 1972;Ben-Yosef et al. 2010, 2014. The identity of the fly's symbiotic microflora is in the focus of scientific controversy. For many decades a mutualistic relationship with the extracellular bacterium Pseudomonas savastanoi has been suspected (Hagen 1966). However, most recently discovered associations involve the bacterium Acetobacter tropicalis (Kounatidis et al. 2009) and the before undescribed gut bacterium Candidatus Erwinia dacicola, which was numerically dominant inhabitant of the fly's oesophagus bulb (Ben-Yosef et al. 2014;Capuzzo et al. 2005;Estes et al. 2009;Sacchetti et al. 2008). Based on the absence of Ca. E. dacicola in other investigated tephritids and the peculiar morphological structure of the bacteria-hosting oesophagus bulb of B. oleae, Capuzzo et al. (2005) proposed a coevolution of B. oleae with Ca.E. dacicola. A total of 16 different bacteria were found to be associated with wild olive flies in Greece, displaying varying bacteria compositions between generations of the same year (Tsiropoulos 1983). It is important to note, that a lower number of different bacteria, as well as a reduced compositional variability was detected in laboratory-reared olive flies (Tsiropoulos 1983).

Life Stages at the Soil
Life stages associated to the soil, such as late instar larvae, pupae and emerging adults, are most exposed to predation (Bateman 1972;Daane and Johnson 2010). Common predators at the soil surface are arthropods like beetles, myriapods and ants (Daane and Johnson 2010;Neuenschwander et al. 1983). Ants are considered to play a major role in the destruction of B. oleae developmental stages. Individual ants or groups are capable of transporting B. oleae to the nest, eventually damaging their prey (Neuenschwander et al. 1983). Ants were also reported to be the main predators of B. oleae pupae in recently invaded habitats in California (Orsini 2006;Orsini et al. 2007). Additionally, food searching birds, e.g. the black bird (Turdus merula), the robin (Erithacus rubecula) or the starling (Sturnus vulgaris), may destroy life stages of B. oleae on the soil surface (Neuenschwander et al. 1983). Pupae removal by birds can account for up to 70%, depending on orchard type (Bigler et al. 1986).

Adults
In contrast to their monophagous larvae, adults of B. oleae feed on a variety of organic sources, including insect honeydews, plant nectar and pollen, fruit exudates, bird dung, bacteria and yeast (Christenson and Foote 1960;Daane and Johnson 2010;Tsiropoulos 1977). As a result, adult olive flies may be found in vegetation other than olive trees, e.g. walnut, apple or plane trees (Economopoulos et al. 1982). Despite this range of food sources, the diet of B. oleae may be poor or unbalanced in its amino acid composition (Ben-Yosef et al. 2010). Experimental evidence suggests that the gut symbiontic bacterium Ca.E. dacicola, that was present in all life stages of B. oleae (Estes et al. 2009) enables adult olive fruit flies to utilize non-essential amino acids and urea as a source of nitrogen (Ben-Yosef et al. 2010, 2014. The authors proposed that B. oleae depends on the symbiont for protein synthesis and egg production.
Birds as well as cursorial and web spiders are considered to be important predators in olive orchards, capable of suppressing B. oleae population densities (Bigler et al. 1986;Picchi et al. 2016). These predators were found to be considerably more numerous around organic olive orchards compared to conventionally managed groves  (Bigler et al. 1986;Picchi et al. 2016Picchi et al. , 2017. However, the general lack of scientific literature on further potential predators of adult olive flies and predation rates highlights avenues for future research.
The life stages of B. oleae and interspecies relations are depicted in Fig. 4.1.

Competition for Olive Fruits
A number of olive tree pests have been identified besides the olive fruit fly. A comprehensive list of pests and diseases occurring on the Maltese Islands is provided by Haber and Mifsud (2007). The authors report different species of insects and eriophyid mites, as well as fungal and bacterial diseases. Most of these species occur primarily on leaves, stems and fresh shoots of the trees (Haber and Mifsud 2007;Martin et al. 2000), which indicates a rather low potential for direct competition for olive fruits with B. oleae. However, some hemiptera species also directly impact olive fruits, e.g. Pollinia pollini, Aspidiotus nerii, Hemiberlesia rapax, Leucaspis riccae (Alford 2014;Daane et al. 2005;Haber and Mifsud 2007). Even though these species do not penetrate the olive fruit skin, potential surface modifications might impede oviposition by B. oleae. In general, no literature references could be detected concerning the fly's competition with other olive pests.
At the same time, different studies highlight aggressive intraspecific interactions for adult olive fruit flies (Benelli 2014;Benelli et al. 2013Benelli et al. , 2015. The intense intraspecific competition indicates that genetically modified flies, which might be released for eradication purposes, need to be capable of competing with wild types in order to mate and successfully spread the modified gene.

Potential Hazards with Regard to Gene Drive Release
Based on the ecological network analysis, a number of potential hazards, associated to the release of gene drive-equipped olive fruit flies to wild populations can be identified and are briefly discussed in this section.

Unintentional Long-Distance Transport of the Genetically Modified Organism
A serious increase of exposure associated with the release of genetically modified organisms is the potential for unintentional transport to non-target locations. Specimens might arrive in new locations via natural dispersal or by human agency. The latter bears potential for long-distance transport of the altered organism. In the receiving habitat, the newcomer may become invasive or co-exist with related species, enabling a potential transfer of modified genes across species boundaries. In this section potential pathways for long-distance transport of the olive fruit fly Bactrocera oleae are evaluated.
There are three main vectors for unintentional long-distance transport of B. oleae: • Transport of adult individuals Living adult insect specimens might be transported as hitchhikers on aircrafts, e.g. in the passenger area or within the airline baggage. A prominent example of insect transport along this pathway is the commonly known airport malaria (Isaäcson 1989). In a similar manner, also agricultural pests may travel by this vector (Liebhold et al. 2006). The increase of public air traffic is expected to enhance the chances of insect transport by this vector in coming decades. • Transport of immature life stages within the olive host The olive fruit fly undergoes egg and larval development within its olive fruit host. During this time, transport of the fruit might, in favourable conditions result in transport of the living insect to distant locations. Inspection records of 1962 indeed revealed that, within its host, the olive fruit fly arrived in the US via ships and aircrafts (Rainwater 1963). Both transport of single olives or entire trees may serve as a vector for unintentional introductions. Especially the trade of olive cultivars may result in unintentional introductions of the olive fruit fly and was suggested for the arrival of B. oleae in the New World (Nardi et al. 2005).
• Transport of immature stages (larvae, pupae) within the soil Life stages associated to the soil, such as larvae and pupae, may be transported within their host material. Soil adhering to shoes of outgoing tourists or soil associated to traded olive trees might serve as vector for immature stages of the olive fruit fly.

Hybridization and Horizontal Gene Transfer Across Species Boundaries
Ecological risk assessment for the release of gene drive-carrying organisms to natural populations is urged to consider the risk of unintentional transfer of the altered gene across species boundaries. Vertical inter-species transfer of altered genes will depend on the organism's ability to produce hybrids with related species. Three fruit flies from the B. (subgenus Bactrocera) dorsalis complex (B. dorsalis, B. papayae, and B. philippinensis) were found capable of cross-mating (Schutze et al. 2013) and sperm transfer was observed between B. dorsalis and B. (Bactrocera) carambolae in outdoor field cages (McInnis et al. 1999). For the olive fruit fly, no studies testing the hybridization potential with related species were detected. B. oleae belongs to the subgenus Daculus, a sister group to the subgenus Bactrocera (Smith et al. 2003;Zhang et al. 2010). Hybridization trials should focus on the nearest African congeners (Daculus), B. biguttula and B. munroi, which are also associated with Oleaceae (Copeland et al. 2004;Drew and Hancock 2000) and exhibited a mean genetic divergence of 3.4% to B. oleae (Bon et al. 2016). Morphological features, i.e. length of terminalia, pheromone composition and variation in courtship signals might play a role for mating compatibility (Iwaizumi et al. 1997;Schutze et al. 2013).
The classification of Tephritidae species and their geographic distribution is shown in Fig. 4.2.

Ecological Niche Filling by Other Species
The declared aim of planned releases of gene drive-equipped insects is the local suppression or eradication of pest species. However, after extinction, the vacant ecological niche might be occupied by a related or competing species. In order to assess this possibility for the case study of the olive fruit fly, potential competitors of the olive fly were investigated. Despite the existence of other olive pests, no studies could be detected on competition with other olive pests. As a consequence of this potential lack of direct competitors, a genetically induced reduction in olive fruit fly densities may be unlikely to cause another pest population to increase. The close co-evolution of B. oleae with symbiotic microorganisms, which is essential for its development inside an unripe fruit, already indicates that the ecological niche of the olive fruit fly might not easily be occupied by another organism.

Concluding Remark
With regard to its physiological and ecological characteristics, the olive fly can be considered a prototypic case to discuss the implications of SPAGE application in an economically significant, European context. Furthermore, B. oleae is a suitable candidate for a potential target organism in suppression drives, as it already has been the object of pest control efforts for several decades. Compared to other cases, the narrow specialisation and the involved network of ecological relations make this case relatively easy to survey. It appears well suitable to expand risk assessment experience with a wild species with a high dispersal potential, high rates of gene flow and even some risk of hybridisation with related species. It is easy to study under laboratory conditions and reproduces quickly. This case study revealed that uncertainties exist with regard to the dispersal capacity of gene drive-bearing olive flies, as well as concerning the high gene flow between different populations and most importantly with regard to the population bottlenecks that regularly occur in winter. These would significantly increase or decrease genetic variability between subpopulations and thereby severely jeopardize the intended outcome of any SPAGEapplication. Nevertheless, there remain significant knowledge gaps in particular on the role of relations to symbiotic and other microflora and wider aspects of the ecological relations. However, experience gained in the study of this species can be useful to compare assessments of organisms with a more complex or more difficult to survey life cycles for which gene drive applications are discussed.
The Control of pest populations is one of the main objectives for possible SPAGE applications. The olive fly, as the main pest in olive cultivation poses a good example of a potential target organism in a European context. This case study revealed that uncertainties exist with regard to the dispersal capacity of gene drive-bearing olive flies, as well as concerning the high gene flow between different populations and most importantly with regard to the population bottlenecks that regularly occur in winter. These would significantly increase or decrease genetic variability between subpopulations and thereby severely jeopardize the intended outcome of any SPAGEapplication. In summary, it can be said that the naturally occurring variability, in contrast to comparatively homogeneous laboratory conditions, leads to considerable and easily underestimated uncertainties about possible post-release effects. This is further exemplified in the stochastic model concerning winter bottlenecks and the individual based model, explored in Chap. 6. These models clearly show that even with a highly efficient gene drive, fluctuations in a single ecological factor may lead to vastly different outcomes.
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.