Classification of Data to Extract Knowledge from Neural Networks

A major drawback of artificial neural networks is their black-box character. Therefore, the rule extraction algorithm is becoming more and more important in explaining the extracted rules from the neural networks. In this paper, we use a method that can be used for symbolic knowledge extraction from neural networks, once they have been trained with desired function. The basis of this method is the weights of the neural network trained. This method allows knowledge extraction from neural networks with continuous inputs and output as well as rule extraction. An example of the application is showed. This example is based on the extraction of average load demand of a power plant.

A villamos erőművek szén-dioxid-kibocsátásának modellezése reálopciók segítségével = Modelling of the carbon dioxide emissions of a power plant, using real options

A szerző egy, a szennyezőanyag-kibocsátás európai kereskedelmi rendszerében megfelelésre kötelezett gázturbinás erőmű szén-dioxid-kibocsátását modellezi négy termékre (völgy- és csúcsidőszaki áramár, gázár, kibocsátási kvóta) vonatkozó reálopciós modell segítségével. A profitmaximalizáló erőmű csak abban az esetben termel és szennyez, ha a megtermelt áramon realizálható fedezete pozitív. A jövőbeli időszak összesített szén-dioxid-kibocsátása megfeleltethető európai típusú bináris különbözetopciók összegének. A modell keretein belül a szén-dioxid-kibocsátás várható értékét és sűrűségfüggvényét becsülhetjük, az utóbbi segítségével a szén-dioxid-kibocsátási pozíció kockáztatott értékét határozhatjuk meg, amely az erőmű számára előírt megfelelési kötelezettség teljesítésének adott konfidenciaszint melletti költségét jelenti. A sztochasztikus modellben az alaptermékek geometriai Ornstein-Uhlenbeck-folyamatot követnek. Ezt illesztette a szerző a német energiatőzsdéről származó publikus piaci adatokra. A szimulációs modellre támaszkodva megvizsgálta, hogy a különböző technológiai és piaci tényezők ceteris paribus megváltozása milyen hatással van a megfelelés költségére, a kockáztatott értékére. ______ The carbon-dioxide emissions of an EU Emissions Trading System participant, gas-fuelled power generator are modelled by using real options for four underlying instruments (peak and off-peak electricity, gas, emission quota). This profit-maximizing power plant operates and emits pollution only if its profit (spread) on energy produced is positive. The future emissions can be estimated by a sum of European binary-spread options. Based on the real-option model, the expected value of emissions and its probability-density function can be deducted. Also calculable is the Value at Risk of emission quota position, which gives the cost of compliance at a given confidence level. To model the prices of the four underlying instruments, the geometric Ornstein-Uhlenbeck process is supposed and matched to public available price data from EEX. Based on the simulation model, the effects of various technological and market factors are analysed for the emissions level and the cost of compliance.

Spatial and temporal patterns of true bug assemblages extracted with Berlese funnels (Data to the knowledge on the ground-living Heteroptera of Hungary, No. 2)

A rich material of Heteroptera extracted with Berlese funnels by Dr. I. Loksa between 1953–1974 in Hungary, has been examined. Altogether 157 true bug species have been identified. The ground-living heteropteran assemblages collected in different plant communities, substrata, phytogeographical provinces and seasons have been compared with multivariate methods. Because of the unequal number of samples, the objects have been standardized with stochastic simulation. There are several true bug species, which have been collected in almost all of the plant communities. However, characteristic ground-living heteropteran assemblages have been found in numerous Hungarian plant community types. Leaf litter and debris seem to have characteristic bug assemblages. Some differences have also been recognised between the bug fauna of mosses growing on different surfaces. Most of the species have been found in all of the great phytogeographical provinces of Hungary. Most high-dominance species, which have been collected, can be found at the ground-level almost throughout the year. Specimens of many other species have been collected with Berlese funnels in spring, autumn and/or winter. The diversities of the ground-living heteropteran assemblages of the examined objects have also been compared.

Ecology of the Acalypta species occurring in Hungary Insecta Heteroptera Tingidae data to the knowledge on the ground-living Heteroptera of Hungary, No 3

As a third part of a series of papers on the ground-living true bugs of Hungary, the species belonging to the lace bug genus Acalypta Westwood, 1840 (Insecta: Heteroptera: Tingidae) were studied. Extensive materials collected with Berlese funnels during about 20 years all over Hungary were identified. Based on these sporadic data of many years, faunistic notes are given on some Hungarian species. The seasonal occurrence of the species are discussed. The numbers of specimens of different Acalypta species collected in diverse plant communities are compared with multivariate methods. Materials collected with pitfall traps between 1979–1982 at Bugac, Kiskunság National Park were also processed. In this area, only A. marginata and A. gracilis occurred, both in great number. The temporal changes of the populations are discussed. Significant differences could be observed between the microhabitat distribution of the two species: both species occurred in very low number in traps placed out in patches colonized by dune-slack purple moorgrass meadow; Acalypta gracilis preferred distinctly the Pannonic dune open grassland patches; A. marginata occurred in almost equal number in Pannonic dune open grassland and in Pannonic sand puszta patches.

Comparative analysis of the relationship between phenological phenomena and meteorological indicators based on insect and plant monitoring

Climate change is one of the biggest environmental problems of the 21st century. The most sensitive indicators of the effects of the climatic changes are phenological processes of the biota. The effects of climate change which were observed the earliest are the remarkable changes in the phenology (i.e. the timing of the phenophases) of the plants and animals, which have been systematically monitored later. In our research we searched for the answer: which meteorological factors show the strongest statistical relationships with phenological phenomena based on some chosen plant and insect species (in case of which large phenological databases are available). Our study was based on two large databases: one of them is the Lepidoptera database of the Hungarian Plant Protection and Forestry Light Trap Network, the other one is the Geophytes Phenology Database of the Botanical Garden of Eötvös Loránd University. In the case of butterflies, statistically defined phenological dates were determined based on the daily collection data, while in the case of plants, observation data on blooming were available. The same meteorological indicators were applied for both groups in our study. On the basis of the data series, analyses of correlation were carried out and a new indicator, the so-called G index was introduced, summing up the number of correlations which were found to be significant on the different levels of significance. In our present study we compare the significant meteorological factors and analyse the differences based on the correlation data on plants and butterflies. Data on butterflies are much more varied regarding the effectiveness of the meteorological factors.

Cut mowing and grazing effects with grey cattle on plant species composition in case of Pannon wet grasslands

Examined area can be found at Balaton Uplands National Park (Hungary). 5 sample areas were examined in Badacsonytördemic: 1: 32 hectare under-grazed pasture, 2: 38 hectare overgrazed pasture, 3: 34 hectare hayfield, 4: trampled area, 5: beaten track. Livestock population was 118 in the monitored pastures. Sampling was executed along five 52m long circular transects, within 5cm × 5cm interlocking quadrates. Based on the data we can state that the curve of the drinking area was the highest of speciesarea examinations however weed appeared because of degradation which provided more species. According to species-area examinations overgrazed areas were richer in species then other examined areas. Based on diversity data drinking area considered degraded, while meadow and overgrazed areas was considered as proper state. Diversity of meadow was larger, but dominance of economically useful species was smaller. The amount of less valuable species – Carex hirta – increased.

The role of disturbance and resource availability in structuring plant communities

Disturbances alter competitive hierarchies by reducing populations and altering resource regimes. The interaction between disturbance and resource availability may strongly influence the structure of plant communities, as observed in the recolonization of seagrass beds in outer Florida Bay that were denuded by sea-urchin overgrazing. There is no consensus concerning the interaction between disturbance and resource availability on competition intensity (CI). On the other hand, species diversity is dependent on both factors. Peaks in species diversity have been observed to occur when both resource availability and disturbance intensity are high, thus implying that CI is low. Based on this supposition of previous models, I presented the resource-disturbance hypothesis as a graphical model to make predictions of CI as a function of both disturbance intensity and the availability of a limiting resource. The predictions of this model were tested in two experiments within a seagrass community in south Florida, in which transplants of Halodule wrightii were placed into near-monocultures of Syringodium filiforme in a full-factorial array. In the first experiment, two measures of relative CI were calculated based on the changes in the short-shoot number (SS) and of rhizome length (RHL) on the transplants. Both light and disturbance were identified as important factors, though the interaction between light * disturbance was not significant. Relative CISS ranged between 0.2 and 1.0 for the high light and high disturbance treatments and the relative CIRHL < 0 for the same treatments, though results were not significantly different due to high variability and low sample size. These results, including a contour schematic using six data points from the different treatment combinations, preliminarily suggests that the resource-disturbance hypothesis may be used may be used as a next step in developing our understanding of the mechanisms involved in structuring plant communities. Furthermore, the focus of the model is on the outcome of CI, which may be a useful predictor of changes in species diversity. Further study is needed to confirm the results of this study and validate the usefulness of this model in other systems. ^

Plant biogeography and conservation on a tropical island: Isla del Coco, Costa Rica

Isla del Coco (Cocos Island) is a small volcanic island located in the Pacific 500 km west of Costa Rica. Three collecting trips to Isla del Coco, in addition to herbarium research, were completed in order to assess the floristic diversity of the island. The current flora of Isla del Coco contains 262 plant species of which 37 (19.4%) are endemic. This study reports 58 species as new to the island. Seventy-one species (27.1%) were identified as introduced by humans. In addition, five potentially invasive plant species are identified. Seven vegetation types are identified on the island: bayshore, coastal cliff, riparian, low elevation humid forest, high elevation cloud forest, landslide and islet. ^ The biogeographic affinities of the native and endemic species are with Central America/northern South America and to a lesser extent, the Caribbean. Endemic species in the genus Epidendrum were investigated to determine whether an insular radiation event had produced two species found on Isla del Coco. Phylogenetic analysis of the internal transcribed spacer (ITS) of nuclear ribosomal DNA was not able to disprove that the endemic species in this genus are not sister species. Molecular biogeographic analyses of ITS sequence data determined that the Isla del Coco endemic species in the genera Epidendrum, Pilea and Psychotria are most closely related to Central American/northern South American taxa. No biogeographical links were found between the floras of Isla del Coco and the Galápagos Islands. ^ The native and endemic plant diversity of Isla del Coco is threatened with habitat degradation by introduced pigs and deer, and to a lesser extent, by exotic plant species. The IUCN Red List and RAREplants criteria were used to assess the extinction threat for the 37 endemic plant taxa found on the island. All of the endemic species are considered threatened with extinction at the Critically Endangered (CR) by the IUCN criteria or either CR or Endangered (EN) using RAREplants methodology. ^

Environment, culture, and medicinal plant knowledge in an indigenous Amazonian community

Diminishing cultural and biological diversity is a current global crisis. Tropical forests and indigenous peoples are adversely affected by social and environmental changes caused by global political and economic systems. The purpose of this thesis was to investigate environmental and livelihood challenges as well as medicinal plant knowledge in a Yagua village in the Peruvian Amazon. Indigenous peoples’ relationships with the environment is an important topic in environmental anthropology, and traditional botanical knowledge is an integral component of ethnobotany. Political ecology provides a useful theoretical perspective for understanding the economic and political dimensions of environmental and social conditions. This research utilized a variety of ethnographic, ethnobotanical, and community-involved methods. Findings include data and analyses about the community’s culture, subsistence and natural resource needs, organizations and institutions, and medicinal plant use. The conclusion discusses the case study in terms of the disciplinary framework and offers suggestions for research and application.

Ecological effects of low-level phosphorus additions on two plant communities in a neotropical freshwater wetland ecosystem

We conducted a low-level phosphorus (P) enrichment study in two oligotrophic freshwater wetland communities (wet prairies [WP] and sawgrass marsh [SAW]) of the neotropical Florida Everglades. The experiment included three P addition levels (0, 3.33, and 33.3 mg P m−2 month−1), added over 2 years, and used in situ mesocosms located in northeastern Everglades National Park, Fla., USA. The calcareous periphyton mat in both communities degraded quickly and was replaced by green algae. In the WP community, we observed significant increases in net aboveground primary production (NAPP) and belowground biomass. Aboveground live standing crop (ALSC) did not show a treatment effect, though, because stem turnover rates of Eleocharis spp., the dominant emergent macrophyte in this community, increased significantly. Eleocharis spp. leaf tissue P content decreased with P additions, causing higher C:P and N:P ratios in enriched versus unenriched plots. In the SAW community, NAPP, ALSC, and belowground biomass all increased significantly in response to P additions. Cladium jamaicense leaf turnover rates and tissue nutrient content did not show treatment effects. The two oligotrophic communities responded differentially to P enrichment. Periphyton which was more abundant in the WP community, appeared to act as a P buffer that delayed the response of other ecosystem components until after the periphyton mat had disappeared. Periphyton played a smaller role in controlling ecosystem dynamics and community structure in the SAW community. Our data suggested a reduced reliance on internal stores of P by emergent macrophytes in the WP that were exposed to P enrichment. Eleocharis spp. rapidly recycled P through more rapid aboveground turnover. In contrast, C. jamaicense stored added P by initially investing in belowground biomass, then shifting growth allocation to aboveground tissue without increasing leaf turnover rates. Our results suggest that calcareous wetland systems throughout the Caribbean, and oligotrophic ecosystems in general, respond rapidly to low-level additions of their limiting nutrient.

Invasive exotic plant indicators for ecosystem restoration: An example from the Everglades restoration program

We have developed a comprehensive ecological indicator for invasive exotic plants, a human-influenced component of the Everglades that could threaten the success of the restoration initiative. Following development of a conceptual ecological model for invasive exotic species, presented as a companion paper in this special issue, we developed criteria to evaluate existing invasive exotic monitoring programs for use in developing invasive exotic performance measures. We then used data from the selected monitoring programs to define specific performance measures, using species presence and abundance as the basis of the indicator for invasive exotic plants. We then developed a series of questions used to evaluate region and/or individual species status with respect to invasion. Finally, we used an expert panel who had answered the questions for invasive exotic plants in the Everglades Lake Okeechobee model to develop a stoplight restoration report card to communicate invasive exotic plant status. The report card system provides a way to effectively evaluate and present indicator data to managers, policy makers, and the public using a uniform format among indicators. Collectively, the model, monitoring assessment, performance measures, and report card enable us to evaluate how invasive plants are impacting the restoration program and how effectively that impact is being managed. Applied through time, our approach also allows us to follow the progress of management actions to control the spread and reduce the impacts of invasive species and can be easily applied and adapted to other large-scale ecosystem projects.

Seasonal plant water uptake patterns in the saline southeast Everglades ecotone

The purpose of this study was to determine the seasonal water use patterns of dominant macrophytes coexisting in the coastal Everglades ecotone. We measured the stable isotope signatures in plant xylem water of Rhizophora mangle, Cladium jamaicense, and Sesuvium portulacastrum during the dry (DS) and wet (WS) seasons in the estuarine ecotone along Taylor River in Everglades National Park, FL, USA. Shallow soilwater and deeper groundwater salinity was also measured to extrapolate the salinity encountered by plants at their rooting zone. Average soil water oxygen isotope ratios (δ 18O) was enriched (4.8 ± 0.2‰) in the DS relative to the WS (0.0 ± 0.1‰), but groundwater δ 18O remained constant between seasons (DS: 2.2 ± 0.4‰; WS: 2.1 ± 0.1‰). There was an inversion in interstitial salinity patterns across the soil profile between seasons. In the DS, shallow water was euhaline [i.e., 43 practical salinity units (PSU)] while groundwater was less saline (18 PSU). In the WS, however, shallow water was fresh (i.e., 0 PSU) but groundwater remained brackish (14 PSU). All plants utilized 100% (shallow) freshwater during the WS, but in the DS R. mangle switched to a soil–groundwater mix (δ 55% groundwater) while C. jamaicense and S. portulacastrum continued to use euhaline shallow water. In the DS, based on δ 18O data, the roots of R. mangle roots were exposed to salinities of 25.4 ± 1.4 PSU, less saline than either C. jamaicense(39.1 ± 2.2 PSU) or S. portulacastrum (38.6 ± 2.5 PSU). Although the salinity tolerance of C. jamaicense is not known, it is unlikely that long-term exposure to high salinity is conducive to the persistence of this freshwater marsh sedge. This study increases our ecological understanding of how water uptake patterns of individual plants can contribute to ecosystem levels changes, not only in the southeast saline Everglades, but also in estuaries in general in response to global sea level rise and human-induced changes in freshwater flows.

Early diagenesis of plant-derived dissolved organic matter along a wetland, mangrove, estuary ecotone

We studied the role of photochemical and microbial processes in contributing to the transformation of dissolved organic matter (DOM) derived from various plants that dominate the Florida Everglades. Plant-derived DOM leachate samples were exposed to photochemical and microbial degradation and the optical, chemical, and molecular weight characteristics measured over time. Optical parameters such as the synchronous fluorescence intensity between 270 and 290 nm (Fnpeak I), a strong indicator of protein and/or polyphenol content, decreased exponentially in all plant leachate samples, with microbial decay constants ranging from 21.0 d21 for seagrass to 20.11 d21 for mangrove (half-life [t1/2] 5 0.7–6.3 d). Similar decreases in polyphenol content and dissolved organic carbon (DOC) concentration also occurred but were generally an order of magnitude lower or did not change significantly over time. The initial molecular weight composition was reflected in the rate of Fnpeak I decay and suggests that plantderived DOM with a large proportion of high molecular weight structures, such as seagrass derived DOM, contain high concentrations of easily microbially degradable proteinaceous components. For samples exposed to extended simulated solar radiation, polyphenol and Fnpeak I photochemical decay constants were on average 20.7 d21 (t1/2 1.0 d). Our data suggest that polyphenol structures of plant-derived DOM are particularly sensitive to photolysis, whereas high molecular weight protein-like structures are degraded primarily through physical–chemical and microbial processes. Furthermore, microbial and physical processes initiated the formation of recalcitrant, highly colored high molecular weight polymeric structures in mangrove-derived DOM. Thus, partial, biogeochemical transformation of plant-derived DOM from coastal areas is rapid and is likely to influence carbon and nutrient cycling, especially in areas dominated by seagrass and mangrove forests.

Fire and grazing in a mesic tallgrass prairie: impacts on plant species and functional traits

Fire is a globally distributed disturbance that impacts terrestrial ecosystems and has been proposed to be a global “herbivore.” Fire, like herbivory, is a top-down driver that converts organic materials into inorganic products, alters community structure, and acts as an evolutionary agent. Though grazing and fire may have some comparable effects in grasslands, they do not have similar impacts on species composition and community structure. However, the concept of fire as a global herbivore implies that fire and herbivory may have similar effects on plant functional traits. Using 22 years of data from a mesic, native tallgrass prairie with a long evolutionary history of fire and grazing, we tested if trait composition between grazed and burned grassland communities would converge, and if the degree of convergence depended on fire frequency. Additionally, we tested if eliminating fire from frequently burned grasslands would result in a state similar to unburned grasslands, and if adding fire into a previously unburned grassland would cause composition to become more similar to that of frequently burned grasslands. We found that grazing and burning once every four years showed the most convergence in traits, suggesting that these communities operate under similar deterministic assembly rules and that fire and herbivory are similar disturbances to grasslands at the trait-group level of organization. Three years after reversal of the fire treatment we found that fire reversal had different effects depending on treatment. The formerly unburned community that was then burned annually became more similar to the annually burned community in trait composition suggesting that function may be rapidly restored if fire is reintroduced. Conversely, after fire was removed from the annually burned community trait composition developed along a unique trajectory indicating hysteresis, or a time lag for structure and function to return following a change in this disturbance regime. We conclude that functional traits and species-based metrics should be considered when determining and evaluating goals for fire management in mesic grassland ecosystems.