This fish can walk and climb trees. Where did it come from?

Today, the Qinghai-Tibet Plateau is the roof of the world, but it was a vast ocean (part of the ancient Tethys Ocean) hundreds of millions of years ago. Paleontologist Zhou Zhonghe said that since the 1970s, the Chinese Academy of Sciences has organized two large-scale comprehensive scientific surveys and studies on the Qinghai-Tibet Plateau, but there are few documentary works on the paleontological part of the scientific expeditions.

This article is a study of the evolution of a very peculiar fish presented to us by Dr. Wu Feixiang through the discovery and research process of fossils: this kind of fish can walk and climb trees, and some of its species will even "drowned" to death if they are only allowed to breathe underwater. Where did such a novel fish come from?

This article is authorized to be excerpted from the scientific expedition documentary "Folding Mountains and Seas: The Epic of Life on the Qinghai-Tibet Plateau" (Science and Technology Press of China, January 2025).

Written by Wu Feixiang (Researcher at the Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences)

In November 1791, Daldorff, a naturalist from the Danish East India Company, discovered a strange fish on a tree (it is said to be a palm tree, but there is no clear record in the initial official description) while exploring Tranquebar (a British colony at the time) in southeastern India. He saw the fish swimming across the stream beside the pond and climbing up through the cracks of a tree by the water, even climbing to about 5 feet (1 foot is about 30.5 cm) above the ground. It opened its gill cover, stuck the spines on the rear edge of the gill cover into the wall of the crack to support itself, swung its tail back and forth, and supported itself with the spines on its anal fin to climb up. In order to avoid being stabbed by the fish's spines, Daldorff caught the fish with a cork. The following scene also opened the eyes of the naturalist. He was shocked by the tenacious vitality of this small fish: after being caught, it stayed in the shade of the tree for several hours and was able to "run" happily on the dry sand.

In 1797, Dahldorf reported this discovery and named the fish Perca scandens, which means "climbing perch". In 1816, French zoologist and naturalist Frédéric Cuvier and physician, anatomist and naturalist Hippolyte Cloquet established the genus Anabas (climbing perch), removing the species of Asian climbing perch from Perca and other groups and placing them in this genus. This classification scheme is still used today.

Dahldorf's discovery caused a great sensation in Europe at the time. In that era of exploration and discovery (when Dahldorf caught climbing perch in India, it was exactly 20 years after Captain Cook ended his circumnavigation of the world with the Endeavour), such anecdotes were quite eye-catching. To this day, people are still enthusiastic about this interesting fish.

One of the topics that everyone talks about is whether they can climb trees. In addition to Daldurf's account, there was once a very popular legend in India: climbing perch can climb up palm trees to suck alcoholic juice. However, many scholars think this seems unlikely. Based on the observation of the "walking" characteristics of climbing perch, opponents believe that they cannot have such a "supernatural" ability at all, and the legend of climbing perch climbing palm trees may just be a rumor. Even if there are climbing perch on palm trees, it is likely caused by some birds, who picked up the climbing perch from the water and placed it (or accidentally dropped it) in the moist depression at the base of the palm leaves. Just like a leopard drags food up a tree to hide it to avoid other opponents sharing it, and then comes back to enjoy it alone when it is hungry. In fact, the phenomenon of modern birds hiding food is not uncommon.

Regardless of whether it can climb trees or not, Dahldorf's description of the details of the climbing perch's behavior on land is indeed consistent with today's observations. The climbing perch's posture when "walking" is quite "arrogant": in addition to the tail swinging to provide drive, the spiny gill covers (primary and secondary gill covers) also play a very important role. In particular, the secondary gill cover below the primary gill cover is the key. This bone piece can rotate flexibly, and there are many spines on the lower edge. When the climbing perch moves forward, the left and right secondary gill covers are alternately inserted into the ground as a fulcrum, and with the twisting of the tail, the climbing perch moves forward like a "pole vault". Climbing perch can also cross obstacles with a height of half its body length through the cooperation of the primary and secondary gill covers. Therefore, when moving on land, they can easily climb over small piles of rocks and uneven river banks. This obstacle-crossing ability gives the climbing perch an added advantage and helps it spread better on land.

The key secret of climbing perch's ability to run rampant on land is its ability to breathe air. The structure that climbing perch use to breathe air is called the labyrinth organ. This organ is developed from a small bone (epibranchial bone) on the dorsal side of the first gill arch. It is a flower-like structure composed of many folds, which increases the area of ​​the respiratory epithelium. The surface of the labyrinth organ is covered with respiratory epithelium and is rich in capillaries. Unlike other normal gills, the blood passing through the labyrinth organ returns to the heart through the veins and is then pumped to other parts of the body by the heart. In this respect, the labyrinth organ of climbing perch is similar to the lungs of terrestrial animals.

Because of this "cheating" organ, climbing perch can live in water environments that other fish cannot tolerate. In the tropical regions of South Asia, Southeast Asia, Central Africa, and West Africa (temperature 18°C~30°C, distribution area is mostly below 500 meters above sea level, and rarely reaches 1200 meters), shallow stagnant water (dissolved oxygen can be as low as 1 mg/L or less, and most fish require dissolved oxygen above 4 mg/L for normal life activities) is a nightmare for most fish, but climbing perch is happy with it. The labyrinth of climbing perch is complex and large in size, occupying a large space in the gill cavity. As a result, the normal gills used for underwater breathing are greatly atrophied, so that the climbing perch cannot meet the oxygen required for survival, so the climbing perch must often stick its head out of the water to breathe in and out air. The climbing perch in Asia is even more exaggerated, because their labyrinth is too large. If they are not allowed to breathe air, they will faint if they rely on underwater breathing alone, and even die from lack of oxygen, or be "drowned" alive.

The Asian climbing perch has another characteristic - through the coordination of bones and muscles, the air they swallow into their mouths can pass through the respiratory cavity (where the labyrinth is located) and then be discharged from behind the gill cover, so the process of breathing air does not require the participation of water; while the African climbing perch needs to swallow water to squeeze out the "waste gas" in the respiratory cavity when breathing through the labyrinth. Therefore, among all the labyrinth fish, only the Asian climbing perch (Anabas) can climb out of the water and "walk" on land. There are records showing that a climbing perch "walked" 180 meters on land in one night. Don't underestimate this distance. On the tropical plains, they may be able to run to the adjacent puddle or nearby stream in one night.

The distribution area of ​​climbing perch is mostly in tropical monsoon areas. The precipitation brought by the summer monsoon expands the water area and lowers the water temperature. This signals the breeding season to the climbing perch, and they swim to shallow waters to mate and lay eggs. The breeding behavior of African climbing perch varies. Some species of male fish have a stinging feeler (contact organ) on the back of the eye socket or on the caudal peduncle. It is composed of scales with long spines on the rear edge. During mating, the male fish "rolls up" the female fish with its body and uses these structures to stimulate the female fish to produce more eggs. Today, Asian climbing perch do not have such behavior, and judging from the fossils in Tibet, this behavior of Asian climbing perch was lost along with the feeler during the evolution of climbing perch.

In 2011, the plateau expedition team of the Institute of Vertebrate Paleontology and Paleoanthropology rushed to the Lunpola Basin east of Selin Co in northern Tibet and found a fossil layer about 26 million years ago, where many fossils of fish, birds (feathers), plants and insects were preserved. Although there were a lot of climbing perch fossils at the beginning, their preservation was not very good. At that time, the fossil layers that could be dug were relatively shallow. Due to the influence of weathering, it was difficult to observe the skeletal details of the specimens clearly. However, this has already encouraged us. Since the first fish fossil was discovered during the first Qinghai-Tibet expedition more than 40 years ago, the fish fossils of the Cenozoic on the plateau have all been species of Cypriniformes (Cyprinidae or Cobitidae). The appearance of Perciformes fish is a breakthrough in the fossil category, and it also suggests that the fish fossils of the same era still have considerable potential, as evidenced by the catfish fossils discovered later.

Therefore, we searched back and forth within the 200-kilometer range around Selin Co for five or six years, and found many better fossils. In 2017, when the second comprehensive scientific survey and research on the Qinghai-Tibet Plateau was launched, we reported the first fossil of Perciformes fish on the Qinghai-Tibet Plateau, Eoanabas thibetana, pushing the history of the climbing perch group back more than 20 million years. Before that, the only confirmed climbing perch fossils were found in a few gill covers in the Pleistocene strata on Java Island. This fish, which often "walked" on land, might have been the food of ancient humans on the island at that time. The climbing perch fossils in Tibet combine the characteristics of Asian climbing perch (such as the morphology of the jaws and the degree of development of the labyrinth) and African climbing perch (such as the tentacles and the sensory tube openings on the top of the skull), providing excellent material for deducing the evolutionary history of climbing perch.

The holotype of the Tibetan climbing perch (above), sketch (middle) and skeleton reconstruction (below) (drawn by Wu Feixiang)

The gill cover, fin spines and bone plates around the eye sockets of the Tibetan climbing perch show typical characteristics of climbing perch, but do these morphological similarities mean that the ecological habits of ancient and modern climbing perches are similar?

Since the discovery of the first fossil of the climbing perch in 2011, we have been looking for fragments of the labyrinth. Most of the specimens we collected had bones that were not dispersed, so there was only a dark mass of debris where the labyrinth was originally located, and no meaningful details could be distinguished.

Things took a turn for the better in 2015. I was studying the lampreys of the Cretaceous Jehol Biota at the time, and the electron microscope scan of the sucker of the Mesozoic eel fossils was very good. I thought, why not try it on the fossil of the climbing perch? Among the several specimens scanned, only the small specimen collected in the Nima Basin in 2012 was very good. In the crushed structure, some round holes can be seen on several thin slices, which is completely consistent with the labyrinth gills of modern climbing perch in tropical Asia. This is very important for our research. Only by confirming the existence of this organ can we say that in addition to the morphological similarities, the ecological habits of the Tibetan climbing perch are completely comparable to those of today's climbing perch, and it is very likely to be more like the Asian climbing perch with the strongest breathing ability.

Therefore, combined with the plant fossils in the same layer as the climbing perch, such as palm, duckweed leaves, Koelreuteria paniculata, calamus, etc., it can be inferred that the habitat environment of such climbing perch and its living counterparts would not be much different. Therefore, we believe that the ancient altitude of the fossil site should be around 1,000 meters. Later, paleobotany colleagues conducted complex quantitative analysis of leaf shapes and determined that the ancient altitude of the fossil site did not exceed 2,300 meters. Although the conclusion on the ancient altitude still needs to be discussed, some of the plateau history reconstructed based on non-biological evidence (such as isotope geochemistry) is seriously inconsistent with paleontological evidence. Some geologists believe that the place where these fossil organisms lived was already a high and cold place similar to today.

Ecological restoration map of Tibet's climbing perch (drawn by Wu Feixiang)

Because today's climbing perch are distributed in tropical plains of Asia and Africa outside the Qinghai-Tibet Plateau, a long part of their story must have happened elsewhere. Unfortunately, there were very few fossil records of climbing perch family before, except for a few gill cover fossils in the Pleistocene strata on Java Island, and there were only some unsubstantiated speculations about their history. The Tibetan Proto-climbing Perch, as the most complete and primitive climbing perch known to date, has contributed to a new narrative.

The distribution area of ​​modern climbing perch spans Asia and Africa, separated by the arid Iranian Plateau, Arabian Peninsula and Sahara Desert. Animal geographers have been wondering, which continent did climbing perch originate from? When did the Asian and African climbing perch separate? Some people believe that they originated from Gondwana, and after the continental split, climbing perch took the "ride" of the Indian plate and drifted all the way to Asia; others believe that about 20 million years ago, when the African-Arabian Peninsula met the Eurasian continent, they spread from one end to the other, and then the drought in North Africa and West Asia caused the Asian and African climbing perch to separate. But these hypotheses need reliable fossil evidence to verify. Using the Tibetan proto-climbing perch to calibrate the "molecular clock" of the climbing perch family, we know that Asian and African climbing perch parted ways about 40 million years ago. Therefore, neither the ancient "Gondwana origin" nor the recent "Middle East land bridge diffusion" hypothesis can explain the history of climbing perch. The reconstruction of the ancestral distribution area shows that after they originated in Southeast Asia, they spread all the way to Tibet, then through India, and finally arrived at the African continent, where they developed "prosperously". Today, there are many more species of climbing perch in Africa than in Asia, and they are also much more beautiful. Among the African climbing perch, species such as "Plum Blossom" (Ctenopoma acutirostre) and "West African Paradise Island" (Microctenopoma ansorgii) are all stars in aquariums.

The distribution area of ​​modern climbing perch (light blue area) and the location of the discovery of climbing perch fossils on the Qinghai-Tibet Plateau

Scientific research is always full of vitality and has no end. Geological colleagues recently published a set of dating data, which shows that the fossils at the Dayu fossil site, including the geological age of the proto-climbing perch, are more than 10 million years earlier than we thought, or that the climbing perch arrived in Tibet earlier than we thought before. This discussion is still going on, and scientists are approaching a consensus. If we use this new age as a constraint point to re-estimate the origin of the climbing perch (the previous result was 42 million years ago, 95% confidence interval: 32 million to 55 million years ago) and the time of its spread from Asia (Indian subcontinent) to the African continent (the previous result was about 39 million years ago, 95% confidence interval: 30 million to 50 million years ago), it may be earlier than the results we have already estimated.

From another perspective, the ever-detailed history of climbing perch and the possible older fossil age mean that there are actually more "missing links" we need to find than we thought. This is really exciting. Isn't it the fun of finding fossils to track down something that must exist but is unknown, just like a detective?

The title of this article is chosen by the editor. The original text comes from "Fish on the Tree" in the fourth chapter "The Ancient Forest beside Selin Co" of "Folding Mountains and Seas: The Epic of Life on the Qinghai-Tibet Plateau".

Special Tips

1. Go to the "Featured Column" at the bottom of the menu of the "Fanpu" WeChat public account to read a series of popular science articles on different topics.

2. Fanpu provides a function to search articles by month. Follow the official account and reply with the four-digit year + month, such as "1903", to get the article index for March 2019, and so on.