How to get great science out of citizen science

Citizen science is when members of the public directly work with scientists on a particular question or issue. Participation can range from a large number of single interactions to repeated and complex sampling that requires substantial training. A new paper now explains how to train participants, validate the collected data, and produce rigorous scientific papers from the outcomes. Key highlights include the need to increase the quality of data when designing a project, and to apply quality control afterwards to check for issues with the collected data. Studies with large numbers of participants will benefit from simple data collection methods combined with advanced statistical methods that check for agreement among scores submitted by multiple participants; studies with fewer participants require a different suite of tactics including advanced training, outstanding explanatory materials, and independent verification of results. These two approaches (large and simplified vs. small, careful, and complex) can both yield valuable scientific information. The study was conducted by SAFS professor Julia Parrish and research coordinator Hillary Burgess, and their coauthors, and is published in the journal Integrative and Comparative Biology.

Necessary steps of science (gray boxes) and additional elements (italics) that are specific to citizen science.

 


Oysters and eelgrass help each other out under increasing carbon dioxide levels

Increasing human output of carbon dioxide results in higher temperatures and in ocean acidification—the lowering of ocean pH and other chemical changes. Oysters are threatened by ocean acidification, while eelgrass may benefit from the higher carbon dioxide levels in the water. A new laboratory study asks whether culturing Pacific oysters (Crassostrea gigas) together with eelgrass (Zostera marina) can help both out. The idea is that the eelgrass uses up carbon dioxide in the water, which should buffer the effects of ocean acidification and increase pH levels; while oysters filter disease-causing organisms out of the water, which should improve eelgrass growth. In a series of recent experiments that tested these ideas, eelgrass wasting disease declined under higher carbon dioxide conditions, and was even more dramatically cut in the presence of oysters, resulting in faster eelgrass growth. In addition, in the presence of eelgrass, pH did increase, but the increase in oyster mass was not statistically greater than expected by chance. The new study was conducted by Maya Groner while at the University of Prince Edward Island, together with coauthors that include SAFS professor Carolyn Friedman, and is published in the journal Ecology.

Growing oysters with eelgrass should have two main effects: eelgrass uses carbon dioxide, thus increasing pH and reducing the effects of ocean acidification on oyster growth; and oysters filter out pathogens including those that cause eelgrass wasting disease, benefiting eelgrass growth.

 

When cultured with oysters, eelgrass grows faster (top panel), and eelgrass wasting disease (EWD, bottom panel) is greatly reduced, especially when carbon dioxide (pCO2) levels are high.

 

Levels of pH increase when eelgrass is grown, reducing the effects of acidification on oysters.

Intensive use of lake water affects freshwater food webs

Many lakes are important sources of water for agriculture and other purposes, while also supporting diverse ecosystems. In a new study, a comparison is made between the food webs of two natural lakes that were dammed early in the 20th century. The neighboring lakes are nearly identical except that one (Lake Keechelus) experiences rapid drawdown of water beginning early summer while the other (Lake Kachess) remains fuller and fluctuates less in water height during summer, but is lowered to a lesser extent beginning early fall. This water management scheme helps balance water needs for irrigation and threatened salmon downstream. A comparison of the food webs in the two lakes reveals that the more intensively used lake has a less diverse food web (termed “trophic compression”). Thus water use, like other human disturbances such as warming, excessive nutrients, and invasive species, results in lake ecosystems that are less able to cope with external stress. The new paper by Adam Hansen, a SAFS alum, two SAFS post-bachelor researchers Jennifer Gardner and Kristin Connelly, Matt Polacek, and David Beauchamp, appears in the journal Ecosphere.


LIME: a new model for assessing the status of data-poor fisheries

Formal stock assessments are conducted for many large and valuable fisheries, but these typically require reliable catch data, estimates of trends in fish numbers, and age data from caught fish. In data-poor fisheries, these kinds of data are not available, resulting in difficulties in assessing whether they are overfished or sustainably fished. Now a new model called LIME has been developed that accounts for variability in recruitment (the number of baby fish produced each year), and can assess status from samples of the lengths of fish in each year, together with whatever additional information is available. Applying LIME to simulated data (representing “truth”) shows that it can estimate fishery status, especially for short-lived species, provided good information is available for growth rate parameters, and there are multiple years of length data. The LIME model was developed as part of Merrill Rudd’s PhD dissertation at SAFS, together with James Thorson of NOAA, and is published in the Canadian Journal of Aquatic and Fishery Sciences.

Performance of the LIME model under different conditions where model parameters are too low, just right, and too high (columns); for species that are short-lived, medium-lived, or long-lived (rows); and for two variants of catch time series (red, blue). Relative errors close to zero indicate the model has good performance for that scenario.

Fish and chips: DNA analysis on a chip can separate endangered wild steelhead trout from hatchery steelhead trout

Many populations of native steelhead trout in the Pacific Northwest US are threatened by disease, habitat loss, poor ocean survival, and genetic mixing with hatchery steelhead trout. Steelhead are a form of rainbow trout that migrate out to the ocean when young, and return to spawn, just like many salmon species. Hatchery-produced steelhead have lower survival in the wild because they become less afraid of predators; one of the resulting concerns is that interbreeding between hatchery and wild steelhead will erode the natural fitness of wild steelhead and hinder their recovery. However, it can be difficult to separate wild and hatchery steelhead genetically. Now a new rapid genetic technique is applied to steelhead that uses a single sequencing chip to gather information on more than 57,000 locations where DNA differs among populations in trout and salmon species. The new technique greatly simplifies genetic analysis of steelhead comparing to whole-genome sequencing, requiring less than 1% as much storage space and computer time for the analysis. The study found 360 DNA locations that differed between wild and hatchery steelhead and could be used to separate these populations. The research was conducted by SAFS postdoc Wesley Larson, SAFS professor James Seeb, Kenneth Warheit of WDFW, and their coauthors, and appears in the Canadian Journal of Fisheries and Aquatic Sciences.


Steepness, lakes, and forest cover are the main factors influencing salmon-stream temperatures

Salmon returning to streams and lakes in Southeast Alaska are affected greatly by water temperatures both in winter and summer, and these temperatures are projected to increase given climate warming. Changes in water temperature affects the time it takes for salmon eggs to hatch and emerge, and the timing of salmon returning to each stream, as they seek to avoid dangerous peak stream temperatures. Thus predictions of the effect of climate warming on salmon populations relies critically on predictions of water temperatures. Now, a new study identifies the most important factors that can be used to predict changes in water temperature in the future in these salmon-producing streams and lakes. In summer, streams in areas with shallow gradients and more lakes had both higher and more variable temperatures; in winter, temperatures were higher in areas with steep gradients that had greater forest cover and more lakes. Thus both landscape features and forest cover influence stream temperature, and hence the productivity of salmon produced by each watershed. The research was led by Michael Winfree; coauthors include SAFS professor Daniel Schindler; and it appears in the journal Environmental Research Letters.


Traditional knowledge about polar bears adds to climate change evidence in Greenland

Interviews with 25 Inuit polar bear hunters in East Greenland provide a wealth of knowledge about changes in sea ice, warming, and polar bear distribution and trends. Evidence of climate change reported by the hunters included receding glaciers, higher temperatures, and the loss of sea ice. These changes made it harder for them to access sea ice, because dog sledges are no longer safe given wide patches of open water during months when sea ice used to be safe to travel over. In addition, about 80% of hunters reported that more polar bears are entering their communities, which they attributed to both the loss of sea ice and the introduction of quota limits on polar bear hunts. The research on traditional Inuit knowledge was conducted by SAFS and Applied Physics Lab professor Kristin Laidre, Allison Northey, and Fernando Ugarte, and is published in Frontiers in Marine Science.

Change in hunting patterns in the region of Ittoqqortoormiit, showing how past hunting areas (green, 10-15 yr ago) are greatly reduced in recent years (pink).

Predicted impacts of ocean acidification on fisheries on the U.S. west coast

Ocean acidification is the a suite of chemical reactions in the ocean caused by climate change that include higher levels of carbon dioxide and lower pH levels, caused by human emissions of carbon dioxide into the atmosphere. A new model now projects the impacts of ocean acidification from California to Washington, finding that species declined most in the southern regions, but economic impacts were highest in the northern regions. The greatest economic impacts were due to declines in Dungeness crab due to declines in prey species affected by ocean acidification. Given the high fishing revenues based on Dungeness crab, a high priority should be devoted to tracking pH, crab prey, and crab biomass. The new study by former SAFS PhD student Emma Hodgson, SAFS professor Timothy Essington, and their coauthors, appears in the journal Ecological Modelling.

Economic impacts of ocean acidification by region, with positive impacts (>20%) in blue and negative impacts (< -20%) in yellow-brown.

Stretching out supplies of fish food to aquaculture species

Some types of aquaculture-raised (farmed) fish and crustaceans rely on wild-caught fish as feed for omega-3 fatty acids and micronutrients. But with the rapid and continuing rise of aquaculture, and the natural limits to the supply of forage fish (anchovies, herring, and their relatives), eventually this supply of feed will be exhausted. A new study now highlights ways in which the supply of fish food can be eked out further by: (1) reducing the proportion of feed that is based on wild-caught fish and switching to crop-based diets such as soy; (2) increasing catches of forage fish to maximum sustainable levels, adding 30% more catch compared to 2012 levels; (3) eliminating the addition of wild-caught feed to non-carnivorous farmed species; (4) eliminating forage fish from pig and poultry diets; (5) using trimmings from the processing of other wild-caught species as food for farmed fish; and (6) increasing the efficiency of farmed fish production. These adjustments offer a variety of pathways to ensure that forage fish are able to support aquaculture growth beyond the year 2050. The new work by Halley Froehlich (UC Santa Barbara, and SAFS alum), SAFS postdoc Nis Jacobsen, SAFS professor Tim Essington, and their coauthors, appears in the journal Nature Sustainability.

(top) Total mass of food produced from pigs, poultry, and aquaculture species fed using wild-caught fish.  (bottom) Dramatic changes over time in the use of fishmeal from wild-caught forage fish, showing how farmed fish are now the largest consumers of this feed source.
Future trends in the use of wild-caught fish under different scenarios, showing how some scenarios exceed by 2050 the maximum sustainable yield (MSY) of forage fish, while a combination of solutions (“All Adj”) allows the supply to extend well beyond 2050.
(left) Consumption of wild-caught forage fish by species groups, showing how pigs and poultry still consume large supplies. (right) The dominance of mainland China as a consumer of forage fish feed.

Substantial decline in polar bear range with sea ice loss

Polar bears rely heavily on sea ice to search for prey, and in Baffin Bay between Canada and Greenland, such ice-associated searching halts during ice-free months. However, since 1979, higher temperatures have resulted in the ice-free season in this region increasing by 12 days per decade in this region. Now, satellite-tags placed on 81 polar bears in Baffin Bay reveal that polar bears greatly reduced the area in which they forage between 1991-95 and 2009-15, by as much as 70% in summer. As a result, this subpopulation is becoming increasingly isolated from other nearby polar bear subpopulations. These results show the clear impacts of sea ice loss on polar bears now, with increasing impacts likely in the future. The research was led by SAFS professor and Applied Physics Lab researcher Kristin Laidre and her coauthors, and appears in the journal Ecology and Evolution.

Geographical range of Baffin Bay polar bears during the 1990s (mauve) and 2000s (green), showing the substantial decline, fragmentation, and isolation of this subpopulation from nearby polar bear groups.
Monthly changes in polar bear range, showing the especially pronounced decline in range during summer months from the 1990s (mauve) to 2000s (green).