Category: 7. Science

Newswise — North America has lost an estimated 3 billion birds since 1970—a nearly 30% drop across species—mostly due to habitat loss and degradation. So when a team of researchers repeated a bird population study they did 30 years earlier in a very large commercial forest landscape in Maine, they were stunned to find more birds than before.

“When we started this project, we expected to add to the pile of bad news,” says Michael Reed, a professor of biology in the School of Arts and Sciences at Tufts University and co-author of the study. “So we were very pleasantly surprised to find that, for most of the bird species in our study, things were actually looking up.”

The research team wanted to see if bird populations and habitat use had changed over the decades, particularly given a shifting forest landscape. “Forest management practices in Maine have changed significantly since the early 1990s,” says Reed. Due to social pressure, clearcutting has become much less common in Maine. Today, most logging operations remove fewer trees per acre—returning to spots every decade or so—and spread their activity across a broader area

The study, published in Biological Conservation, found that 26—or more than half—of 47 species counted by the researchers had significantly increased in numbers since the early 1990s, while populations for 13 species (or 28%) had remained stable. That’s contrary to what happened across much of the continent, with the North American Breeding Bird Survey showing that 35—or 75%—of the same species analyzed had seen their numbers decline, both regionally and continentally, for the same timeframe.

But what makes the commercial forests of Maine so different from other forests in the northern Atlantic states and North America? And can we learn anything from them to bolster bird populations regionally and nationally?

“Numerous factors are likely behind the abundance of bird populations we see in northcentral Maine,” says Reed. “We can’t know what they all are, but we know at least one: It’s all forest up there.”

The original and current study took place in a 588,000-acre commercial forest nestled within 10 million acres of commercial, public, and protected forest landscape. Together, these woodlands create the largest contiguous tract of non-developed forest east of the Mississippi. The habitat is recognized worldwide as an Important Bird Area, an area officially recognized as critical for protecting bird species and biodiversity by a coalition of international bird conservation groups.

The remote forestland is also one of the darkest places left on the East Coast. “Most bird species migrate at night, orienting by the stars,” says John Hagan, the study’s senior author and the founder and president of Our Climate Common. “So it may be when birds flying at night get tired, they look down, spot a vast patch of darkness, and decide it’s a good place to land and raise young.”

Many of the bird species observed seemed more flexible in their habitat use than previously thought; the researchers have another paper on these findings in the works. This suggests that high-quality forest next to more average forest may still be appealing enough to attract and support more birds over time. Individual birds often return to breed where they were hatched, and many migratory species are drawn to areas where others of their kind are already present. As Reed puts it, this may mean that “the rich get richer” when it comes to birds in Maine’s commercial forests.

Much of the population growth came from more birds per acre, not more habitat. “You’d expect bird numbers to go up if there’s more habitat,” says Reed. “But we actually saw increased numbers for some species whose main habitat acreage stayed the same size or even decreased from the 1990s. For example, in places where we previously counted two ovenbirds singing before, we now counted four.”

Data from one species—the golden-crowned kinglet—suggests that how forests are managed may affect species’ ability to thrive. These tiny, round songbirds are declining sharply across much of North America, including in the commercial forests of New Brunswick, Canada. But just over the border in Maine, their numbers are rising.

In Canada, commercial forests are commonly replanted in neat plantation-style rows, creating simpler forests, with less understory and trees that are all the same age. In contrast, Maine’s commercial forests rely on natural regrowth, creating denser forests with a broader mix of tree species and ages. Reed and Hagan hypothesize this more natural approach may offer better shelter or support more of the insects that kinglets need to raise their young.

Despite the widespread increases, bird numbers for 14 species—about 30%—still declined in the study area. The researchers hope to more closely examine the pressures faced by these decreasing bird species, including species like the winter wren and the Canada warbler, to see if commercial forestry could do something differently to better support them. They are particularly worried about steep declines in mature trees—some more than 200 years old—and its impact on the bird populations. Hagan is now leading additional research to assess this conservation threat. The threat could be on their migration or overwintering grounds, in which case little can be done in Maine to improve their numbers.

Even though about two-thirds of U.S. forestland is available or used to produce industrial wood products, the research team believes theirs is the first bird survey to compare species population numbers in a commercial forest over a long period of time. And given the 521 million acres of commercial U.S. forestland, they hope it certainly will not be the last.

In the face of ongoing human habitat expansion and continental declines in bird populations, the team says it’s important to understand how all types of forest ownership may help create important sanctuaries for birds. “Birds also may be thriving in commercial forests in Michigan, Wisconsin, and Minnesota, which are managed similarly to those in Maine,” Hagan says. “Hopefully, someone will look to see.”

“Many people don’t expect places where you harvest wood to serve as valuable habitat because they are cutting trees down,” adds Reed. “But nobody thinks the goal of a farm is to cut down corn—it’s to grow it. And commercial forests grow trees.”

_______

How to Help the Birds at Home

You may not own an acre of land—never mind hundreds of acres—but you can still bring some qualities of Maine forestland to your own yard.

• Plant a tree that is native to your area.
• Add native shrubs, which provide vital food and shelter for birds while offering multi-season visual appeal for you.
• “Leave the leaves” in the parts of your yard where you are creating understory. Nature abhors a leaf blower—and it means less yardwork and free mulch for you.
• Fight light pollution by reducing outdoor lighting to what’s truly necessary.
• Where safe to do so, leave dead trees standing. “Snags” provide valuable habitat for owls, woodpeckers, and cavity-nesting birds.
• Replace part of your lawn with native plants. Check out Douglas Tallamy’s book Nature’s Best Hope for ideas.

 

About 56 million years ago, Earth slipped into the Paleocene-Eocene Thermal Maximum (PETM), a sudden global warming pulse that pushed temperatures up by at least 5°C (9°F). In the badlands of what is now Wyoming, one mid-sized predator, Dissacus praenuntius, found itself confronting a landscape in flux.

A new study led by Rutgers University suggests the animal’s solution was unexpected: when prey became scarce, it started crunching bones.

“What happened during the PETM very much mirrors what’s happening today and what will happen in the future,” said Andrew Schwartz, the Rutgers anthropology doctoral student who led the research.

The work combines field excavations in the Bighorn Basin with a forensic technique called dental microwear texture analysis (DMTA).

By reading microscopic pits and scratches on fossil teeth, the team reconstructed what the animal was chewing in the weeks before it died; tougher, deeper marks usually point to hard items such as bone.

Dissacus praenuntius ate bones

Dissacus praenuntius was roughly the size of a modern coyote and, at first glance, looked like a lanky wolf. But it walked on tiny hooves and belonged to an archaic lineage known as mesonychids.

“They looked superficially like wolves with oversized heads,” Schwartz said, describing them as super weird mammals “Their teeth were kind of like hyenas. But they had little tiny hooves on each of their toes.”

Before the planet heated up, DMTA shows the predator had a diet which resembled that of today’s cheetahs: mostly soft but sinewy meat. Once the PETM began, everything changed.

“We found that their dental microwear looked more like that of lions and hyenas,” Schwartz said. “That suggests they were eating more brittle food, which were probably bones, because their usual prey was smaller or less available.”

The microwear evidence dovetails with other signs of environmental stress. Earlier work has shown many mammal species shrank during the PETM, and Dissacus appears to have followed that trend – smaller bodies need less food.

While scientists once blamed this shrinkage solely on heat, the new findings strengthen the case that limited nutrition was a major driver.

Adaptability drove survival success

“One of the best ways to know what’s going to happen in the future is to look back at the past,” Schwartz said. “How did animals change? How did ecosystems respond?”

For Schwartz, the PETM offers a natural experiment on how warming reshapes food webs. The lesson, he argues, is that dietary flexibility can make or break a species.

“In the short term, it’s great to be the best at what you do,” he said. “But in the long term, it’s risky. Generalists, meaning animals that are good at a lot of things, are more likely to survive when the environment changes.”

That principle is visible today. “We already see this happening,” Schwartz said. “In my earlier research, jackals in Africa started eating more bones and insects over time, probably because of habitat loss and climate stress.”

Pandas, koalas, and other extreme specialists lack that wiggle room and could therefore be more vulnerable as habitats fragment or shift.

Wyoming’s ancient badlands

To trace Dissacus praenuntius through time, the team zeroed in on the Bighorn Basin, one of the rare places where sedimentary layers preserve an almost year-by-year record across the PETM.

Field crews collected dozens of jaws and isolated teeth from successive strata, then scanned their chewing surfaces under high-resolution confocal microscopes.

Sophisticated software converted the textures into numerical values capturing roughness, complexity, and orientation – clues to the animal’s final meals.

Professor Robert Scott, a co-author on the paper, notes that DMTA is revolutionizing paleontology because it captures dietary snapshots just weeks or months before death, unlike isotope analyses that average years.

The method revealed a clear pattern: before the thermal maximum, tooth surfaces carried long, shallow scratches characteristic of slicing flesh; during and after, they showed deeper pits and gouges typical of bone fragmentation.

Dissacus praenuntius faded away

Despite this adaptive flair for bones, Dissacus praenuntius did not make it through the next 15 million years. Bigger, better-equipped carnivores eventually displaced it, and the shifting vegetation of post-PETM North America favored more agile hunters.

Still, its bone-crunching episode underscores how rapidly predators can recalibrate their behavior when climate jolts prey supply.

Schwartz hopes insights like these will aid modern conservation biology by spotlighting which species could be climate winners or losers.

The study also signals the importance of preserving continuous fossil sites: without the Bighorn Basin’s layered rocks, the dietary flip would have remained invisible.

Local fossils, global lessons

Schwartz traces his fascination with fossils to childhood trips along New Jersey’s waterways with his father, an amateur collector. Now close to completing his Ph.D., he is eager to show the broader public how deep time research can illuminate future challenges.

 “If I see a kid in a museum looking at a dinosaur, I say, ‘Hey, I’m a paleontologist. You can do this, too.’”

That encouragement matters because, as Earth barrels toward PETM-like CO₂ levels, society will need scientists who can decode past crises to steer us through the next one.

The tale of a jackal-sized mammal gnawing harder fare in a hothouse world is more than a curiosity; it is a mirror held up to our own warming century.

The study is published in the journal Palaeogeography, Palaeoclimatology, Palaeoecology

Image Credit: ДиБгд, CC BY 4.0 , via Wikimedia Commons

—–

Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates. 

Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.

—–

Companion cockatoos are renowned for their problem-solving and intriguing characters. It’s no surprise these large, long-lived and intelligent parrots are known to display complex behaviour.

Owners often film their birds dancing to music and post the videos to social media. Snowball, a famous dancing cockatoo, has been shown to have 14 different dance moves.

We wanted to find out more about the dance repertoire of cockatoos and why they might be doing this. In our new research, we examined videos of dance behaviour and played dance music to six cockatoos at an Australian zoo.

These birds weren’t just doing a side step or bobbing up and down. Between them, they had a rich repertoire of at least 30 distinct moves. Some birds coordinated their head bobbing with foot movements, while others undertook body rolls. Our research shows at least 10 of the 21 cockatoo species dance.

If we saw this behaviour in humans, we would draw a clear link between music and dancing and interpret the behaviour as enjoyable. After watching cockatoos voluntarily begin dancing for reasonable lengths of time, it was difficult to reach any conclusion other than cockatoos most likely dance because it’s fun.

How many moves does a cockatoo have?

Dancing is complicated. To dance to music, animals need to be able to learn from others, imitate movements and synchronise their movements. These complex cognitive processes are only known to exist in humans – but evidence is emerging for its presence in chimpanzees and parrots such as cockatoos.

To catalogue the dance moves of cockatoos, we began by studying videos of the behaviour. We analysed 45 dancing videos and recorded all distinct moves.

The five species in these videos were the familiar sulfur-crested cockatoos and little corellas, as well as Indonesian species such as Goffin’s cockatoos, white cockatoos and Moluccan cockatoos.

Across the videos, we spotted 30 movements, including 17 that hadn’t been described scientifically. We also observed 17 other movements, which we classified as “rare” because they were only seen in a single bird.

Head movements were the most common dance move, especially the downward bobbing motion. Half of all videoed cockatoos performed this move.

Dancing – but not to music

Once we catalogued the moves, we then tested whether music could elicit this behaviour in captive cockatoos who weren’t kept as companions.

We undertook a playback experiment with six adult cockatoos at Wagga Wagga Zoo in New South Wales, comprising two sulfur-crested cockatoos, two pink cockatoos and two galahs.

Over three sessions, we played a piece of electronic dance music on repeat for 20 minutes and recorded any responses on video. We repeated our experiment with no music and again with a podcast featuring people talking.

All six cockatoos we studied showed some dancing behaviour at least once over the three sessions. But the rates of dancing weren’t any higher during the playing of music – it was similar to dancing during silence and the podcast.

We don’t fully know why this is. One possibility could be because we played music to existing male-female pairs, and the social environment alone was sufficient to trigger dance behaviour.

Why do cockatoos dance at all?

To find out whether the cockatoo species most prone to dancing were those most closely related, we analysed similarities across species. Goffin’s cockatoos and white cockatoos had the most similar moves, while Goffin’s cockatoo and little corella were the furthest apart.

But this clashed with genetics, as Goffin’s cockatoos are most closely related to little corellas. This suggests dancing behaviour may not be connected to genetic links.

Interestingly, these behaviours are mainly recorded in companion birds. Music playback in the online videos does seem to encourage the bird to keep it going for longer than likely to be seen in zoo or wild birds. These dance moves might represent an adaptation of courtship display movements as a way to connect with their human owners.

Other researchers report being able to trigger dancing behaviour in an African grey parrot and a sulfur-crested cockatoo with music. But the zoo cockatoos in our playback study didn’t respond the same way. This suggests there may be an element of learning to respond to humans.

It’s usually easy to tell if a human behaviour is play or not. But in animals, it can be much more difficult. Researchers define a behaviour as play if it meets four criteria: it occurs while animals are relaxed, it’s begun voluntarily, has no obvious function and appears rewarding. Cockatoo dancing would meet all four of these criteria.

By contrast, repetitive behaviours such as pacing seen in animals kept alone in small cages would not be play – it’s not rewarding and the animals don’t seem relaxed. Parrots kept in poor conditions exhibit self-harming behaviours such as constant screeching and feather pulling.

Captive parrots have complex needs and can experience welfare problems in captivity. Playing music may help enrich their lives.

For cockatoo owners, this suggests that if their birds are dancing, they’re feeling good. And if they’re busting out many different moves in response to music, even better – they might be showing creativity and a willingness to interact.

---

Acknowledgement: Honours student Natasha Lubke is the lead author of the research on which this article is based.

The thyroid, a vital endocrine organ in vertebrates, plays a key role in regulating metabolism and supporting growth. The first gland of both the nervous system and endocrine system to mature during an embryo’s development, it initially evolved more than 500 million years ago out of a “primitive” precursor organ in chordates known as the endostyle. Now, using lamprey as a model organism, Caltech researchers have discovered how the evolutionary acquisition of a certain kind of stem cell, called a neural crest cell, facilitated the evolution of the endostyle into the thyroid.

The research is described in a paper appearing in the journal Science Advances on August 6. The work was conducted primarily in the laboratory of Marianne Bronner , the Edward B. Lewis Professor of Biology and director of Caltech’s Beckman Institute.

Bronner’s lab has long focused on neural crest cells and their role in vertebrate development and evolution. For example, the team previously examined the role of neural crest cells in forming the bony scales that protect sturgeon and other primitive fish, heart tissue in zebrafish and chickens , and neurons of the peripheral nervous system in lamprey.

“Neural crest cells seem to promote evolution,” Bronner says. “When Darwin first proposed the theory of evolution, he was looking at the different shapes of beaks of finches on the Galapagos Islands. Beaks, in addition to other parts of the facial skeleton, happen to arise from neural crest cells. These cells seem to be able to change faster in evolutionary time than cells that are more ancient.”

Vertebrates have neural crest cells, while invertebrates do not, further suggesting that these cells contribute to the evolution of complex body forms. The Bronner lab uses lamprey, slimy parasitic eel-like fish, as a model organism, because modern lamprey share some characteristics with the earliest vertebrates.

The new work, led by Senior Postdoctoral Scholar Research Associate Jan Stundl, examines how neural crest cells contribute to the development of the endostyle in the lamprey. The endostyle is an evolutionary novelty of chordates (animals in the phylum Chordata, which includes vertebrates), and lampreys are the only vertebrates that retain this organ, whose primary function is associated with filter feeding. In lampreys, the larval endostyle, composed of two lobes in a butterfly-like shape, transforms into thyroid follicles during metamorphosis. Stundl and the team traced how neural crest cells give rise to the five different cell types of the endostyle, two of which give rise to the thyroid follicles. Using the gene-editing technology CRISPR, they then genetically deleted genes associated with the neural crest developmental program in lamprey embryos. These modified lampreys failed to develop a fully formed endostyle, exhibiting instead only a primitive lobe resembling the simplified endostyle of invertebrate chordates. The findings suggested that neural crest cells are essential for driving the evolutionary transition from the chordate endostyle to the vertebrate thyroid gland.

“Mother Nature is ‘smart,’” Stundl says. “Instead of evolving something new, you can rebuild from something already present, like the endostyle. Neural crest cells seem to play an important role in enabling this transition to happen. Without the neural crest, we might still be filter feeders.”

The paper is titled “Acquisition of neural crest promoted thyroid evolution from chordate endostyle.” In addition to Stundl and Bronner, Caltech co-authors are postdoctoral scholars Ayyappa Raja Desingu Rajan and Tatiana Solovieva; graduate student Hugo Urrutia; research associate Jana Stundlova; and former postdoctoral scholar Megan Martik now of UC Berkeley. Additional co-authors are Jake Leyhr, Tatjana Haitina, and Sophie Sanchez of Uppsala University; and Zuzana Musilova of Charles University in Prague, Czech Republic. Funding was provided by the National Institutes of Health, the European Union, Alex’s Lemonade Stand Foundation, the American Heart Association, the Helen Hay Whitney Foundation, Swedish Research Council Vetenskapsrådet, and the European Synchrotron Radiation Facility. Marianne Bronner is an affiliated faculty member with the Tianqiao and Chrissy Chen Institute for Neuroscience at Caltech .

A picturesque Scottish peninsula immortalized in a hit Paul McCartney song from the 1970s will host a new U.K. rocket development hub as the country works toward its goal of becoming a major player in European space launch.

The Mull of Kintyre peninsula in southwestern Scotland once offered refuge to the famous ex-Beatle, who lived there on a farm in the aftermath of the legendary band’s acrimonious split. The peninsula’s misty coastline and rolling hills inspired the namesake Wings tune that became the U.K.’s best-selling hit of the 1970s. Now, that wild landscape will become the backdrop for a different kind of history-making.

Winners and losers

About five years ago NASA awarded initial space station development contracts to four different companies: Northrop Grumman, Blue Origin, Axiom Space, and Voyager Space. Since then Northrop has dropped its effort and joined Voyager’s team. There has also been some interest from other companies, most notably Vast, which is working with SpaceX to develop its initial space station.

Probably the most striking thing about the new directive is that it seems to favor Vast over NASA’s original contractors. Specifically, Vast’s Haven-1 module is designed for four astronauts to spend two weeks in orbit, and the company has a more straightforward pathway to building a station that would meet NASA’s minimum requirements.

The other companies had been planning larger stations that would have more permanence in orbit, which matched NASA’s original desires for a successor to the International Space Station. The new directive favors a company building up capabilities through interim steps, including stations with a limited lifespan on orbit.

“All the current players are going to have to do some kind of pivot, at least revisit their current configuration,” McAlister said. “Certain players are going to have to do a harder pivot.”

One industry official, speaking anonymously, put it more bluntly: “Only Haven-1 can succeed in this environment. That is our read.”

The chief executive officer of Vast, Max Haot, told Ars that the company bet that starting with a minimum viable product was the best business strategy and fully funded that approach without government money.

“It seems like NASA is now leaning into an approach for the future of CLDs that is led by what industry believes it can achieve technically and build a credible business model around,” Haot told Ars. “Seeing that information from contractors before committing to buying services can help increase long-term risk reduction. This seems similar to the successful approaches used by NASA for cargo and crew.”

Vast has worked closely with SpaceX in the development of its station, to the extent that Haven-1 will largely rely on the Dragon spacecraft for life support and propulsion. Future stations, such as Haven-2, will have more independent capabilities.

“To consult the statistician after an experiment is finished is often merely to ask him to conduct a post mortem examination. He can perhaps say what the experiment died of.” – Ronald A. Fisher

The modern biology toolbox is larger than ever, as a widening array of cutting-edge molecular techniques supplements the classic approaches that still drive the field forward. Teams of researchers with complementary expertise may combine these tools in countless creative ways to produce novel research. Regardless of which methods are chosen for data collection, all empirical research projects share a common foundation: the principles of good experimental design. Far from eliminating the need for statistical literacy, -omics technologies make careful, sound experimental design more important than ever^1,2.

This Perspective was motivated by the observation that many biology projects are doomed to fail by experimental design errors that make rigorous inference impossible. The results of such errors can range from waste to, in the worst case, the introduction of misleading conclusions into the scientific literature, including those with clinical consequences. Our goal is to highlight common experimental design errors and how to avoid them. Although methods for analyzing challenging datasets are improving^3,4,5,6, even advanced statistical techniques cannot rescue a poorly designed experiment. For this reason, we focus on choices that must be made before an experiment or study is conducted, rather than on data analysis choices (however, if you are working on microbiomes, we highly recommend that you read this article in conjunction with Willis and Clausen⁷, who highlight important points for planning and reporting on statistical analyses). In particular, we address four key elements of a well-designed experiment: adequate replication, inclusion of appropriate controls, noise reduction, and randomization.

First, we discuss how in -omics research, many errors arise because of the misconception that a large quantity of data (e.g., deep sequencing or the measurement of thousands of genes, molecules, or microbes) ensures precision and statistical validity. In reality, it is the number of biological replicates that matters. We also explain how to recognize and avoid the problem of pseudoreplication, and we introduce power analysis as a useful method for optimizing sample size. Second, we introduce several strategies (blocking, pooling, and covariates) for minimizing noise, or variation in the data caused by randomness or other unplanned factors. Third, we briefly review how missing positive and negative controls can compromise experimental results, and provide examples from both -omics and non-omics research. Finally, we describe two critical functions of experimental randomization: preventing the influence of confounding factors, and empowering researchers to rigorously test for interactions between two variables.

While these practices are covered in undergraduate and graduate level classes and in textbooks, as reviewers we—along with editors and some colleagues—have observed basic experimental design errors in submitted manuscripts. Indeed, the authors of this Perspective have also made and learned from some of these errors the hard way. This fact inspired this Perspective to provide a succinct overview of important experimental design principles that can be used in training of early-career scientists and as a refresher for seasoned biologists. We present examples from projects that use high-throughput DNA sequencing; however, unlike best methods for statistical analysis, these experimental design principles apply equally to any experiment regardless of the type of data being collected, including proteomics, metabolomics, and non-omics data. We also provide a list of additional resources on the topics discussed (Table 1) and practical steps for designing a rigorous experiment (Box 1).

Empowerment through replication

Many researchers intuitively understand that any individual data point might be an outlier or a fluke. As a result, we have very low confidence in any conclusion that was reached based on one isolated observation. With each additional data point that we collect, however, we get a better sense of how representative that first data point really was, and our confidence in our conclusion grows. Statistics empower us to measure and communicate that confidence in a way that other researchers will understand.

Why biological replication is more important than sequencing depth

Most biologists have heard that having more data will empower them to test their hypotheses. But what does it really mean to have more data? High-throughput technologies that generate millions to billions of DNA sequence reads, and counts of thousands of different genes or microbes, can create the illusion of a big dataset even if the number of replicates, or sample size, remains small. Although deeper sequencing per replicate can improve power in some cases, it is primarily the number of biological replicates that enables researchers to obtain clear answers to their questions.

To illustrate why this is, consider the hypothesis that two species of plants host different ratios of two microbial taxa in their roots. We can estimate this ratio for the two groups or populations of interest (i.e., all existing individuals of the two species) by collecting random samples from those populations. A sample size of 1 plant per species would be essentially useless, because we would have no way of knowing whether that plant is representative of the rest of its population, or instead is an anomaly. This is true regardless of the amount of data we have for that plant; whether it is based on 10³ sequence reads or 10⁷ sequence reads, it is still an observation of a single individual and cannot be extrapolated to make inferences about the population as a whole. Similarly, if we measure the abundances of thousands of microbes per plant, this same problem would apply to our estimates for each of those microbes. In contrast, measuring more plants per species will provide an increasingly better sense of how variable the trait of interest is in each population.

To what extent does the amount of data per replicate matter? Deeper sequencing can modestly increase power to detect differential abundance or expression, but those gains quickly plateau after a moderate sequencing depth is achieved^3,8. Extra sequencing is most beneficial for the detection of less-abundant features, such as rare microbes or low-expression transcripts, and features with high variance⁸. Projects that focus specifically on such features will require deeper sequencing than those that do not, or else may benefit from a more targeted approach. Finally, it is worth highlighting the related problem of treating -omics features (e.g., genes or microbial taxa) as the units of replication, as is common in gene set enrichment and pathway analyses. Such analyses describe a pattern within the existing dataset, but they are entirely uninformative about whether that pattern would hold in another group of replicates⁹. Instead, they only allow inference about whether that pattern would hold for a newly-observed feature in the already-measured group of replicates, which is often not the researcher’s intended purpose.

Replication at the right level

Biological replicates are crucial to statistical inference precisely because they are randomly and independently selected to be representatives of their larger population. The failure to maintain independence among replicates is a common experimental error known as pseudoreplication¹⁰ (Box 2). When experimental units are truly independent, no two of them are expected to be more similar to each other than any other two. Pseudoreplication becomes a problem when the incorrect unit of replication is used for a given statistical inference, which artificially inflates the sample size and leads to false positives and invalid conclusions (Fig. 1). In other words, not all data points are necessarily true replicates.

Although pseudoreplication is occasionally unpreventable (particularly in large-scale field studies), it can and should be anticipated and avoided whenever possible. The correct units of replication are those that can be randomly assigned to receive one of the treatment conditions that the experiment aims to compare. In experimental evolution, for instance, the replicates are random subsets of the starting population, assumed to be identical, each of which may be assigned to a different selective environment^11,12. Failure to include enough independent sub-populations, or to keep them independent throughout the experiment (e.g., by pooling replicates; Fig. 1C–D), will cause pseudoreplication of the evolutionary process of interest¹³. In some cases, however, mixed-effects modeling techniques can adequately account for the non-independence of replicates¹⁰.

Optimizing sample size

If most measurements in a dataset are similar to each other, this indicates that we are measuring individuals from a low-variance population with respect to the dependent variable. In contrast, a wide range of trait values signals a high-variance population. This within-group variance (i.e., the variance within one population) is central to determining how many biological replicates are necessary to achieve a clear answer to a hypothesis test: when within-group variance is high relative to the between-group variance, more replicates are required to achieve a given level of confidence (Fig. 2A). However, when the budget for sequencing is fixed, then increasing the sample size is costly—not only because wet lab costs increase, but also because the amount of data per observation decreases. Too many replicates can waste time and money, while too few can waste an entire experiment. How can a biologist know ahead of time how many replicates are enough?

A flexible but underused solution to this problem—power analysis—has existed for nearly a century^14,15. Power analysis is a method to calculate how many biological replicates are needed to detect a certain effect with a certain probability, if the effect exists (Fig. 2A). It has five components: (1) sample size, (2) the expected effect size, (3) the within-group variance, (4) false discovery rate, and (5) statistical power, or the probability that a false null hypothesis will be successfully rejected. By defining four of these, a researcher can calculate the fifth.

Usually, both the effect size and within-group variance are unknown because the data have not yet been collected. The choice of effect size for a power analysis is not always obvious: researchers must decide a priori what magnitude of effect should be considered biologically important. Acceptable solutions to this problem include expectations based on the results of small pilot experiments, using values from comparable published studies or meta-analyses, and reasoning from first principles. For example, a biologist planning to test for differential gene transcription may define the minimum interesting effect size as a 2-fold change in transcript abundance, based on a published study showing that transcripts stochastically fluctuate up to 1.5-fold in a similar system. In this scenario, the stochastic 1.5-fold fluctuations in transcript abundance suggest a reasonable within-group variance for the power analysis. As another example, a bioengineer may only be interested in mutations that increase cellulolytic enzyme activity by at least 0.3 IU/mL relative to the wild type, because that is known to be the minimum necessary for a newly designed bioreactor. To determine the within-group variance for a power analysis, the bioengineer could conduct a small pilot study to measure enzyme activity in wild-type colonies. As a final example, if a preliminary PerMANOVA for a very small amount of metabolomics data (say, N = 2 per group) from a pilot study estimated that R² = 0.41, then a researcher could use that value as the target effect size for which he/she aims to obtain statistical support.

Freely available software facilitates power analysis for basic statistical tests of normally-distributed variables, including correlation, regression, t-tests, chi-squared tests, and ANOVA^16,17,18,19. Power analysis for -omics studies, however, is more complex for several reasons. -Omics data comprise counts that are not normally distributed and may contain many zeroes, and some features may be inherently correlated with each other; these properties must be modeled appropriately for a power analysis to be useful²⁰. The large number of features often requires adjustment of P-values to correct for inflated Type I error rates, which in turn decreases power. Furthermore, statistical power varies among features because they differ not only in within-group variance, but also in their overall abundance. In general, more replicates are required to detect changes in low-abundance features than in high-abundance ones⁸. Statistical power also varies among different alpha- and beta-diversity metrics²¹ that are commonly used in amplicon-sequencing analysis of microbiomes. For power analysis of multivariate tests (e.g., PERMANOVA), which incorporate information from all features simultaneously to make broad comparisons between groups, simulations are necessary to account for patterns of co-variation among features. In such cases, pilot data or other comparable data are crucial for accurate power analysis.

Fortunately, tools are available to estimate power for common analyses used in proteomics²², RNA-seq^3,20,23,24, and microbiome studies^21,25,26. Recent reviews^27,28 demonstrate this process for various forms of amplicon-sequencing data, including taxon abundances, alpha- and beta-diversity metrics, and taxon presence/absence. They also consider several types of hypothesis tests, including comparisons between groups and correlations with continuous predictors. Although power analysis may seem daunting at first, it is an investment with excellent returns. This skill empowers biologists to budget more accurately, write more convincing grant proposals, and minimize their risk of wasting effort on experiments that cannot generate conclusive results.

Empowerment through noise reduction

As explained above, statistical power is positively related to the sample size and negatively related to the within-group variance. Thus, we can increase power either by including more biological replicates or by reducing within-group variance. Because our budgets do not allow us to increase replication indefinitely, it is useful to consider: What methods exist to increase power by decreasing within-group variance?

A classic way to minimize within-group variance is to remove as many variables as possible²⁹. Common examples include using a single strain instead of multiple; strictly controlling the lab environment; and using only male host animals³⁰. Such practices minimize noise, or variance contributed by unplanned, unmeasured variables. However, they come with a drawback: the loss of generalizability^30,31,32. If a mutation’s phenotype cannot be detected in the face of minor environmental variation, is it likely to be relevant in nature? If an interesting function occurs only in a lab strain, is it important for the species in general? Such limitations should always be considered when interpreting results.

Another technique to reduce within-group variance is blocking, the physical and/or temporal arrangement of replicates into sets based on a known or suspected source of noise (Fig. 2B). For instance, clinical trials may block by sex, so that results will be based on comparisons of males to males and females to females. Crucially, all experimental treatments must be randomly assigned within each block. Blocking is also useful for large experiments with more replicates than can be measured at once; as long as the replicates within each block are treated identically, sources of noise such as multiple observers can be controlled. The most powerful form of blocking is paired design, in which experimental units are kept in pairs from the beginning. One unit per pair is randomly assigned to the treatment group, the other to the control; the difference between units is then calculated directly, automatically accounting for sources of noise that are shared within the pair (Fig. 2B, bottom panel). A possible downside of highly-structured blocking designs is that they can complicate the re-use of the data to answer questions other than the one for which the design was optimized, whereas a simple randomized design is highly flexible but not optimized for any particular purpose.

When sources of noise are known beforehand, they ideally should be measured during the experiment so that they can be used as covariates. A covariate is any independent variable that is not the focus of an analysis but might influence the dependent variable. When using regression, ANOVA, or similar methods, one or more covariates may be included in the statistical model to control for their effects on the variable of interest (Fig. 2C). For instance, controlling for replicates’ spatial locations can dramatically reduce noise in field studies^33,34. Similarly, dozens of covariates related to lifestyle and medical history contribute to the variation in human fecal microbiome composition³⁵. The authors showed that by controlling for just three of these covariates, the minimum sample size needed to detect a microbiome contrast between lean and obese individuals would decrease from 865 to 535 individuals per group.

Finally, pooling–combining two or more replicates from the same group prior to measurement–can reduce within-group variance and the influence of outliers³⁶. For example, pooling RNA can empower biologists to detect differential gene expression from fewer replicates, reducing costs of library preparation and sequencing³⁷. This approach is especially helpful for detecting features that are low-abundance and/or have large within-group variance. Its main drawbacks are that it reduces sample size and results in the loss of information about specific individuals, which may be necessary for unambiguously connecting one dataset to another or linking the response variable to covariates. In fact, excessive or unnecessary pooling is a common experimental design error that can eliminate replication (Fig. 1C–D; Box 2), but is easily avoided by remembering that the pools themselves are the biological replicates.

Empowerment through inclusion of appropriate controls

Key to any experimental design is the inclusion of appropriate positive and negative controls to strengthen conclusions and enable correct interpretation of the experimental results (Box 2). Positive controls can confirm that experimental treatments and measurement approaches work as expected. Positive controls can, for example, be samples with known properties that are carried through a procedure from start to end alongside the experimental samples. For microbiome-related protocols, mock community samples can often be a good positive control³⁸. In addition to serving as positive controls, spike-ins can serve as internal calibration and normalization standards^39,40. Negative controls allow detection of artifacts introduced during the experiment or measurement. Negative controls can be samples without the addition of the organism/analyte/treatment of interest that are carried alongside the experimental samples through all experimental steps. They often reveal artifacts caused by the matrix/medium in which samples are embedded; for example, to analyze secreted proteins in the culture supernatant of a bacterium, measuring the proteins in non-inoculated culture medium would be a critical negative control. In microbiome and metagenomics studies, negative controls are particularly crucial when working with low-biomass samples because contaminants from reagents can become prominent features in the resulting datasets⁴¹.

Empowerment through randomized design

The random arrangement of biological replicates with regards to space, time, and experimental treatments is a crucial research practice for several reasons.

Randomization protects against confounding variables

For a given level of replication, the researcher must decide how to distribute those replicates in time and space. The importance of randomization—the arrangement of biological replicates in a random manner with respect to the variables being tested—has long been appreciated in ecology, clinical research, and other fields where external influences are impossible to control. Even in relatively homogenous lab settings, failure to randomize can lead to ambiguous or misleading results. This is because randomization minimizes the possibility that unplanned, unmeasured variables will be confounded with the treatment groups⁴². Unlike experimental noise—random deviations that decrease power by increasing variance—confounding variables cause a subset of replicates to deviate systematically from the others in a way that is unrelated to the intended experimental design. Thus, they cause biased results. Fortunately, confounding variables that are structured in time and/or space can be controlled through randomization (Fig. 3). Even for complex experimental designs, randomization can be easily achieved by sorting a spreadsheet based on a randomly generated number to assign the position of each replicate. In a fully randomized design, all replicates in an experiment are randomized as a single group. In structured designs such as an experiment that employs blocking (see “Empowerment through noise reduction”), however, the best approach is to randomize the replicates within each block, independently of the other blocks.

Projects using -omics techniques are particularly vulnerable to batch effects, a common type of confounding variable with the potential to invalidate an experiment^{1,43,44,45,46} (Fig. 3C). When not all replicates can be processed simultaneously, they must be divided into batches that often vary in their exposure to not only chronological factors but also variation among reagent lots, sequencing runs, and other technical factors. While batch effects can be minimized through careful control of experimental conditions, they are difficult to avoid entirely. Although some tools are available to cleanse datasets of batch-related patterns⁴⁷, the biological effect cannot be disentangled from the batch effect if the two are severely confounded. Therefore, it is always wise to randomize replicates among analytical batches (Fig. 3C).

Randomization allows testing for interactions

Finally, randomization is necessary to rigorously test for interactions between experimental factors (Fig. 3D–H). An interaction is present if one independent variable moderates the effect of another independent variable on the dependent variable. Examples of interactions include epistasis between genes, temperature influencing the phenotypic effect of a mutation, and host genotypes differing in susceptibility to a pathogen.

Many biologists are interested in such context-dependency, but proper experimental design is crucial for testing interactions rigorously. It can be tempting to conduct a series of simple trials that alter one variable at a time and then compare the results of those trials; however, this approach is invalid for testing interactions⁴⁸. Instead, when multiple independent variables are of interest, the biological replicates must be randomized with respect to all of them, simultaneously (Fig. 3D).

Occasionally full randomization is impossible and the experiment may require a split-plot design, where one variable is randomly assigned to groups of replicates rather than individual replicates. For instance, tubes cannot be randomized with respect to temperature within a single incubator; therefore, they must be distributed across multiple incubators, each of which is randomly assigned to a temperature treatment. The main challenge of split-plot designs is that two plots can differ from each other in ways other than the applied treatment, leading to uncertainty about any observed effects of the treatment. In the above scenario, to minimize the possibility that unintentional differences between incubators are the true cause of any observed differences, ideally at least 2 incubators would be used per temperature (either in parallel or in sequential replications of the experiment). As long as the tubes are randomized within incubators with respect to other variables, tests for interactions are valid. However, split-plot designs have less statistical power than fully-randomized designs.

Simulating “biology, chemistry, and physics coming together” in even one pocket along Greenland’s 27,000 miles (43,000 kilometers) of coastline is a massive math problem, noted lead author Michael Wood, a computational oceanographer at San José State University. To break it down, he said the team built a “model within a model within a model” to zoom in on the details of the fjord at the foot of the glacier.

Using supercomputers at NASA’s Ames Research Center in Silicon Valley, they calculated that deepwater nutrients buoyed upward by glacial runoff would be sufficient to boost summertime phytoplankton growth by 15 to 40% in the study area.

More Changes in Store

Could increased phytoplankton be a boon for Greenland’s marine animals and fisheries? Carroll said that untangling impacts to the ecosystem will take time. Melt on the Greenland ice sheet is projected to accelerate in coming decades, affecting everything from sea level and land vegetation to the saltiness of coastal waters.

“We reconstructed what’s happening in one key system, but there’s more than 250 such glaciers around Greenland,” Carroll said. He noted that the team plans to extend their simulations to the whole Greenland coast and beyond.

Some changes appear to be impacting the carbon cycle both positively and negatively: The team calculated how runoff from the glacier alters the temperature and chemistry of seawater in the fjord, making it less able to dissolve carbon dioxide. That loss is canceled out, however, by the bigger blooms of phytoplankton taking up more carbon dioxide from the air as they photosynthesize.

Wood added: “We didn’t build these tools for one specific application. Our approach is applicable to any region, from the Texas Gulf to Alaska. Like a Swiss Army knife, we can apply it to lots of different scenarios.”

United Launch Alliance’s (ULA) new Vulcan Centaur rocket will conduct its first-ever national security launch next week, if all goes according to plan.

ULA announced on Tuesday (Aug. 5) that it’s targeting Aug. 12 for USSF-106, a U.S. Space Force mission that will lift off from Cape Canaveral Space Force Station in Florida.