Cave-dwelling fish that can survive on surface found in Meghalaya
The limestone cave is located just 15 kilometres from Mawsynram, one of the wettest places on earth. Most hypogean (underground-dwelling) species lose their eyesight and pigmentation over generations, relying on heightened non-visual senses to navigate pitch-dark waters. Unlike them, this unique species of stone loach has eyes and shows off its colours too.
Published in the Journal of Fish Biology, a prestigious journal by the Fisheries Society of the British Isles, the discovery has drawn international attention and national applause. Assam's education minister Ranoj Pegu took to social media platform X to congratulate the team, calling it 'a proud moment for Northeast India's scientific community.'
With a pale yellow-green body and bold black bars numbering between 14 and 20, the fish sports a distinctive thick stripe near its dorsal fin — the very feature that earned it the species name densiclava, meaning 'dense stripe' in Latin.
'This discovery highlights the untapped biodiversity thriving in Meghalaya's underground ecosystems. It's the ninth new fish species we've described from Northeast India, and the sixth known cave-associated fish from Meghalaya' said Professor Dandadhar Sarma, head of Zoology department at Gauhati University.
Unlike fully cave-adapted species such as Schistura papulifera and Neolissochilus pnar — which are pigmentless, blind, and cannot survive in a surface environment— Schistura densiclava thrives inside caves but isn't entirely dependent on subterranean life.
Researchers said that the fish was found exclusively in a cool, fast-flowing stream some 60 metres inside the cave, where water temperatures hover at a chilly 18°C and oxygen levels are low. The fish's resilience in such nutrient-scarce conditions is remarkable.
Its diet includes copepods, tiny shrimp, insect fragments, and even bat guano.
According to Kangkan Sarma, one of the study's lead authors, the species shows clear sexual dimorphism. 'Males are slimmer with irregular patterns and puffier cheeks, while females are more robust and display more uniform markings,' he noted, adding, 'DNA sequencing confirms that this is a genetically distinct species, unlike any other Schistura found in the region.'
The East Khasi Hills region, where Krem Mawjingbuiñ lies, is a part of Meghalaya's celebrated cave systems — one of the richest and least explored subterranean landscapes in the world. With over 1,700 caves and cave locations documented, but only a fraction thoroughly studied, Meghalaya is considered a global hotspot for cave biodiversity.
'There were no visible signs of human disturbance inside Krem Mawjingbuiñ,' said one researcher, noting how seasonal access and the surrounding dense forest have kept the cave largely untouched by tourism or development.This ecological isolation, researchers said, is a double-edged sword — protecting species for now, but making them vulnerable to even minor intrusions.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
NDTV
4 hours ago
- NDTV
Google's New AI Can Read Ancient Roman Inscriptions; Historians Call It "Jaw-Dropping"
Google DeepMind has introduced a new artificial intelligence tool called Aeneas, designed to assist historians in studying ancient Roman inscriptions, according to a report by The Guardian. The tool helps identify the origin and date of the inscriptions and suggests missing words in damaged or incomplete texts. Aeneas, named after the Trojan hero from Roman mythology, was developed in collaboration with historians, including Dr Thea Sommerschield from the University of Nottingham. The AI was trained on a database of nearly 200,000 Latin inscriptions, containing more than 16 million characters. The inscriptions, found on monuments, tombs, and even everyday items, are valuable records of life in ancient Rome. However, many are fragmented or worn, making them difficult to interpret. Aeneas analyses the text and sometimes images from an inscription and compares it with similar examples from the 7th century BC to the 8th century AD. It uses deeper linguistic and historical patterns, not just keyword matches, to find connections, according to The Guardian. The AI can determine the likely Roman province where an inscription was created and estimate its date within about 13 years. It also proposes possible words to fill in missing parts, tested so far on texts where the original wording is already known. In tests, Aeneas provided accurate insights into famous inscriptions such as the Res Gestae Divi Augusti and linked similar texts across different regions of the Roman Empire. Historians called the tool "transformative," with 23 researchers finding it useful in 90% of cases. Experts believe Aeneas will open new opportunities in the study of Latin inscriptions, making it easier for more people to contribute to historical research without needing access to rare materials or deep prior expertise. However, scholars emphasise the importance of using the tool thoughtfully and critically.
India Today
9 hours ago
- India Today
Google DeepMind unveils Aeneas AI model, claims it can decipher ancient inscriptions in seconds
Google DeepMind has recently introduced the Aeneas model, a new approach in historical research that utilises artificial intelligence to aid the interpretation of ancient Roman inscriptions. Developed in partnership with leading universities, Aeneas accelerates the process by which historians can identify textual "parallels" in Latin is adept at processing fragmentary or damaged inscriptions, providing historians with newfound capabilities to contextualise ancient texts. While the model primarily focuses on Latin, there is potential for adaptation to other ancient languages, expanding its reach further. This adaptability promises to enhance historical inquiry across different cultures and Aeneas is tailored to work with inscriptions that often lack comprehensive contextual details. By leveraging a combination of textual and visual data, the model can offer insights into Roman life and society. According to the blogpost, Google claims that this model achieves a 73 per cent accuracy rate in restoring gaps within inscriptions up to ten characters long, and a notable capability for dating texts, placing them within 13 years of historians' estimates. These capabilities make it a vital tool for exploring the Roman world through various inscriptions, from political graffiti to business model has been tested on the Res Gestae Divi Augusti, a famous Roman inscription attributed to Emperor Augustus. Aeneas provided a distribution of possible dates, capturing different prevailing hypotheses in a quantitative manner. This demonstrates the model's ability to transform historical questions into a probabilistic framework, offering historians a new approach to longstanding debates. The use of "embeddings" helps in drawing connections, allowing historians to uncover deeper insights into historical excels in geographical attribution through a multimodal generative neural network that analyses both text and images. Utilising the Latin Epigraphic Dataset (LED) with over 1,76,000 inscriptions, Aeneas offers a more precise grouping of texts by date than other models. This integration of AI into historical workflows exemplifies the synergy between machine learning and expert knowledge, fostering a collaborative research model is accessible through an interactive platform available to researchers and educators, aligning with initiatives to improve AI literacy. Aeneas supports the restoration of inscriptions with unknown gap lengths, a critical feature for managing severely damaged texts. Its ability to search for "parallels" enriches the understanding of Roman society and its geographical expanse, significantly enhancing historical in collaboration with the University of Nottingham, and partners from the Universities of Warwick, Oxford, and Athens University of Economics and Business (AUEB), Aeneas represents a concerted effort to harness AI for historical research. This collaboration highlights a broader initiative to enable historians to identify and interpret parallels at scale, providing starting points for inquiry and developers are committed to enhancing the model's versatility through a new teaching syllabus designed to integrate technical skills with historical analysis. This initiative supports AI literacy and aligns with European educational frameworks, providing educational resources that bridge technical and historical Aeneas becomes an integral part of historical research, its capabilities in processing multimodal inputs and restoring texts of unknown length will prove indispensable. By transforming historical analysis into a more quantitative and interpretable process, Aeneas promises to unlock new perspectives on ancient societies, ensuring that the legacies of past civilisations can be explored and understood in innovative ways.- EndsMust Watch
Hindustan Times
17 hours ago
- Hindustan Times
Ludhiana: ICAR hosts workshop to boost maize yields in push for bioethanol
With the Indian government pushing for a significant increase in ethanol blending with petrol, a three-day training-cum-workshop on the theme 'Empowering Field Implementation for Maize-Based Bioethanol and Catchment Development' kicked off at the Indian Council of Agricultural Research (ICAR)–Indian Institute of Maize Research (IIMR) at Ladhowal here on Wednesday. A total of 27 field personnel from the project took part. (HT Photo) A total of 27 field personnel associated with the project, working across 78 districts in 15 states, participated in the programme. The participants shared progress from their respective regions, including successful field interventions, increasing farmer participation, and dissemination of advanced maize production technologies. Director, ICAR-IIMR, HS Jat emphasised the need to enhance maize productivity through scientific and cost-effective methods. He highlighted that the Government of India aims to achieve 30 per cent ethanol blending (E30) in petrol by the year 2030, for which maize is emerging as a key feedstock. He stressed the importance of field-based innovations and greater farmer participation in achieving this goal. To accomplish the E30 mission by 2030, India needs to produce 65–70 million tonnes of maize annually, with an annual growth rate of 8–9 per cent. This will require the deployment of high-yielding hybrids and mechanisation. The event is being held under the project titled 'Enhancing Maize Production in the Catchment Areas of Ethanol Industries.' The workshop was inaugurated under the chairmanship of HS Jat. On this occasion, SL Jat , Ph Romen Sharma (co-principal investigator), and other scientists of the institute were present. Principal investigator, SL Jat, while highlighting the objectives of the workshop, stated that the main aim of the project is to ensure a stable supply of raw material for ethanol production through enhanced maize productivity. He underscored the need for greater coordination and experience-sharing among regional scientists and field staff. Status of Maize in India In the year 2024–25, the total maize area in India stood at approximately 12 million hectares, with an average productivity reaching 3.5 tonnes per hectare. Maize is no longer limited to animal feed and industrial uses; it is now emerging as a promising alternative crop for ethanol production. From one tonne of maize, around 380 litres of ethanol can be produced, making it a viable option alongside sugarcane for ethanol production. This workshop will serve as an effective platform for knowledge and experience exchange, and will play a crucial role in strengthening the national mission of sustainable maize production with a focus on ethanol. ICAR–Indian Institute of Maize Research (ICAR-IIMR) is actively promoting Kharif maize cultivation in Punjab as a strategic intervention to address the challenges of crop diversification, regional maize deficit, and environmental conservation. As part of its ongoing efforts, the institute successfully demonstrated improved maize varieties and production technologies across more than 450 acres during the Kharif season 2025 in key maize potential districts including Jalandhar, Ludhiana, Gurdaspur, Amritsar, Pathankot, Moga, Hoshiarpur, Patiala, SBS Nagar, Sangrur, Fazlika, Bathinda and Sri Muktsar Sahib. To strengthen farmer engagement and adoption, capacity-building programmes and awareness campaigns are being conducted regularly. These initiatives aim to promote sustainable maize-based cropping systems that can help reduce dependence on water-intensive paddy, bridge the maize supply gap for feed and ethanol industries, and contribute to ecological balance through reduced groundwater extraction and improved soil health.
