In a paper published in Science in 1990, Mary Power showed, through experiments conducted in a Northern Californian river, that fish have large effects on river food webs. Power’s results provided support, from a river community, for the Hairston-Smith-Slobokin — Fretwell theory of trophic control. Twenty-six years after the paper was published I spoke to Power about her motivation to do this study, her memories of field work and what we have learnt since about the effects of fish on riverine food webs.
Citation: Power, M. E. (1990). Effects of fish in river food webs. Science, 250(4982), 811-814.
Date of interviews: 8th December, 2016 (via Skype)
Hari Sridhar: I’d like to start by asking you a little about your motivation to do this study. I looked at your publication profile and learnt that, at the time of this study, you had already been interested in fishes for a long time -you did a PhD on a fish species in Panama – and the interest in food webs had also already been there for a long time. How did this particular study come about?
Mary Power: Personally, my history with water is that I was not recognized as being very near-sighted when I was a child, until I was about eight years old. The first time I saw things clearly was when I had a mask on my face under water, when I was maybe about four or five. It was what is sometimes called the flashbulb moment. I saw these creatures, which, if you’re just quiet under water, do what they do in front of your face. But also there’s that index of refraction. So it was blinding clarity and a great overwhelming vision of the beauty of the world under water. This happened to be in a marine environment off Cape Cod. There were kelps and whelks (large snails) gliding over the surface and little crabs skittering around and fish. I knew then that I had to be under water, my life had to be under water from that moment. And that’s not uncommon with aquatic ecologists. John Endler once asked a bunch of us at a conference, “Could your mother keep your feet dry when you were three?”, and the answer was “No” all around the table.
Then when it came time to think about grad. school and on projects, I’d already by then been exposed to coral reefs in a field course, but I realized that there were many more really good ecologists working in coral reefs and in the ocean than there were in clear fresh water. I hadn’t chosen between ponds or lakes or rivers at that time, but it had to be clear, so you could see directly, be under water with the organisms. And then I noticed that rivers especially were understudied. And there are these clear flowing rivers – there were more at the time in Central and South America – and I was entranced by tropical biology too, so that’s why I ended up doing a dissertation in Panama on Armoured catfish (Loricariidae). There were four loricariids that lived in the river I studied, the Rio Frijoles.
What else to tell you… how did my personal interest in food webs come? I did a Master’s at Woods Hole Oceanographic Institution in Boston University and I was paid – I was on somebody’s grant to work in aquaria – and I remember noticing that fish in aquaria didn’t behave at all like the same fish– same species, size class and sex– in natural ponds. So I started to see the artificiality; for a fish, an aquarium might be worse than a prison , because fish are probably getting all these mechano-vibration signals bouncing off the glass and coming back to hit their lateral lines. This constant chaos and noise has to change their behaviour. So I started watching free-swimming fish. I was first interested in ethology and behaviour; then I realised that what the fish chose to do was not just a matter of finding its food but also avoiding its predators. I saw them being fearful and being conservative in their behaviour. So that was kind of from the fish side, recognizing how important top-down and bottom-up controls those are – you need to eat, you also need to avoid becoming somebody else’s dinner. I went to University of Washington, initially to work with Tom Schoener, who did foraging ecology. But after Tom left for Davis, I ended up being adopted as a finishing student by Bob Paine. Bob’s world view really worked well in rivers—most life in rivers is benthic, and like the intertidal at low tide, you can cruise over landscapes while snorkelling, and seeing patterns. Paine’s concept of strong food chains is a very useful world view, and I kept seeing patterns play out again and again in the rivers that I’ve studied – in Panama, in the Ozarks, in Oklahoma prairies, and in California.
HS: How did you get interested in this particular system in California?
MP: Well, to be honest, I have been happily married since 1977, and we spent very little of the first decade of our marriage in the same time zone. And so, when Bill got a job at Berkeley, and I, about 6 years later, was lucky enough to get a job there, I just decided to start working in California. There was also a little bit of carbon sensitivity. I would have loved, and I still may try, to do more work in the tropics or in the Midwest where the fish are more charismatic and diverse. Salmon biologists wouldn’t like to hear me say this, but there just isn’t the fish diversity west of the Rockies that there is in the Midwest or down in the tropics. I just wanted to study something local and to invest in where I lived, and also maybe not be flying round in jets and being away from home all the time.
HS: Was this the first paper based on your work in this system?
MP: Oh, there were several papers before that. When I started at Berkeley, the bureaucracy required to study vertebrates was formidable, and still is. – I think that’s true in many countries and I’m curious how it is in India. This was one reason I avoided studying fish in California river food webs for my first three years at Berkeley. Also, I was interested in learning more about the bottom of the food web, particularly nutrient controls on algae. I got interested in nitrogen from friends in ecosystem science, like Nancy Grimm and Stuart Fisher in Arizona, who were showing how important nitrogen was. I tried, that but I kept getting these weak or inconsistent results. And then my husband said, “Mary, why don’t you do what you’ve done everywhere else?”By that time, I had done these experiments that showed strong top-down control by fish in Panama, in Ozark Rivers, and in prairie-margin streams of south-central Oklahoma. I didn’t anticipate that fish would have a strong effect in the Eel River of California, because fish feeding in the upper watershed weren’t very diverse, dense, or large. The biggest resident fish are 30-cm long resident rainbow trout, but most fishes are < 10 cm long: juvenile salmonids and minnows—California roach. It’s not like, you know, meter-long Midwestern gar or Neotropical armored catfish. But Bill provoked me, so I filed all the paperwork and with students and friends, built 12 large in-river enclosures to study fish effects. In retrospect, I was very lucky with that 1990 paper because, after working in the Eel three years, I had some appreciation of the Mediterranean seasonality. In a normal winter, we have one or more scouring floods that reach or exceed bankfull levels. As you can read in that paper, the following spring, algae recover first and it can get very green after a scouring winter flood, the reason being that huge numbers of a big stonecase grazer, a caddisfly Dicosmoecus, get scoured away. In the subsequent papers, you may have seen that we don’t get these long food chains after drought years because, if the Dicosmoecus aren’t removed over the winter, they have already by summer outgrown the fish predators, so they mow down algae as it starts to grow in the spring, and it’s a two-level food chain. You don’t get the large algal recovery unless disturbances (or experiments) take the Dicosmoecus out of the system.
All that came after this 1990 paper, though. With a lot of help from good friends and great students, I built 12, (then in subsequent years, 24) enclosures, each fencing off 6 m2 of the river bed…It’s huge effort; you literally have to put a ton of sediment around each – 50 twenty-pound buckets- because otherwise river otters will dig under them and mess up your experiments. You also have to put bird net flaring up and away from the top edge to keep the otters from climbing in over the walls. It was a huge amount of work but I had great help from the students who were cheerful in uncomfortable situations.
In five weeks we got these huge, blindingly clear effects in the opposite direction of what we expected. Fish (juvenile salmonids and the California roach minnow) that ate many algivorous insects were nonetheless bad for algae! When juvenile steelhead and adult roach minnows were present in enclosures, the filamentous algae on enclosed boulders collapsed from 60-80 cm streamers down to 1-2 cm high lumpy tufts, within 5 weeks. In exclosures without these larger fish, algae didn’t collapse, but green streamers got overgrown by other algae that turned out to be good nitrogen fixers. The surprise was that fish were at the top of a four-level, not a three-level web (although diets alone would tell us three levels, because they were eating mostly algivorous insects). The grazer that collapsed the algae when larger fish were present was a little midge larva that is often the most common macro-invertebrate grazer in the summer food web. This midge builds itself a secure retreat– I call it a “tuft’– out of algae. Big midge infestations change the architecture of the algal streamers and can collapse them. I’d used this same larva in “midges in bondage” experiments to explore what Joel Brown calls the ‘landscape of fear’…I’d tie midges around their middle with cotton thread. You could make a little knot that would hold the midge firm but not harm it. You could dangle them out in various positions in the river and see where their lives are more dangerous due to fish predation. When these midges were wiggling exposed on the end of the thread, fishes eat them right away –but our fish didn’t seem to be able to recognize them as prey when they were inside their algal tufts. But other, smaller predators could—particularly lestid damselfly nymphs and other smaller predators like the 1 cm long fry of minnows or stickleback fish. We followed this up in a 1992 Ecology paper by simply watching these little predators feed on midges in their tufts – here is an unusual case, perhaps, where the experimental result motivated observation rather than vice versa. No sane person would stare at a damselfly nymph or a baby roach underwater for 7 or 8 minutes; they’re usually not doing anything obvious. But they may be observing their worlds. Damselflies will gaze at a tuft in the algae for literally seven or eight minutes, and then send those extensible mouthparts straight into the tuft to grab the midge in kind of a surgical strike. So we did staged feeding trials and watched long enough to see them do that. All three small predators were equally effective in suppressing midge infestation of clean algae we set out in flow through buckets in the river. The fish fry may get the midges by poking into the tuft or capturing them as the young dispersing instars are just settling out. Fry may be small enough to eat such tiny prey.
HS: Was the choice of this study site dictated mainly by convenience, in terms of how close it was to your university? Were other people already working there when you started?
MP: The Angelo Reserve was a Nature Conservancy site when I first worked there. I was very happy in Panama where you get deep in the forest and only occasional hunters or GIs practicing for Vietnam will come by, but my experiments were left alone. When I worked in the Midwestern Ozarks, sometimes you had to ask permission of somebody who was staring down a rifle sight at your partner down in the river, but after you talked to those folks you can work, and the experiments weren’t sabotaged. But in California with so many “wreckreationalists” (spelled with a w) that wasn’t the case, and I spent three years trying to set up small, simple unobtrusive manipulations. For example, I’d put little photosynthesis chambers in National Forest lands, but the guys in their all-terrain vehicles get as deep or deeper into the forest, so my little chambers were always smashed up. Bill Trush, when he was a forestry student at Berkeley, told me about this Nature Conservancy preserve around the upper South Fork Eel River. It was 3.5 hours north of Berkeley, and it had a fence across the one road that went into 8000 acres of forest that protected five kms of river. When I started working, there had already been Nature Conservancy scientists in the area. The Angelo was the first US gift to Nature Conservancy west of Mississippi. Some important work by Cameron Barrows was done there to figure out why spotted owls need old-growth forest. They need it for thermo-regulation, to warm up in the canopy, and cool off on the forest floor. That work had been done there, but nobody was working in the rivers. I just had to ask permission to do small simple things at first. Peter Steel, the grandson of the Angelos, who gave that gift to the Nature Conservancy, was the Nature Conservancy Steward. Then in 1994, the reserve was transferred to the University of California Natural Reserve System. Peter remained as the only resident staff member who has ever been there to take care of the 8000 acres and subsequently the labs and buildings we have built with grants. I’ve been the UC Berkeley faculty manager. Now we are both up in years, and looking for other people to work at the site and help with its care and feeding. This wonderful protected site was crucial for my own research at Berkeley, and has enabled a lot of other research and teaching programs in biology and Earth sciences. So I’ve become passionate about what the University of California Reserve System is doing. Berkeley now manages five active reserves protected for university-level teaching and research and public outreach. The UC system as a whole protects 40 such Natural Reserves. That’s a huge system of sites that are protected for field research. You can’t usually do prolonged manipulative field research on federal, state or private lands. Nature Conservancy doesn’t want the visual impact, and the state or federal agencies like the Park Service can’t guarantee protection for long-term field research from even what other agencies do, let alone the public.
When was the first time you visited this site?
That would be 1987, just after I was getting pretty discouraged about vandalism of my work in the Mendocino National Forest. I started doing work there in 1988, and in 1989 I did the experiments that were written up in the 1990 paper. And I kept doing those same experiments, but in years with different winter and summer hydrology. I’ve repeated that experiment five times over five years – three winter-flood years and two years that I call drought years, when we didn’t get enough winter rain to scour the bed. There’s a 2008 Ecological Monographs paper that summarizes the results from five summer experiments—three following winter floods, and two following winters with no flood scour. The story is also described in some book chapters. The 1989 results were amazingly clear – fish had top-down effects from the 4thtrophic level – but then in 1990-91, I couldn’t repeat these results at all. The fish had no impact on persistence of algal biomass. Whether fish were present or absent in a river enclosure, the algae collapsed within 2-3 weeks. Then we (this was Tim Wootton and myself) – noticed we had entered a drought, and that a large armored grazer, Dicosmoecus (a caddisfly about the size of a cigarette butt), was 1-2 orders of magnitude denser than it was during post-flood summers. So we did the right experiment for having entered a drought year. We removed the Dicosmoecus from enclosures, and we could recover algal biomass. And now we have a lot more long-term understanding of top-down and bottom-up controls. We now have a marine sediment core that goes back 83 years. It was collected off the mouth of the Eel by an oceanographer Chuck Nittrouer at University of Washington in Washington. And we can see the correspondence. I’ve surveyed algae for, now, over 30 years, across transects just in the upper South Fork Eel within the Angelo reserve, this 8000-acre postage stamp over a ~10000 square kilometre basin. That huge basin empties into a deep marine canyon. Chuck took sediment cores at several hundred sites with a remotely operated vehicle. He was interested in sediment transport and deposition on the canyon slope off the continental margin there. Three of these cores had excellent lamination of the layers so that you could annually resolve, pretty closely, the sediment that came from one year versus another one water year versus another. Of those years, 13 corresponded with years when I had census data. I asked Chuck if he’d ever looked at diatoms, and he had not, and generously let us look at portions of these three cores. My PhD student Jack Sculley counted freshwater diatoms sampling down the core in intervals of mm that corresponding to different years (starting in 1918, when the layering rebuilt after the 1906 earthquake, until 2001 when the core was collected). Jack found a significant relationship with the years that I got big algal blooms, even with only 13 points from years of overlap with my upstream transects. When I measured a big algae year in the upper part South Fork Eel, the whole river apparently pumped out more of freshwater diatoms. Remember that the green macro-algae Cladophora proliferates during post-scour winters. When this happens, surface area for colonization by smaller algae (largely diatoms) can increase by 5-6 orders of magnitude over that afforded by stony substrates. The common benthic freshwater diatoms don’t look like any marine diatom, so they’re relatively easy to distinguish and count. A very common epiphytic diatom that colonize Cladophora after mid-summer, Epithemia, so it preserves well.
Jack’s study also taught me a lesson in patience with noisy data. If you look at that 2008 paper you’ll see that although there’s decent support for big algae summers requiring preceding winter flood scour there are some big flood years when we didn’t get very big blooms the following summer. That’s usually because we had a second late spate that took out a lot of algae after it started growing. Jack’s data were better correlated with my big algae years than with peak winter discharge. So the marine record preserves the actual summer algal proliferation magnitude, rather than responding to hydrologic factors alone, giving us what Jack calls a ‘paleo proxy’. It’s a short record for paleo stuff, you know, only 83 years, but still it gives us a encouragement that the pattern from the whole drainage can be explained by what we measured in a very small part of it.
It also addresses a problem we’ve faced in inferring a top-down (consumer-driven) mechanism. Big winter floods could also produce big algae summers if more rain flushed more nutrients into the river from watershed. Also, if higher flows persist late into the summer, they would prolong favourable growth conditions…So there are bottom-up (resource/condition) mechanisms that would produce the same correlation. Bottom up and top down controls aren’t, as many have written, mutually exclusive, obviously, because there are inputs and outputs for change in storage. But often one forcer is more important. The only evidence we had of important top-down control, initially, was taking those big grazers out and getting some recovery of algae. But we’ve seen some larger scale whole-basin contrasts that support the top down forcing, and then this long oceanographic record showed the same result, when Jack analyzed freshwater diatom abundance versus total annual rainfall for long records available within the Eel basin. He found a noisy, slightly negative relationship of total rain during the preceding winter to algal abundance the following summer. -. And so, all the inputs from the watershed and persistence of the discharge of summer weren’t driving big algae years. It was really the release from grazing that seems more important. This work was published in PNAS in 2017.
HS: It’s now more than 26 years since you did this study. Has the site changed in any significant way in terms of the community, from the time you did this study to now?
MP: Well, that tuft-weaving midge is important during some years but not others. The pattern that seems to be emerging is that that is the creature that makes the salmon-bearing food chain in the Eel a four level food chain. In one year, one of those flood years when we did experiments, the midge was not at all abundant. I think a spring spate swept away floating algal mats that were important for midges as floating incubator nurseries. That year, there was a three-level effect, because the fishes could eat all of the herbivores – mayflies mostly – that had the potential to control algae. So if you get a flood year you don’t get the armored caddisflies that can outgrow them, but then if you have a spate that wipes away floating mats where midges oviposit, then you don’t get heavy midge infestation. The Canadian climate change model predicts that we’ll get more rain on the north coast, and that the rain will fall later into the spring and maybe even early summer. Under those conditions, you might get export of algae after it started growing and dynamics could change. We’ve seen a handful of years with those conditions. So really, there are four types of years, hydrologically – there are two types of winter flows (did it scour or didn’t it?) -and two summer flows: are low summer base flows high to connect and gently flush sunny main stem river pools, and to keep them from warming much above 25o C? If so, then attached algae will be dominated by good edible algae (diatoms, mostly): the green macroalgae and their diatom epiphytes, and the rocks and their epilithic algae. This food web provides great nutrition for fish prey—fish rear well, out-migrate over the summer, and then if there’s algae left over, this highly edible algae senesces, detaches and is exported to the estuary. Down there, we found that the little isopods crustaceans in the estuary love the river algae, so it disappears right away. They devour it much more rapidly than they eat all the local soft green seaweeds that were thought to be the base of the estuarine food web – that remain simply because they’re not getting eaten as fast as the imported river algae. So we are excited to have a harder look at these diatom-covered river exports as a hidden but possibly important energy source to estuarine food webs in these kinds of rivers.
Now, though, let’s discuss what happens in a very dry summer, when sunny channel pools stagnate. This severe drought state is bad after either kind of winter, but the worst case would be if you had a big flood winter. So now you have a lot of biomass of green algae getting covered with diatoms in the early summer but then dying and rotting as the summer base flow drops below critical, and warms and stagnates. This can be brought on either by natural drought and or by human water extraction. Water is getting extracted a lot by for marijuana cultivation. Some folks living along the California North Coast also blame “Doug firification” fire suppression in the basins has allowed dense growths of younger Douglas fir trees to fill up spaces in the older more mature forest that seems to have been maintained by native burning. Whatever the cause, if summer flow is insufficient to keep the mainstem big pools from stagnating and warming up, and then cold-adapted animals like salmonids die, and the nutritious algae rot. Hot, rotting algal mats can be incubators for toxic cyanobacteria (Power et al. 2015). So, toxic cyanobacteria are getting more and more notoriety in the big rivers. They used to be thought of primarily as a problem in lakes and ponds maybe, but they are becoming problematic all over the world actually in rivers that are subject to massive flow reductions and warming. We tentatively think of this as a third food web state—not a three or four-level salmon bearing web, not a two-level chain in which Dicosmoecus gets most of the primary production, but possibly a one-level food chain. It’s not clear who gets to grow on these cyanobacteria (we want to study that more), but if the neighbour’s dog runs through the wrong backwater and licks its fur, it can die of convulsions in 20 or 30 minutes. So that’s getting folk’s attention. The silver lining for us is that through teaching some algal identification short courses, we’ve gotten to meet, work with, and then become friends with both euro-Americans and tribes who really love this river and care about its future. We can teach them algal identification and they can teach us how the river works and what they’ve seen over their lifetimes and long histories in river reaches where they’ve lived. So there are strong friendships and partnerships now; it’s really been quite wonderful.
HS: At the time when you were doing this study, were you staying on the reserve?
MP: Yes, primarily because that’s the only place you could put up these big net enclosures and not have them messed with. But for the estuarine studies, I think now that we know people, we may have more luck. We have access to places that are otherwise kind of guarded. You could do several little experiments. Like my student Charlene Ng, who found out how much the crustacean in the estuary love the river algae. She simply took these plastic disposable pipettes, slit the bulbs, pinched a tuft of algae and put those little things down into the gravel bed of the estuary with all the diatom covered river algae or local sea weeds: cafeteria experiments. You could do a lot. We’d like to learn a lot more about energy sources and food webs in the estuaries, and how these might change with different river flow regimes. That’s hard, but I’m optimistic because I’m getting to know more and more people who care about the river and are starting to understand why it’s interesting to study it scientifically.
HS: Could you give us a sense of what field work was like during this study, compared to now? What was your daily routine when you were doing these experiments?
MP: If you want real change why don’t I take you back to ‘75 when I was working in Panama?
HS: Yes, I’d like to hear about that too.
MP: Okay, so we had miracle inventions but we didn’t have Ziploc bags then. We didn’t have cable ties. We didn’t have Vexar. We had to work with hardware cloth and chicken wire that tear up your hands underwater. There were so many things we didn’t have, but we did have write-in-the-rain notebooks and we had these wonderful Casio watches which I still use. You probably use these too for your behavioural work. They cost $13 and they’ve got a stop watch, a timer…all these wonderful things. I remember those things just come out. And then the mosquito repellent would eat them, so I learnt not to use mosquito repellent. Just all these little technologies. Now leaping forward to 1990, which skips work in the 80s. I worked with a good friend of mine who’s actually a bat biologist and he suggested, I think, that I use Vexar. Vexar is so much easier than when I did my first enclosures in Panama with my 72-year old father. We used hardware cloth and filled steel poles that we had scavenged from the Panama Canal dredging division. I didn’t have any money as a graduate student. The material was very heavy, and we carried them three km up a Panamanian river and put in enclosures there. You cut yourself to ribbons on hardware cloth. Whenever I have hurt myself in streams it has generally been from a piece of rebar or a piece of hardware cloth -which I’ve installed. Vexar was meant to keep chickens from cutting their feet and it was wonderful. And then Bill Rainey suggested that we use these things called hog rings. It’s a ring that you clamp shut, so it pinches through the nostrils of a pig and then you can lead the animal (cruel, actually). We repurposed them, and they were very great for closing and sealing things on the Vexar. And then my father, who grew up on a dirt floor homestead with gravity-fed water in Idaho, taught me something that has never been surpassed in technology for stream enclosures or other enclosures, which is– if nothing else works, cut up an inner tube. There are places where you can’t tie PVC pipe and screens together under water, at some impossible angle, but you can always tightly wrap bicycle inner tube around it. I think all of us who work in the field enjoy improvising technologies that go from the stone age, through 19th century hardware stores, to, these days, electronic technologies (but we older ecologists are terrified by electronics). The other stuff that is so much fun is the arts and crafts aspect of doing field manipulations. It seems you always have to invent new stuff, you always have to go into some local hardware store saying, ‘what can I do with these dog dishes?’ (a question Paul Dayton made famous). There’s a lot of that and a lot of it hasn’t changed. That’s the other thing I really like about the experience of working so long at the Eel, but really about any kind of field ecology where you collaborate with people who love it as well -you’re always trading these fun technologies, and many of them take you back to the nineteenth century, and then you realize life wasn’t as tough then as one might think, because those guys had really ingenious manually operated tools or water power tools that really worked. We just don’t know how to do it now, but they had many ways that were so ingenious, and it made doing things back then easier than you might expect.
HS: Do you still use a similar design and similar material to build cages for your experiment?
MP: Well, we used those boat-shaped enclosures for five years, and I made 24 of them for the years two through five. In the next four years, I doubled the number of enclosures so I could look at the roach and steelheads separately and tease those differences apart. And then, these floating mats of algae that detach are hotspots of emergence and insect densities, for the taxa of insects that can tolerate their hot daytime temperatures and low oxygen. So I was wondering if they were more important (in terms of bug production) as floating incubators or refuges for fish predation or both, you know. I started to test that and it was almost starting to yield results, and then the otters got in and tipped over all the instruments. Vexar’s pretty immortal; these things will probably outlast me. But what we do now for experiments is we often do parallel channels with fibreglass panels called wiggly board. We found ways to stick that in and make parallel stream channels for manipulating nutrients, stream bed texture or biota. And gutters are good for microbiota and very small insects. As I mentioned earlier, though, I have very strong negative opinions of studies that stuff fish into a one square foot enclosure or a bakery bucket of the kind I use for the small experiments with invertebrates or algae. If you are lazy or clueless about the scale in your manipulation relative to your organisms’ operating scales, you won’t learn much about nature.
HS: If you don’t mind, I want to go over the names of the people you acknowledge, just to get a sense of who these people were and how they helped.
MP: Sure.
HS: Okay, you first thank B. Harvey and W. Dietrich for discussion.
MP: Bill’s my husband. And Bret is a friend who I knew when he was a graduate student in Oklahoma. He was a graduate student with Bill Matthews. He now is a research scientist with the US Forest Service in Arcata Ca., and is a great field ecologist who really knows fish and rivers. Bill Matthews and Art Stewart were two young professors at Univ. Oklahoma, based at the UO Biological Station on the north shore of Lake Texoma. The three of us studied the Bass-Campostoma Minnow-algae food chain in Brier Creek, and another river food web in the southwest Ozarks. They were incredibly fun to work with, and very supportive of me as an unemployed postdoc.
Jennifer Nielsen had been a fisheries field technician for US Forest Service, then came to graduate school at Berkeley. She was a fish-whisperer, and an expert in electro-fishing. I knew nothing about electro-fishing, so she advised me on which ones to buy and then she taught me how to do this so we don’t harm the fish. You have to be very light in your touch. Jennifer Nielsen – we called her ‘the goddess of shock’.
Peter Steel is the grandson of Marjorie Angelo, and is the sole manager of that entire 8000 acre reserve
Habte Kifle was an undergraduate chemical engineer at Berkeley, who worked with me two summers as a field assistant. He got assistance to get to the US from a Catholic relief agency that first put him in Vermont. He found it too cold and made his way to Berkeley. When I met him he was he’d been admitted to Berkeley. Berkeley has been a kind of “worm hole” through the world for people who have a lot of talent but not much opportunity. Habte, for example became a chemical engineering major at Berkeley and worked for me for two summers. He now works for the regional water quality control board here in Marin county. He basically lost 10 years to the civil war in Eritrea – but he came out of Berkeley with the opportunities that any middle-class American kid would have. That’s just a plug for what Berkeley has done. We know other dramatic stories of people who have suffered injustice in the world, come here and gotten back on the track.
I don’t remember Stephanie Grimm too well. She was an undergraduate who worked for us briefly that summer.
Sarah’s a very deep thinking natural historian and community-level ecologist who has focussed on frogs, and the impacts of thermal and flow effects of dams on frog conservation biology. She was one of my first two graduate students. She has become dear friend after all of these years and still lives in Berkeley. She has done incredible long-term studies of frogs throughout Northern California, and has become a leading voice for their conservation throughout northern California.
Jane Marks was the other graduate student in that first cohort, and is now a professor at northern Arizona. Jane got a master’s with another dear friend—our algal guru Rex Lowe, so she was the one who recognized Epithemia, our keystone diatom that is particularly nutritious because of its endosymbiotic cyanobacteria that fix nitrogen, and turn the nitrogen into all the protein amino acids. Rex and Jane are still dear friends and the sources of my fascination with the diversity of algae at the base of the food web. My disappointment in not having charismatic tropical fishes in the Eel has been well compensated by the fascination of learning about the diversity, beauty, and magical natural history one can find under the microscope once you look at 200 to 400 X. It’s interesting in our algal short courses how quickly our citizen scientists learn that too. They come shy, never having looked under a microscope, but after a few hours, you can’t even get them to break for beer. Seriously!
Jane has a wicked sense of humour. Installing enclosures took long days and some uncomfortable work. The river was not that warm in June. There’s no snow melt but nights are very cold even through the summer at the Eel. We would have to be chest or waist deep in water, sometimes bucket brigading all these buckets of fine sediment. Get 12 students, hand each of them 20-30 pound buckets and say clench your butts and keep your back straight. Once, I was under water carefully placing the sediment around the skirts of the enclosure so they’d stay anchored to this very irregular rock. You need fine gravel to do that and sometimes the supply of fine gravel would be rather remote from where you needed to put your enclosure. So it would be a long chain of hauling heavy buckets. At one point, Jane came up with her bucket and said confidentially, “Mary, the men are saying “Shoot the otters””. You know, it’s like any great fun field operation; lots of funny memories, happy memories. And sore backs.
Bill Rainey is a dear friend and bat biologist, a very ingenious field biologist and unlike me, not scared of technology. He was a source of a lot of the practical knowledge for making these enclosures, like the hog rings, for example.
Kelty. I don’t remember his first name. He was an undergraduate student from Oakland who had an African first name. He was the one who actually suggested that we build these enclosures to be boat-shaped so that we could put the pointy end into the flow and streamline them a little bit. He was great.
Whiting was a graduate student of Bill Dietrich – my husband – who helped us for a day or two with field work and then there’s Brett Harvey. So, it looks like they helped a little bit with field work too at that time.
Wayne Getz was a theorist at Berkeley, and has been a mind-stretching colleague in the best sense of the word.
Tom Schoener was my first advisor. He was an editor, one of the first ecology editors for Science magazine, after the mid-1990s, when Science got friendlier to ecology. Very high quality editor.
Andrea Worthington was one of my intellectual siblings in the first Schoener cohort from the University of Washington, and she read the manuscript. Andrea and I were in Panama – she studied frugivorous manakins on small islands off Barro Colorado Island.
Bill Hanson was the editor at Science who did a great job handling this paper.
HS: Who was T. Steel?
MP: Trish Steel was married to Peter, so I think I just thanked her too. She didn’t get too involved in the work and Peter basically keeps the roads open. When I first started working there I used to sleep on the floor in a little house rented by Bill Trush, the graduate student who showed me the Angelo reserve. Some years later Bill Dietrich and I bought that little house and 20 acres. But like most natural reserves, Angelo started with rustic housing – there’s just a bare board with a roof, and you put a pad and sleeping bag on it. And you bring a flash light because it is completely off the grid. Since the late eighties, Peter and I have gotten grants together to build up a laboratory there that’s pretty nice actually, and a classroom. We finally have some more comfortable housing. Over the last 20 years, Bill Dietrich has brought research collaborations there g to do hydrology and geomorphology, and our funding levels increased with NSF grants. All of the lab, housing, environment sensor networks and wireless communications infrastructure were externally funded by research grants and several private donations.
HS: I want to ask you a little bit about the writing of the paper. It’s a long time ago but do you remember when and where you did most of the writing of this paper?
MP: It’s interesting that I remember it very well. I’m determined to do this again, and I suggest that if you ever have something that’s short, simple and clear, you give this a try; but maybe it won’t mean as much for somebody of your age. I wrote this paper on a thin old growth Doug fir slab table that was put up on two legs that cross. They are hammered together and support a crude table. I was in our little Eel cottage, the walls are starting to separate from the floor, and every night at 11 pm, these interesting solfugid arachnids would run across the floor, diagonally from one corner of the 10×10 foot room to the other. I could almost set my clock by them. They looked like decapods because their jaws were so big. I wrote the paper on a yellow legal pad of paper – the paper was yellow with thin blue lines and I wrote it out longhand cursive with a pen or pencil. It was a very meditative experience, I knew exactly what I wanted to say. When I write something on a computer there are like 45 drafts. But with this there were maybe four or five. Doing it by hand I think helps. Many people my age are wondering what’s being missed actually because people aren’t writing stuff anymore. From the mind to the hands of the page and back to the mind in a slow contemplative rhythm is different from typing on a keyboard. We wonder what we are losing in terms of lyricism or clarity or carefulness. If you ever have the time, try that as an experiment. Although I don’t even know if young people, like really young people in elementary school in the US, are even learning how to write cursive these days.
HS: Do you remember how long it took you?
MP: Maybe a week, because I already had all the data analyzed. I’m very grateful to be at Berkeley, but it is a very hectic urban setting; there are few quiet refuges. So I’ve done all my writing up at the reserve where it’s quiet. And I remember looking out the window – I have this old growth rustic kind of rough board table and a pad of paper in front of me – and I look out of a broken dirty window, down a rolling meadow that goes to a riparian gallery forest, which has the Eel flowing just out of sight behind that forest. A huge long view down the hill and a little wood fence that my father, when he was in his eighties- like late eighties – insisted on putting up with his axe and sledgehammers. It was on the edge of a steep hill. My mother and I were sure he was going to kill himself, but he got it up just fine because he had made a lot of split-rail fences when he was a kid. That fence is still standing, and along that beautiful rustic weathered wood fence, quail -usually California quail, sometimes mountain quail, a bigger and more vividly colored species – march up and down. The males will march up and down making sure their hens and chicks are safe and sounding alarms. There would be deer and, occasionally, foxes and coyotes in the field. It was quite wonderful.
HS: That sounds like a wonderful setting to do your writing! Can you tell us a little about the publication of this paper? Was Science the first place you submitted it to? Did it have a smooth ride through peer-review?
MP: In this case, yes. That’s very unusual for me. I think I have had only one other smooth ride; that’s the Oecologia Bass-Campostoma paper in ‘83. I think it’s because ecology had started in the mid-1990s being represented by really good ecologists on the board of Science. The previous editor-in-chief was not a friend of ecology. But I think I was very lucky that Tom Schoener saw this study; I didn’t know he was a Science editor at that point. Actually it was Bill Rainey who advised me to submit to Science, because the results were so clear. I might not have thought to do that, without his advice.
HS: Did this paper receive a lot of attention when it was published?
MP: Yeah, it did, positive and negative. Some other stream ecologists were upset that the densities of fish (10 individuals m-2) were pretty high —average densities in these kinds of rivers are typically < 2individuals m-2. When I repeated the experiment I used two fish per square meter, which is reasonable for the average density. Fish do aggregate (you often see 20- 30 fish per square meter) to forage in and around the bedrock islands where turfs grow longest because they didn’t completely scour the previous winter. In a gravel bedded river, these boulder-bedrock formations are islands, and I would say 80-90 percent of the fishes and most invertebrate biomass is there during big algae summers. It’s like a kelp forest. Now they may not be foraging day and night, but they may not be doing that in the enclosures either, so we’re not sure how apologetic we have to be for ten fish per square meter. We did get similar results with two fish per square meter. In the Oklahoma work, we re-arranged fish among whole habitats, and we found that one bass was enough to flip the food web state. You’d have pools that were 70-100 meters long and maybe 10 meters wide. But fish are so potentially voracious and plastic in how much they can feed that even one individual is enough to completely either consume or drive out all the minnows in a pool, when you release one bass experimentally into a minnow pool.
HS: You said that there were also positive responses to the paper. Could you tell us a little about that?
MP: I think this was a big home run and I was lucky to have it early in my career; this is pre-tenure. But when I tell the story of the next year, I always make a little joke out of the fact that I’d gotten this really slam dunk result in ’89, wrote it up and it was published in ’90. In summer ’90, I repeated the experiment and it totally failed. As I tell this in talks, I watch the faces of students or young academics .. you know, first everybody’s horrified if we can’t repeat our results. But in ecology, context can change everything. And that’s a lesson I want to convey: if you can’t repeat your work, why not? Figuring that out will enlarge your understanding of nature. And so what we recognized – I and Tim Wootton who was postdocing with me at the time – we both recognized about the same time that there are all these huge, like, over an inch long stonecase caddisflies around in summers of 1990 and 1991. The algae would start to bloom in the summer, but you would get these bald spots nibbled into it. There would be herds of these little caddisflies moving almost as a front through the algae and oblivious to fish. So it was a two-level food chain. And then we did a couple of years of experiments where we needed to again build 24 enclosures – steelhead, this caddisfly Dicosmoecus – both-neither – and we got no algae with Dicosmoecus, and release of algae when Dicosmoecus was removed. There was also slightly more algae when steelhead were absent, as if they were still doing fourth trophic level stuff. But the big effect was the two-level effect. And so that’s when I became really preoccupied by context-dependent thinking. Seems like only Jim Estes has a food web that can last5 million years and cover, you know, the north Pacific and make good predictions. Bob Paine, in his 1980 Tansley lecture, said that in Alaska, Pisaster is just another starfish because you’ve got freezing that gnaws holes in the mussel bed; so there you don’t need the starfish. So there’s all this context-dependency. And rather than be distressed and say, ‘Oh, you know, community ecology can’t generalize’, what we need to do is take context-dependency head on. And we have more & more tools these days, remote-sense based and mapping tools and the tracers in environmental sensing, that we should be able to discover, hopefully, manageable ranges of conditions that change the performance of web members and the strength of some of their links, and change, then, the paths that control plants or release plants. So I think that’s what that really amounts to. This sounds really big – big as in unmanageable – but it’s just merging sensible study of the fundamental or maybe the realized niche with food chain dynamics, in a way that gives us predictive context-dependence for which of the potentially powerful food chains will control how populations are limited. So it’s all of ecology basically, going back to all of ecology, but with more depth in tools and knowledge than we used to have.
HS: In the abstract of the paper you say, “These manipulations provide evidence that the Hairston, Smith, Slobodkin-Fretwell theory of trophic control, which predicts that plants will be alternately limited by resources or herbivores in food webs with odd and even numbers of trophic levels, has application to river communities”. Now, in retrospect, if you had had the opportunity to frame that again, would you change anything?
MP: Well, I do know more about the history of the ideas about trophic cascades than I used to. And I know some of this just by conversations rather than stuff that’s published. We can all notice that there’s somebody who’s not even easy to find on the internet. I’ve kind of promised myself I’m going to scan his articles that are in English and then I’ll post them, but I don’t have the skills to take a fold out hand drawn graph and get it into a PDF that is easy to look at. But the first guy who thought as we do now about top down back controls might have been Hrbacek, a Czech working in flood plain lakes and rivers in Bohemia; I think in eastern Czechoslovakia. And then Fred Smith, who was Bob Paine’s advisor, met him and heard him. And then Fred came back, and in an article published for the Journal of Sanitary Engineering, he published an idea of top-down and bottom-up controls that was pretty close to what I was then thinking of as the Fretwell theory. I admired Fretwell very much. I discovered the paper because of Tom Schoener’s MacArthur lecture; Tom said there’s this obscurely published paper – it was Fretwell 1974. I read that, and even though it was an armchair thought experiment, it was magic for me. I hadn’t really been motivated by a strong theoretical framework until the 1990s, but on learning about it, realized it could explain a lot of what I’d seen in four rivers–a two-level food chain in Panama, except where the birds could get the fish, and then it became three-level and then a two- level chain or (we don’t know) maybe 4 or 6-levels in the Ozarks and a three-level food chain in Oklahoma Prairies and then a four-level food chain in Eel river except after drought. So suddenly, there was that theory of ‘n’ level food chain. Ecologists have also been interested in what controls chain length. They are supposed to shorten with disturbance, and lengthen with environmental productivity. But that’s only if you ignore that over the course of succession, lower trophic levels fill up with predator-resistant taxa. We know from succession that late succession plants or animals were defended usually; that’s why you plough a field, to get rid of the woody stuff so you can grow your lettuce and spinach, right? That’s well known, but not recalled when food web theorists found their computer simulated food webs were less robust if they’d modelled longer food chains. And then Bruce Menge’s is based on natural history, but it’s a fairly specialized case of predators that have to grip in the intertidal and can’t feed while hanging on with their tube feet. So, in general, disturbance should maybe, after time, – intermediate levels of disturbance – should lengthen food chains, as Connell showed it to increase species diversity. A lot of our kind of early nineties to mid nineties work made that argument. Productivity might shorten food chains if it allows these resistant taxa to takeover sooner. We have a little bit of limited data for that, but I’d like to actually do more work to test that idea.
But back to the history. Long after I’d written these papers, Bob Paine told me that he’d visited Fretwell in Kansas and shared Fred Smith’s food chain view. So the Fretwell theory may have come from conversations with Bob that transmitted Fred Smith’s – Bob’s advisor’s – ideas. Fretwell wrote really nice thought papers, and I actually like his ‘74 paper better than his ‘84 paper. The latter is assertive, but the ‘74 paper is more contemplative – well, maybe, you have this one level trophic system on the east slopes of the Rockies, and then you get into the prairie where it is two-level, then you get to the deciduous forest of the Midwest where it is three-level and then you get the salt marshes where maybe it is rich enough to be a four-level chain. This is speculative, but thence thing is that there are some nice ancillary predictions, like you get your colours changing in the autumn when you have a three-level system and that the plants are going to be nutrient limited, so they withdraw some pigments into their stems before shedding leaves. There are some fun spin offs. It’s a case where soft thinking is still very valuable for inspiring a more critical look.
HS: In your paper, you cite ‘Fretwell 1977’. Is that the paper you are talking about?
MP: Yeah I was wrong about 74, it is 77. But let me tell you why I was also so interested in Fretwell. I went to the, I think it was the 1990 INTECOL meetings that were in Yokohama Japan, and I met Laurie Oksanen there and Gary Polis and also Lennart Persson. Laurie was so inspiring. He was Fretwell’s last graduate student before Fretwell dropped out of science. Laurie and his colleagues put the Fretwell theory into mathematical formulations. But more than the mathematics, Lauri and his wife Tarja are such cosmic thinkers. Laurie is really interested in paleobiology, gradients of food chain control and differences in food chain control across the entire globe – he would think about the Seychelles or the African Serengeti food webs or the arctic food webs. Getting a chance to interact with him and we corresponded extensively after these meetings and then he and Tarja came over and visited me in the in the early nineties – that was a huge inspiration for my excitement about the potential global-scale explanatory power of this simple food chain idea
HS: What impact do you think this paper had on your career?
MP: Well the work in Oklahoma was what…As I told Bill Matthews, I never would have gotten the job at Berkeley without the research I was able to do with him and Arthur Stewart. We were a great team – and still great friends. Bill is a fish expert, Art was an ecosystem ecologist and algal expert and I was the community ecologist bridge. We had a wonderful four years looking at Ozark mountain and prairie streams. That was really what turned me from a pretty frustrated Berkeley captive spouse captive spell into somebody who could compete for a job. I was actually the second choice for this job at Berkeley, but the first (Rich Lenski), fortunately for me, said no, so I lucked out . At that time, I had under my belt the Panama story, the Ozark story and the Brier Creek story and that got me a job at Berkeley, but when this 1990 paper came out it was pre-tenure; tenure is still a big deal so I wasn’t that secure in my career. And that’s why I like to dramatize that story about getting wildly different results in 1990-1991 and then getting back to the three-level in 1993 and then four-level again in 1997. In ecology in particular, well maybe in all of science. you’re not always going to repeat your work, and it’s not always going to be because what you initially discovered is wrong; just limited. So if you take that view that it was limited and what else is happening now, then you grow in your understanding.
HS: If you were to repeat this experiment today, what would you change? You already spoke about how the design of the cages is different now. Would you change anything else about the experiment itself?
MP: I’d still use those 6 m2 enclosures, made out of immortal Vexar screen. It’s always preferable to use whole natural habitats as your experimental unit. In the Oklahoma prairie streams, we could re-arrange predators and prey among whole stream pools without doing anything except sometimes splitting pools down the middle. In the Eel river mainstem channel pools are too big (e.g., 30 m * 150 m* 7 m deep) to manipulate at the whole habitat scale. Enclosures in pools are not too bad because the pools are pretty lentic in the summer. For productivity manipulations, we’ve done fairly successful experiments with different opacities of shade cloth covering 0.5 x 2 m flow-through channels, looking at food chain assembly (Wootton and Power 1993 PNAS). If I could repeat that experiment, I’d do more macro-invertebrate taxonomy to figure out why a change happened between day 30 and day 55, which I think was due to the predator gleaning. This change bears on this long and short food chain versus productivity argument. You have five different productivities that are successfully manipulated by different shading levels of light in parallel stream channels, built over the natural stream bed, then the food web assembles naturally through, 6-mm mesh walls across the upstream and downstream channel ends. In these 5 channel blocks, with 5 replicates, you’ve got your natural detritus and algae, you’ve got bugs and larval fish coming in. But we didn’t want to admit the larger fish because they would have just hung out under the dark canopy using it as an inner shady cave and then gone and foraged in the more productive channels. The enclosure scale would be too small for the big fish. So we excluded them with those upstream and downstream mesh walls, and just looked at food webs assembling across a productivity gradient that could be only maximum three trophic levels. And what we found was exactly what the Oksanen-Fretwell 1981 model predicted for a three-level chain. It was that algal biomass increased across the light gradient, herbivore biomass was flat, and predators increased. The data were remarkably convergent with what the simple models – no predator interference or inefficiencies – predicted. That was after day 30. Day 55, the herbivores are starting to look a little bit more positive, looking more like a food chain where there’s leakage of productivity so that each level gets some, either due to predator interference, or, more likely, my hunch is that the remaining herbivores were getting dominated by more predator resistant forms; the edible ones had been gleaned out. But we didn’t have the taxonomic resolution in the way we processed the samples to really test that.
HS: In the in the last sentence of the paper you say “Cladophora blooms occur in sunlit rivers worldwide [], and fish-mediated effects on their accrual, architecture, and persistence will affect energy flow within these rivers and from rivers to watersheds [].” I wanted to ask you whether we now know about these effects in other rivers that have Cladophora?
MP: The algae that we studied in the prairie streams was Rhizoclonium, very similar and a close relative of Cladophora with the same tendency to get overgrown with diatoms. We saw three-level food chains there—Rhizoclonium was grazed to a barren state if grazing minnows were the top of a food chain, and bass predators turned the pool into a green, three-level world. In terms of Cladophora diverting river production to upland terrestrial consumers, much of this occurs because primary producers, often Cladophora and its epiphytes, feed larval aquatic bugs that eventually fly out of the rivers as winged adults. There’s a lot of famous work on this done in Japan by Nikano, Murakami and their colleagues. I didn’t know about Nikano’s work in 1990 when I started working on rivers feeding forests. I was inspired by the work of Stuart Fisher and Larry Gray in a desert stream, Sycamore Creek AZ, where export of bugs and direct export of algae are feeding the desert. There they have Cladophora and Epithemiaas important components of their producer system.
HS: In the 26 years since the paper was published have you ever read it again, apart from today?
MP: I don’t think so. I’ve got a lot of other things that are not different; they build on this kind of thinking. There are a bunch of things in it that I actually disagree with, one of them being that this is the first demonstration. But there is something in the end that I didn’t like. I may need to search for this and email it to you later
HS: Okay. In general, did anything else strike you when you read this paper again? . Do you think your writing has changed ?
MP: In the latter part of my career, I’ve spent a huge amount of my time editing the writing of students who were never taught to write well. We have a deficit in our high schools. I was so fortunate that in junior high Janet O’Neill made us write one two-page book report every week and it’d come back with often more red ink than blue ink. That was huge for me, but today students aren’t getting that, so I spend so much time heavily editing bad writing, although I’m not sure my writing is really good.
HS: This paper is very well-written and easy to read.
MP: Thanks very much for that.
HS: Did you take the photographs in the paper?
MP: Yes. I used a single-lens reflex underwater Nikonos. So you have to judge distance, you know, unlike a point and shoot. It’s easier these days to get a good underwater photograph. In that photograph the algae is too dark, unfortunately, at least in my copy of it. I don’t remember, I don’t think this was published in colour; it should have been. Color is an important field clue to the state of the Cladophora. It grows many meters after scour winters, so we call it Rapunzel’s hair- as in that fairytale? Yes? Rapunzel’s hair is green when she’s young, and then she becomes a blond in middle age, and redhead in older age as her hair is more thickly colonized by diatoms. We have another 2008 Freshwater Biology paper proposing that rates of nitrogen fixation and bug emergence might be estimated photogrammetrically over large areas from aerial extent of the green, yellow, and red Cladophora patches. .
HS: Do you recognize the places in the photographs?
MP: Yeah I do. One thing that has changed in the river, which I forgot to mention, is that some of these boulders- like you see in ‘A’ there is a boulder emerging from the bed – it’s kind of nearly buried now. The Eel is a steep river and the geology is very fragile–made of mudstones and sandstones from the sea bed that was shoved up tectonically into the west coast of North America. So you get these landslides in head waters that sends loads of sediment down the canyon. We’re now at a point where some of the deeper pools with clean exposed boulder bed rock formations are a little bit buried in sediment, and in other pools the sediments move beyond that. So that’s a slow change in the long profile or the bathymetry of the river. The only change in the forest, which is happening throughout California because of fire suppression, is that, without the native American legacy of tending the forests with cool fires in the fall, we’re getting encroachment of Douglas fir. Even in the dark old growth old forests the cohorts are punching up and when they reach the canopy will just fill in the forest. That seems to maybe affecting hydrology too as I mentioned earlier. The Euro-Californians are starting to listen to the native Californians after about 100 years.
Would you count this as one of your favourite pieces of work?
Yeah I think I would, along with the 1985 paper with Bill Matthews and Art Stewart. But like many others, I have what my colleague Dave Ackerly calls: “my most important widely unread paper”. It wasn’t a trophic cascade study. Rather than asking what are these fish doing to their communities, I was asking, how do these fish cope with changing opportunities and risks in their world? I’d followed 1300+ individually marked armored catfish for 2.5 years through 16pools spanning ~ 3 km of the Rio Frijoles in central Panama. I censused them in each pool, quantified their feeding rates,, and tracked individuals’ growth and survival, as well as productivity and biomass of their attached algal food. I found that catfish tracked variation in algal growth rates among these pools so closely that their collective grazing damped out all biomass variation in algae. This was despite huge (16-fold) differences in light-limited primary productivity among sunlit and shaded pools. Algae in sunny, crowded pools were just as scant as algae in dark pools less crowded with grazers. As a result, the armored catfish growth rates, dry and rainy season, were similar among dark, half-shaded and sunny pools, and so were their survivorships. In other words, these armored catfish had maintained Fretwell’s ideal free distribution over scales in nature that had not previously been documented. That paper was one of the first out of my dissertation. I was naive about getting it published. It was rejected (I thought) from Ecology because a referee said it wasn’t experimental. But in fact, my paper used experiments to calibrate algal growth in dark and sunny pools, for the ideal free model. I had to measure algal productivity without grazing, so I elevated tiles above the river bed. That year, Ecology published a paper in which the authors had just elevated tiles above a river bed, and found higher accrual of algae. It was published because the experiment was emphasized. The reviewers of my paper didn’t notice that I’d done the same thing under ranges of canopy cover, and thought I’d just done an observational study. The other thing that happened to that paper, has to do with writing. After it was rejected from Ecology, I sent it to Journal of Animal Ecology in Britain. Remember this is all by snail mail. I got an elderly reviewer who was out of touch with revised punctuation rules of the journal. He also thought it was ridiculous to say that fish had topographic memories. I’d watched marked individuals that I’d moved among pools go back, travelling kilometres across dangerous riffles, to return to the same pools, sometimes the same little alcoves, where I’d originally caught and marked them. Catfish need memories to sample the environment enough to evaluate relative habitat quality, for the ideal free distribution. To counter that reviewer, I found (much to my delight) a 1913 reference published in French showing that fish had excellent topographic memories. There’s other good work that shows fish remember their habitats, very interesting behavioural work showing they can remember where the tide pools are at low tide because they have surveyed them at high tide. In Hawaii, I did experiments which I’ve not published, inspired by work on gobies by Aaronson – both blennies in Hawaii and gobies in the Bahamas can’t see surrounding tide pools when they’re trapped in one at low tide. But if you bully them into a corner of their little tide pool, they can unerringly flip over hot rocks to pools upslope that they can’t see, but must remember from high tide.. That J.Animal Ecology paper went back and forth six times across the Atlantic. Each time, the reviewer threatened that if I didn’t get the punctuation right, it would just be rejected. But I couldn’t see any rules I’d broken. Finally, I just wrote to the managing editor saying I can’t figure out why this reviewer thinks I’m not following the rules because here are the rules and here’s what I’m doing and the reviewer is out of date. It finally got published in 1984, but was clearly never read. The Brits subsequently published a number of book chapter reviews on the Ideal Free Distributions (it’s quite an elegant theory), pronouncing that it had never been demonstrated at any large scale in nature. When Nils Christian Stenseth and his group analyzed pike tracking perch across two basins in Windermere Lake, that paper was heralded as the first large-scale demonstration of the theory in nature.
So that paper is I think the best work I’ve ever been able to do, when I was able to live in the system continuously, before getting a day job. I think the most important thing to tell students about the writing experience would be: don’t let reviewers or editors be too heavy handed with your voice. You don’t want to say silly things, you don’t make jokes that may not be accessible to an international audience, but don’t let reviewers or editors remove all the fun. There are certain journals that are better than others in preserving the voice of the author, which I think is important. I’d say Oikos is very good at keeping the voice.
HS: What would you say to a student who is about to read this paper published 26 years ago? Would you guide their reading in anyway? Would you add any caveats? Would you point them to other papers they should read along with this?
MP: I’m answering the question not I think the way you intended. I’d say that please go out into nature and test ideas, but it will come at a cost, because this kind of work is slow and sometimes physically hard, but it’s also incredible fun. Don’t just sit back and review a lot of published papers. Certainly, that’s valuable sometimes; meta-analyses can be valuable, if done thoughtfully. But they can’t substitute for going out and seeing what’s happening in nature, particularly as it’s changing so fast. We may feel a little bad that it’s hard to replicate results in ecology, but it’s very important we keep trying so we learn to anticipate how different forcers from the outside and different internal states are going to change e.g. the outcomes of these food web interactions. The muddy boots field work, sometimes motivated by theory – not always but sometimes – is so important. I get really discouraged when I visit universities, even great universities, and I ask the post-doc who needs to get 4-5 papers out in the next 6 months, “Oh, so you’re analyzing British bird survey data (and I don’t say yet again)? And the post-doc says, “I’ve come up with a hypothesis that insects are limiting their over-winter survival. I say, “Oh, are there any data on insects?” “No, there are no surveys of insects.” What distresses me is the finality in their voice, because they don’t have any intention to even consider surveying insects in limiting parts of the range themselves, or searching for other factors might limit their bird populations. They have let go of their curiosity to go out there and do the fieldwork. So I guess what I would plead is this: If you read this 1990 paper–guess what! The rules are going to change. If you go out to nature (please) and study the same system, you’ll likely find it changed. And you can add tremendously if you find a different outcome and can figure out why it changed. But you can’t get that by just reading old literature, especially when change is accelerating in the world.
0 Comments