AUTOMATIC FISH DETECTION FROM DIFFERENT MARINE ENVIRONMENTS VIDEO USING DEEP LEARNING

The marine environment provides many ecosystems that support habitats biodiversity. Benthic habitats and fish species associations are investigated using underwater gears to secure and manage these marine ecosystems in a sustainable manner. The current study evaluates the possibility of using deep learning methods in particular the You Only Look Once version 3 algorithm to detect fish in different environments such as; different shading, low light, and high noise within images and by each frame within an underwater video, recorded in the Atlantic Coast of Morocco. The training dataset was collected from Open Images Dataset V6, a total of 1295 Fish images were captured and split into a training set and a test set. An optimization approach was applied to the YOLOv3 algorithm which is data augmentation transformation to provide more learning samples. The mean average precision (mAP) metric was applied to measure the YOLOv3 model’s performance. Results of this study revealed with a mAP of 91,3% the proposed method is proved to have the capability of detecting fish species in different natural marine environments also it has the potential to be applied to detect other underwater species and substratum.

Differential Effects of Transgenerational Plasticity on Morphological and Photosynthetic Properties Between an Invasive Plant and Its Congeneric Native One

Transgenerational plasticity allows offsprings to be more adaptive in the environmental conditions experienced by their parents. It is suggested that differential effects of transgenerational plasticity on growth performance of offspring ramets may help to understand successful invasion of invasive plant with clonal growth comparing with its congeneric native one. A pot experiment using invasive herb Wedelia trilobata and its congeneric native species Wedelia chinensis was conducted to investigate differential effects of high/low light treatment experienced by mother ramets on morphological and photosynthetic properties of offspring ramets subjected to stressful low light treatment. For W. chinensis, stolon length and maximum carboxylation rate (Vmax) in offspring ramets from mother ramets subjected to low light treatment were significantly greater than those in offspring ramets from mother ramets subjected to high light treatment. For W. trilobata, leaf area and potential maximum net photosynthetic rate (Pmax) in offspring ramets from mother ramets subjected to low light treatment were significantly greater than those in offspring ramets from mother ramets subjected to high light treatment. We tentatively concluded that effects of transgenerational plasticity on morphological and photosynthetic properties among clonal plants could be species-specific. In addition, more favorable effect of transgenerational plasticity on growth performance was observed in the invasive plant than in its congeneric native species. It is suggested that transgenerational plasticity may be very important for successful invasion of invasive plant with clonal growth, especially in maternal environmental conditions. So, our experiment provides new insight into invasive mechanism of invasive plants.

Defoliation Significantly Suppressed Plant Growth Under Low Light Conditions in Two Leguminosae Species

Seedlings in regenerating layer are frequently attacked by herbivorous insects, while the combined effects of defoliation and shading are not fully understood. In the present study, two Leguminosae species (Robinia pseudoacacia and Amorpha fruticosa) were selected to study their responses to combined light and defoliation treatments. In a greenhouse experiment, light treatments (L+, 88% vs L−, 8% full sunlight) and defoliation treatments (CK, without defoliation vs DE, defoliation 50% of the upper crown) were applied at the same time. The seedlings’ physiological and growth traits were determined at 1, 10, 30, and 70 days after the combined treatment. Our results showed that the effects of defoliation on growth and carbon allocation under high light treatments in both species were mainly concentrated in the early stage (days 1–10). R. pseudoacacia can achieve growth recovery within 10 days after defoliation, while A. fruticosa needs 30 days. Seedlings increased SLA and total chlorophyll concentration to improve light capture efficiency under low light treatments in both species, at the expense of reduced leaf thickness and leaf lignin concentration. The negative effects of defoliation treatment on plant growth and non-structural carbohydrates (NSCs) concentration in low light treatment were significantly higher than that in high light treatment after recovery for 70 days in R. pseudoacacia, suggesting sufficient production of carbohydrate would be crucial for seedling growth after defoliation. Plant growth was more sensitive to defoliation and low light stress than photosynthesis, resulting in NSCs accumulating during the early period of treatment. These results illustrated that although seedlings could adjust their resource allocation strategy and carbon dynamics in response to combined defoliation and light treatments, individuals grown in low light conditions will be more suppressed by defoliation. Our results indicate that we should pay more attention to understory seedlings’ regeneration under the pressure of herbivorous insects.

Production and cross-feeding of nitrite within Prochlorococcus populations

Prochlorococcus is an abundant photosynthetic bacterium in the oligotrophic open ocean where nitrogen (N) often limits the growth of phytoplankton. Prochlorococcus has evolved into multiple phylogenetic clades of high-light (HL) adapted and low-light (LL) adapted cells. Within these clades, cells encode a variety of N assimilation traits that are differentially distributed among members of the population. Among these traits, nitrate (NO3-) assimilation is generally restricted to a few clades of high-light adapted cells (the HLI, HLII, and HLVI clades) and a single clade of low-light adapted cells (the LLI clade). Most, if not all, cells belonging to the LLI clade have the ability to assimilate nitrite (NO2-), with a subset of this clade capable of assimilating both NO3- and NO2-. Cells belonging to the LLI clade are maximally abundant at the top of the nitracline and near the primary NO2- maximum layer. In some ecosystems, this peak in NO2- concentration may be a consequence of incomplete assimilatory NO3- reduction by phytoplankton. This phenomenon is characterized by a bottleneck in the downstream half of the NO3- assimilation pathway and the concomitant accumulation and release of NO2- by phytoplankton cells. Given the association between LLI Prochlorococcus and the primary NO2- maximum layer, we hypothesized that some Prochlorococcus exhibit incomplete assimilatory NO3- reduction. To assess this, we monitored NO2- accumulation in batch culture for 3 Prochlorococcus strains (MIT0915, MIT0917, and SB) and 2 Synechococcus strains (WH8102 and WH7803) when grown on NO3- as the sole N source. Only MIT0917 and SB accumulated external NO2- during growth on NO3-. Approximately 20-30% of the NO3- transported into the cell by MIT0917 was released as NO2-, with the balance assimilated into biomass. We further observed that co-cultures using NO3- as the sole N source could be established for MIT0917 and a Prochlorococcus strain that can assimilate NO2- but not NO3-. In these co-cultures, the NO2- released by MIT0917 was efficiently consumed by its partner strain during balanced exponential growth. Our findings highlight the potential for emergent metabolic partnerships within Prochlorococcus populations that are mediated by the production and consumption of the N cycle intermediate, NO2-.

An Internet of Things (IoT)-Based Master-Slave Regionalized Intelligent LED-Light-Controlling System

Reducing residential and industrial electricity consumption has been a goal of governments around the world. Lighting sources account for a large portion of the whole energy/power consumption. Unfortunately, most of the existing installed lighting systems are ancient and have poor energy efficiency. Today, many manufacturers have introduced light-controlling systems into the current market. However, existing light controlling systems may not be successfully applied to buildings, streets, and industrial buildings due to high costs and difficult installation and maintenance. To combat this issue, this article presents an easy-to-install, low-cost, Master-Slave intelligent LED light-controlling system based on Internet of Things (IoT) techniques. The benefit of using the proposed system is that the brightness of the LED lights in the same zone can be changed simultaneously to save in energy consumption. Furthermore, the parameters of the LED lights can be directly set. Moreover, the related data are collected and uploaded to a cloud platform. In this article, we use 15 W T8 LED tubes (non-induction lamps) as a case study. When the proposed system is installed in a zone with few people, the energy-saving rate is as high as 90%. Furthermore, when 12 people pass by a zone within one hour, its energy-saving rate can reach 81%. Therefore, the advantages of using the proposed system include: (1) the original lamp holder can be retained; (2) no wiring is required; and (3) no server is set up. Moreover, the goal of energy saving can also be achieved. As a result, the proposed system changes the full-dark mode of the available sensor lamp to the low power low-light mode for standby. Further, it makes the sensor lamps in the same zone brighten or low-light way simultaneously, which can quickly complete large-scale energy-saving and convenient control functions of intelligent LED lighting controlling system.

Adjustable light source for low light level with nearly constant correlated color temperature

Low light level (LLL) calibration becomes more and more important since the rapid growth of remote sensing. The spectral radiance at normal higher light levels can be calibrated with good accuracy, while LLL spectral radiance cannot. If an adjustable light source can be designed at nearly constant correlated color temperature (CCT) covering several orders of magnitude, low light level spectral radiance can be obtained with the help of a photodetector. Whether or not the spectral distribution of an integrating sphere based light source is nearly constant is investigated. By adjusting the diameter of the variable aperture between the integrating sphere and tungsten lamp, the spectral radiance can be varied over 6 orders of magnitude. However, the relative spectrum in the red region increases notably when the spectral radiance is decreased to 1/100000. If the spectral radiance is decreased further, the spectral difference can be more than 300% and CCT decreases more than 250 K. By using baffles and another integrating sphere, low light level radiation source at nearly constant spectral distribution is obtained. The variation of CCT is less than 50 K over 6 orders of magnitude.