top of page
PKX (aka PKD) Monitoring


eDNA testing for PKX (Tetracapsuloides bryosalmonae) on the Upper Yellowstone arose in part due to 2016 Yellowstone River fish kill and closure​. From the event, we were reminded that we do not have enough water monitoring data in order to leverage modern scientific analytics and computer machine learning techniques to predict, identify and resolve water quality issues in the Upper Yellowstone Watershed. 

This web page, along with the rest of the RiverNET monitoring efforts, is intended to share peer-reviewed science on the on-going effort to understand root causes of the 2016 Fish kill and better understand the watershed dynamics to prevent future incidents and, more importantly, maintain the Upper Yellowstone as some of cleanest water in the world. Citizen scientists and researchers are showing that low-cost autonomous water monitoring is possible. One recent study published in the credible Nature journal even shows that we can monitor down to the DNA level through scaleable robotic methods like some have done on the Upper Yellowstone to test for PKD.

As detailed in this research report, "we have found evidence of  T. bryosalmonae in salmonid fish from nearly all waters across the Rocky Mountains where we have received samples, including most rivers in southwest Montana (see below Figure 1 in Hutchins et al. 2019). Many of these rivers had hydrologically stressful conditions (i.e., warm water temperatures, low discharge) similar or worse than the Yellowstone River in 2016; yet there were no documented PKD fish kills. These results indicate that T. bryosalmonae is widely distributed, and suggest that warm temperatures, low flow conditions, and agricultural/irrigation practices cannot alone explain PKD-caused fish kills." That said, water consumption practices in July-August might have been altered to mitigate the impact. We just don't know yet. RiverNET, and the community science and volunteers behind it, is devoted to sharing water monitoring data (regardless of the source -- government agency or local citizen) so as to promote a healthy debate over the root causes. We all share the common goal of setting aside emotion, discovering the truth and then acting on that truth.

Again in 2020, upwards of 200 fish were reported dead on varying stretches of the Upper Yellowstone but concentrated downstream from Livingston to Big Timber. An FWP press release is here.

Stream gauge data for the relevant time period from Carter's Bridge is below.


Authors: Adam Sepulveda (USGS Northern Rocky Mountain Science Center), Lacey Hopper (USFWS Bozeman Fish Health Center), Ken Staigmiller (MT FWP), Scott Opitz (MT FWP), and Travis Horton (MT FWP)

For any additional questions please contact:

What is PKD?

Proliferative kidney disease (PKD) is a disease of salmonid fish caused by the endoparasitic myxozoan,(Okamura et al. 2011).

Clinical PKD is characterized by a massive inflammatory response caused by cells proliferating in the kidney and spleen in salmonid fish infected with T. bryosalmonae. However many salmonid fish infected with T. bryosalmonae do not develop clinical PKD, especially when water temperatures are below 9° C. Thus, occurrence of T. bryosalmonae in fish or in the water does not mean that a fish will develop clinical PKD and die. Fish with subclinical infections may also be able to carry the parasite indefinitely, allowing the continuous production of spores (Okamura et al. 2011).  


What is Tetracapsuloides bryosalmonae?

T. bryosalmonae is a eukaryotic myxozoan parasite in the phylum cnidaria (the same phylum that contains coral and jellyfish). It has a complex life cycle, infecting fish and bryozoans and having both endoparasitic and spore-like planktonic life stages associated with both fish and bryozoan hosts.


What is PKX?

PKD has been recognized since the early 20th century, but the disease source of PKD was a mystery until 1999 when was identified as the causative agent. Prior to its identification in 1999, the organism that caused the disease was simply referred to as the proliferative kidney organism “X”, hence PKX. However, the disease-causing agent is now referred to by its scientific name, .


What is a bryozoan?

Bryozoans are sessile, colonial, aquatic invertebrates that are found throughout the world in both marine and freshwater systems. Freshwater bryozoans (Phylactolaemates) generally experience optimal growth at temperatures between 15-28 °C (59-83 °F; Wood 2009). Phylactolaemate populations are often subject to seasonal dynamics:  a spore-like statoblast (a reproductive body comparable to a seed) affixes to a hard substrate and establishes a colony, the colony grows throughout the warm season and produces new statoblasts. The colony then dies in the cold months and the population is perpetuated with new statoblasts. However, some species are capable of overwintering as a mature colony.


Is Tetracapsuloides bryosalmonae an invasive species?

We do not know and likely will never know if . By definition, an invasive species is one that has been introduced to an environment where it is non-native, or alien, and whose introduction causes environmental or economic damage or harm to human health (Pagad et al. 2015). Prior to August 2016, PKD or T. bryosalmonae had not been documented in the Yellowstone River subbasin.   


To learn about PKD, the US Geological Survey’s Northern Rocky Mountain Science Center has partnered with the USFWS Bozeman Fish Health Center and Montana FWP to develop molecular tools for surveillance and to use these tools to describe the occurrence and distribution of in regional rivers.


Using these molecular tools, we have detected DNA in seemingly healthy salmonids in most major rivers and tributaries in western Montana (see answer to next question). We have also detected DNA in archived fish samples collected in 2012 in the Yellowstone River near Big Timber, MT. PKD was also documented in Middle Creek Reservoir (Smith River drainage) in 1990 and 1991 and in Cherry Creek (Madison River) in the 1990s(Macconnell & Peterson 1992).


These data indicate that is broadly distributed and has been in the region for decades. We do not have data to determine when first inhabited the region so cannot definitively label as a native or invasive species.


Regardless of ’s status as a native or invasive species, it critical to clean, drain and dry all gear and equipment since unintentional movement of the parasite could extend the geographical range of the disease and facilitate novel and potentially deadly combinations of fish, bryozoan, and parasite strains.



Where in Montana, other than the Yellowstone River, has PKD and/or Tetracapsuloides bryosalmonae been documented?

In Montana, PKD was first documented in cutthroat trout in Middle Creek Reservoir (Smith River drainage) in 1990 and 1991 (Macconnell & Peterson 1992) and in Cherry Creek (Madison River) in the 1996 (FWP Fish Health Lab, pers. comm.).


We used molecular DNA tools to screen fish tissue samples collected in 2016 and 2017 for presence of DNA. It is important to underscore that presence of the parasite’s DNA in fish tissue is not indicative of presence of the clinical PKD disease or fish mortality and that our findings likely do not represent the full extent of ’s distribution. Major tributaries where we have detected DNA include:


  • Big Hole R.

  • Bighorn R.

  • Blackfoot R.

  • Boulder R.

  • Clark Fork R.

  • East Gallatin R.

  • Flathead R.

  • Gallatin R.

  • Jefferson R.

  • Madison R.

  • Middle Fork of Flathead R.

  • North Fork of Flathead R.

  • Missouri R.

  • Ruby R.

  • Shields R.

  • Smith R.

  • Stillwater R.

  • Sun R.


How long has Tetracapsuloides bryosalmonae been in the area?

PKD-like disease signs in North America were first documented in the 1960s in the American River Hatchery (CA; Hedrick et al. 1985). The first verified outbreak of PKD in North America was in 1981, at an Idaho fish hatchery (Smith et al. 1984), and was subsequently detected in salmonids in California, Washington and British Columbia.  In Montana, PKD was first documented in cutthroat trout in Middle Creek Reservoir (Smith River drainage) in 1990 and 1991 (MacConnell and Peterson 1992). T.bryosalmonae was detected in Cherry Creek (tributary to the lower Madison) in 1996 (FWP Fish Health Lab, pers. comm.).


Where else has PKD resulted in fish kills and how do they compare to the Yellowstone fish kill?

The magnitude of the Yellowstone River fish kill appears to be unprecedented, as no other published report or study describes events were > 1000 fish died. Because the population size of Mountain whitefish in the Yellowstone River is not known, it is unclear what percent of this population died. Reports from fish farms and hatcheries do indicate instances with 85% - 100% mortality. In North America, PKD fish kills have been infrequently reported in technical or peer-reviewed literature. Specific examples include:


European examples:

  • 85% reduction in Atlantic salmon parr densities in the River Aelva in central Norway (Sterud et al. 2007). The number of dead fish sampled each day varied from 3 – 120.

  • Swiss lowland rivers, models indicate PKD can cause brown trout mortality of >25% (Borsuk et al. 2006)

North American examples:

  • Cutthroat trout in Middle Creek Reservoir, MT (MacConnell and Peterson 1992)

  • Rainbow trout in Hagerman Fish Hatchery, ID (Smith et al. 1984)

  • Mountain whitefish in South Fork of Snake River, ID

  • Trout in Silver Creek, ID

  • Chinook salmon in Merced & Tuolumne  rivers, CA (Foott, Stone & Nichols 2007)

    • First detected in the 1980s

    • In hatchery, 27% cumulative mortality of juvenile chinook salmon

  • Wild sockeye salmon in Vancouver Island, British Columbia (Kent et al. 1995)

  • Pink salmon in the Quinsam River, British Columbia (Braden et al. 2010)


Since Tetracapsuloides bryosalmonae occurs in other rivers in the region, why did a PKD fish die-off only occur in the Yellowstone River?

We do not know. Many regional rivers had stressful (e.g., high water temperature and low discharge) conditions similar or worse than the Yellowstone River in 2016, yet there were no documented PKD fish kills. Despite higher flows and cooler water temperatures in July, another documented PKD fish kill occurred in the Yellowstone River in 2017, though mortality was much lower. This unanticipated die-off further underscores how little we understand about PKD disease dynamics. Taken together, these results indicate that T. bryosalmonae is widely distributed and suggest that warm temperatures and low flow conditions cannot alone explain PKD-caused fish kills though these stressful conditions are clearly important to PKD.


Currently, we are collecting data to test two hypotheses that may explain why the Yellowstone River was more vulnerable to PKD. First, we are testing if variations in infection patterns across rivers are a result of different genetic types of and that genetic types in the Yellowstone River are distinct and potentially more virulent. Second, we are testing if bryozoan hosts for are more abundant in the Yellowstone River, though we do not know what past densities of bryozoans were like. We anticipate that there will not be a single answer that explains PKD fish die-offs in all places across all times.

Will a PKD fish die-off happen again in the Yellowstone River or other rivers?

We don’t know, but suspect that future die-offs are possible. 


Why did PKD primarily kill Mountain whitefish in the Yellowstone River? Why not trout?

We don’t know why PKD primarily killed Mountain whitefish in the Yellowstone River and also in the South Fork of the Snake River (ID). No previous research has assessed Mountain whitefish vulnerability to PKD.


What research is being done to prevent, predict, or control PKD?

  • Our collaborative team is currently doing research on the following topics related to PKD:

  • Testing if variations in infection and disease are a result of different genetic types of T. bryosalmonae and that the 2016 outbreak was the result of a host-specific type.

  • Validating environmental DNA (eDNA) tools for T. bryosalmonae surveillance and associate eDNA measurements with infection prevalence based on lethal fish sampling.

  • Developing and validating quantitative PCR (qPCR) methods to characterize infection status and benchmark method against histological measures of disease state.

  • Identify bryozoan species that act as hosts for T. bryosalmonae at sentinel sites in the Yellowstone River.

  • Collect temporal data at sentinel sites on the target bryozoan hosts, fish hosts, parasite eDNA and environmental parameters that will eventually feed into epidemiological models to predict PKD outbreaks.


We are also collaborating with European PKD experts (B. Okamara and H. Hartikainen) to better describe T. bryosalmonae genomics and the factors that influence the contemporary distribution and potential virulence of different T. bryosalmonae genetic types.


How can private citizens and businesses help?

  • Communicate the need for PKD research to local, state and federal lawmakers and legislative staff.

  • If you see something suspicious like dead fish, disease signs or fish behavioral changes, please contact Montana FWP.

  • Conserve water, low water causes stress to aquatic communities

  • Minimize stress to fish during any fish kills.

  • Follow aquatic invasive species prevention and containment best practices and clean, drain, and dry all gear and equipment.



  • Borsuk, M.E., Reichert, P., Peter, A., Schager, E. & Burkhardt-Holm, P. (2006) Assessing the decline of brown trout (Salmo trutta) in Swiss rivers using a Bayesian probability network. 224-244.

  • Braden, L., Prosperi‐Porta, G., Kim, E. & Jones, S. (2010) Tetracapsuloides bryosalmonae in spawning pink salmon, Oncorhynchus gorbuscha (Walbaum), in the Quinsam River, British Columbia, Canada. Journal of Fish Diseases, 33, 617-621.

  • Foott, J.S., Stone, R. & Nichols, K. (2007) Proliferative kidney disease (Tetracapsuloides bryosalmonae) in Merced River Hatchery juvenile Chinook salmon: Mortality and performance impairment in 2005 smolts. 57.

  • Kent, M., Higgins, M., Whitaker, D. & Yokoyama, H. (1995) Proliferative kidney disease and in wild-caught salmonids from the Puntledge River system, Vancouver Island, British Columbia. 13-17.

  • Macconnell, E. & Peterson, J.E. (1992) Proliferative kidney disease in feral cutthroat trout from a remote Montana reservoir: a first case. 182-187.

  • Okamura, B., Hartikainen, H., SCHMIDT‐POSTHAUS, H. & Wahli, T. (2011) Life cycle complexity, environmental change and the emerging status of salmonid proliferative kidney disease. Freshwater Biology, 56, 735-753.

  • Pagad, S., Genovesi, P., Carnevali, L., Scalera, R. & Clout, M. (2015) IUCN SSC Invasive Species Specialist Group: invasive alien species information management supporting practitioners, policy makers and decision takers.

  • Smith, C., Morrison, J., Ramsey, H. & Ferguson, H. (1984) Proliferative kidney disease: first reported outbreak in North America. 207-216.

  • Sterud, E., Forseth, T., Ugedal, O., Poppe, T.T., Jørgensen, A., Bruheim, T., Fjeldstad, H.-P. & Mo, T.A. (2007) Severe mortality in wild Atlantic salmon Salmo salar due to proliferative kidney disease (PKD) caused by (Myxozoa). 191-198.

  • Wood, T.S. (2009) Bryozoans. , pp. 437-454. Elsevier.

Collection and Filtration Field Protocol

In order to support the great research being done by the USGS and FWP, and to educate others regarding this ongoing study of PKD on the Upper Yellowstone, on May 24th, 2018 members of the Upper Yellowstone Watershed and Guides for Conservation collected 6 samples from 6 sites on the Upper Yellowstone River based on the USGS field protocol. Below is the field protocol as well as pictures and video from our experience. We thank the USGS for allowing us to participate in their important study, and to help advance the scientific knowledge of this amazing watershed. (Note: During our collection, we used a more expensive, electric peristaltic pump, rather than the described below. Once the lab analysis is done, we will report the results.)

Prior to Embarking:

Check your inventory for…

  • Field Blank bags (one with water and one unopened)

  • Sample kits (see Appendix I)

  • Bag of extra filters

  • Pump kit (see Appendix II)

  • Charged drill batteries


Step 1. Sample Site Selection

Select a location in the stream to collect the eDNA sample that has moderate flow (not a pool or eddy). eDNA sample collection should precede any activity in the stream or be positioned upstream of any other activity in-stream. Samples can be collected from a dock or levee that is protruding into the stream. Try to be consistent with sample location for subsequent samplings.

Example of labeling format for a field blank taken from Carter’s Bridge on May 1st 2018 at 9am:

20180501 CB FB 09:00

Step 2. Setup Filtration Site

  1. Put on one pair of nitrile gloves.

  2. Prepare sampling area by setting up peristaltic pump, milk crate, drill, tubing, and sample kit (FIGURE 1).

  3. To prepare peristaltic pump

    • Using the Philips bit and drill, unscrew the four bolts that hold the pump head together on the mounting board and removed the top section.

    • Position the tube in the peristaltic pump head such that the pump head is approximately in the center of the length of tubing and the inlet (with the plastic adaptor) is on the left and the outlet is on the right (FIGURE 2).

    • Replace the top section of the pump head and remount the pump head on board with the bolts.

  4. Remove the seven small whirl-paks with silica desiccant from the quart size bag and label each as follows with a permanent marker on the bottom of the whirl-pak bag if not already done so:

    • Date: YearMonthDay (YYYYMMDD)

    • Fish Access Site name: Brogan’s Landing (BL), Grey Owl (GO), Dan Bailey (DB), Mallard’s Rest (MR), Pine Creek (PC), Carter’s Bridge (CB).

    • Sample: FB, 1, 2, 3, 4, 5, 6

    • Time: HH:MM

Step 3. Filter Field Blank sample

  1. Find the filled field blank (FB) bag for the site that you are sampling.

  2. Follow the instructions from Step 5 and process this filter prior to collecting field samples.

Step 4. Collect Field Samples

  1. Take the six unopened whirl-pak bags to the sample collection location (a 5-gallon bucket is helpful here).

  2. Collect the eDNA water sample at an arm’s reach into the stream while standing on dry ground. We strongly recommend that you keep your feet and legs out of the water so that you do not have to decontaminate boots or waders.

  3. Insert open bag into water upstream of yourself and fill each bag to the marked fill line.

  4. Seal each bag by rolling the end at least three times and twisting the ends together.

  5. Place full bags in a shaded location to prevent UV degradation of the sample while filtering.

Collecting Water from the River

Step 5. Filter the Sample

  1. Remove a filter assembly from the eDNA kit and snap onto the plastic adapter at the inlet end of the tubing. The filter is pre-loaded in the filter assembly (FIGURE 3).

  2. Check that there are no visible holes, cracks, or gaps in the filter that water can easily flow through. You may need to reseat or replace the filter.

  3. Place filter assembly in holder (e.g., milk crate, flask clamp, cardboard box).

  4. Point outlet end of tubing towards the ground and away from any sampling gear.

  5. Open one of the sample bags and carefully pour 250 ml into filter funnel, measuring using the graduated lines on the filter cup.

  6. Apply the drill to the pump head (drill should turn clockwise) until the sample is filtered. 


*If glove or supplies become soiled or contaminated (e.g., drop filter cup), put on new gloves. When in doubt error on the side of caution and replace gloves or supplies.​

One Type of Peristaltic Pump

Running Water through the Filter

Step 6. Removing and preserving filter

  1. Dry the filter by continuing to pump for approximately 1 minute after all water has been passed through the filter.

  2. Unsnap the filter cup and, using new sterile tweezers for each filter, fold filter in quarters, dirty side in.

  3. Place filter in labeled desiccant whirl-pak bag.

  4. Be sure desiccant bag is labeled with date, site, sample #, and time (see Step 2).

  5. Place all desiccant bags collected at a site in the provided quart size ziplock bag and label with the site name.​

Storing Filter which Contains DNA

Step 7. Cleanup

  1. Place all used forceps, filter cups, and filter holders back into the gallon bag.

  2. Gloves and used whirl-pak bags can be thrown away.

  3. Repack the drill and pump head and proceed to the next site. 


Step 8. Temporary Storage

  1. Place samples in a dry, clean location out of sunlight. Ideally, place these samples in a cooler with ice. 

  2. Place in a refrigerator until picked up by USGS NOROCK. Both the quart-size bag with the samples and the gallon-size bag with used items should be returned.


Step 9. Pickup

  1. For pick-up, contact Patrick HutchinsCell

Appendix I: Pump Kit Contents

  • 1 pair of sterile gloves

  • 7 Filter assemblies, each with a filter cup, filter, and filter holder

  • 6 large whirl-pak bags with fill line

  • 1 filled medium whirl-pak bag

  • 1 unopened whirl-pak bag

  • 7 Dessicant bags

  • 7 Plastic forceps (only 1 pictured)

  • 1 Quart-size zip lock bag

Appendix II: Pump Kit contents

  • Peristaltic pump head on mounting board (or other alternatives)

  • 3ft length of tubing with plastic adaptor

  • If using the standard peristaltic pump, you will need a cordless drill with battery and battery charger, with 12mm socket ¼” socket adaptor and Philips drill bit (not pictured)

bottom of page