implementing invertebrate-based … stores/data libraries/files/oweb...implementing...

1

Project completion report to the Oregon Watershed Enhancement Board

IMPLEMENTING INVERTEBRATE-BASED BIOMONITORING

IN SELECTED WILLAMETTE VALLEY WETLANDS

Clean Water Services, Hillsboro OR West Hayden Island, Portland OR

Submitted to: Wendy Hudson Oregon Watershed Enhancement Board

775 Summer St., Suite 360 Salem, OR 97301

Submitted by: Celeste Mazzacano

Staff Scientist / Aquatic Program Director

The Xerces Society for Invertebrate Conservation

628 NE Broadway, suite 200

Portland, OR 97232

2

Table of Contents

Project background and summary……………………………………………………………………...…………………...…..………………..pg. 3

Methods…………………………………………………………………………………………………………………………………………………………pg. 4

Site Selection…………………………………………….……………………………………………………………………………………….pg. 4

Volunteer training and sampling………………………………………………………………………………………………………..pg. 6

Habitat Assessment……………………………………………………………………………………………………………………………pg. 6

Environmental data……………………………………………………………………………………………………………………………pg. 7

Macroinvertebrate sampling……………………………………………………………………………………………………………..pg. 7

Statistical Methods…………………………………………………………………………………………………………………………….pg. 8

Results and Discussion……………………………………………………………………………………………………………………………………pg. 9

Wetland macroinvertebrate taxa……………………………………………………………………………………………………….pg. 9

Habitat Assessment……………………………………………………………………………………………………………………………pg. 9

Environmental data………………………………………………………………………………………………………………………….pg. 10

Macroinvertebrate community structure…………………………………………………………………………………………pg. 11

Regression analysis of macroinvertebrate community attributes………………………………………………….…pg. 12

Conclusions………………………………………………………………………………………………………………………………………………..…pg. 14

References……………………………………………………………………………………………………………………………………………………pg. 15

Appendix A. Wetland Human Disturbance Assessment form……………………………………………………………………….pg. 16

Appendix B. Wetland macroinvertebrate taxa list…………………………………………………………………………………………pg. 19

3

Project background and summary

This project was undertaken to evaluate a suite of macroinvertebrate-based indicators of wetland biological

health developed previously by Xerces for wetlands in the Willamette Valley. Invertebrates are used routinely

to assess the biological health of streams (Rosenberg & Resh 1996; Karr & Chu 1999), but similar tools for

wetlands have been lacking, and the extreme variability and dynamic nature of wetland invertebrate

communities has been identified as a significant barrier to developing invertebrate indicators (Batzer 2013).

From 2007 through 2010, Xerces worked to develop an invertebrate-based biological assessment tool that could

be used reliably across wetlands in the Willamette Valley to assess wetland quality, detect responses to

anthropogenic stressors, and evaluate restoration success (OWEB, 2011). Our intent was to develop an

invertebrate-based Index of biological Integrity (I-IBI), based on macroinvertebrate community attributes

(metrics) that were significantly different between least-disturbed and most-disturbed sites, and had a sufficient

range of values within the attribute that a scoring system could be devised. Our analysis of the data collected

during these four years showed large variation in macroinvertebrate community composition at the same sites

across consecutive years, and this high level of dynamism was ultimately too great to allow a set of scored

rubrics to be developed. However, data analysis revealed two different suites of attributes that were

significantly different at most-impaired versus least-impaired riverine and flats wetlands. Our goal in this project

was to work with volunteers to conduct additional sampling at wetlands in the Willamette Valley to investigate

the ease of use and reliability of wetland sampling and habitat assessment protocols by non-professional users

and test the functionality of this suite of indicator attributes in additional wetlands. We also worked with a

variety of partners to engage in outreach about the importance of wetland habitats and their associated

invertebrate communities.

Major findings from this study include:

Most partners and volunteers who had conducted macroinvertebrate-based monitoring in streams were

unaware that completely different indicators of biological health are required for wetland monitoring.

A high level of consistency in conducting wetland macroinveretrbate sampling was seen between

volunteers and trained professionals.

Much less consistency was seen in completing the Human Disturbance Assessment (HDA) that provides

an indication of anthropogenic impairment at a site (resulting in an overall ranking of least-disturbed,

intermediate disturbance, or most-disturbed). Not only did scores vary among different people rating

the same site, indivduals who had been involved in site restoration activities in the past were more likely

to rate the wetland as experiencing lower levels of human disturbance. More detailed training will be

required to enable non-expert users to implement the HDA successfully, and the links between the land

use checklist on the HDA form and the appropriate resulting scoring should be made more explicit.

Two new taxa were added the Willamette Valley wetland invertebrate taxa database developed during

Xerces’ previous 2007-2010 wetland study: Elodes (a genus of scirtid marsh beetle) and Ptychoptera (a

genus of phantom crane fly).

A weak correlation between increased site impairment and higher conductivity and total phosphorus

was observed, consistent with the findings of Xerces’ previous wetlands research (2007-2010).

4

Analysis of similarity of macroinvertebrate community composition showed that all sites sampled by

both trained volunteers and Xerces staff were most similar to each other, indicating that volunteers

were able to conduct the wetland macroinvertebrate sampling protocol consistently and successfully.

Higher levels of anthropogenic-induced impairment appeared to be a stronger driver of similarity in

community composition among different wetlands, regardless of hydrogeomorphic type (i.e. riverine vs.

flats).

Macroinvertebrate community attributes identified previously as potential indicators of biological health

for riverine wetlands performed more robustly than those identified for flats wetlands. Six of the 12

attributes identified previously for riverine wetlands showed a strong to moderate correlation with HDA

score, while only two of the seven attributes identified previously for flats wetlands showed a

correlation with HDA score.

Nine of the 12 macroinvertebrate community attributes calculated for riverine wetlands in this study

were within the ranges expected based on their disturbance class when compared to mean attribute

values from Xerces’ work in 2007-2010. This suggests that even within the framework of extreme

annual variability in wetland invertebrate communities, these attributes may have predictive value for

wetland biological health at riverine sites. Conversely, only one of the macroinvertebrate community

attributes for flats wetland in this study had a mean value within the range seen in Xerces 2007-2011

work.

Methods

Site selection

Xerces worked with partners at the Port of Portland, Portland Environmental Services Clean Rivers Program,

Willamette Resources & Education Network, Friends of Tideman Johnson, and Columbia Slough Watershed

Council to determine suitable sites for macroinvertebrate sampling and/or wetland outreach opportunities to

the public. Sampling sites copnsisted of both riverine and flats types of wetlands, according to the

hydrogeomorphic (HGM) classification system (Brinson 1993, Adamus 2001). Sites at which volunteer-based

macroinvertebrate sampling and/or outreach and education occurred included West Eugene Wetlands

(Tsanchefin; 2 different sampling sites), Columbia Slough (Whitaker Pond), Tideman Johnson, Brookside, West

Hayden Island (two different sampling sites), Killarney, Kingfisher Marsh, and Bobcat Marsh. A map depicting

the site locations is shown in Figure 1.

5

Figure 1. Map of wetland sampling and public outreach sites.

I. Portland area

II. Eugene area

6

Volunteer Training and Sampling

Habitat assessment and macroinvertebrate sampling were done in conjunction with volunteers and/or staff

associated with site management at Tsanchefin, West Hayden, and Tideman Johnson wetlands. Trained

volunteers and Xerces staff did replicate habitat assessments and macroinvertebrate sampling at these sites to

examine consistency of results. In addition, where sufficient numbers of volunteers were available, multiple

different individuals conducted an HDA assessment of the same site, to look at consistency of disturbance rating

among volunteers.

At sites such as Killarney, Bobcat Marsh, Clean Water Services Marsh, and Kingfisher Marsh, site managers were

educated about wetland macroinvertebrates, bioassessment, and sampling techniques but did not take part in

replicate sampling, and all samples were taken solely by Xerces staff.

At Brookside and Whitaker Pond, extensive education about wetland invertebrates was conducted, including a

guided field session, but formal sampling was not done. At Whitaker Pond, Xerces staff participated in the

Columbia Slough Watershed Council’s annual Slough 101 training day, working with about 40 local residents and

high school student participants. At Brookside, we worked with BES staff during a field session of the Clean

Rivers Program for grade school children, but the number and age of the students precluded formal sampling.

Habitat Assessment

Determining the range of human-induced stressors impacting a given wetland during a site visit is challenging.

Rapid wetland assessment techniques have been developed for Oregon (ORWAP; Adamus et al. 2009), but these

require trained professionals with specialized knowledge and take several hours to complete. To make basic

wetland assessment more accessible to a variety of users, Xerces previously implemented a wetland Human

Disturbance Assessment (HDA) form, modified from a rubric developed by Gernes & Helgen for wetlands in

Minnesota (U.S. EPA 2002). HDA scores were shown previously to correlate well with the more comprehensive

ORWAP stressor scores when both analyses were performed at the same site (OWEB 2011). The HDA examines

five aspects of the site (see Appendix A for the full form):

Buffer landscape disturbance (land use within 50 ft/15 m of wetland)

Immediate landscape influence (500 ft/150 m of surrounding land)

Habitat alteration, immediate landscape (500 ft/150 m of surrounding land)

Hydrologic alteration, immediate landscape (500 ft/150 m of surrounding land)

Chemical & Sediment Inputs

Each aspect can be rated as Excellent (0 points), Moderate (5 points), Fair (10 points), or Poor (15 points), and is

accompanied by a checklist to guide the user rating, allowing notation of elements such as road density;

industrial, agricultural, or residential development; proportion of non-native plant species; logging, grazing,

construction, foot traffic and vehicle use; dams or culverts; etc. The site HDA score is calculated by summing the

rating for each section, with a lower score indicating better conditions, such that an utterly pristine site would

receive a total score of 0, and a completely disturbed site would score 75 points. Because the Chemical &

Sediment Inputs section includes nutrient levels, final scores for each site were not calculated until water

chemistry data were returned by the contracted lab (see Environmental data below).

7

Environmental data

The location of the sampling site within each wetland was photographed and recorded using a Garmin Rino 120

GPS unit (NAD 83 datum). Measurements such as turbidity, pH, and temperature were taken adjacent to the

sampling region prior to macroinvertebrate sampling, to avoid trampling or disturbing the region from which

macroinvertebrates would be netted. Water temperature, conductivity and pH were measured using a Hach

SensIon 156 multiparameter meter. Turbidity was measured using a polycarbonate transparency tube with a

Secchi disk at the bottom. Calibration of the pH and conductivity probes was checked at the beginning of each

sampling day.

Additional water samples were taken for off-site determination of total Kjeldahl nitrogen, total phosphorus, and

chloride. Samples to be analyzed for nitrogen and phosphorus were placed in acid-washed 1-liter containers

provided by the lab, and a separate sample for chloride determination was taken in a 250 mL container. All

samples were immediately placed in a cooler, and refrigerated afterwards until being delivered within 7 days to

Alexin Analytical Laboratory (Portland, OR) for analysis.

Macroinvertebrate sampling

All macroinvertebrate sampling was done during May, the standard period used by Xerces for wetland studies.

This index period was used previously because it is late enough in the spring that more macroinvertebrates are

mature enough to identify to genus and species, but early enough that there is less risk of sampling sites having

dried down to the point where water levels are too low for sampling nets to be used.

Macroinvertebrates were sampled using a D-frame dip net with 500 m mesh in the near-shore zone of

emergent vegetation, in water 1.6 – 3.2 ft. (0.5 to 1 m) deep. Sampling transects were 24-30 ft. (7-9 m) long,

and were delineated using three 4-foot cedar stakes driven into the substrate. For sites sampled by both

volunteers and Xerces staff, adjacent replicate sampling transects were established. The water depth at each

stake was measured and recorded. Two composite dip net samples were taken at each transect. Each

composite sample consisted of three sets of 1-meter sweeps taken through the top 1-3 in. (2.5-7.5 cm) of the

benthos and up through the water column on one side of each of three cedar stakes (“shore” side and “open

water” side). Thus, each composite sample was comprised of nine individual 1-meter sweeps, three sweeps

each on one “side” of each cedar stake (see Figure 2 for illustration).

8

Figure 2. Wetland macroinvertebrate sampling design. Yellow lines indicate 4-foot cedar stakes used to mark

the transect. Numbers indicate each of the nine 1-meter dipnet sweep areas that comprise a single composite

sample. The 2nd composite sample is taken in the same way on the opposite side of the stakes.

If the volume of sediment in the net bag after three consecutive sweeps was excessive, sample volume was

reduced by submerging the bottom of the net bag in the water in a region of the wetland away from the

sampling site, and stirring the contents with one hand while gently swirling and bouncing the net in the water.

This also allowed large pieces of debris to be rinsed and removed, along with any captured amphibians and fish.

All nine sweeps comprising a single composite sample were pooled in a 10-gallon bucket. Any remaining fish

and amphibians were removed and replaced in the wetland, and larger pieces of debris were rinsed and

discarded. Pooled material was poured through a sieve with 500 m mesh, and rinsed further to remove

sediment. All rinse water was poured through a 500 m mesh sieve prior to use, to avoid accidentally

introducing additional invertebrates into the sample. Sample material was transferred to 1-L Nalgene jars and

80% ethanol was added as a preservative. For maximum preservation, sample volume comprised no more than

75% of the jar, and samples that contained large amounts of filamentous algae comprised no more than 50% of

the jar volume. At the end of each day, the ethanol in each sample was poured off and replaced with fresh 80%

ethanol.

All samples were delivered to the taxonomic lab (ABR, Inc., Forest Grove, OR) for identification. Each composite

sample was randomly subsampled to a target count of 500 organisms, and “large and rare” invertebrates were

picked and identified after the target subsample number was reached. Organisms were identified to the lowest

taxonomic level possible, usually genus.

Statistical methods

The PRIMER v6 software package (Clarke & Gorley 2006) was used to examine overall invertebrate community

structure at each site and the degree of community similarity between sites and replicate samples taken by

volunteers and Xerces staff at the same site. Resemblance matrices for taxa abundance were created for sites

using square root-transformed abundance data (Bray–Curtis distance measure), and patterns in taxa

9

aggregations at each site were examined using CLUSTER analysis and non-metric multidimensional scaling

(MDS).

Xerces previous work identified a suite of macroinvertebrate community attributes that differed significantly

among least-impaired and most-impaired sites. To assess the reliability of the attributes at new sampling sites,

the value of attribute was calculated for all macroinvertebrate samples taken. These attributes included:

Mean significantly greater at most-impaired riverine sites: abundance; #, % diversity, and % abundance

of highly tolerant taxa (Modified Hilsenhoff Biotic Index (MHBI) = 8-10); # of non-insect taxa; # genera

and % diversity of (Crustacea + Mollusca); % diversity of Crustacea; % Chironomus of total Chironomidae

(a tolerant genus of non-biting midge); % abundance Sphaeriidae (pea clams)

Mean significantly greater at least-impaired riverine sites: # genera ECOT (Ephemeroptera, Coleoptera,

Odonata, and Trichoptera); # Coleoptera taxa

Mean significantly greater at most-impaired flats sites: % diversity collector/gatherers; % abundance

Chironomini (tribe of non-biting midges with many tolerant genera); % abundance Chironomus

Mean significantly greater at least-impaired flats sites: # of taxa; % diversity, and % abundance of ETSD

(Ephemeroptera, Trichoptera, Sphaeriidae, and dragonflies); % abundance Sphaeriidae

Linear regression analysis was done in Excel to assess the relationship between the above invertebrate

community attributes and site HDA score (as calculated by Xerces staff). HDA scores were also used to assign

each site to an overall class: class 1 (least-disturbed, HDA score = 0-22), intermediate disturbance (HDA score =

22.1-42), and most-disturbed (HDA = 42.1-75). The mean value of each attribute among all sites in the same

HDA class was then compared to the mean values for the same attributes calculated from Xerces 2007-2010

study to determine whether they fell within a similar range.

Results & Discussion

Wetland macroinvertebrate taxa

A total of 69 taxa were found among all the wetlands samples (see Appendix B for a complete list). Only two

had not been found previously at any wetland during the four years of Xerces sampling in the Willamette Valley,

namely Elodes (a genus of scirtid marsh beetle) and Ptychoptera (a genus of phantom crane fly), both of which

were part of the community at Tideman Johnson. The number of unique taxa at each site ranged from 20 to 31.

Overall abundance was dominated numerically by a few tolerant, cosmopolitan taxa including aquatic

earthworms (Oligochaeta), snails (Lymnaea, Physa), microcrustacea (cladocerans and copepods), and aquatic

sowbugs (Crangonyx, Hyalella).

Habitat Assessment

Xerces past work showed a high degree of consistency of site HDA scoring among multiple trained users of the

rubric (OWEB 2011). Different Xerces sampling teams scored the same wetlands across different years and,

while the HDA score was generally not identical for the same site in different years, the magnitude of the change

was small enough that the overall site classification as least-impaired, intermediate-impaired, or most-impaired

changed for only four sites of the 50 sites sampled, and those were at the upper or lower score limit for a given

10

impairment class. We had also found consistent correlation between our site HDA score and the more

comprehensive and detailed stressor score generated via ORWAP analysis, indicating that HDA assessment was

a robust and reliable tool. However, because all users from 2007-2010 had been trained Xerces staff, we

wanted to assess the ease of use and consistency of the HDA assessment among volunteers who lacked previous

experience with wetland assessment.

Our results from this project indicate that HDA assessment varies greatly among new users. Staff of agencies

that manage wetlands, such as Port of Portland, tended to be more consistent and generated assessment scores

more similar to those generated by Xerces staff at the same site, but both staff and volunteers that had been

involved in restoration activities at a site were strongly biased towards lower scoring indicating less disturbed

conditions. However, all were conscientious about using the associated checklist of site characteristics intended

to help guide users to a realistic score. These results indicate that additional or more detailed training will be

required to enable non-expert users to implement the HDA form successfully at their wetland, and that the links

between the land use checklist on the HDA form and the appropriate resulting scoring need to be strengthened

and made more explicit.

Overall, site HDA scores generated by Xerces staff ranged from 35 to 62.5; thus, all sites were either moderately

or most-impaired, with no least-impaired sites sampled.

Environmental data

All values for water chemistry parameters measured (pH, conductivity, turbidity, and nutrients) fell within the

range of minimum and maximum values seen for each in Xerces previous 2007-2010 study. The only exception

was the total phosphorus level at Tideman Johnson (2.56 mg/L), which was an order of magnitude higher than

total phosphorus at the other sites in this study and two times greater than the maximum value for total

phosphorus seen in the 2007-2010 study.

Values for pH were similar among all sites, ranging from 7.1 to 8.6, and the median value (7.6) was the same as

the median pH for all wetlands during Xerces previous work. Conductivity showed more variation, ranging from

42.1 to 245 uS/cm, with a median value of 127 uS/cm. Turbidity was low at all sites, ranging from 10 – 15.7 NTU

(nephelometric turbidity units). Nutrient values ranged widely among sites; highs and lows for chloride ranged

from a low of 1.3 mg/L (West Hayden Island) to a high of 9.6 mg/L (Killarney), with a median of 3.7 mg/L; total

Kjeldahl nitrogen had a low of 1.2 mg/L (Tsanchefin and Kingfisher Marsh) and a high of 20.9 mg/L (Tideman

Johnson), with a median of 1.8 mg/L; and total phosphorus ranged from a low of 0.09 mg/L (Tsanchefin) to a

high of 2.56 mg/L (Tideman Johnson), with a median value of 0.23 mg/L.

Xerces’ previous work in Willamette Valley wetlands did not find strong correlations between water chemistry

parameters and site impairment or macroinvertebrate community composition, and the values for conductivity,

Cl-, total Kjeldahl N, and total P varied greatly from year to year the same site. However, a weak correlation was

seen between site impairment and higher levels of N, P, and conductivity. In this study, a similar weak

correlation was seen between site impairment and higher conductivity (R2 = 0.3832) and higher total

phosphorus (R2 = 0.13), but no correlation with total nitrogen was found.

11

Macroinvertebrate community structure

CLUSTER analysis

A CLUSTER analysis was done on a Bray-Curtis similarity matrix of square root-transformed macroinvertebrate

taxa abundance data in PRIMER v6. All sites that were sampled by both trained volunteers and Xerces staff

clustered most closely with each other. Samples taken at Tideman Johnson by volunteers and Xerces staff

clustered at 80.1% similarity, while volunteer and Xerces samples taken at Tsanchefin 1 and Tsanchefin 2

clustered with each other at 72.2% and 86.7% similarity, respectively. This indicates that volunteers trained in

Xerces’ standardized wetland invertebrate sampling techniques were able to implement the sampling protocols

consistently and successfully.

The mitigation wetlands sampled in the Jackson Bottoms complex (Kingfisher Marsh, Bobcat Marsh, and Clean

Water Services wetland) and the Port of Portland West Hayden Island sites formed a separate cluster, with an

overall similarity of 58.5%, but clustering was not based solely on geographic proximity. Bobcat and Kingfisher

had a community similarity of 72.8%, but the next most similar invertebrate community was at West Hayden

Island site A, which was 61.5% similar to the Kingfisher/Bobcat cluster. The community at West Hayden Island B

was most similar to that at the Clean Water Services wetland (68.8% similarity), and this site pair shared a 58.5%

similarity with the Kingfisher/Bobcat/West Hayden A cluster.

Figure 3. Similarity of macroinvertebrate community composition among wetland sampling sites

12

MDS ordination

MDS ordination of the macroinvertebrate communities at the wetland sampling sites revealed additional

community similarities based on level of human disturbance (HDA class) and hydrogeomorphic type (riverine vs.

flats). All of the sites sampled in this project experienced substantial degrees of human impact, such that no

sites were considered minimally disturbed (HDA Class 1). Although some separation was seen between riverine

and flats sites, the level of human disturbance affected the ordination more, as both riverine and flats class 3

sites were separated from the remaining flats sites, which were all HDA class 2. This suggests that as human-

induced stressors increase, the macroinvertebrate community becomes more uniform among equally impaired

sites, regardless of hydrogeomorphic type.

Figure 4. MDS ordination of wetland sampling sites. HDA class 2 = intermediate disturbance; HDA class 3 =

most-disturbed. R = HGM-riverine wetlands, F = HGM flats wetlands.

Regression analysis of macroinvertebrate community attributes

The challenges of finding ecologically meaningful signals in wetland macroinvertebrates in the face of the

extreme variability of their habitats and community composition from year to year were recently summarized by

Batzer (2013), whose assessment of extensive research done in 14 different areas of North America was

accompanied by commentary on “…the frustrating lack of consistency in invertebrate response that occurs

within and among wetlands.” This variability has been evident in the work done by Xerces in Willamette Valley

wetlands from 2007-2010. In 2009, Xerces developed a preliminary invertebrate-based IBI consisting of six

metrics, based on analysis of data from riverine wetland reference sites sampled in 2007 and 2008. However,

this IBI was not supported by the 2009 and 2010 datasets, and final analyses of sites sampled across 2-4 years

revealed instead a suite of macroinvertebrate community attributes whose means were significantly different

between most-impaired (class 3) and least impaired (class 1) sites. Different sets of attributes were identified

13

for riverine and flats wetlands, although there was some overlap; these attributes are presented in Table 1. This

study sought to test how well these attributes performed in additional Willamette Valley wetlands.

Table 1. Macroinvertebrate community attributes that differed significantly between least- and most-impaired

natural wetlands in Xerces 2007-2010 study (*p value <0.05; **p value between 0.05 and 0.1). a MHBI = modified Hilsenhoff Biotic Index; b ECOT = Ephemeroptera, Coleoptera, Odonata, and Trichoptera; c tolerant genus

of chironomid midge; d fingernail clams; e chironomid midge tribe with many tolerant genera; f ETSD = Ephemeroptera,

Trichoptera, Sphaeriidae, and dragonflies

Riverine wetlands Flats wetlands

Attribute Mean greater at

Attribute Mean greater at

Abundance class 3* % diversity collector/gatherers Class 3*

# of highly tolerant taxa (MHBI 8-10)a class 3* % abundance Chironominie Class 3*

# of non-insect taxa class 3** # taxa ETSDf Class 1*

# genera (Crustacea + Mollusca) class 3** % diversity ETSDf Class 1*

# genera ECOTb class 1* % abundance ETSDf Class 1*

% diversity highly tolerant (MHBI 8-10) a class 3* % abundance Chironomusc Class 3*

% abundance highly tolerant (MHBI 8-10) a class 3* % abundance Sphaeriidaed Class 1*

% div. Crustacea class 3*

% diversity (Crustacea + Mollusca) class 3*

% Chironomusc of total Chironomidae class 3**

# taxa Coleoptera class 1*

% abundance Sphaeriidaed class 3*

The value of each of the above attributes was calculated for the macroinvertebrate community at each sampling

site and plotted against the site HDA score. The mean of each attribute among all sampling sites in the same

class in this study was also compared to the mean attribute values for all class 1 and class 3 sites in Xerces 2007-

2010 study, to determine whether they fell within the same range.

Overall, attributes identified previously for riverine wetland performed more robustly than those identified for

flats wetlands. Six of the 12 attributes for riverine wetlands showed a strong to moderate correlation with HDA

score (abundance, R2 = 0.61; # taxa Coleoptera, R2=0.56; # genera ECOT, R2=0.52; % Chironomus of total

Chironomidae, R2=0.51; % abundance Sphaeriidae, R2=0.35; and # non-insect taxa, R2=0.31); three showed a

weak correlation (% abundance highly tolerant (MHBI 8-10), R2=0.25; % div highly tolerant (MHBI 8-10), R2=0.11;

and % diversity (Crustacea + Mollusca), R2=011); and three were not correlated with HDA score (# highly

tolerant taxa (MHBI 8-10); # genera (Crustacea + Mollusca); and % diversity Crustacea).

In contrast, only two of the seven attributes identified previously for flats wetlands showed a correlation with

HDA score, although the relationships were relatively strong (% abundance Sphaeriidae, R2=0.62; and

% abundance Chironomini, R2=0.59).

Interestingly, nine of the 12 attributes calculated for riverine wetlands in this study, all of which had HDA scores

indicating most-disturbed (class 3) conditions, were within the ranges calculated from Xerces work in 2007-2010

14

(see Table 2 below). This suggests that even with the framework of the normal extreme annual variability in

wetland invertebrate communities, these attributes may have predictive value for wetland biological health at

riverine sites. Conversely, the means for the macroinvertebrate community attributes for flats wetland in this

study fell within the range seen in Xerces 2007-2011 work for only a single attribute, % abundance Chironomini.

Table 2. Comparison of the means for the values of macroinvertebrate community attributes from riverine sites

sampled in 2011 (all class 3) to class 1 and class 3 means for the same attributes from Xerces’ 2007-2010 study.

Attribute Mean, class 3 riverine sites, 2011

Mean, class 3 riverine sites, 2007-2010

Mean, class 1 riverine sites, 2007-2010

# highly tolerant taxa (MHBI 8-10) 14.2 + 2.6 17 + 4.7 11.5 + 4.5

# non-insect taxa 12.4 + 2.8 12.8 + 3.2 9.4 + 3.1

# genera (Crustacea + Mollusca) 8.4 + 2.9 9.3 + 2.8 6.2 + 2.7

% diversity highly tolerant (MHBI 8-10) 63.7 + 12.8 63.0 + 5.8 39.3 + 14

% abundance highly tolerant (MHBI 8-10) 93.7 + 4.9 84.9 + 9.4 48.2 + 29.5

% diversity Crustacea 23.4 + 8.9 20.7 + 6.7 12.8 + 6.4

% diversity (Crustacea + Mollusca) 37.9 + 14.7 34.1 + 4.0 22.6 + 10.2

% Chironomus of total Chironomidae 60 + 15.4 34.6 + 25.2 12.0 + 20.5

# taxa Coleoptera 0.8 + 0.8 0.75 + 0.96 2.8 + 1.8

Conclusions

Many land managers as well as stakeholders and volunteers associated with regional wetlands are

interested in wetland invertebrates and biological assessment but are unfamiliar with the

macroinvertebrate community expected to occupy in these habitats.

The wetland macroinvertebrate sampling protocol is robust, reliable, and consistent, based on high levels of

community similarity between samples taken by newly-trained volunteers and Xerces staff at the same sites.

The Human Site Disturbance Assessment (HDA) rubric used to provide a measure of the degree of

anthropogenic-induced stressors at a site should be revised to make the links between the surrounding land

uses that are rated and the corresponding score easier to understand. This may also help remediate the

observed effects of unintentional bias towards lower scoring on the part of staff or volunteers who have

been involved in restoration projects at the site.

The ecoregion-specific dataset of Willamette Valley wetland macroinvertebrates built by Xerces over four

years of previous work (2007-2010) is extensive but not entirely complete, with only two new taxa found in

this study that had not been seen in wetlands in previous years.

Macroinvertebrate community attributes identified previously by Xerces’ as having potential predictive

value for biological health at flats wetlands did not perform well in this study. Greater variation in

hydroperiod and the presence of an overall more tolerant macroinvertebrate community compared to

riverine sites may make flats wetlands less amenable to biological assessment.

15

Macroinvertebrate community attributes identified in Xerces’ previous work (2007-2010) as being

significantly different between most- and least-impaired wetlands and whose performance was tested in

this study were much more consistent for riverine wetlands than for flats. Nine of the twelve

macroinvertebrate community attributes previously identified as having predictive value for wetland

biological health in riverine sites showed correlated appropriately with disturbance level. The mean values

for the majority of attributes identified for riverine communities were also in the same ranges seen in

Xerces’ previous work.

Macroinvertebrate community attributes identified as having predictive value for biological health of

riverine wetlands in the Willamette Valley have initially performed well, although more extensive testing at

additional sites would be needed to confirm their continued performance.

References

Adamus P. R. 2001. Guidebook for the hydrogeomorphic (HGM)-based assessment of Oregon wetland and

riparian sites: statewide classification and profiles. Oregon Division of State Lands, Salem, OR.

Adamus P., J. Morlan and K.Verble. 2009. Manual for the Oregon Rapid Wetland Assessment Protocol (ORWAP),

version 2.0. Oregon Dept. of State Lands, Salem, OR.

Batzer, D. P. 2013. The seemingly intractable ecological responses of invertebrates in North American wetlands:

a review. Wetlands 33; 1-15.

Brinson M. M. 1993. A hydrogeomorphic classification of wetlands. Technical Report WRP-DE-4. U. S. Army

Corps of Engineers Waterways Experiment Station, Vicksburg, MS.

Clarke K. R. and R. N. Gorley. 2006. PRIMER-E v6 User Manual/Tutorial. Primer-E, Plymouth, UK.

Karr J. R. and E. W. Chu. 1999. Restoring life in running waters: better biological monitoring. Island Press,

Washington DC.

Oregon Watershed Enhancement Board. 2011. Oregon’s wetland monitoring and assessment program: a pilot

study. Final report to US EPA Wetland Program Development Grant Program, compiled by K. Abraham. 61 pp.

Available at http://www.oregon.gov/OWEB/MONITOR/docs/final_wetland_report.pdf.

Rosenberg, D. M. and Resh, V. H. 1996. Use of aquatic insects in biomonitoring. In: An Introduction to the

Aquatic Insects of North America, Merritt, R. W and Cummins, K. W. (eds.), Kendall/Hunt Publishing Company,

Dubuque IA, pg. 87-97.

U. S. EPA. 2002. Methods for evaluating wetland condition: developing metrics and indexes of biological

integrity. EPA-822-R-02-016, Office of Water, U. S. Environmental Protection Agency, Washington DC.

http://www.oregon.gov/OWEB/MONITOR/docs/final_wetland_report.pdf

16

Appendix A. Wetland Human Disturbance Assessment form

Site name: Date: County/City: Rated by:

Total HDA score (75 possible) =

1. Buffer landscape disturbance (land use within 50 ft/15 m of wetland): ______ points

Excellent: reference-quality; little to no evidence of disturbance in buffer (0)

Mod.: mainly undisturbed, some evidence of human use in buffer (5)

Fair: significant human influence; large proportion of buffer filled with human use (10)

Poor: intense human influence; all or almost all of buffer filled with human use (15)

Use the checklist below to guide your rating:

Excellent Moderate

Mature woodlot (>20 yr.), forested Old field, rangeland, conservation reserve

Mature prairie Restored prairie (>10 yr)

Other wetlands Young 2nd growth woodlot (<20 yr)

Other long-recovered area Shrubland

Fair Poor

Residential with unmowed areas Urban development

Active pasture/grazing Industrial development

Less intensive agriculture Intensive residential, mowed

Park turf or golf course Intensive agriculture or grazing

Newly fallowed agricultural fields Mining in/adjacent to wetland

High road density/other impervious surface Active construction activity

Comments:

Immediate landscape influence (500 ft/150 m of surrounding land): ______ points

Excellent: reference-quality; natural landscape; little/no evidence of human use (0)

Mod.: mainly undisturbed, some evidence of human use influence (5)

Fair: significant human influence; large proportion of landscape filled with human use (10)

Poor: all or most of landscape area filled with human use, isolating the wetland (15)


Excellent Moderate

Mature woodlot (>20 yr.), forested Old field, rangeland, conservation reserve

Mature prairie Restored prairie (>10 yr)

Other wetlands Young 2nd growth woodlot (<20 yr)

Other long-recovered area Shrubland

17

Fair Poor

Residential with unmowed areas Urban development

Active pasture/grazing Industrial development

Less intensive agriculture Intensive residential, mowed

Park turf or golf course Intensive agriculture or grazing

Newly fallowed agricultural fields Mining in/adjacent to wetland

High road density/other impervious surface Active construction activity

Comments:

3. Habitat alteration, immediate landscape (500 ft/150 m of surrounding land): _____ points

Excellent: reference-quality; natural landscape; no evidence of alteration (0)

Mod.: low intensity alteration or past alteration not currently affecting wetland (5)

Fair: highly altered but with some recovery from previous alterations (10)

Poor: little natural habitat present, highly altered habitat (15)


Vegetation removal/disturbances present

Excessive mowing Shrub removal

Tree plantations Woody debris removal

Tree removal/logging/clearcutting Emergent vegetation/aquatic bed removal

Low spp diversity and/or predominance of

nonnative or disturbance-tolerant native

spp

Excessive grazing/herbivory

Livestock hooves Vehicle use

Cultivation Other:

Comments:

4. Hydrologic alteration, immediate landscape (500 ft/150 m of surrounding land): _____ points

Excellent: reference-quality; natural landscape; no evidence of alteration (0)

Mod.: low intensity alteration or past alteration not currently affecting wetland (5)

Fair: current or active alteration at significant levels (10)

Poor: current or active alterations with major hydrologic disturbance (15)


18

Ditch inlet/outlet Berm, levee or dike

Tile drain Road or railroad bed

Point source input Drainage

Weir or dam Unnatural connection to other waters

Dredging Dewatering in/near wetland

Grading or filling in/near wetland Source water alteration

Other:

Comments:

5. Chemical & Sediment Inputs: ______ points

Excellent: as expected for natural site, little/no evidence of additional human-related input (0)

Mod.: inputs in low range, little/slight evidence of additional human-related input (5)

Fair: inputs in mid-range, significant evidence of additional human-related input (10)

Poor: high levels of human-related inputs, high potential for biological harm (15)


High [Cl]* High conductivity

High [total P]* Unnaturally high or low pH

High [total N]* High turbidity reading

Excessive algal growth/density Soil disturbance in immediate buffer

Eroding banks/slopes Other:

* These can only be assessed after the site visit, when the water chemistry analysis results have been received

from the lab

Comments:

19

Appendix B. Wetland macroinvertebrate taxa list. This is a cumulative list representing all taxa found at any

site sampled in 2011. Taxa in shaded rows represent organisms not found in Xerces’ previous Willamette Valley

wetland surveys (2007-2010).

Taxon

Phylum:Class or other

Order

Family

Common name

Hydra Cnidaria: Hydrozoa

Hydroida Hydridae Hydra

Turbellaria Turbellaria flatworm

Nemata Nematoda round worm

Oligochaeta Annelida: Oligochaeta

segmented worm

Erpobdellidae Annelida: Hirudinea

Erpobdellidae leech

Helobdella stagnalis Annelida: Hirudinea

Glossiphoniidae leech

Musculium Mollusca: Bivalvia

Pisidiidae pea clam

Pisidiidae Mollusca: Bivalvia

Pisidiidae pea clam

Lymnaea Mollusca: Gastropoda

Lymnaeidae pond snail

Physa Mollusca: Gastropoda

Physidae tadpole snail

Gyraulus Mollusca: Gastropoda

Planorbidae ramshorn snail

Helisoma trivolvis Mollusca: Gastropoda


Promenetus exacuous Mollusca: Gastropoda


Chydoridae Crustacea Cladocera Chydoridae waterflea

Ostracoda Crustacea Ostracoda seed shrimp

Copepoda Crustacea Copepoda copepod

Crangonyx Arthropoda: Crustacea

Amphipoda Crangonyctidae scuds

Hyalella Arthropoda: Crustacea

Amphipoda Hyalellidae scuds

Caecidotea occidentalis Arthropoda: Crustacea

Isopoda Asellidae aquatic sow bugs

Oribatida Arthropoda: Arachnida

Sarcoptiformes aquatic mite

Trombidiformes Arthropoda: Arachnida

Trombidiformes aquatic mite

Sympetrum Arthropoda: Insecta

Odonata Libellulidae skimmer dragonfly

Coenagrion/Enallagma Arthropoda: Insecta

Odonata Coenagrionidae pond damselfly

Ischnura Arthropoda: Insecta

Odonata Coenagrionidae pond damselfly

Lestes Arthropoda: Insecta

Odonata Lestidae spreadwing damselfly

Callibaetis Arthropoda: Ephemeroptera Baetidae small minnow

20

Insecta mayfly

Belostoma Arthropoda: Insecta

Heteroptera Belostomatidae giant water bug

Corixidae Arthropoda: Insecta

Heteroptera Corixidae water boatman

Buenoa Arthropoda: Insecta

Heteroptera Notonectidae backswimmer

Notonecta Arthropoda: Insecta

Heteroptera Notonectidae backswimmer

Microvelia Arthropoda: Insecta

Heteroptera Veliidae shortlegged strider

Lepidostoma Arthropoda: Insecta

Trichoptera Lepidostomatidae casemaking caddisfly

Limnephilus Arthropoda: Insecta

Trichoptera Limnephilidae northern caddisfly

Agabus Arthropoda: Insecta

Coleoptera Dytiscidae predaceous diving beetle

Dytiscus Arthropoda: Insecta


Hydroporus Arthropoda: Insecta


Hygrotus Arthropoda: Insecta


Neoporus Arthropoda: Insecta


Rhantus Arthropoda: Insecta


Haliplus Arthropoda: Insecta

Coleoptera Haliplidae crawling water beetle

Hydraena Arthropoda: Insecta

Coleoptera Hydraenidae minute moss beetle

Berosus Arthropoda: Insecta

Coleoptera Hydrophilidae water scavenger beetle

Tropisternus Arthropoda: Insecta

Coleoptera Hydrophilidae water scavenger beetle

Elodes Arthropoda: Insecta

Coleoptera Scirtidae marsh beetle

Ceratopogoninae Arthropoda: Insecta

Diptera Ceratopogonidae biting midge

Dasyhelea Arthropoda: Insecta

Diptera Ceratopogonidae biting midge

Chaoborus Arthropoda: Insecta

Diptera Chaoboridae phantom midge

Dixella Arthropoda: Insecta

Diptera Dixidae dixid midge

Ptychoptera Arthropoda: Insecta

Diptera Ptychopteridae phantom crane fly

Sciomyzidae Arthropoda: Insecta

Diptera Sciomyzidae marsh fly

Simulium Arthropoda: Insecta

Diptera Simuliidae black fly

Odontomyia Arthropoda: Insecta

Diptera Stratiomyidae soldier fly

21

Pilaria Arthropoda: Insecta

Diptera Tipulidae crane fly

Chironomidae pupae Arthropoda: Insecta

Diptera Chironomidae nonbiting midge

Ablabesmyia Arthropoda: Insecta


Chironomus Arthropoda: Insecta


Corynoneura Arthropoda: Insecta


Cricotopus Arthropoda: Insecta


Dicrotendipes Arthropoda: Insecta


Limnophyes Arthropoda: Insecta


Metriocnemus Arthropoda: Insecta


Micropsectra Arthropoda: Insecta


Microtendipes Arthropoda: Insecta


Orthocladius Complex Arthropoda: Insecta


Paratanytarsus Arthropoda: Insecta


Paratendipes Arthropoda: Insecta


Procladius Arthropoda: Insecta


Prodiamesa Arthropoda: Insecta


Psectrocladius Arthropoda: Insecta


Pseudochironomus Arthropoda: Insecta


Tanytarsus Arthropoda: Insecta


Tribelos Arthropoda: Insecta


implementing invertebrate-based … stores/data libraries/files/oweb...implementing...

Documents