<div> <p><a>Biotic interactions play key roles in determining invasive species’ establishment success in receiving ecosystems (Tilman 2004). The invasional meltdown hypothesis suggests that initial invaders can facilitate subsequent invasions through direct (e.g., commensalism, mutualism) and indirect (e.g., changes in habitat and energy flow) pathways (Simberloff & Holle 1999; Simberloff 2006). Such positive interactions among invaders can alter community-level processes, but little research on this has been done in aquatic-terrestrial landscapes. My dissertation explores the links between reciprocal facilitation of invasive species and ecosystem change in a desert river system in the southwest USA. </a></p> <p> Like most rivers in the southwestern United States, the San Juan River has been altered by hydrologic regulation and biological invasions that affect ecosystem function and act synergistically to induce substantial ecosystem change. Invasion of channel catfish (<i>Ictalurus punctatus</i>) has drastically altered the fish assemblage of the San Juan River, yet the impacts of riparian invasion by a fruit-bearing tree, Russian olive (<i>Elaeaganus agustifolia</i>) have largely been ignored. Channel catfish have been observed consuming Russian olive fruits, but the level of facilitation between species and corresponding impact on the ecosystem is unclear. </p> <p>Channel catfish may benefit directly from Russian olive invasion by feeding on fallen fruits and/or indirectly from habitat alteration and invertebrate prey production from Russian olive detritus. Additionally, channel catfish may facilitate germination, growth, and seed dispersal of Russian olive. Mutualism between these invaders could increase the fitness of each species, thereby facilitating invasion success. Plant-animal mutualism is the most common form of facilitation among invaders, but no studies, to our knowledge, have evaluated facilitation between invasive riparian plants and aquatic invaders and their combined impact on ecosystem function. My goal preparing this dissertation is to determine whether mutual facilitation between riparian and aquatic invasive species influences ecosystem change through biotic interactions. </p> <p>To test for mutual facilitation, I first determined the contribution of Russian olive fruits to channel catfish growth by evaluating seasonal diet composition across four sites and six time periods. I then used replicated growth experiments to determine assimilation rates of Russian olive fruits consumed by channel catfish. Using bioenergetics models, I then determined how Russian olive subsidies in San Juan River contribute to channel catfish biomass. To determine whether channel catfish benefit Russian olive reproduction, I compared germination rates of seeds consumed by channel catfish to seeds consumed by terrestrial mammals and control seeds that had not been eaten. </p> <p>Russian olive fruits were the most important diet item for channel catfish during the fall and spring, comprising up to 57 and 70% of stomach contents by mass, respectively, and were consumed throughout the year. Feeding trials revealed that Russian olive fruits contributed little to growth or lipid deposition, but they did provide metabolic energy allowing channel catfish fed exclusively Russian olive fruits to maintain weight. In addition, Russian olive trees received a reproductive benefit through increased germination success of seeds consumed by channel catfish over those transported by water. Using bioenergetic models, I showed that Russian olive fruits subsidized 46% of San Juan River channel catfish biomass, indicating that the subsidy from Russian olive fruits had a population-level impact. This dissertation thus establishes mutual facilitation by non-native species across ecosystem boundaries, a phenomenon that few studies heretofore have demonstrated in the ecology or invasion biology literature.</p> </div> <br>

Doctor of Philosophy(PhD), / Until recently, Northwestern Australia was thought to be relatively free of serious fruit fly pests. Although a noxious strain, present in Darwin since 1985, was widely believed to be an infestation of the Queensland fruit fly, Bactrocera tryoni, from the East coast, the fruit flies present outside this area were believed to be the benign endemic species, B. aquilonis. However, during the year 2000, infestations of fruit flies were discovered on major commercial crops in both Western Australia and the Northern Territory. It was not known whether these outbreaks were due to an invasion of the major pest species, Bactrocera tryoni, a change in the behaviour of B. aquilonis, or a hybridisation event between the two species. Finding the source of these outbreaks has been complicated by the fact that, since B. tryoni and B. aquilonis are virtually indistinguishable morphologically, it was not known which species are present in the region. Traditionally any tryoni complex fly caught in the Northwest was called B. aquilonis based solely on location. In order to get a good population profile of the region, an extensive trapping program was set up to include flies from urban areas, commercial crops and natural areas where the benign strain is thought to remain. Tests of genetic differentiation and clustering analyses revealed a high degree of homogeneity in the Northwest samples, suggesting that just one species is present in the region. The Northwest samples were genetically differentiated from the Queensland samples but only to a small degree (FST =0.0153). MtDNA sequencing results also showed a small degree of differentiation between these regions. A morphological study of wing shape indicated that there are some minor identifiable morphological differences between East coast and Northwest laboratory reared flies. This difference was greater than that seen between B. jarvisi populations across the same geographic range. The results suggest that the flies caught in the Northwest are a separate population of B. tryoni. Soon after pest flies were discovered in Darwin, a population became established in Alice Springs. This population had a low genetic diversity compared with Queensland and Darwin populations, and showed evidence of being heavily founded. In 2000, an outbreak was discovered in the nearby town of Ti Tree. Due to the geographic and genetic similarity of these populations, Alice Springs was determined to be the source of the Ti Tree outbreak. To investigate the founding of these populations, a program was developed to estimate the propagule size. Using a simulation method seven different statistics were tested for estimating the propagule size of an outbreak population. For outbreaks originating from populations with high genetic diversity, the number of alleles was a good estimator of propagule size. When, however, the genetic diversity of the source population was already reduced, allele frequency measures, particularly the likelihood of obtaining the outbreak population from the source population, gave more accurate estimates. Applying this information to the Alice Springs samples, it was estimated that just five flies were needed to found the major population in and around Alice Springs. For Ti Tree, the propagule size was estimated to be 27 flies (minimum 10). In 2000, a much larger outbreak occurred in the developing horticultural region of Kununurra in northern Western Australia. An important question for the management of the problem is whether there is an established fly population or the flies are reinvading each year. This population was found to have a large amount of gene flow from the Northern Territory. Within the Kununurra samples, one group of flies was genetically differentiated from all the other samples. This group came from a small geographic area on the periphery of Kununurra and appeared to be the result of an invasion into this area at the time when the population was building up following the dry season. A further threat to the Northwest horticultural regions comes from B. jarvisi. A recent increase in the host range of this species has lead to speculation that it may become a greater pest in Northwestern Australia. At the present time, protocols for the population monitoring and disinfestation of this species are not in place. Here it is shown that B. jarvisi eggs are more heat tolerant than B. tryoni eggs and that monitoring of B. jarvisi populations is possible using cue lure traps placed according to fruiting time and location of their favoured host, Planchonia careya.

Doctor of Philosophy(PhD), / Until recently, Northwestern Australia was thought to be relatively free of serious fruit fly pests. Although a noxious strain, present in Darwin since 1985, was widely believed to be an infestation of the Queensland fruit fly, Bactrocera tryoni, from the East coast, the fruit flies present outside this area were believed to be the benign endemic species, B. aquilonis. However, during the year 2000, infestations of fruit flies were discovered on major commercial crops in both Western Australia and the Northern Territory. It was not known whether these outbreaks were due to an invasion of the major pest species, Bactrocera tryoni, a change in the behaviour of B. aquilonis, or a hybridisation event between the two species. Finding the source of these outbreaks has been complicated by the fact that, since B. tryoni and B. aquilonis are virtually indistinguishable morphologically, it was not known which species are present in the region. Traditionally any tryoni complex fly caught in the Northwest was called B. aquilonis based solely on location. In order to get a good population profile of the region, an extensive trapping program was set up to include flies from urban areas, commercial crops and natural areas where the benign strain is thought to remain. Tests of genetic differentiation and clustering analyses revealed a high degree of homogeneity in the Northwest samples, suggesting that just one species is present in the region. The Northwest samples were genetically differentiated from the Queensland samples but only to a small degree (FST =0.0153). MtDNA sequencing results also showed a small degree of differentiation between these regions. A morphological study of wing shape indicated that there are some minor identifiable morphological differences between East coast and Northwest laboratory reared flies. This difference was greater than that seen between B. jarvisi populations across the same geographic range. The results suggest that the flies caught in the Northwest are a separate population of B. tryoni. Soon after pest flies were discovered in Darwin, a population became established in Alice Springs. This population had a low genetic diversity compared with Queensland and Darwin populations, and showed evidence of being heavily founded. In 2000, an outbreak was discovered in the nearby town of Ti Tree. Due to the geographic and genetic similarity of these populations, Alice Springs was determined to be the source of the Ti Tree outbreak. To investigate the founding of these populations, a program was developed to estimate the propagule size. Using a simulation method seven different statistics were tested for estimating the propagule size of an outbreak population. For outbreaks originating from populations with high genetic diversity, the number of alleles was a good estimator of propagule size. When, however, the genetic diversity of the source population was already reduced, allele frequency measures, particularly the likelihood of obtaining the outbreak population from the source population, gave more accurate estimates. Applying this information to the Alice Springs samples, it was estimated that just five flies were needed to found the major population in and around Alice Springs. For Ti Tree, the propagule size was estimated to be 27 flies (minimum 10). In 2000, a much larger outbreak occurred in the developing horticultural region of Kununurra in northern Western Australia. An important question for the management of the problem is whether there is an established fly population or the flies are reinvading each year. This population was found to have a large amount of gene flow from the Northern Territory. Within the Kununurra samples, one group of flies was genetically differentiated from all the other samples. This group came from a small geographic area on the periphery of Kununurra and appeared to be the result of an invasion into this area at the time when the population was building up following the dry season. A further threat to the Northwest horticultural regions comes from B. jarvisi. A recent increase in the host range of this species has lead to speculation that it may become a greater pest in Northwestern Australia. At the present time, protocols for the population monitoring and disinfestation of this species are not in place. Here it is shown that B. jarvisi eggs are more heat tolerant than B. tryoni eggs and that monitoring of B. jarvisi populations is possible using cue lure traps placed according to fruiting time and location of their favoured host, Planchonia careya.