Machine learning highlights the importance of primary and secondary production in determining habitat for marine fish and macroinvertebrates

1. Species distribution models for marine organisms are increasingly used for a range of applications, including spatial planning, conservation, and fisheries management. These models have been constructed using a variety of mathematical forms and drawing on both physical and biological independent variables; however, what might be called first-generation models have mainly followed the form of linear models, or smoothing splines, informed by data collected in the context of fish surveys.
2. The performance of different classes of variables were tested in a series of species occurrence models built with machine learning methods, specifically evaluating the potential contribution of lower trophic level data. Random forest models were fitted based on the classification of the absence/presence for fish and macroinvertebrates surveyed on the US Northeast Continental Shelf.
3. The potential variables included physical, primary production, secondary production, and terrain variables. For accepted model fits, six variable importance measures were computed, which collectively showed that physical and secondary production variables make the greatest contribution across all models. In contrast, terrain variables made the least contribution to these models.
4. Multivariable analyses that account for all performance measures reinforce the role of water depth and temperature in defining species presence and absence; however, chlorophyll concentration and some specific zooplankton taxa, such as Metridia lucens and Paracalanus parvus, also make important contributions with strong seasonal variations.
5. Our results suggest that lower trophic level variables, if available, are valuable in the creation of species distribution models for marine organisms.

     

Effects of thermal-time artificial scarification on the germination dynamics of black locust (<Emphasis Type="Italic">Robinia pseudoacacia</Emphasis> L.) seeds

The aim of this study was to determine the efficiency of methods for the artificial scarification of black locust seeds based on the influence of the air and liquid nitrogen temperatures. Three populations were selected for the analyses: a managed tree stand, a selected seed stand and a seed orchard. Samples of 200 seeds (4 × 50) were randomly selected. Before scarification, the seeds were stored at ?70 °C for 24 h. Nine different scarification treatments were evaluated. To calculate the germination capacity and the dynamics of germination, observations were made on days 3, 5, 7, 10 and 14 after treatment. For statistical analyses, we used repeated measures analysis of variance. The obtained results indicated that the highest scarification efficiency for black locust seeds occurred in the treatment in liquid nitrogen and in an air temperature range of 85–90 °C. The provenance of seeds had a significant effect on the dynamics of germination. Thermal methods and the use of liquid nitrogen can be alternatives to the labour intensive methods of cutting the seeds or using measures that require specialised equipment.     

Depression and later enhancement of the critical flicker frequency during prolonged monocular deprivation

One eye was visually deprived for 1 day, and the critical flicker frequency in the other eye was determined at the start of the deprivation period and then at intervals of 3, 6, 9, 15, and 24 hours. There was an initial depression in performance, followed by an enhancement effect. No significant changes in the critical flicker frequency were observed in the occluded eye at corresponding times; thus the depression-enhancement phenomenon is specific to the nonoccluded eye.     

Evaluation of hybrid and population cultivars on standard and high-input technology in winter oilseed rape

The problem raised here is the intensification of technology in field crops of winter oilseed rape. Is there a justification for this by explaining what relationship between the yield and individual yielding traits are formed in hybrid and population cultivars. During years 2012–2015 field studies to evaluate five hybrid and five population cultivars response on the standard and high-input technology in winter oilseed rape production have been executed in Poland. High-input technology included: double foliar application (in autumn, and in spring) of micro-nutrients, additional top fertilisation of 40?kg N per ha in spring, double application of the amino-acid bio-stimulators, and additional growth regulator. According to the principle component analysis made for both technologies it was found that hybrids and population cultivars displayed varied response which is explained here on the basis of morphological, growth, and yield elements. The high-input technology improved the yield responded elements e.g. number of siliques and number of seeds per silique as well as winter hardiness. This impact was more evident for population cultivars than for hybrids. We state that high-input technology, as the proposal in this paper, that provided 9.6% increase in seed yield, is biologically justified for winter oilseed rape crop and can be recommended in North and central European countries.     

Do the impacts of alien invasive plants differ from expansive native ones? An experimental study on arbuscular mycorrhizal fungi communities

No studies have compared so far the effects of alien invasive and expansive native (widespread, mono-dominant) plants on arbuscular mycorrhizal fungi (AMF). Four global or European most successful invaders (Impatiens glandulifera, Reynoutria japonica, Rudbeckia laciniata, Solidago gigantea) and two expansive plants native to Europe (Artemisia vulgaris, Phalaris arundinacea) were grown in pots to elucidate the magnitude and direction of changes in AMF abundance, species richness, and species composition in soils from under multispecies native vegetation. In a second stage, the effects of these changes on a native plant, Plantago lanceolata, were assessed. Plant species identity had larger impact on AMF abundance, species richness, and species composition as well as on P. lanceolata than origin of the species (alien vs. native). This could be due to the character of AMF relationships with the plants, i.e., their mycorrhizal status and dependency on AMF. However, the alterations induced by the plant species in soil chemical properties rather than in AMF community were the major drivers of differences in shoot mass and photosynthetic performance of P. lanceolata. We determined that the plants produced species-specific effects on soil properties that, in turn, resulted in species-specific soil feedbacks on the native plant. These effects were not consistent within groups of invaders or natives.     

Differential influence of four invasive plant species on soil physicochemical properties in a pot experiment

Purpose

This study compared the effects of four invasive plants, namely Impatiens glandulifera, Reynoutria japonica, Rudbeckia laciniata, and Solidago gigantea, as well as two native species—Artemisia vulgaris, Phalaris arundinacea, and their mixture on soil physicochemical properties in a pot experiment.

Materials and methods

Plants were planted in pots in two loamy sand soils. The soils were collected from fallows located outside (fallow soil) and within river valley (valley soil) under native plant communities. Aboveground plant biomass, cover, and soil physicochemical properties such as nutrient concentrations, pH, and water holding capacity (WHC) were measured after two growing seasons. Discriminant analysis (DA) was used to identify soil variables responsible for the discrimination between plant treatments. Identified variables were further compared between treatments using one-way ANOVA followed by Tukey’s HSD test.

Results and discussion

Plant biomass, cover, and soil parameters depended on species and soil type. DA effectively separated soils under different plant species. DA on fallow soil data separated R. laciniata from all other treatments, especially I. glandulifera, native species and bare soil, along axis 1 (related mainly to exchangeable K, N-NH₄, total P, N-NO_3, and WHC). Large differences were found between R. laciniata and S. gigantea as indicated by axis 2 (S-SO₄, exchangeable Mg, total P, exchangeable Ca, and total Mg). DA on valley soil data separated R. japonica from all other treatments, particularly S. gigantea, R. laciniata, and native mixture, along axis 1 (N-NO₃, total N, S-SO₄, total P, pH). Along axis 2 (N-NO₃, N-NH₄, Olsen P, exchangeable K, WHC), large differences were observed between I. glandulifera and all other invaders.

Conclusions

Plant influence on soil differed both among invasive species and between invasive and native species. Impatiens glandulifera had a relatively weak effect and its soil was similar to both native and bare soils. Multidirectional effects of different invaders resulted in a considerable divergence in soil characteristics. Invasion-driven changes in the soil environment may trigger feedbacks that stabilize or accelerate invasion and hinder re-colonization by native vegetation, which has implications for the restoration of invaded habitats.

     

Are cadmium, lead and mercury concentrations in mosses across Europe primarily determined by atmospheric deposition of these metals?

Purpose

This study aimed at investigating correlations between heavy metal concentrations in mosses and modelled deposition values as well as other site-specific and regional characteristics to determine which factors primarily affect cadmium, lead and mercury concentrations in mosses. The resulting relationships could potentially be used to enhance the spatial resolution of heavy metal deposition maps across Europe.

Materials and methods

Modelled heavy metal deposition data and data on the concentration of heavy metals in naturally growing mosses were integrated into a geographic information system and analysed by means of bivariate rank correlation analysis and multivariate decision trees. Modelled deposition data were validated annually with deposition measurements at up to 63 EMEP measurement stations within the European Monitoring and Evaluation Programme (EMEP), and mosses were collected at up to 7,000 sites at 5-year intervals between 1990 and 2005.

Results and discussion

Moderate to high correlations were found between cadmium and lead concentrations in mosses and modelled atmospheric deposition of these metals: Spearman rank correlation coefficients were between 0.62 and 0.67, and 0.67 and 0.73 for cadmium and lead, respectively (p?<?0.001). Multivariate decision tree analyses showed that cadmium and lead concentrations in mosses were primarily determined by the atmospheric deposition of these metals, followed by emissions of the metals. Low to very low correlations were observed between mercury concentrations in mosses and modelled atmospheric deposition of mercury. According to the multivariate analyses, spatial variations of the mercury concentration in mosses was primarily associated with the sampled moss species and not with the modelled deposition, but regional differences in the atmospheric chemistry of mercury and corresponding interactions with the moss may also be involved.

Conclusions

At least for cadmium and lead, concentrations in mosses are a valuable tool in determining and mapping the spatial variation in atmospheric deposition across Europe at a high spatial resolution. For mercury, more studies are needed to elucidate interactions of different chemical species with the moss.     

Predicting the effects of whale population recovery on Northeast Pacific food webs and fisheries: an ecosystem modelling approach

The recovery of whale populations from historical depletion may have the potential to noticeably affect Northeast Pacific ecosystems and fisheries. Surplus production models based on whaling catch records were used to reconstruct the historical abundances of five large whale species in the waters surrounding Haida Gwaii, British Columbia. The results suggest that the local abundances of all five species were vastly higher before the onset of modern whaling. A comparison of ecosystem models representing the states of the local marine food web before and after full whale recovery indicates that abundant whales could consume large proportions of the annual production of their principal prey, ranging up to 87% for Pacific herring (Clupea pallasii) and 72% for piscivorous rockfish (Sebastes spp.). Dynamic modelling of the food web effects of whale recovery, including simulations of simultaneous top‐down and bottom‐up forcing and a Monte Carlo sensitivity analysis, revealed noticeable (～6–12%) top‐down effects on Pacific herring biomass owing to increased predation by humpback and fin whales. However, these effects cannot explain the magnitude of recent declines in local herring biomass. The dynamic modelling results also suggest that top‐down effects of whale recovery could result in reduced biomasses of large rockfish as a result of predation by sperm whales, as well as potential cascading effects on many demersal fish groups. These findings have numerous practical implications for ecosystem‐based fisheries management and whale conservation strategies in Northeast Pacific waters.