Adopted by the NAACC Steering Committee November 10, 2015
INTRODUCTION
The North Atlantic Aquatic Connectivity Collaborative (NAACC) was launched in 2015 with a rapid assessment protocol for evaluating aquatic passability at road‐stream crossings and an online database (https://www.streamcontinuity.org/cdb2) for storing and scoring data collected using this protocol. Two scoring systems are used to evaluate aquatic passability at non-tidal road‐stream crossings. The first is a coarse screen for use in classifying crossings into one of three categories: “Full AOP” (Aquatic Organism Passage), “Partial AOP,” and “No AOP.” The second system is an algorithm for computing an aquatic passability score, ranging from 0 (low) to 1 (high), for each road‐stream crossing. These two scoring systems are not particular to any taxonomic or functional group but instead seek to evaluate passability for the full range of aquatic organisms likely to be found in rivers and streams.
NAACC COARSE SCREEN
Table 1 below identifies characteristics and conditions that allow crossings to be classified as providing “Full AOP,” “Reduced AOP,” or “No AOP.”
| Metric | Flow Condition | Crossing Classification | ||
|---|---|---|---|---|
|
Full AOP (if all are true) |
Reduced AOP (if any are true) |
No AOP (if any are true) |
||
| Inlet Grade | At stream grade | inlet drop or perched | ||
| Outlet Grade | At stream grade | Cascade, Free Fall onto cascade | ||
| Outlet Drop to water surface | =0 | >=1 ft | ||
| Outlet Drop to water surface/Outlet Drop to stream bottom | >0.5 | |||
| Inlet or Outlet water depth | Typical -low | >0.3 ft | <0.3 ft w/Outlet Drop to water surface >0 | |
| Moderate | >0.4 ft | <0.4 ft w/Outlet Drop to water surface >0 | ||
| Structure substrate matches stream | 100% | <100% | ||
| Physical barrier severity | none | minor or moderate | severe | |
The primary objective of the coarse screen is to identify those crossings that are likely to be a barrier to most or all species and those that are likely to provide something close to full aquatic organism passage. If it is necessary to get a better feel for how bad those crossing are that are labeled as “reduced AOP” one can use the numeric scoring system.
NAACC NUMERIC SCORING SYSTEM
The numeric scoring algorithm is based on the opinions of experts who decided both the relative importance of all the available predictors of passability as well as a way to score each predictor. Scoring involves three steps: (1) generating a component score for each predictor variable, (2) combining these predictions with a weighted average to generate a composite score for the crossing, and (3) assigning a final score based on the minimum of the composite score or the component score for the outlet drop variable.
Variables Used
Crossing assessments are generally done during “typical low‐flow conditions.” Some variables are important for assessing conditions at the time of the survey; others provide indirect evidence of likely conditions at higher flows.
- Inlet Grade: The position of the structure invert relative to the stream bottom at the inlet.
- Outlet Drop: Outlet drop is based on the variable Outlet Drop to Water Surface unless the value for Water Depth Matches Stream = “Dry” in which case outlet drop is based on the variable Outlet Drop to Stream Bottom.
- Physical Barriers: This variable covers a wide variety of circumstances ranging from obstructions to dewatered culverts or bridge cells that represent physical barriers to aquatic organism passage.
- Constriction: The relative width of the crossing compared to the width of the stream. “Severe” = <50%, “Moderate” = 50‐100%; other options include “Spans Only Bankfull/Active Channel” and “Spans Full Channel & Banks.” Constriction is an indirect indicator of potential velocity issues at higher flows.
- Water Depth: Water depth in the structure relative to water depths found in the natural channel at the time of survey.
- Water Velocity: Water velocity in the structure relative to water velocities found in the natural channel at the time of survey.
- Scour Pool: Presence/absence of a scour pool at the crossing outlet and size relative to the natural stream channel. Scour Pool is an indirect indicator of potential velocity issues at higher flows. Scour pool is included solely as an indicator of velocities at higher flows. It is not based on the effects of the pool itself which can actually be positive for fish passage.
- Substrate Matches Stream: An assessment of whether the substrate in the structure matches the substrate in the natural stream channel. Substrate Matches Stream is used to evaluate how a discontinuity in substrate might inhibit passage for species that either use substrate as the medium for travel (e.g., mussels) or require certain types of substrate for cover during movements (e.g., crayfish, salamanders, juvenile fish).
- Substrate Coverage: Degree to which a crossing structure is covered by substrate. Substrate Coverage is directly related to passability for some aquatic species that require substrate or that tend to avoid areas that lack cover. It is also an important element of roughness that can create areas of low‐velocity water (boundary layers) utilized by weak‐swimming organisms. Substrate Coverage is also an indirect indicator of potential velocity issues at higher flows.
- Openness: Cross‐sectional area of the structure opening divided by the structure length (distance between inlet and outlet) measured in feet. Openness is calculated for both the inlet and outlet and the lower value is assigned to the structure. If there are multiple structures at a crossing the value for the structure with the highest Openness is assigned to the crossing as a whole. Turtles are believed to be affected by the Openness of a crossing structure; other species may be affected as well.
- Height: Maximum height of the crossing structure. This variable is parameterized so that it only comes into play for very small structures.
- Outlet Armoring: Presence/absence of streambed armoring (e.g., riprap, asphalt, concrete) at the outlet and the relative amount of armoring. Armoring is considered “extensive” if the length (upstream to downstream) of the streambed that is armored is greater or equal to half the bankfull width of the natural stream channel. Outlet Armoring is an indirect indicator of potential velocity issues at higher flows.
- Internal Structures: Presence/absence of structures inside a culvert or bridge (e.g. weirs, baffles, supports). The Internal Structures variable is used in the scoring algorithm as it relates to the potential for creating turbulence within a crossing structure. To the extent that Internal Structures physically block the movement of aquatic organisms it is covered by the Physical Barriers variable.
Step 1: Component Scores
The component scores are not meant to equate to passability. In each case the component score is intended the cover the full range of problems (assessable by our protocol) associated with that variable: from 0 (worst case) to 1 (best case). For inlet grade, having an inlet drop or perched inlet is the worst case among the options, thus they score "0." This is not meant to say that all structures with inlet drops are impassible. The effect of inlet grade on passability scores is controlled by the weight it is given in computing the composite score (see Step 2 below).
Scoring categorical predictors is simply a matter of assigning a score for each possible category. Table 2 lists all of the categorical predictors and the scores associated with each category.
Scoring continuous predictors requires a function to convert the predictor to a score. There are three continuous predictors and three associated functions. The functional forms used were chosen because they have shapes desired by the expert team or because they fit the series of points specified by the expert team. Appendix A includes the r code defining each of these functions (“x” is the measured value for each variable).
The scoring equation for Openness is:
(1) So = a(1-e-kx(1-d ))1/(1-d)
Where So is the score for openness, a=1, k=15, and d = 0.62
when openness is recorded in feet.
The equation for Height is:
(2) Sh = min {ax2/(b2 +x2) ,1}
Where Sh is the component score for height, a = 1.1, and b=2.2
when height is recorded in feet.
The equation for Outlet Drop is:
(3) Sod =1- ax2/(b2 + x2)
Where Sod is the Outlet Drop component score, a=1.029412, and b=0.51449575
when outlet drop is recorded in feet.
| Parameter | level | score |
|---|---|---|
| Constriction | severe | 0 |
| moderate | 0.5 | |
| spans only bankfull/active channel | 0.9 | |
| spans full channel and banks | 1 | |
| Inlet grade | at stream grade | 1 |
| inlet drop | 0 | |
| perched | 0 | |
| clogged/collapsed/submerged | 1 | |
| unknown | 1 | |
| Internal structures | none | 1 |
| baffles/weirs | 0 | |
| supports | 0.8 | |
| other | 1 | |
| Outlet armoring | extensive | 0 |
| not extensive | 0.5 | |
| none | 1 | |
| Physical barriers | none | 1 |
| minor | 0.8 | |
| moderate | 0.5 | |
| severe | 0 | |
| Scour pool | large | 0 |
| small | 0.8 | |
| none | 1 | |
| Substrate coverage | none | 0 |
| 25% | 0.3 | |
| 50% | 0.5 | |
| 75% | 0.7 | |
| 100% | 1 | |
| Substrate matches stream | none | 0 |
| not appropriate | 0.25 | |
| contrasting | 0.75 | |
| comparable | 1 | |
| Water depth | no (significantly deeper) | 0.5 |
| no (significantly shallower) | 0 | |
| yes (comparable) | 1 | |
| dry (stream also dry) | 1 | |
| Water velocity | no (significantly faster) | 0 |
| no (significantly slower) | 0.5 | |
| yes (comparable) | 1 | |
| dry (stream also dry) | 1 |
Some notes about the component scores
- The option "clogged/collapsed/submerged" for inlet grade is an option surveyors use to indicate that it was not possible to measure the structure's dimensions. If the inlet is clogged or collapsed enough to affect passability it will be covered under physical barriers. This is why it receives a "1" instead of a "0", because problems associated with this option are covered by the physical barriers variable.
- The rationale for giving a component score of "1" to "unknown" for inlet grade is similar to that for "clogged/collapsed/submerged." It is hard to know how to interpret "unknown." However, if conditions at the inlet are creating a physical barrier to passage it will be covered under physical barriers.
- We included inlet grade as a variable in addition to physical barriers because inlet drops create both velocity and physical barrier (jump barrier) issues. The physical barrier issues are covered by the physical barriers variable. The inlet grade variable captures the velocity issues at the inlet. Perched inlets can create depth issues at low flows (if water can't get into the structure inlet). These may not be apparent at the time of the survey. Thus, the presence of a perched inlet is a concern even if it doesn't represent a physical barrier ("dry") at the time when the survey is conducted.
- The variable internal structures is included to account for turbulence issues. There is likely to be turbulence associated with weirs and baffles when these are included inside crossing structures. If they also create physical barriers they will be covered by the physical barriers variable. They are often included in structures to help aquatic organism passage but they sometimes do more harm than good and may be good for some species while creating problems for others. The inclusion of well‐designed weirs or baffles is likely to improve the component scores for water depth and water velocity. They get docked a little in our scoring system for introducing turbulence.
- It is difficult to know how to score the "other" option under internal structures because it is difficult to know what, if any, impact these other structures will have on turbulence. If, however, they represent a physical barrier they will be covered under the physical barriers variable.
Step 2: Weighted Composite Scores
An expert team of nine people provided input on how the variables should be weighted based on best professional judgement. The weights used with the component scores are listed in table 3. The weights are simply the means of the nine weights for each variable provided by the experts. We display the weights out to three decimal places not to suggest that we know the weights to this level of precision but to reduce overall error in the model by not introducing an additional source of error (rounding error). The composite score is the sum of the products of each component score and its weight.
| parameter | weight |
|---|---|
|
Outlet drop |
0.161 |
|
Physical barriers |
0.135 |
|
Constriction |
0.090 |
|
Inlet grade |
0.088 |
|
Water depth |
0.082 |
|
Water velocity |
0.080 |
|
Scour pool |
0.071 |
|
Substrate matches stream |
0.070 |
|
Substrate coverage |
0.057 |
|
Openness |
0.052 |
|
Height |
0.045 |
|
Outlet armoring |
0.037 |
|
Internal structures |
0.032 |
Step 3: Final Aquatic Passability Score
The final Aquatic Passability Score is the lower of either the composite score or the Outlet Drop component score. The rationale for this is that although many factors can affect aquatic organism passage, when an outlet drop is above a certain size it becomes the predominant factor that determines passability.
Aquatic Passability Score = Min[Composite Score, Outlet Drop score]
Mapping Aquatic Passability Scores
For mapping purposes, we assigned narrative descriptors for different ranges of aquatic passability as follows.
|
Descriptor |
Aquatic Passability Score(s) |
|---|---|
|
No barrier |
1.0 |
|
Insignificant barrier |
0.80 – 0.99 |
|
Minor barrier |
0.60 – 0.79 |
|
Moderate barrier |
0.40 – 0.59 |
|
Significant barrier |
0.20 – 0.39 |
|
Severe barrier |
0.00 – 0.19 |
People often ask about the relationship between these categories and actual passability for fish and other aquatic organisms. At this point the relationship is unknown and we regard it as a fruitful area for future research. The concept of aquatic passability is complicated and includes: variation in the swimming and leaping abilities of individuals within a species (what proportion of the population can pass), variability in passage requirements for a broad diversity of species that inhabit rivers and streams (what proportion of species can pass), and the timing of passability (for what proportion of the year is the structure passable).
For now, the best way to consider the aquatic passability scores is that they represent the degree to which crossings deviate from an ideal. We assume that those crossings that are very close to the ideal (scores > 0.6) will present only a minor or insignificant barrier to aquatic organisms. Those structures that are farthest from the ideal (scores < 0.4) are likely to be either significant or severe barriers. These are, however, arbitrary distinctions imposed on a continuous scoring system and should be used with that in mind.
APPENDIX A ‐ R code for continuous scoring functions.
#‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐#
# define function for Openness score calculation #‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐#
calc.openness.score <‐ function(x){
# Using von Bertalanffy functional form (Bolker pg 97)
a = 1
k = 15
d=0.62
return(a * (1‐exp(‐k*(1‐d)*x))^(1/(1‐d)))
# note exp is based on e not 10.
}
#‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐#
# Define Function for Calculating Height Scores #‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐#
calc.height.score <‐ function(x){
a<‐ 1.1
b<‐ 2.2
# Use Holling Type II function (Bolker pg 92):
result <‐ a*x^2/(b^2 + x^2)
result[result > 1] <‐ 1 # Truncate results to 1
return(result)
}
#‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐#
# Define Function for Calculating Outlet Drop Scores #‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐#
calc.outlet.drop.score <‐ function(x){
a<‐ 1.029412
b<‐ 0.51449575
score <‐ 1 ‐ a*x^2/(b^2 + x^2)
score[x > 36] <‐ 0
return(score)
}