i
ASSESSING PERSISTENCE OF TWO RARE DARTER SPECIES  
USING POPULATION VIABILITY ANALYSIS MODELS 
 
Except where reference is made to the work of others, the work described in this thesis 
is my own or was done in collaboration with my advisory committee.  
This thesis does not include proprietary or classified information. 
 
 
______________________________________ 
Wendi Winter Hartup 
 
 
Certificate of Approval: 
 
___________________________ ___________________________ 
Carol E. Johnston, Chair J. Barry Grand 
Associate Professor Associate Professor 
Fisheries and Allied Aquacultures Forestry and Wildlife Sciences 
 
 
 
___________________________ ___________________________ 
Craig Guyer  Stephen L. McFarland 
Professor   Acting Dean 
Biological Sciences Graduate School
 ii
 
ASSESSING PERSISTENCE OF TWO RARE DARTER SPECIES 
USING POPULATION VIABILITY ANALYSIS MODELS 
 
Wendi Winter Hartup 
 
 
 
A Thesis 
Submitted to 
the Graduate Faculty of 
Auburn University  
in Partial Fulfillment of the  
Requirements for the  
Degree of 
Master of Science 
 
 
 
Auburn, Alabama 
August 8, 2005 
 iii
ASSESSING PERSISTENCE OF TWO RARE DARTER SPECIES  
USING POPULATION VIABILITY ANALYSIS MODELS 
 
 
 
Wendi Winter Hartup 
Permission is granted to Auburn University to make copies of this thesis 
 at its discretion, upon request of individuals or institutions and at their expense.  
 The author reserves all publication rights. 
 
 
 
 
     
   ______________________________ 
    Signature of Author 
 
 
      
  ______________________________ 
    Date 
 
 
 
 
Copy sent to: 
 
Name _______________________________________ Date_____________ 
 
 iv
VITA 
 Wendi Winter Hartup, the daughter of Richard Lee and Miriam (Werner) Winter, was 
born July 27, 1974 in Emporia, Kansas.  She graduated from Emporia High School in 
1992.  She attended Cottey College, Nevada, Missouri, for two years and graduated with 
an Associates of Art degree in Marine Biology in August 1994.  She then attended Troy 
State University, Troy, Alabama in 1994 and graduated with her Bachelor of Science 
degree in Marine Biology and a Minor in Mathematics in November 1996.  In 1997, she 
began working for the Alabama Water Watch program, based in the Department of 
Fisheries and Allied Aquacultures of Auburn University.  She married Chad Allen 
Hartup, son of Larry and Verlene (Howell) Hartup, on April 22, 2000.  While working 
full-time, she began graduate school in the Department of Fisheries and Allied 
Aquacultures of Auburn University in January, 2001.  
 v
THESIS ABSTRACT 
ASSESSING PERSISTENCE OF TWO RARE DARTER SPECIES 
USING POPULATION VIABILITY ANALYSIS MODELS 
 
Wendi Winter Hartup 
Master of Science, August 8, 2005 
(B. S., Troy State University, 1996) 
(A.A., Cottey College, 1994) 
 
86 Typed Pages 
Directed by Carol E. Johnston 
 The objective of this study was to assess whether results predicted from population 
viability analysis (PVA) models resemble historical and current presence/absence data for 
Etheostoma boschungi and Etheostoma brevirostrum.  Etheostoma boschungi is a 
migratory species and has two distinct, but adjacent habitats: non-breeding and breeding.  
Etheostoma boschungi spawn from late February to late March and are known from 30 
sites in tributaries of the Tennessee River drainage of Alabama and Tennessee.  
Etheostoma brevirostrum prefers rocky runs and riffles with a fast current and spawn 
from early April to late May.  Etheostoma brevirostrum is known to occur in disjunct 
populations located in tributaries of the Coosa River drainage and are believed to live 
three years.  We used information on fecundity, life span, and population size to 
determine fertility and survival vital rates for a three-stage, pre-breeding, PVA model for
 vi
each species.  The age-structured model used Leslie projection matrices to calculate 
deterministic population growth rates (?) and relative elasticities of the vital rates.   
 The population size for E. boschungi decreases when adult fertility rates are below 
0.896.  The population size for E. boschungi falls below one individual after five years 
when ? = 0.309 for a one-batch fecundity and after five years when ? = 0.322 for a two-
batch fecundity.  The E. boschungi population size increases to more than 1,000 
individuals after ten years when ? = 1.158 and after nine years when ? = 1.184 for one-
batch and two-batch fecundities, respectively.  The population size for E. brevirostrum 
decreases when adult fertility rates are below 0.737, 0.929, and 0.818 for the upstream, 
downstream, and combined populations, respectively.  Etheostoma brevirostrum 
population size projected for upstream, downstream, and combined stream segments 
(using E. coosae and E. pyrrhogaster survivals) resulted in more than 50 times the initial 
population size after one year.  As long as ? < 1, then the population will decrease but 
when ? > 1, then the population increases.  The elasticity analysis of the matrix indicated 
the fertilities made the largest relative contribution to ?, while the survivals were smaller 
for all sites and for both species. 
 Despite some of the limitations in gathering data for rare species, we believe PVA 
models are useful for studying fishes.  State agencies should focus on improving the 
habitat within each of the watersheds to improve survival of the juveniles to year one, in 
addition to monitoring the watershed more closely for possible land use or pollution 
impacts. 
 vii
ACKNOWLEDGEMENTS 
 I would like to thank each of my committee members and my advisor for their 
guidance in my graduate career, their contributions to this thesis, and their understanding 
of my challenges with working full-time while attending graduate school.  My friend and 
advisor, Dr. Carol Johnston, helped me to broaden my knowledge of conservation and 
assisted with many of the theoretical aspects of this document.  Dr. Barry Grand provided 
assistance with the construction of matrix models.  The United States Fish and Wildlife 
Service and the Alabama Department of Conservation and Natural Resources supported 
research for this project. 
  Special thanks go to the staff of Alabama Water Watch for their support and the 
following people who provided invaluable field assistance: Ben Beck, Jeff Garner, Dr. 
Carol Johnston, Kevin Kleiner, Steven Harrington, Chad Hartup (my husband), Andrew 
Henderson, Stuart McGregor, Bryan Phillips, and Catherine (Nordfelt) Phillips.  I 
dedicate this thesis to my husband, my family, and extended family for their never-
ending support and encouragement.  
  
 viii
Style manual or journal used: Copeia 
Computer software used: Windows Microsoft Word 2002, Windows Microsoft Excel 
2002, PopTools version 2.5.9., MapInfo Professional 6.0 for Windows 
 
 ix
TABLE OF CONTENTS 
    Page 
LIST OF TABLES................................................................................................................x 
 
LIST OF FIGURES ............................................................................................................xii 
 
INTRODUCTION.................................................................................................................1 
 
Objectives....................................................................................................................7 
 
METHODS ........................................................................................................................9 
 
Study species ...............................................................................................................9 
 
Threats to each species ..............................................................................................11 
 
Field data collection..................................................................................................12 
 
Laboratory procedures and data collection................................................................14 
 
Analysis of data .........................................................................................................17 
 
The age-based matrix model ......................................................................................18 
 
RESULTS ........................................................................................................................21 
 
Field Data .................................................................................................................21 
 
Laboratory Data........................................................................................................24 
 
Vital rates..................................................................................................................25 
 
DISCUSSION ...................................................................................................................31 
 
LITERATURE CITED ........................................................................................................37 
 
APPENDIX I: DESCRIPTION AND MAP OF SAMPLING SITES FOR EACH SPECIES.....................70 
 
 
 
 
 x
LIST OF TABLES 
    Page 
TABLE 1.  MARK-RECAPTURE SAMPLING RESULTS FROM THE 2001 AND 2002 BREEDING 
SEASONS OF Etheostoma boschungi.  Confidence limits in parentheses are calculated 
with Poisson distribution..........................................................................................48 
 
TABLE 2.  MARK-RECAPTURE RESULTS FOR FEMALE Etheostoma boschungi DATA IN 2001 
AND 2002.  n
0
 = juveniles, n
1
 = one-year-old adults, n
2
 = two-year-old adults, n
3
 = 
three-year-old adults, ? = estimated population size per age class with 95 percent 
confidence limits in parentheses, mean size in mm SL is followed by ? 1 standard 
deviation (SD) with range in parentheses. ................................................................49 
 
TABLE 3.  SUMMARY OF 2002 TRANSECT DATA, RIFFLE VOLUME, AND STREAM LENGTH 
ON SHOAL CREEK.  Length, width and depth from transects were measured in meters 
and stream segments were measured with mapping software.  US = upstream of 
Highrock Lake, DS = downstream of Highrock Lake, Combo = upstream and 
downstream of Highrock Lake combined, SL = standard length, SD = standard 
deviation..................................................................................................................50 
 
TABLE 4.  NUMBER OF Etheostoma brevirostrum COLLECTED IN THE FIELD, THE 
ESTIMATED NUMBER OF FISH PER M? AND THE EXTRAPOLATED POPULATION SIZE FOR 
EACH AGE CLASS.  Values in parentheses are 95 percent confidence limits.  US = 
upstream of Highrock Lake, DS = downstream of Highrock Lake, Combo = upstream 
and downstream of Highrock Lake combined, n
1
 = one-year-old adults, n
2
 = two-
year-old adults, n
3
 = three-year-old adults, ? = extrapolated population size per age 
class.........................................................................................................................51 
 
TABLE 5.  LIFE HISTORY TABLE FOR Etheostoma boschungi.  This table uses one-batch 
fecundity.  All parameters are the same with the two-batch fecundity.  See text for 
sources and definitions.............................................................................................52 
 
TABLE 6.  SPECIES CLOSELY RELATED TO Etheostoma boschungi.  P
i
 = survival rates for 
each age class, f
i
 = mature ova or fecundity, % = assumption that each age class was 
collected in proportion to its relative number in the population, that the population 
was neither increasing nor decreasing, and that the number of juveniles entering the 
population each year was constant. ..........................................................................53 
 
TABLE 7.  ESTIMATED ANNUAL SURVIVAL RATES FROM FAST WITH Etheostoma 
boschungi  DATA.  F
i
 = adult fertility rates and P
0
 = juvenile survival rate (derived 
from one-batch and two-batch fecundities)...............................................................54 
 
 xi
TABLE 8.  ESTIMATED POPULATION SIZE PROJECTIONS FOR Etheostoma boschungi.  F
i
 = 
adult fertility rates, f
i
 = one-batch (92 ova) and two-batch (197 ova) fecundity, P
i
 = 
survival rates, ? = population growth rate, and ? = total population size...................55 
 
TABLE 9.  LIFE HISTORY TABLE FOR Etheostoma brevirostrum.  This table uses two-batch 
fecundity.  US = upstream of Highrock Lake, DS = downstream of Highrock Lake, 
Combo = upstream and downstream of Highrock Lake combined.  See text for 
sources and definitions.............................................................................................56 
 
TABLE 10.  ELASTICITY VALUES FOR Etheostoma brevirostrum.  See text for sources and 
definitions................................................................................................................57 
 
TABLE 11.  ESTIMATED ANNUAL SURVIVAL RATES FROM FAST WITH Etheostoma 
brevirostrum  DATA.  F
i
 = adult fertility rates and P
0
 = juvenile survival rate (derived 
from two-batch fecundities). ....................................................................................58 
 
TABLE 12.  SPECIES CLOSELY RELATED TO Etheostoma brevirostrum.  P
i
 = survival rates 
for each age class, f
i
 = mature ova or fecundity, % = assumption that each age class 
was collected in proportion to its relative number in the population, that the 
population was neither increasing nor decreasing, and that the number of juveniles 
entering the population each year was constant. .......................................................59 
 
TABLE 13.  ESTIMATED FERTILITY RATES AND GROWTH RATE FROM CLOSELY RELATED 
SPECIES WITH Etheostoma brevirostrum.  P
0
 = juvenile survival, F
i
 = estimated adult 
fertility rate, ? = growth rate, SAD = stable age distribution and Reprod = 
reproductive values..................................................................................................60 
 
TABLE 14.  ESTIMATED POPULATION SIZE PROJECTIONS FOR Etheostoma brevirostrum.  F
i
 
= adult fertility rates, f
i
 = two-batch fecundity, P
i
 = survival rates, ? = population 
growth rate, and ? = total population size. ...............................................................61 
 
 
 xii
LIST OF FIGURES 
 Page 
Fig. 1.  Historical locations (circles) and sampling site (star) of Etheostoma boschungi 
within tributaries of the Tennessee River drainage, USA. Letters on the map 
correspond to the following: a) South Fork Buffalo River and Chief Creek of Buffalo 
River, b) Shoal Creek, c) Cypress Creek, d) Swan Creek, and e) Flint River. ...........62 
 
Fig. 2.  Historical locations (circles) and sampling sites (star) of Etheostoma brevirostrum 
within tributaries of the Coosa River drainage, USA.  Letters on the map correspond 
to the following: a) Conasauga River, b) Coosawattee River, c) Etowah River, and d) 
Choccolocco Creek..................................................................................................63 
 
Fig. 3.  Map of the Dodd Site, Middle Cypress Creek on Dodd Road, Wayne County, 
Tennessee (S = small seepage stream, M = area where the larger seepage stream joins 
with Middle Cypress Creek) (TERRASERVER-USA: An online provider of USGS 
digital maps [web application], http://terraservice.net/, 2004, unpubl.). ....................64 
 
Fig. 4.  Sampling locations (circles) of E. brevirostrum within Shoal Creek of the Coosa 
River drainage in Alabama, USA. ............................................................................65 
 
Fig. 5.  A three-stage, birth-pulsed, pre-breeding model for adult populations of E. 
boschungi and E. brevirostrum.  Circles denote stages in the age-structured model and 
arrows indicate transition probabilities.....................................................................66 
 
Fig. 6.  Length-frequency distribution of collected Etheostoma boschungi from selected 
years.  A = University of Alabama museum specimens, B = (McGregor and Shepard, 
1995) C = 2001-2002 census data. ...........................................................................67 
 
Fig. 7. Length-frequency distribution of collected Etheostoma brevirostrum from 2001 
and 2002 in Shoal Creek. .........................................................................................68 
 
Fig. 8. Example of projected population size per age class of Etheostoma boschungi when 
P
0
 = 0.004, P
1
 = 0.074, P
2
 = 0.571, P
3
 = 0, and F
i
 = 0.184, using a one-batch 
fecundity..................................................................................................................69 
 
  
 
 1
INTRODUCTION 
 Many populations of fishes appear to be declining at rates warranting their protection.  
These declines could be due to factors that occur from natural disasters or anthropogenic 
activities.  Natural factors, such as floods, drought and unseasonable water temperatures 
(either too cold or too warm), can displace fishes, increasing the chance of local 
extirpation at multiple scales (Kuehne and Barbour, 1983). 
 Anthropogenic activities may contribute to population declines through habitat 
alteration, water table recession, and surface/groundwater contamination.  Habitats are 
often altered directly during construction of roads, ditches, and/or dams, and affected 
indirectly by sedimentation from urban development or logging practices (Boschung and 
Nieland, 1986; Allan, 1995; McGregor and Shepard, 1995; Beissinger and McCullough, 
2002; Primack, 2002).  Habitat alteration is usually permanent and is considered a 
significant cause of population decline (Boyce, 1992; Primack, 2002).  Drawing from the 
water table for drinking purposes or agriculture practices can affect stream levels, which 
can mirror drought conditions in some small streams.  Runoff of animal wastes from 
feedlots and pesticides from cropland can overload surface water with nutrients, which 
can cause algal blooms and low dissolved oxygen levels (Allan, 1995). 
 Most darters, particularly those of the genus Etheostoma, have short life spans (2-4 
years), have limited distributions, and frequently live in restricted habitats.  Fishes with 
limited distributions may be confined to a single stream system and/or have very few 
 2
populations.  Many darters that live in restricted habitats also have water quality 
requirements which include high dissolved oxygen levels (> 8 ppm) with inversely low 
temperatures (< 10 ?C).  Streams with these values are indicative of near-pristine waters, 
as most aquatic organisms need dissolved oxygen levels over 5 ppm to thrive (Ultsch et 
al., 1978; Mettee et al., 1996).  The vulnerability to water chemistry changes make the 
presence of species of Etheostoma an indicator of good water quality but also make them 
vulnerable to changes in environmental quality. 
 Any number of the previous factors alone or combined can be limiting to a species? 
survival.  In order to protect a species from drastic declines in population size or eventual 
extinction, conservationists must determine the stability of populations under a variety of 
circumstances so that conservation efforts can be targeted appropriately.  To do this, 
conservationists try to conduct comprehensive studies of a species by examining 
available life history information (growth, survival, and reproduction), collecting 
additional life history information (population size, life span, and fecundity), and 
monitoring existing population levels.  Often policy decisions are based on an estimation 
of how long a population can persist and the rate at which the size of a population is 
changing (Boyce, 1992; Heppell et al., 2000b; Primack, 2002).  These data are frequently 
lacking, especially for endangered species (Beissinger, 2002).  Meanwhile, resource 
managers may have to make judgments that will affect numerous species and recommend 
alternatives for imperiled species, sometimes using incomplete data sets.  By utilizing 
short-term predictions, early feedback on a management action can be evaluated so that 
future decisions are easier to make (Burgman et al., 1993; Fagan et al., 2001; Hanski, 
2002; Primack, 2002). 
 3
 One tool utilized by resource managers to determine a population?s persistence is 
population viability analysis (PVA).  A variety of computer software packages are now 
available that construct simple and complex models with uses ranging from 
determination of the risk of endangerment for a population to providing guidance on land 
use planning strategies (Ak?akaya, 2000; Fagan et al., 2001).  Detailed, complex models 
are not necessary if the primary variable causing population decline is loss or altered 
habitat, which is the situation with most species under the Endangered Species Act 
(Boyce, 2002).  Rarely are there enough data available for endangered species to allow 
estimation of parameters required for complex models, however, workable models (for 
shorter-lived species) can still be built even when there are gaps in data for a population 
(Burgman et. al, 1993; Beissinger, 2002).  Simple PVA models can predict dynamics of 
imperiled species, which may be adequate for conservation planning (Belovsky et al., 
2002). 
 There are numerous types of PVAs with some based on deterministic and others on 
stochastic projections.  Many PVAs utilize matrix models to estimate the likelihood that a 
population will persist for a certain period of time; these models can be used to provide 
resource management with insights into what parameters influence survival or extinction 
rates (Boyce, 1992; Williams et al., 1999; Ferrara, 2001; Beissinger and McCullough, 
2002).  In most matrix-based PVA models, males are ignored because including both 
sexes underestimates the risk of extinction (Brook et. al, 2000).  Additionally, females are 
usually the limiting sex during breeding so abundances are expressed in terms of females 
and productivity (Burgman et. al, 1993; Brook et. al, 2000).  Unless males are the 
 4
limiting sex, then vital rates of males will not influence population growth rates as 
significantly.   
 Models are constructed by first classifying individuals into age classes or life history 
stages (i) at time (t) to determine the size of the matrix.  Leslie models are the simplest 
matrices and integrate age-specific survival and fertility rates (Leslie, 1945).  Leslie 
matrices can be used to estimate the probability of reaching a critical population level and 
to calculate average extinction times.  Population size and fecundity are used to derive 
estimates of fertility (F
i
) and survival (P
i
) rates, which are needed for each age class to be 
modeled.  Generally, age classes are determined by annual growth (Caswell, 2001).  The 
fertility rate of a population is the number of live births produced over some time period 
and is often expressed as young females produced per female in the population (Krebs, 
1985).  The fecundity of a population is the potential level of reproductive performance 
and is usually greater than fertility (Krebs, 1985).  Survival rates represent the fraction of 
those individuals of age i that survive to be of age i+1 (Krebs, 1985; Burgman et al., 
1993; Heppell et al., 2000a; Caswell, 2001). 
 These vital rates are incorporated into a matrix model to determine population growth 
trends.  An example Leslie Matrix for four ages looks like the following 4x4 projection 
matrix. 
 Projection Matrix (A) = 
?
?
?
?
?
?
?
?
?
?
?
?
0P00
00P0
000P
FFFF
3
2
1
4321
 
 5
The projection matrix is multiplied by a population vector (the number of individuals n in 
each stage i at time t) to project the population through time.  The vector must have the 
same number of rows as the matrix (see example below). 
 Population Vector = 
?
?
?
?
?
?
?
?
?
?
?
?
it
it
it
it
n
n
n
n
 
 If all the parameters in the model never varied, it would be deterministic.  The 
expected population growth rate (?) is the largest real positive (dominant) eigenvalue of 
the projection matrix.  The left eigenvector (reproductive values) corresponds with the 
dominant eigenvalue and measures the value of an individual in each of the age classes 
based on the anticipated total number of offspring (Burgman et al., 1993).  The unit sum 
of the right eigenvector (stable age distribution) estimates the proportion of individuals in 
each age class even though the total population density grows or declines (Caswell, 2001; 
Morris and Doak, 2002).   
 Sensitivity analysis has become an essential component of demographic analyses.  
Each element in a sensitivity matrix measures how small a change in the corresponding 
element of the projection matrix A would change the dominant eigenvalue, while keeping 
all other elements in A constant (Caswell, 1994; Morris and Doak, 2002; Wilson, 2002).  
The sensitivity (s
ij
) of ? to changes in the elements of A is defined by: 
 
v,w
wv
a
s
ji
ij
ij =
?
?
=
?
 
where i refers to the row and j refers to the column of the matrix element a, w and v are 
the left and right eigenvectors, respectively of A corresponding to ?, and the v,w  
 6
denotes the scalar product of w and v (Caswell, 1994, 2001).  The most common form of 
sensitivity analysis is elasticity analysis or proportional sensitivities (Caswell, 1994; de 
Kroon et. al., 2000).  Elasticities are rescaled sensitivities to account for the magnitude of 
? and the matrix element (Caswell, 2001): 
 
ij
ij
ijij
alog
loga
se
?
?
==
?
?
 
Simply put, the elasticity analysis of A compares the relative influence of each vital rate 
on ?.  The largest elasticity values represent the life history stage with the strongest 
influence on population growth rates (Burgman et al., 1993; Mills and Lindberg, 2002).  
This information can be used to infer how management that affects a particular vital rate 
will affect the trend of ? (Heppell et al., 2000a).   
 Various government agencies now use PVA models to evaluate species proposed for 
listing under the Endangered Species Act (Boyce, 1992).  For the past two decades, over 
50 PVAs have aided in the conservation efforts of species of plants, tortoises, bears, and 
birds (McCarthy et al., 2001).  For example, several age-structured models have been 
constructed for long-lived seabirds and resulted in adult survival rates as the most 
influential parameter on population growth (Cuthbert et al., 2001).  Desert tortoise and 
sea turtle models indicated diminishing the mortality of large individuals would have the 
greatest influence on population growth (Doak et al., 1994; Heppell et al., 2000a).   
 Many species commonly analyzed with PVA models are long-lived, have decades of 
life history information available and can easily be studied for population size, behavior 
and dynamics.  For example, determination of the number of offspring for vertebrates is 
observed in the field.  Few fishes have been analyzed with PVA models other than large, 
 7
more long-lived fish, like sturgeon, or more abundant fish, such as salmon and trout 
(Boyce, 1992; Williams et al., 1999). 
 Short-lived species, such as fishes, are more difficult to observe in the field.  For 
example, to determine offspring numbers in fishes, especially Etheostoma species, 
fecundity must be determined by dissecting ovaries, staging ova into size classes, and 
counting the number of ova.  Prior to the early 1990s, fecundity was determined from 
counts of all eggs produced by individual females or from guesses as to the number of 
ripe ova produced by females (Heins and Baker, 1993a; 1993b).  Heins and Baker (1988; 
1989) provided science with a standard way to examine ova and classify them by 
developmental stage and clutch size.  Problems can arise when researchers try to compare 
historical published fecundity with fecundity determined by the current standard.  
Additionally, whether or not a fish spawns multiple times during a breeding season or 
just once can affect values for fecundity.  There are many conflicting methods to 
determine if a fish spawns multiple times (Weddle and Burr, 1991).  The age of fishes 
can be determined empirically by direct observation, statistically with length frequencies 
and anatomically from certain fish structures (Jearld, 1983).  However, for many darters 
it is difficult to gather dependable estimates of age, especially if age is determined by size 
or growth controlled by environmental variation (Burgman et al., 1993).     
Objectives 
 The objectives of this study were to: a) assess population levels at historic sites for 
Etheostoma boschungi, the Slackwater Darter and Etheostoma brevirostrum, the Holiday 
Darter, b) assess which parameters have the greatest relative influence on the population 
growth rate, c) predict possible extinction time periods and d) determine the population 
 8
trend.  The results from this study will address the utility of PVA models for rare fishes 
and provide information that can be utilized for conservation of these fishes. 
 9
METHODS 
Study species 
 The study species are Etheostoma boschungi, the Slackwater Darter, and Etheostoma 
brevirostrum, the Holiday Darter.  Although these two darter species are in the same 
family (Percidae), their life histories differ.  The following provides information on the 
breeding habits, historical locations, length, life span, and published population size for 
each species. 
 Etheostoma boschungi is a migratory species and has two distinct, but adjacent 
habitats: non-breeding and breeding.  For most of the year, they live in gravel-bottomed 
pools of creeks where they burrow under piles of old leaf litter or detritus that accumulate 
where water flow is slow (Wall and Williams, 1974; Boschung, 1976a; 1976b; 1979; 
McGregor and Shepard, 1995).  Streams inhabited during most of the year are usually 
0.6-12 m wide and 0.15-2 m deep with high dissolved oxygen levels (8.1-12.4 ppm).  In 
November, E. boschungi migrates approximately 3-6 km to the breeding habitat.  The 
breeding habitat is shallow water (5-10 cm deep), which originates in seeps, boils or 
flooded fields that slowly runs off into adjacent streams.  Once winter rains increase 
water levels in the streams, the darters have access to these shallow waters.  Etheostoma 
boschungi spawn from late February to late March (Boschung, 1976a; 1976b; 1979).  
McGregor (pers. comm.) stated E. boschungi is opportunistic and will spawn in 
vegetation found in stream channels as long as there is enough flowing water to keep
 10
eggs oxygenated, even in landscaped yards, lawns, or similar areas.  In April to early 
May, juveniles migrate to the non-breeding habitat (Boschung, 1976a). 
 Historically, E. boschungi has been collected from 30 sites in tributaries of the 
Tennessee River drainage of Alabama and Tennessee (Fig. 1) (Boschung, 1976a; 1976b; 
McGregor and Shepard, 1995).  The total population size of the species was not estimated 
prior to my study but a 1976 study estimated the largest population (found within the 
Cypress Creek watershed) to contain 3,600 individuals (Boschung, 1976a).  Boschung 
determined that the life span of E. boschungi is no more than four years and they reach a 
maximum size of 65 mm standard length (SL) (Boschung, 1976a; Etnier and Starnes, 
1993; Page, 1983).  Standard length is a typical measurement for fishes and is the 
distance from the tip of the snout to the hypural plate.  Juvenile E. boschungi range in 
size from 10-12 mm SL by early April and by the end of their first year are 30 mm ?  2 
mm SL (Boschung, 1976a).  Boschung estimated fecundity to be an average of 320 ripe 
eggs from three specimens (Boschung, 1976a; Boschung and Nieland, 1986).   
 Etheostoma brevirostrum, unlike E. boschungi, does not migrate in order to spawn 
and prefers rocky runs and riffles with a fast current (Suttkus and Etnier, 1991; Johnston 
and Phillips, 2001).  Streams in which they are found are 3-12 m wide and 7-44 cm deep.  
Etheostoma brevirostrum spawn from early April to late May. 
 Etheostoma brevirostrum is known to occur in disjunct populations located in 
tributaries of the Coosa River drainage (Suttkus and Etnier, 1991; Mettee et al., 1996; 
Johnston and Phillips, 2001).  These populations may represent two or more taxa and the 
Shoal Creek population is a new species (B. R. Kuhadja, pers. comm.).  This study 
concentrates on the population that is restricted to Shoal Creek between Sweetwater and 
 11
Whiteside Mill Lakes in the Talladega National Forest (Fig. 2).  Little is known about the 
life history of E. brevirostrum but they are believed to live three years and reach a 
maximum size of 53 mm SL (Suttkus and Etnier, 1991).  Fecundity analysis has not been 
conducted for E. brevirostrum.   
Threats to each species 
 Etheostoma boschungi is federally listed as threatened (U.S. Fish and Wildlife 
Service, 1984; 1990).  Their habitat and complex cycles are ultimately, what limit their 
abundance and distribution (U.S. Fish and Wildlife Service, 1984).  Potential threats 
include: ditching to drain areas with shallow groundwater, urban development, surface 
and groundwater contamination from point and nonpoint source pollution, and seepage 
areas diked to form ponds (U.S. Fish and Wildlife Service, 1984).  Additionally, 
reproduction of E. boschungi is dependent upon rainfall and water temperature.  Rainfall 
must be heavy enough to cause the main channel to flood into spawning grounds and 
temperatures must be more than 14 ?C for spawning to occur (U.S. Fish and Wildlife 
Service, 1984).  If weather conditions lower the water temperature, then the spawning 
season could be shortened (Boschung, 1979).    
 Etheostoma brevirostrum is not federally protected; however, living within the 
boundaries of the Talladega National Forest does provide some protection.  Every 
national forest has a forest plan that includes streamside management zones, watershed 
assessment and aquatic viability from a habitat standpoint (USDA Forest Service, 2004).  
In Alabama, the focus of the forest plan is to use a watershed approach for management 
decisions and to ensure activities conducted within the National Forest do not have 
negative effects on water quality, and are suitable to maintain native aquatic communities 
 12
(D. Thurmond, pers. comm).  The forest plan implements State best management 
practices, such as streamside management zones (SMZs), to protect water quality from 
upslope land use practices (USDA Forest Service, 2004).  Streamside management zones 
are areas adjacent to waterbodies and typically contain sediment filter strips to restrict 
ground disturbance and protect stream banks (USDA Forest Service, 2004).  Since the 
forest plan was implemented, policies concerning clear cutting and ecosystem 
management have caused the numbers of acres of regeneration harvesting to decline 
(USDA Forest Service, 2000). 
 The city of Heflin, Alabama could impact the quantity of water available for aquatic 
species because Heflin is growing rapidly compared to the State average and there is a 
potential to need additional water resources from the impoundments of Shoal Creek 
(USDA Forest Service, 2000).  It is also possible that SMZs are not always properly 
implemented or maintained and could introduce sediment loads and runoff to the streams 
(D. Thurmond, pers. comm.).  Currently, the main threat to the E. brevirostrum 
population is division into two subpopulations by a small impoundment that blocks 
movement (Johnston and Phillips, 2001).  Habitat fragmentation, reduced water quality 
(due to pollution and/or poor land use), drought, and unseasonable temperatures could all 
have an effect on spawning capabilities of either population, ultimately affecting 
population levels. 
Field data collection 
 Two darter species were assessed in relation to life history traits (fecundity, number 
of age classes, age specific population sizes, and age specific survival rates) during the 
 13
breeding season.  Data for E. boschungi were collected from one site in Cypress Creek 
and data for E. brevirostrum data were collected from four sites in Shoal Creek. 
 In November 2000, January 2001, March 2001, and March 2002 a survey for the 
presence of sizable populations of E. boschungi was conducted at 30 historical sites 
located throughout the Tennessee Valley, representing both breeding and non-breeding 
habitats (Johnston and Hartup, 2001; 2002) (Fig. 1).  Two to three collectors used dip 
nets and 3.3 m seines to sample one to two hours for E. boschungi at each site, however, 
E. boschungi was present at only two sites.  Darters present at the Speedway site 
(McGregor and Shepard, 1995) near the headwaters of Shoal Creek, Tennessee, consisted 
of less than ten individuals.  We concluded that sampling this population further could be 
detrimental to their persistence in the future.   
 The other population, found near the headwaters of Middle Cypress Creek, appeared 
to be large enough to study.  This particular location is a historical breeding site of E. 
boschungi, known in many publications as the Dodd site (Boschung, 1976a; 1976b; 
McGregor and Shepard, 1995).  The darters migrate to and breed in a seepage stream 
created by boils or springs that gather enough water to create two small streams (each < 
0.5 m wide).  These tiny streams combine (1 m wide) and flow 75 m across a rural farm 
to join Middle Cypress Creek (Fig. 3). 
 A mark-recapture study was conducted during the breeding seasons of 2001 and 2002 
for E. boschungi at the Dodd site.  Fish were collected for 45 min with dip nets and 
placed immediately into ice chests with aerated stream water.  Each individual was sexed 
(male, female, or undeterminable/juvenile; Wall and Williams, 1974), measured with a 
 14
caliper to the nearest 0.01 mm SL, clipped on the top portion of the caudal fin (Murphy 
and Willis, 1996), and then released.  Attempts at recapture occurred within a week.   
 Etheostoma brevirostrum were collected during the 2001 and 2002 breeding seasons 
in Shoal Creek, Alabama, from two sites above and two sites below Highrock Lake (Fig. 
4).  Transect data (measurements of riffle length, three widths and three depths per site) 
were collected from each site for use in determining stream volume (length x mean width 
x mean depth).  Two to three collectors used 3.3 m seines for up to two hours at each site.  
Each site was sampled with at least six passes of seining or until no more E. brevirostrum 
were found.  Not enough specimens were available for mark-recapture at all sites, so E. 
brevirostrum were measured with a caliper to the nearest 0.01 mm SL, sexed (male, 
female, or undeterminable/juvenile; Suttkus and Etnier, 1991) and released.  I assumed 
all available fish were caught. 
Laboratory procedures and data collection 
 To reduce population impacts, I attempted to use museum specimens for fecundity 
analysis (this requires sacrifice of females).  However, existing E. boschungi specimens 
from the University of Alabama were not useable.  No large series of museum species 
collected during the breeding season was available for E. brevirostrum.  Therefore, I 
generated original data for fecundity of both species.  
 Specimens were fixed in 5% buffered formalin and then transferred to water.  The SL 
of each specimen was measured to the nearest 0.1 mm with a digital caliper.  Each 
specimen was weighed to the nearest 0.01 g.  The right ovary was then removed from the 
specimen, air dried, and weighed to the nearest 0.01 g.   
 15
 Fecundity was determined by dissecting the right ovary and classifying ova into one 
of six developmental stages described by Heins and Baker (1988): ripe (RE), ripening 
(MR), mature (MA), late maturing (LM), early maturing (EM), and latent (LA).  All ova 
were classified by developmental stage under a dissecting microscope and counted.  Ten 
ova were measured (digital caliper under a dissecting microscope) for each of the three 
largest size classes (RE, MR, and MA).  Although the RE developmental stage is 
sometimes present, it usually excluded from fecundity analysis because it is difficult to 
determine if any eggs have been oviposited or not (Heins and Baker, 1989; 1993a).  The 
most important developmental stages of ova for determining fecundity are the MR and 
MA stages.  Specimens were vouchered in the Auburn University museum collection. 
 The number of ova per batch (developmental stage) or batch fecundity spawned 
during the breeding season remains unpublished for both of these species.  Some species 
are multiple spawners, meaning they spawn more than one batch of ova per breeding 
season.  The average number of ova in the MR developmental stage is called one-batch 
fecundity.  Two-batch fecundity is the average number of MR ova plus the average 
number of MA ova.  For E. boschungi, we compare models using one-batch and two-
batch fecundities.  The spawning period for this species is short, and it is not known for 
any species of Ozarka how many clutches are spawned.  Like other members of the 
subgenus Ulocentra, E. brevirostrum is likely a multiple spawner (O?Neil, 1981; Page 
and Madden, 1981; Erickson and Mahan, 1982; Carney and Burr, 1989; Weddle and 
Burr, 1991; Johnston and Haag, 1996).  I constructed all models for E. brevirostrum with 
two-batch fecundities. 
 16
 To determine age classes for each species, we used histograms of length frequencies.  
For E. boschungi, I used all historical length data in addition to what we collected to 
determine size ranges for each age class.  Historical length data are not available for E. 
brevirostrum, so only data collected during this study were used for length-frequency 
analysis.  
 To estimate the population size for each age class (n
it
) of E. boschungi, we used 
mark-recapture data from females captured during the study.  The adjusted Petersen 
estimate (Chapman, 1951) was used to calculate population size, where m is the number 
of females captured and marked in the initial sample; c is the total number of females 
captured in a subsequent sample; and r is the number of marked females recaptured in c. 
 
)r(
)c)(m(
nit
1
11
+
++
=  (1) 
 Equation (1) represents the number of females in the population at time t, given that 
no changes have occurred in the population size and that adequate mixing has occurred 
between sampling periods.  This formula provides an unbiased estimate for smaller 
populations (Ricker, 1975).  Ricker (1975) recommends replacing r in equation 1 with 
variables from a Poisson distribution chart to calculate confidence limits.  The estimated 
population size is then multiplied by the distribution of each age class.     
 Transect data from E. brevirostrum sampling were separated into three data sets to 
determine the volume of the riffles because we wanted to compare population levels from 
two sites above and two sites below Highrock Lake as well as with a hypothetical 
population that combines data from all sites.  The density of fish/ m
3
 was estimated by 
 17
dividing the number of females per age class (n
i
) by riffle volume (equation 2) (Gordon 
et. al., 1992; U. S. Environmental Protection Agency, 1997).  
 Density of Fish/ m
3
 = n
i 
/ riffle volume (2) 
 Three segments of Shoal Creek were measured in meters with computer software 
(MAPINFO PROFESSIONAL, version 6.0, MapInfo Professional Corporation, Troy, NY, 
2000, unpubl.).  Tributaries were not included in the measurements because results from 
a 2001 movement study in Shoal Creek showed E. brevirostrum had access to many of 
the tributaries but did not utilize them (Johnston and Phillips, 2001).  Therefore, only the 
mainstem of Shoal Creek (from Sweetwater Lake to Whiteside Mill Lake) was used for 
these measurements.  Measurements were separated into the following segments to 
correspond with each of the models: upstream of Highrock Lake to Sweetwater Lake; 
downstream of Highrock Lake to Whiteside Mill Lake; and from Sweetwater Lake to 
Whiteside Mill Lake.  The last measurement included the length of Highrock Lake.  
 To estimate the population size for each age class of E. brevirostrum, we multiplied 
density of fish/ m
3
 in equation 2 by the estimated stream segment lengths measured with 
MapInfo (equation 3). 
 Total population for each age class = (equation 2)(stream segment length) (3) 
Analysis of data 
 The mean size, standard deviation (SD), size range and population 95% confidence 
limits (95% CL) were calculated for all field data.  For laboratory data, the mean size and 
number of ova, SD, and range of size and number of ova per female per developmental 
stage were calculated for both species. 
 
 18
 
The age-based matrix model 
 First we constructed a life history table using the following basic ecology formulas 
(Krebs, 1985): 
 n
i+1
 = n
i
 - d
i  
(4) 
 q
i
 = d
i
 / n
i  
(5) 
 P
i
 = n
i+1
/ n
i
    (6) 
 L
i
 = (n
i+1
+ n
i
)/ 2 (7) 
 
?
?
=
i
ii LT     (8) 
 e
i
 = T
i
/ n
i
    (9) 
 F
i
 = (P
0
)(f
i
)/ 2   (10) 
where i = age interval, n
i
 = number of females at start of age interval i, P
i
 = proportion of 
organisms surviving to start of age interval i, d
i
 = number dying during the age interval i 
to i+1, q
i
 = rate of mortality during the age interval i to i+1, L
i
 = number of individuals 
alive on average during the age interval i to i+1, T
i
 = cumulative sum of L
i
 in units of 
individuals times time units, e
i
 = mean expectation of life for organisms alive at start of 
age i, F
i
 = fertility rate or number of female offspring per female aged i per year, P
0
 = 
proportion of organisms surviving to start of age one and f
i
 = batch fecundity (number of 
eggs produced per batch).  Fertilities of all mature ages classes are assumed to be equal. 
 Since this was a breeding census for both species, we did not collect enough juvenile 
to one-year-old data to calculate survival.  Although some of the historical data showed 
population levels for E. boschungi (3,600 individuals; Boschung, 1976a), these were not 
 19
distributed by age.  Individuals up to one-year-old could not be determined.  We collected 
survival data on closely related species to E. boschungi and E. brevirostrum and 
substituted the P
0
 of sister species to determine F
i
.  We also estimated instantaneous 
annual mortality rates (Z) from the fishery analysis software package, FAST (SLIPKE, J. 
W. AND M. J. MACEINA. 2000.  Fishery Analysis and Simulation Tools, version 2.1.  
Auburn University, AL.).  Total annual mortality rates are derived by adding fishing 
mortality (F
*
) to natural mortality (M).  Fishing mortality is the result of fish harvesting 
and would be zero for both our species because darters are not usually harvested.  Natural 
mortality is the result of old age, predation, parasites, diseases, and abiotic causes.  For 
our purposes, Z is equal to M.  FAST calculates M from a selection of formulas (Pauly, 
1980; Hoenig, 1983; Peterson and Wroblewski, 1984; Chen and Watanabe, 1989; Jensen, 
1996; Quinn and Deriso, 1999), which is then converted to annual survival rates with the 
following formula: 
 
MZ
eeP
??
==  (11) 
 PopTools was used to construct the models used in this study (POPTOOLS: software 
for the analysis of ecological models, version 2.5.9., G. M. Hood, 
http://www.cse.csiro.au/poptools/, 2003, unpubl.).  PopTools is an add-in to the computer 
program, Microsoft Excel, which has tools for analyzing a Leslie projection matrix, an 
elasticity analysis of the projection matrix, and a population projection.  We constructed a 
three-stage, pre-breeding, birth-pulsed population model (Fig. 5).  The three age classes 
used in our model for each species include one-year-old adults (n
1
), two-year-old adults 
(n
2
), and three -year-old adults (n
3
).  We assumed for the model that these darters at birth 
are zero years old until one year later, at which time they have the capability to 
 20
reproduce.  The PVA matrix model assumes the fish live three years, no migration occurs 
into or out of a population, sampling was conducted just before breeding season, and n
3
 is 
the last reproductive stage (individuals die after they breed at this age).  
 The fertility and survival rates are put into a 3x3, Leslie projection matrix, which is 
deterministic.  I used the basic analysis function in PopTools (Hood 2003) to calculate ? 
(expected population growth rate), the stable age distribution (right eigenvector) and 
reproductive value (left eigenvector) for the deterministic projection matrix.   
 I could not estimate the fertility rates for the matrix directly from the data we 
collected because we did not sample juvenile survival.  Sister species of E. boschungi in 
the subgenus Ozarka lacked juvenile survival as well.  Some of the sister species for E. 
brevirostrum in the subgenus Ulocentra had published juvenile survival rates, which I 
substituted into equation 10 for E. brevirostrum fertility rates.   
 In order to estimate fertility rates for E. brevirostrum and E. brevirostrum, I used 
instantaneous survival rates for each age class from my study as well as estimated 
survival rates from FAST in the Leslie projection matrix.  I then set the SOLVER tool in 
Windows Excel to ? = 1.  The SOLVER tool calculates what the fertility rates should be 
with the provided survival rates when ? = 1.  I then used the elasticity function in 
PopTools to conduct an analysis of the matrix to assess which parameters have the 
greatest relative influence on the population growth rate.  I also projected the population 
over time to determine when and if the population approaches one and to determine the 
population trend. 
 
 21
RESULTS 
Field Data 
 Etheostoma boschungi --In 2001 a total of 274 E. boschungi (152 males: 26-50 mm 
SL; x SL = 34 mm; SD = 4.45; and 122 females: 27-50 mm SL; x SL = 34 mm; SD = 
4.70) were collected from the Dodd site on Middle Cypress Creek, Tennessee.  The 
overall sex ratio, 0.8 females:1 male, was not significantly different (?? = 3.28, p>0.05).  
In 2002 a total of 79 E. boschungi (21 males: 38-49 mm SL; x SL = 43 mm; SD = 2.80; 
and 58 females: 32-52 mm SL; x SL = 42 mm; SD = 4.86) were collected from the Dodd 
site on Middle Cypress Creek, Tennessee.  The overall sex ratio, 2.76:1, was significantly 
female biased (?? = 17.33, p < 0.001). 
 Length frequencies were used to determine age classes for E. boschungi (Fig. 6).  
Data for Fig. 5A were from University of Alabama museum specimens collected in the 
1970s (unpubl.).  Data for Fig. 5B were from a Geological Survey of Alabama study in 
the 1990s (McGregor and Shepard, 1995) and data for Fig. 5C were collected in 2001 and 
2002 (Johnston and Hartup, 2001; 2002).  Historical collections depicted in Fig. 5A and 
Fig. 5B was sampled for other purposes than population abundance.  The color scheme 
for the graphs depicts the pre-breeding time period (blue, red and yellow), where only 
adults are present and the post-breeding period (green), where mostly juveniles are 
present with very few adults.  Based on these graphs, the data indicates the following four 
 22
age classes: juveniles (n
0
) = <26 mm SL, one-year-old adults (n
1
) = 27-39 mm SL, two-
year-old adults (n
2
) = 40-48 mm SL and three-year-old adults (n
3
) = >49 mm SL. 
 Mark-recapture data from 2001 resulted in an estimated female population size of 234 
(95% CL = 145-398), a male population size of 331 (95% CL = 209-552) and a total 
population size for the Dodd Site at 581 individuals (95% CL = 413-846) (Table 1).  
Mark-recapture data from 2002 resulted in a significantly larger female population size of 
450 (95% CL = 136-818; ?? = 68.21, p < 0.001), a significantly smaller male population 
size of 63 (95% CL = 19-115; ?? = 182.3, p < 0.001) and a total population size for the 
Dodd Site at 545 individuals (95% CL = 199-1362) that was not statistically different.     
 The 2001 mark-recapture census resulted in the following mean number of 
individuals in each age class: n
1
 = 48, n
2
 = 4 and n
3
 = 2.  The proportion of individuals 
per age class resulted in a decreasing population size, as the fish become older (Table 2).  
The 2002 mark-recapture census resulted in the following mean number of individuals in 
each age class: n
1
 = 9, n
2
 = 17 and n
3
 = 3, which resulted in a population size that does 
not decrease as the fish grow older.   
 Etheostoma brevirostrum -- In 2001 a total of 143 E. brevirostrum were collected 
from four sites on Shoal Creek.  Exactly 45 males (35-49 mm SL; x SL = 42 mm; SD = 
3.51) and 41 females (30-48 mm SL; x SL = 39 mm; SD = 3.66) were collected from the 
two sites upstream of Highrock Lake.  A total of 57 E. brevirostrum (29 males: 34-44 
mm SL; x SL = 39 mm; SD = 2.46; and 28 females: 33-44 mm SL; x SL = 37 mm; SD 
= 2.60) were collected from the two sites downstream of Highrock Lake. 
 
 23
 In 2002 a total of 74 E. brevirostrum were collected from the same four sites on Shoal 
Creek.  Eight males (37-49 mm SL; x SL = 43 mm; SD = 3.96) and 9 females (24-45 
mm SL; x SL = 37 mm; SD = 7.42) were collected from the two sites upstream of 
Highrock Lake.  A total of 57 E. brevirostrum (21 males: 36-46 mm SL; x SL = 40 mm; 
SD = 2.99; and 36 females: 22-50 mm SL; x SL = 38 mm; SD = 4.49) were collected 
from the two sites downstream of Highrock Lake.  None of the sex ratios for 2001 was 
significantly different from a 1:1 relationship, which would be expected for most darters.  
In 2002, the sex ratio, 1.7:1, was significantly female biased (?? = 3.95, p < 0.05) for the 
downstream sites of Highrock Lake but was not statistically different for the upstream 
sites of Highrock Lake or for Shoal Creek. 
 Length frequencies were used to determine age classes for E. brevirostrum (Fig. 7).  
Collection data were from a 2001 movement study (Johnston and Phillips, 2001) and our 
2002 collection in Shoal Creek.  The graph indicates four age classes: juveniles (n
0
) = 
<30 mm SL, one-year-old adults (n
1
) = 31-40 mm SL, two-year-old adults (n
2
) = 41-45 
mm SL, and three -year-old adults (n
3
) = >46 mm SL. 
 Riffle volume determined from transect data resulted in 198 m
3
 for the sites upstream 
of Highrock Lake, 174 m
3
 for the sites downstream of Highrock Lake and 372 m
3
 for all 
sites combined (Table 3).  The lengths for segments of Shoal Creek mainstem were 
determined as follows: upstream of Highrock Lake to Sweetwater Lake= 3.8 km, 
downstream of Highrock Lake to Whiteside Mill Lake= 2.5 km, and from Sweetwater 
Lake to Whiteside Mill Lake = 6.5 km.  
 
 24
 Although the number of fish collected per sampling season ranged from 2-54 
individuals per age class in 2001 and 1-32 individuals per age class in 2002, the density 
of fish per stream section resulted in < 0.2 fish/m? for all sites in 2001 and 2002 (Table 
4).  We did not use the juvenile data in the extrapolated populations because the models 
are set up as pre-breeding censuses.  The distribution of population size for each age class 
in 2001 showed a decrease in number as the fish became older for both populations 
upstream and downstream of Highrock Lake.  In 2002, the extrapolated population of the 
upstream segment did not decrease, as the fish grew older.  The upstream population size 
with 732 E. brevirostrum in 2001 was almost double that of the downstream population 
size with 400 E. brevirostrum.  In 2002, the upstream population of 135 E. brevirostrum 
was much less abundant than those collected in 2001 and was smaller than the 2002 
downstream population.  The downstream population size increased slightly from 2001 to 
2002 with 400 E. brevirostrum to 498 E. brevirostrum.   
Laboratory Data 
 Etheostoma boschungi --Ten females (40.6 ? 48.8 mm SL; x SL = 44.2 mm; SD = 
2.53) were sampled during the breeding season for fecundity analysis.  The mean number 
and diameter of ova in the three largest developmental size classes were, from largest to 
smallest: RE ( x = 8, SD = 4.56 and x = 1.23 mm, SD = 0.09); MR ( x = 92, SD = 54.53 
and x = 0.95 mm, SD = 0.09); and MA ( x = 105, SD = 52.46 and x = 0.73 mm, SD = 
0.07).  The largest egg class (RE) was present in 50% of the fish but was not used in 
determining fecundity.  The one-batch fecundity was the mean number of ova from the 
 25
MR stage ( x = 92) and the two-batch fecundity was the mean number of ova from the 
MR stage added to the MA stage (197). 
 Etheostoma brevirostrum --Five females (33.1-42.8 mm SL; x SL = 36.4 mm; SD = 
3.87) were sampled for fecundity analysis.  The mean number and diameter of ova in the 
three largest developmental size classes were, from largest to smallest: RE ( x = 3, SD = 
1.15 and x = 1.34 mm, SD = 0.19); MR ( x = 66, SD = 26.83 and x = 0.98 mm, SD = 
0.09); and MA ( x = 109, SD = 28.59 and x = 0.61 mm, SD = 0.16).  The largest egg 
class (RE) was present in 60% of the fish but was not used in determining fecundity.  The 
two-batch fecundity was the mean number of ova from the MR stage added to the MA 
stage (175). 
Vital rates 
 Etheostoma boschungi --We used the population size determined from 2001 data in 
Table 2 for each age class because the information fit in the best with the life history 
table (Table 5).  Fertility rates could not be calculated for the basic life history table from 
the existing data we collected because we did not collect any information on juvenile 
survival.  We were able to calculate survival for the PVA model for each age class of 
adults (P
1
 = 0.074, CL = 0.073-0.076; P
2
 = 0.571, CL = 0.5-0.58; P
3
 = 0).  The rate of 
mortality during the interval from one age to the next was q
1
 = 0.926; q
2
 = 0.429; q
3
 = 1.0 
and the average expectation of life for each age class was e
1
 = 0.62 yrs; e
2
 = 1.07 yrs; e
3
 = 
0.5 yrs. 
 Using the E. boschungi survival from Table 5, the adult fertility rates would have to 
be 0.896 for the population to be stable (stable age distribution: n
1
 = 0.896, n
2
 = 0.066, n
3
 
 26
= 0.038 and reproductive values: n
1
 = 0.303, n
2
 = 0.426, n
3
 = 0.271).  By back-calculating 
with equation 10, the juvenile survival rate would be 0.019 for a one-batch fecundity of 
92 ova and 0.009 for a two-batch fecundity of 197 ova.  The elasticity analysis of the 
matrix using our calculated survival rates and the SOLVER fertility indicated the 
fertilities made the largest relative contribution to ? with elasticities of 0.79, 0.06, and 
0.03 while survivals were smaller at 0.09 and 0.03.   
 None of the species closely related to E. boschungi had estimates for juvenile survival 
to year one (Table 6).  Of the species closely related to E. boschungi that had available 
information, survival rates ranged from 0.396 to 0.549 ( x = 0.38 ? 0.23 SD) for one-
year-old adults and 0.429 to 0.855 ( x = 0.64 ? 0.3 SD) for two-year-old adults.  For each 
of these species with published survival rates, the authors assume that each age class was 
collected in proportion to its relative number in the population, the population is 
stationary, and the number of juveniles entering the population is constant.  The species 
with the closest percentage of individuals in each age class to our data from Table 2 was 
E. cragini.  Five of the closely related species die sometime before their third or fourth 
year, like E. boschungi, while E. punctulatum lives up to five years.   
 Of the closely related species from Table 6, E. cragini was the only species with 
population data compatible with our data.  Using E. cragini survivals and setting ? = 1, 
results in F
i
 = 0.937 with the SOLVER tool (stable age distribution: n
1
 = 0.937, n
2
 = 
0.044, n
3
 = 0.019 and reproductive values: n
1
 = 0.305, n
2
 = 0.409, n
3
 = 0.286).  Using 
equation 10, results in juvenile survival rates of 0.020 and 0.010 when using one-batch 
and two-batch fecundities, respectively.   
 27
 Annual survival rates estimated with the program FAST ranged from 0.153 to 0.368 
( x = 0.23 ? 0.07 SD) (Table 7).  By using the FAST annual survival rates for both P
1
 and 
P
2
 and setting ? = 1, the SOLVER tool estimated fertility rates ranging from 0.665 to 
0.850 ( x = 0.781 ? 0.06 SD).  One-batch fecundities resulted in juvenile survival rates 
ranging from 0.014 to 0.018 ( x = 0.017 ? 0.001 SD) and two-batch fecundities resulted 
in juvenile survival rates ranging from 0.007 to 0.009 ( x = 0. 008 ? 0.001 SD).  Juvenile 
survival rates derived from two-batch fecundities were almost half the juvenile survival 
rates derived from one-batch fecundities. 
 The population size decreases when adult fertility rates are below 0.896 and juvenile 
survival rates are below 0.019 and 0.009 for one- and two-batch fecundities, respectively 
(Table 8).  The population size falls below one individual after five years when P
0
 = 
0.004, adult fertility rates are 0.184 and ? = 0.309 for a one-batch fecundity of 92 ova 
(Table 8, Fig. 8).  For a two-batch fecundity of 197, the population size falls below one 
individual after five years when P
0
 = 0.002, adult fertility rates are 0.197 and ? = 0.322.  
The population size increases to 1,000 individuals after ten years when ? = 1.158 (one-
batch fecundity used) and after nine years when ? = 1.184 (two-batch fecundity used).  
As long as ? < 1, then the population will decrease but when ? > 1, then the population 
increases.   
 Etheostoma brevirostrum -- We used the population size for each age class 
determined from 2001 data in Table 4 because the proportion of individuals fit in the best 
with the life history table (Table 9).  Fertility rates could not be calculated for the basic 
life history table from the existing data we collected because we did not collect any 
 28
information on juvenile survival.  Survival rates were highest among the population 
upstream of Highrock Lake and lowest from the population downstream of Highrock 
Lake.  The average expectation of life for each age class was highest for the upstream 
population as well. 
 The SOLVER tool along with E. brevirostrum survivals and two-batch fecundities 
from Table 9 and setting ? = 1, resulted in estimated adult fertility rates of 0.737 (stable 
age distribution: n
1
 = 0.737, n
2
 = 0.211, n
3
 = 0.053 and reproductive values: n
1
 = 0.376, 
n
2
 = 0.347, n
3
 = 0.277) for the upstream population and a back-calculated juvenile 
survival rate of 0.008.  A stable downstream population results in adult fertility rates of 
0.929 (stable age distribution: n
1
 = 0.929, n
2
 = 0.071, n
3
 = 0 and reproductive values: n
1
 = 
0.350, n
2
 = 0.325, n
3
 = 0.325) and a back-calculated juvenile survival rate of 0.011.  For 
the entire Shoal Creek population, adult fertility rates would have to be 0.818 (stable age 
distribution: n
1
 = 0.818, n
2
 = 0.152, n
3
 = 0.30 and reproductive values: n
1
 = 0.357, n
2
 = 
0.351, n
3
 = 0.292) to be stable with a back-calculated juvenile survival rate of 0.009.  The 
elasticity analysis of the matrix using our calculated survival rates and the SOLVER 
fertility rates indicated the fertilities made the largest relative contribution to ?, while the 
survivals were smaller for all sites (Table 10).  The highest elasticity value for fertility 
rates came from the downstream matrix.   
 Many of the annual survival rates estimated with FAST were the same for E. 
brevirostrum as for the E. boschungi because many of the formulas are based on size and 
maximum age (Table 11).  Annual survivals ranged from 0.177 to 0.425 ( x = 0.308 ? 
0.10 SD).  By using the FAST annual survival rates for both P
1
 and P
2
 and setting ? = 1,  
 29
the SOLVER tool estimated fertility rates ranging from 0.623 to 0.827 ( x = 0.716 ? 0.08 
SD).  Two-batch fecundities resulted in juvenile survival rates ranging from 0.007 to 
0.009 ( x = 0.008 ? 0.001 SD). 
  Four closely related species to E. brevirostrum had estimates for juvenile survival to 
year one and die sometime before their third or fourth year, like E. brevirostrum, while E. 
simoterum lives up to two years and E. zonale lives up to five years (Table 12).  Of the 
species that had survival information available, survivals ranged from 0.123-0.761 for P
0
 
( x = 0.54 ? 0.29 SD), 0-0.286 for P
1
 (x = 0.16 ? 0.10 SD) and 0-0.182 for P
2
 ( x = 0.06 
? 0.08 SD).  For each of the species with published survival rates, the authors assume that 
each age class was collected in proportion to its relative number in the population, the 
population is stationary, and the number of juveniles entering the population is constant. 
 Of the closely related species from Table 12, E. raneyi and E. zonale were the only 
species with population data for the same number of age classes but neither of them had 
juvenile survival.  I used the published juvenile survivals from the other species to 
determine results for the upstream, downstream, and combined stream segments (Table 
13).  The juvenile survival rates used for E. brevirostrum from E. coosae and E. 
pyrrhogaster cause the fertility rates to be more than five times higher than the E. 
simoterum fertility rates.  Population size projected for upstream, downstream, and 
combined stream segments (using E. coosae and E. pyrrhogaster survivals) resulted in 
more than 50 times the initial population size after one year.  The calculated right 
eigenvectors were similar among species as were the left eigenvectors. 
 30
 The downstream population size falls below one individual after six years when P
0
 = 
0.004, adult fertility rates are 0.350 and ? = 0.415 and the upstream population falls 
below one individual after six years when P
0
 = 0.001, adult fertility rates are 0.079 and ? 
= 0.254 (Table 14).  The population size for E. brevirostrum falls below one individual 
after four years when P
0
 = 0.001, adult fertility rates are 0.088 and ? = 0.224 for the 
combined population.  The population size increases to 1,000 individuals after eight years 
for the downstream population when ? = 1.122, and after three years for the upstream 
population when ? = 1.142.  The combined population is already above 1,000 individuals 
when ? = 1. 
 31
DISCUSSION 
 Currently both species appear to be present in very low population levels within their 
respective watersheds.  Thirty years ago, E. boschungi was scattered throughout the 
Tennessee Valley in small, disjunct populations (Wall and Williams, 1974; Boschung, 
1976a; 1976b; 1979).  A study in 1995 discovered new breeding/non-breeding sites, 
indicating a larger distribution (McGregor and Shepard, 1995).  However, after 
assessment of population levels at these historic sites, only two populations of E. 
boschungi were found to currently exist (Johnston and Hartup, 2002) indicating within 
the last ten years population levels have decreased drastically.  The last mark-recapture in 
2002 showed the population consisted primarily of two- to three-year-old adults at the 
Dodd site, which has been a fairly, stable stronghold over the years.  Historical data 
collected for E. boschungi was for presence/absence rather than to collect population size 
data.  Based on these data as well as our population data, the entire population is 
declining.  In many studies with limited data, count-based information is best viewed as a 
tool that provides relative measures of the health of two or more populations (Morris and 
Doak, 2002).  
 Few data exist for E. brevirostrum and historical population levels were not assessed.  
The total range of the species has been reduced to the mainstem of Shoal Creek.  
Johnston and Phillips (2001) suggest the two populations found in the upstream and 
downstream sites of Highrock Lake are at levels less than 1,000 individuals.  For darters 
 32
when the number of individuals is more than 1,000, it is considered an abundant and 
viable population (Primack, 2002; Shaffer et al., 2002) 
 To assure resiliency in a population, the population needs to be large enough to avoid 
genetic problems, persist for extended periods of time and have the ability to respond to 
favorable conditions in the environment when they occur (Primack, 2002; Shaffer et al., 
2002).  Low numbers for both these species could be attributed to annual variation in 
population sizes, but sampling or data compilation error could also be factors.  Both 
species were difficult to sample.  Etheostoma boschungi appeared to disappear when they 
left the breeding habitat and migrated to the non-breeding habitat.  They were difficult to 
find or sample outside of the breeding season.  This was consistent with published 
monitoring of the species (Boschung, 1976a, 1976b; McGregor and Shepard, 1995).  
Populations could have gone into deeper pools than our equipment or collectors could 
handle or they could have migrated further downstream than presumed.  Many of the 
streams containing historical breeding sites have also been altered (Johnston and Hartup, 
2002).  While E. boschungi is more difficult to find and sample outside of the breeding 
season, determining a way to monitor population levels year-round would provide better 
estimates for survival rates and population size.  
 Etheostoma brevirostrum live in wide streams in the strong currents of riffle 
mesohabitat.  Large, slippery boulders make sampling a challenge.  Populations could 
easily move into crevices of rocks too heavy for collectors to move so it is possible they 
exist in larger numbers than we estimated for this study.  Additionally the rainy, wet 
season of the year is from February to June, which is when both species breed (Kuehne 
and Barbour, 1983).  Rainfall during this time can limit chances of collecting either 
 33
species by flooding sampling sites.  Collectors have to sample larger areas that are less 
confined since stream banks and flood plains often overflow.  This situation can allow 
fish easier escape from collectors, transport them downstream to breed in new locations 
or allow the fish to become prey. 
 Estimates of population size were possibly underestimated for E. boschungi, because 
mark-recaptures were only conducted once per season.  Multiple mark-recapture events 
can help provide a better picture of the population.  For E. brevirostrum, densities from 
defined areas were extrapolated to the upper, lower, and combined reach of Shoal Creek 
occupied by the species.  In doing this, we assumed each segment was one large riffle, 
which could make our numbers higher than in reality.  To solve this, we could estimate 
the volume for all riffles per segment, sum the totals, and calculate density. 
 Estimates of fecundity for both species were based on egg counts for a one-batch 
spawn and two-batch spawn per female.  We determined E. boschungi had a one-batch 
fecundity of 92 ova, which is comparable to the mature ova (102) of a closely related 
species and another member of the Ozarka subgenus, E. trisella (Ryon, 1985).  However, 
two-batch fecundity for E. trisella (256) is higher than our estimate.  Boschung?s (1976) 
fecundity estimates were more comparable to the two-batch fecundity of E. trisella than 
to any of our results.  We suspect Boschung?s estimates represent counts of several 
developmental stages, resulting in larger counts than ours.  We also compared a closely 
related species, E. coosae, which occurs sympatrically with E. brevirostrum in the Coosa 
River.  Etheostoma coosae has a lower fecundity (126; O?Neil, 1981) than what we 
determined for E. brevirostrum.  Historical E. boschungi, E. trisella and E. coosae 
fecundities were likely not as comparable to our results because the fecundities were 
 34
evaluated with varying ova classification methods prior to the current standard by Heins 
and Baker (1988). 
 There are many conflicting methods to determine if a fish spawns multiple times 
during a breeding season, which also affects the fecundity of fishes.  Multiple spawning 
may represent a bet-hedging strategy in warm water streams (Weddle and Burr, 1991).  
Bet-hedging species are those that produce large numbers of eggs and invest very little 
parental care (Promislow and Harvey, 1990).  Although the spawning period for E. 
boschungi is approximately one month, it is unlikely there is enough time to spawn 
multiple batches (Boschung, 1976).   
 Although results from the elasticity analyses of both species show fertilities make the 
largest relative contribution to ?, the component of equation 10 that most changes is the 
survival of the juveniles.  This illustrates that fecundity is not the influential parameter on 
population growth.  Therefore, if the timing of spawning was not long enough and the 
fecundity was not the major determinant of population persistence, it is not important 
whether E. boschungi is a single- or multiple-batch spawner.  For a multiple-batch 
spawner, like E. brevirostrum, fecundity also does not contribute as much to persistence 
of the population.   
 Etheostoma boschungi has persisted as a species at some level for the last 30 years 
even though many of the populations have declined.  It is possible the population at the 
Dodd site could persist for another 30 years and we could rely on the juvenile survival 
rate of 0.016 from Table 8.  The loss of so many locations for the distribution of E. 
boschungi would appear to make it the more imperiled species.  Etheostoma brevirostrum 
appears to be persisting but in small numbers.  It is possible the juvenile survival rates 
 35
could be different for upstream vs. the downstream populations.  If we were to rely on 
published juvenile survival rates of sister species, the number of individuals observed in 
the field ought to be much higher.  It is possible this species has persisted for at least 10 
years (since it was first described) and we could rely on a juvenile survival rate of 0.003 
for the upstream population and 0.005 for the downstream population from Table 14.  It 
is also doubtful the juvenile survival rate is as high as those published for other species in 
Table 12.  Both E. boschungi and E. brevirostrum are egg attachers, which mean there is 
no parental care for larvae and the larvae are not protected from predation or disease 
(Page, 1983).  
 These models are fairly straightforward and with enough information about a fish 
species, they could be useful to resource managers in predicting possible extinction time 
periods and the population trend.  The essential information needed for PVA models is 
the survival rates and population levels broken down by age classes.  Journals are 
publishing less life history papers because many of the species appear to be the same in 
all aspects, however, each species is unique when analyzing survival and fertility rates.  If 
researchers were to continue studying the life history of a species before it became rare, it 
might be possible to prevent population declines.  The more available information about a 
species makes constructing models easier. 
 State agencies should focus on improving the habitat within each of the watersheds to 
improve survival of the juveniles to their first year, while monitoring the watershed more 
closely for possible land use or pollution impacts.  This will correspond with improved 
population growth rates and larger population levels.  More comprehensive status surveys 
should be conducted for each species and historical sites should be revisited on a regular 
 36
basis.  Once more information is gathered, a stochastic model would be more realistic to 
project future population growth even though deterministic models are simpler to design 
and run.  Stochasticity can incorporate chaos and randomness into the model, which can 
mimic drastic changes in the environment as well as predict the effect resource 
management has on the population growth rate.   
  Resource managers should explore new methods or techniques (captive breeding; if a 
method can be found that preserves genetic diversity) to introduce more native species to 
the population or reintroduce species back to the population.  Further studies can provide 
more information for future modeling, which can be incorporated into management plans.  
More than ever the importance of public education for imperiled species should be 
implemented by state agencies.  The public rarely has knowledge of such information as 
to why a species is important to the ecosystem and even to human survival.  We all need 
water to exist and if aquatic species are indicators of the trend of our water quality, 
resource managers should strive to educate and provide more information about 
imperiled species as well as how to assist recovery.  The best case scenario would be to 
inform the public before the species has a high probability of extinction.  If enough 
interest is generated within a community, the public will volunteer to be involved in the 
planning process to help a population recover.  Any one of these efforts would start the 
process for allowing the persistence of these species to improve but immediate 
conservation efforts are needed. 
 37
LITERATURE CITED 
AK?AKAYA, H. R.  2000.  Viability analyses with habitat-based metapopulation models.  
Pop. Ecol. 42:45-53. 
ALLAN, J. D.  1995.  Stream Ecology: Structure and Function of Running Waters.  
Chapman and Hall, NY. 
BART, JR., H. L. AND L. M. PAGE. 1992. The influence of size and phylogeny on life 
history variation in North American Percids. p. 553-572. In: Mayden, R. L. (ed.). 
Systematics, historical ecology, and North American freshwater fishes. Stanford Univ 
Press, Stanford, CA.  
BEISSINGER, S. R. AND M. I. WESTPHAL.  1998.  On the use of demographic models of 
population viability in endangered species management.  J. Wildl. Manage. 62:821-
841. 
----- AND D. R. MCCULLOUGH (EDS.).  2002.  Population Viability Analysis.  Univ. of 
Chicago Press, Chicago. 
BELOVSKY, G. E., C. MELLISON, C. LARSON, AND P. A. VAN ZANDT.  2002.  How good 
are PVA models?  Testing their predictions with experimental data on brine shrimp.  
p. 257-283.  In: Population Viability Analysis.  S. R. Beissinger and D. R. 
McCullough (eds.).  Univ. of Chicago Press, Chicago.  
 38
BOSCHUNG, H. T.  1976a.  An evaluation of the Slackwater Darter, Etheostoma 
boschungi relative to its range, critical habitat, and reproductive habits in the Cypress 
Creek Watershed and adjacent stream systems.  An assessment of the probable 
impacts of the Cypress Creek Watershed project on the Slackwater Darter and its 
critical habitat.  USDA - Soil Conserv. Service, Auburn, AL. 
-----.  1976b.  Amplification of final environmental impact statement for Cypress Creek 
Watershed: Lauderdale County, Alabama and Wayne County, Tennessee.  USDA - 
Soil Conserv. Service, Auburn, AL. 
-----.  1979.  Report on the breeding habitats of the Slackwater Darter (Percidae: 
Etheostoma boschungi) in the Cypress Creek Watershed.  USDA - Soil Conserv. 
Service, Auburn, AL. 
----- AND D. NIELAND. 1986.  Biology and conservation of the Slackwater Darter, 
Etheostoma boschungi (Pisces: Percidae).  Southeast. Fish. Counc. Proc. 4:1-4. 
BOYCE, M. S.  1992.  Population Viability Analysis.  Annu. Rev. Ecol. System. 23:481-
506. 
-----.  2002.  Reconciling the small-population and declining-population paradigms.  p. 
41-49.  In: Population Viability Analysis.  S. R. Beissinger and D. R. McCullough 
(eds.).  Univ. of Chicago Press, Chicago. 
BROOK, B. W., M. A. BURGMAN AND R. FRANKHAM.  2000.  Differences and 
congruencies between PVA packages: the importance of sex ratio for predictions of 
extinction risk.  Conserv. Ecol. 4:6. 
BURGMAN, M. A., S. FERSON AND H. R. AK?AKAYA.  1993.  Risk Assessment in 
Conservation Biology.  Chapman and Hall, NY. 
 39
CARNEY, D. A. AND B. M. BURR. 1989. Life histories of the Bandfin Darter, Etheostoma 
zonistium, and the Firebelly Darter, Etheostoma pyrrhogaster, in Western Kentucky. 
Ill. Nat. Hist. Sur. Biol. Notes 134:1-16. 
CASWELL, H.  1994.  Sensitivity analysis of periodic matrix models.  Ecology 75:1299-
1303. 
-----.  2001.  Matrix Population Models: Construction, Analysis and Interpretation, 2d ed.  
Sinauer Associates, Inc., Sunderland, MA. 
CHAPMAN, D. E. 1951.  Some properties of the hypergeometric distribution with 
applications to zoologic sample censuses.  Univ. Calif. Publ. Stat. 1:131-160. 
CHEN, S. AND S. WATANABE.  1989.  Age dependence and natural mortality coefficient in 
fish population dynamics.  Nippon Suisan Gakkaishi 55:205-208. 
CUMMINGS, K. S., J. M. GRADY, AND B. M. BURR.  1985.  The life history of the Mud 
Darter, Etheostoma asprigene, in Lake Creek, Illinois.  Ill. Nat. Hist. Sur. Biol. Notes 
122:1-16. 
CUTHBERT, R., D. FLETCHER AND L. S. DAVIS.  2001.  A sensitivity analysis of Hutton?s 
shearwater: prioritizing conservation research and management.  Biol. Conserv. 
100:163-172. 
DE KROON, H., J. VAN GROENENDAEL AND J. EHRLEN.  2000.  Elasticities: a review of 
methods and model limitations.  Ecology 81:607-618. 
DOAK, D., P. KAREIVA AND B. KLEPETKA.  1994.  Modeling population viability for the 
desert tortoise in the western Mojave Desert.  Ecol. Appl. 4:446-460. 
ERICKSON, J. E. AND D. C. MAHAN.  1982.  Biology, distribution, and status of the 
Banded Darter, Etheostoma zonale, in Michigan.  Mich. Acad. 14:347-358. 
 40
ETNIER, D. A. AND R. M. BAILEY.  1989.  Etheostoma (Ulocentra) flavum, a new darter 
from the Tennessee and Cumberland River drainages.  Occ. Papers Mus. Zoo. Univ. 
Mich. 717:1-24. 
ETNIER, D. A. AND W. C. STARNES.  1993.  The fishes of Tennessee.  Univ. Tenn. Press, 
Knoxville, TN. 
FAGAN, W. F., E. MEIR, J. PRENDERGAST, A. FOLARIN AND P. KARIEVA.  2001.  
Characterizing population vulnerability for 758 species.  Ecol. Lett. 4:132-138. 
FERRARA, A. M.  2001.  Life-history strategy of Lepisosteidae: implications for the 
conservation and management of alligator gar.  Unpubl.  Ph.D. diss., Auburn Univ., 
Auburn, AL. 
GORDON, N. D., T. A. MCMAHON AND B. L. FINLAYSON.  1992.  Stream Hydrology: an 
Introduction for Ecologists.  John Wiley and Sons, Inc., NY. 
HAMBRICK, P. S. AND H. W. ROBISON.  1979.  Life history aspects of the Paleback Darter, 
Etheostoma pallididorsum (Pisces: Percidae), in the Caddo River system, Arkansas.  
Southwest. Natur. 24:475-484. 
HANSKI, I.  2002.  Metapopulations of animals in highly fragmented landscapes and 
population viability analysis.  p. 86-108.  In: Population Viability Analysis.  S. R. 
Beissinger and D. R. McCullough (eds.).  Univ. of Chicago Press, Chicago. 
HEINS, D. C. AND J. A. BAKER.  1988.  Egg sizes in fishes: do mature oocytes accurately 
demonstrate size statistics of ripe ova?  Copeia 1988:238-240. 
----- AND -----.  1989.  Growth, population structure, and reproduction of the percid fish 
Percina vigil.  Copeia 1989:727-736. 
 41
----- AND -----.  1993a.  Reproductive biology of the brighteye darter, Etheostoma 
lynceum, (Teleostei: Percidae) from the Homochitto River, MS.  Icthyol. Explor. 
Fresh. 4:11-20. 
----- AND -----.  1993b.  Clutch production in the darter Etheostoma lynceum and its 
implications for life-history study.  J. Fish Biol. 42:819-829. 
HEPPELL, S. S., J. R. WALTERS AND L. B. CROWDER.  1994.  Evaluating management 
alternatives for red-cockaded woodpeckers: a modeling approach.  J. Wildl. Manage. 
58:479-487. 
-----, D. T. CROUSE AND L. B. CROWDER.  2000a.  Using matrix models to focus research 
and management efforts in conservation.  p 148-168.  In: Quantitative methods for 
conservation biology.  S. Ferson and M. Burgman (eds.).  Springer-Verlag New York, 
Inc., NY.   
-----, C. PFISTER AND H. DE KROON.  2000b.  Elasticity analysis in population biology: 
methods and applications.  Ecology 81:605. 
HEYDE, C. C., AND J.E. COHEN.  1985.  Confidence intervals for demographic projections 
based on products of lambda matrices.  Theor. Pop. Biol. 27:122-153. 
HOENIG, J. M.  1983.  Empirical use of longevity data to estimate mortality rates.  Fish. 
Bull. 82:898-903. 
HOTALLING, D. R. AND C. A. TABER.  1986.  Aspects of the life history of the Stippled 
Darter, Etheostoma punctulatum.  Amer. Midland Natur. 117:428-434. 
JEARLD, JR., A.  1983.  Age determination.  p 301-324.  In: Fisheries Techniques.  L. A. 
Nielsen and D. L. Johnson (eds.).  American Fisheries Society, MD. 
 42
JENSEN, A. L.  1996.  Beverton and Holt life history invariants result from optimal trade-
off of reproduction and survival.  Can. J. Fish. Aqu. Sci. 53:820-822. 
JOHNSTON, C. E. 1995.  Spawning behavior of the Paleback Darter, Etheostoma 
pallididorsum, (Percidae).  Southwest. Natur. 40:422-424. 
----- AND W. R. HAAG.  1996.  Life history of the Yazoo Darter (Percidae: Etheostoma 
raneyi), a species endemic to north-central Mississippi.  Tul. Stud. Zool. Bot. 30:47-
60. 
----- AND C. L. KNIGHT.  1999.  Life-history traits of the bluenose shiner, Pteronotropis 
welaka (Cypriniformes: Cyprinidae).  Copeia 1999:200-205. 
-----.  2000.  Movement patterns of imperiled blue shiners (Pisces: Cyprinidae) among 
habitat patches.  Ecol. Fresh. Fish. 9:170-176. 
----- AND B. W. PHILLIPS.  2001.  Distribution and habitat of the Holiday Darter in Shoal 
Creek, Talladega National Forest, Alabama.  Ala. Div. Wildl. Freshwater Fish. 
Section 6 Project, Montgomery, AL. 
----- AND W. W. HARTUP.  2001.  Proximity to spawning habitat and reproductive success 
in Etheostoma boschungi: how far can darters migrate?  Ala. Div. Wildl. Freshwater 
Fish. Section 6 Project, Montgomery, AL. 
----- AND -----.  2002.  Status survey for Etheostoma boschungi, Slackwater Darter.  
USFWS, Jackson, MS. 
KREBS, C. J.  1985.  Ecology: The Experimental Analysis of Distribution and Abundance, 
3d ed.  Harper and Row, Publishers, Inc., NY.   
KUEHNE, R. A. AND R. W. BARBOUR.  1983.  The American Darters.  Univ. Press of 
Kentucky, Lexington, KY. 
 43
LABBE, T. R. AND K. D. FAUSCH.  2000.  Dynamics of intermittent stream habitat regulate 
persistence of a threatened fish at multiple scales.  Ecol. Appl. 10:1774-1791. 
LESLIE, P. H.  1945.  On the use of matrices in certain population mathematics.  
Biometrika 33:183-212. 
MANN, C. C. AND M. L. PLUMMER.  1999.  A species' fate, by the numbers.  Science 
284:36. 
MCCARTHY, M. A., H. P. POSSINGHAM, J. R. DAY AND A. J. TYRE.  2001.  Testing the 
accuracy of population viability analysis.  Conserv. Biol. 15:1030-1038. 
MCGREGOR, S. W. AND T. E. SHEPARD.  1995.  Investigations of Slackwater Darter, 
Etheostoma boschungi, populations in the Cypress Creek, Swan Creek, and Flint 
River systems of north Alabama, 1992-1994.  Ala. Dept. Conserv. Nat. Resources, 
Montgomery, AL. 
METTEE, M. F., P. E. O?NEIL, AND J. M. PIERSON.  1996.  Fishes of Alabama and the 
Mobile Basin.  Oxmoor House, Birmingham, AL. 
MILLS, L. S. AND M. S. LINDBERG.  2002.  Sensitivity analysis to evaluate the 
consequences of conservation actions.  p. 338-366.  In: Population Viability Analysis.  
S. R. Beissinger and D. R. McCullough (eds.).  Univ. of Chicago Press, Chicago. 
MORRIS, W. F. AND D. F. DOAK.  2002.  Quantitative Conservation Biology: Theory and 
Practice of Population Viability Analysis.  Sinauer Associates, Inc., Sunderland, MA. 
MURPHY, B. R. AND D. W. WILLIS (EDS.).  1996.  Fisheries Techniques, 2d ed.  American 
Fisheries Society, MD. 
O?NEIL, P. E.  1981.  Life history of Etheostoma coosae (Pisces: Percidae) in Barbaree 
Creek, Alabama.  Tul. Stud. Zool. Bot. 23:75-83. 
 44
PAGE, L.M. AND P. W. SMITH.  1970.  The life history of the Dusky Darter, Percina 
sciera, in Embarras River, Illinois.  Ill. Nat. Hist. Sur. Biol. Notes 69:1-15. 
-----.  1975.  The life history of the Stripetail Darter, Etheostoma kennicotti, in Big Creek, 
Illinois.  Ill. Nat. Hist. Sur. Biol. Notes 93:1-15. 
----- AND B. M. BURR.  1976.  The life history of the Slabrock Darter, Etheostoma smithi, 
in Ferguson Creek, Kentucky.  Ill. Nat. Hist. Sur. Biol. Notes 99:1-12. 
----- AND R. L. MAYDEN.  1981.  The life history of the Tennessee Snubnose Darter, 
Etheostoma simoterum, in Brush Creek, Tennessee.  Ill. Nat. Hist. Sur. Biol. Notes 
117:1-11. 
-----.  1983.  Handbook of darters.  TFH Publications, Inc., Neptune City, NJ. 271 pp. 
-----.  1985.  Evolution of reproductive behaviors in percid fishes.  Ill. Nat. Hist. Surv. 
Bull. 33:275-295. 
PAULY, D.  1980.  On the interrelationship between natural mortality, growth parameters, 
and mean environmental temperature in 175 fish stocks.  J. du Conseil International 
pour l?Exploration del la Mer 39:175-192. 
PETERSON, I. AND J. S. WROBLEWSKI.  1984.  Mortality rate of fishes in the pelagic 
ecosystem. Can. J. Fish. Aqu. Sci. 41:1117-1120. 
PORTER, B. A., T. M. CAVENDER AND P. A. FUERST.  2002.  Molecular phylogeny of the 
Snubnose Darters, Subgenus Ulocentra (Genus Etheostoma, Family Percidae).  Mol. 
Phylogenet. Evol. 22:364-374. 
POWERS, S. L. AND R. L. MAYDEN.  2003.  Etheostoma cervus: a new species from the 
Forked Deer River system in western Tennessee with comparison to Etheostoma 
pyrrhogaster (Percidae: Subgenus Ulocentra).  Copeia 2003:576-582. 
 45
PRIMACK, R. B.  2002.  Essentials of Conservation Biology, 3d ed.  Sinauer Associates, 
Inc., Sunderland, MA. 
PROMISLOW, D. W. L. AND P. H. HARVEY.  1990.  Living fast and dying young: a 
comparative analysis of life-history variation among mammals.  J. Zool. 220:417-437. 
QUINN, T. J., II. AND R. B. DERISO.  1989.  Quantitative Fish Dynamics.  Oxford 
University Press, NY. 
RICKER, W. E.  1975.  Computation and interpretation of biological statistics of fish 
populations.  Bull. Fish. Res. Board Canada. 191. 
RYON, M. G.  1986.  The life history and ecology of Etheostoma trisella (Pisces: 
Percidae).  Am. Mid. Nat. 115:73-86. 
SHAFFER, M., L. H. WATCHMAN, W. L. SNAPE III, AND I. K. LATCHIS.  2002.  Population 
viability analysis and conservation policy.  p. 123-142.  In: Population Viability 
Analysis.  S. R. Beissinger and D. R. McCullough (eds.).  Univ. of Chicago Press, 
Chicago. 
SLACK, W. T., S. T. ROSS AND J. A. EWING, III.  2004.  Ecology and population structure 
of the bayou darter, Etheostoma rubrum: disjunct riffle habitats and downstream 
transport of larvae.  Enviro. Biol. Fish. 71:151-164. 
SLOOTEN, E., D. FLETCHER AND B. L. TAYLOR.  2000.  Accounting for uncertainty in risk 
assessment: case study of Hector's dolphin mortality rate due to gillnet entanglement.  
Conserv. Biol. 14:1264-1270. 
SMITH, R. K. AND K. D. FAUSCH.  1997.  Thermal tolerance and vegetation preference of 
Arkansas Darter and Johnny Darter from Colorado Plains Streams.  Trans. Am. Fish. 
Soc. 126:676-686. 
 46
SPELLERBERG, I. F.  1991.  Monitoring Ecological Change.  Cambridge Univ. Press, NY. 
SUTTKUS, R. D. AND D. A. ETNIER.  1991.  Etheostoma tallapoosae and E. brevirostrum, 
two new darters, subgenus Ulocentra, from the Alabama River drainage.  Tul. Stud. 
Zool. Bot. 28:1-24.   
TABER, C. A., B. A. TABER, AND M. S. TOPPING.  1986.  Population structure, growth and 
reproduction of the Arkansas Darter, Etheostoma cragini (Percidae).  Southwest. 
Natur. 31:207-214. 
TOEPFER, C. S.  1997.  Population and conservation biology of the threatened Leopard 
Darter.  Unpubl. Ph.D. diss., Okla. State Univ. 
U.S. ENVIRONMENTAL PROTECTION AGENCY.  1997.  Volunteer stream monitoring: a 
methods manual.  EPA 841-B97-003. 
USDA FOREST SERVICE.  2000.  Annual monitoring and evaluation report for FY 2000.  
National Forests in Alabama, Montgomery, AL. 
-----.  2004.  Revised land and resource management plan.  National Forests in Alabama, 
Montgomery, AL. 
U.S. FISH AND WILDLIFE SERVICE.  1984.  Slackwater Darter recovery plan. U.S. Fish and 
Wildlife Service, Atlanta, GA. 
-----.  1990.  Endangered and threatened species recovery program: report to Congress. 
ULTSCH, G. R., H. T. BOSCHUNG, AND M. J. ROSS.  1978.  Metabolism, critical oxygen 
tension, and habitat selection in darters (Etheostoma).  Ecology 59:99-107. 
VIVES, S. P.  1987.  Life history of Etheostoma punctulatum (Pisces: Percidae) in 
northeastern Oklahoma.  Southwest. Natur. 32:439-447. 
 47
WALL, B. R. AND J. D. WILLIAMS.  1974.  Etheostoma boschungi, a new percid fish from 
the Tennessee River drainage in northern Alabama and western Tennessee.  Tul. 
Stud. Zool. Bot. 18:172-172. 
WEDDLE, G. K. AND B. M. BURR.  1991.  Fecundity and dynamics of multiple spawning 
in darters: an in-stream study of Etheostoma rafinesquei.  Copeia 1991:419-433. 
WILLIAMS, J. D. AND H. W. ROBISON.  1980.  Ozarka, a new subgenus of Etheostoma 
(Pisces: Percidae).  Brimleyana 4:149-156. 
WILLIAMS, L. R., A. A. ECHELLE, C. S. TOEPFER, M. G. WILLIAMS AND W. L. FISHER.  
1999.  Simulation modeling of population viability for the Leopard Darter (Percidae: 
Percina Pantherina).  Southwest. Nat. 44:470-477. 
WILSON, P. H.  2002.  Using population projection matrices to evaluate recovery 
strategies for Snake River Spring and Summer Chinook Salmon.  Conserv. Biol. 
17:782-794. 
 48
TABLE 1.  MARK-RECAPTURE SAMPLING RESULTS FROM THE 2001 AND 2002 BREEDING 
SEASONS OF Etheostoma boschungi.  Confidence limits in parentheses are calculated with 
Poisson distribution. 
 
 
 
 
 
Females Males Total Females Males Total
Marked (m )67 63 130 29 8 37
Caught (c ) 54 87 141 29 13 42
Recapture (r )15 16 31 1 1 2
(8.4-24.8) (9.2-26) (21-44) (0.1-5.6) (0.1-5.6) (0.2-7.2)
Population (? ) 234 331 581 450 63 545
(145-397.9) (208.6-552.2) (413.4-845.5) (136.4-818.2) (19.1-114.5) (199.3-1361.7)
2001 2002
 49
TABLE 2.  MARK-RECAPTURE RESULTS FOR FEMALE Etheostoma boschungi DATA IN 2001 
AND 2002.  n
0
 = juveniles, n
1
 = one-year-old adults, n
2
 = two-year-old adults, n
3
 = three-
year-old adults, ? = estimated population size per age class with 95 percent confidence 
limits in parentheses, mean size in mm SL is followed by ? 1 standard deviation (SD) 
with range in parentheses. 
 
 
 
 
 
Proportion ? Mean Size ? SD Proportion ? Mean Size ? SD
n
0
-- --
n
1
89.6% 209 33.2 ? 2.77 31.0% 140 36.2 ? 1.90
(129.9-356.5) (27-39) (42.3-253.9) (32-39)
n
2
6.6% 15 42.9 ? 2.65 58.6% 264 43.7 ? 2.41
(9.6-26.3) (40-48) (79.9-479.6) (40-48)
n
3
3.8% 9 50 ? 0 10.3% 47 50.3 ? 1.35
(5-15) (49-54) (14.1-84.6) (49-54)
Total 234 450
2001 Females 2002 Females
 50
TABLE 3.  SUMMARY OF 2002 TRANSECT DATA, RIFFLE VOLUME, AND STREAM LENGTH 
ON SHOAL CREEK.  Length, width and depth from transects were measured in meters and 
stream segments were measured with mapping software.  US = upstream of Highrock 
Lake, DS = downstream of Highrock Lake, Combo = upstream and downstream of 
Highrock Lake combined, SL = standard length, SD = standard deviation. 
 
 
 
 
 
Length ? SD Width ? SD Depth ? SD
US 159.96 ? 3.60 4.29 ? 2.24 0.29 ? 0.27 198.28 3818.5
DS 124.43 ? 3.87 6.19 ? 3.34 0.23 ? 0.20 173.64 2475.3
Combo 284.39 371.92 6531.7
Transect Summary
Length of Stream 
Segment (m)
Riffle Volume (m
3
)
 51
TABLE 4.  NUMBER OF Etheostoma brevirostrum COLLECTED IN THE FIELD, THE 
ESTIMATED NUMBER OF FISH PER M? AND THE EXTRAPOLATED POPULATION SIZE FOR 
EACH AGE CLASS.  Values in parentheses are 95 percent confidence limits.  US = 
upstream of Highrock Lake, DS = downstream of Highrock Lake, Combo = upstream and 
downstream of Highrock Lake combined, n
1
 = one-year-old adults, n
2
 = two-year-old 
adults, n
3
 = three-year-old adults, ? = extrapolated population size per age class. 
 
 
 
 
Collected
Fish/m
3
? Collected Fish/m
3
?
US n
1
28 0.141 539 3 0.015 58
(27.06-28.94) (0.136-0.146) (521-557) (0.39-5.61) (0.002-0.028) (7-108)
n
2
8 0.040 154 4 0.020 77
(7.36-8.64) (0.037-0.044) (142-166) (2.53-5.47) (0.013-0.028) (49-105)
n
3
2 0.010 39 -
(0.14-3.86) (0.001-0.019) (3-74)
DS n
1
26 0.150 371 29 0.167 413
(25.31-26.69) (0.146-0.154) (361-381) (28.16-29.84) (0.162-0.172) (401-425)
n
2
2 0.012 29 5 0.029 71
(0.73-3.27) (0.004-0.019) (10-47) (3.70-6.30) (0.021-0.036) (53-90)
n
3
- 1 0.006 14
Combo n
1
54 0.145 948 32 0.086 562
(53.40-54.60) (0.144-0.147) (938-959) (31.21-32.79) (0.084-0.088) (548-576)
n
2
10 0.027 176 9 0.024 158
(9.40-10.60) (0.025-0.029) (165-186) (8.09-9.91) (0.022-0.027) (142-174)
n
3
2 0.005 35 1 0.003 18
(0.14-3.86) (0.000-0.010) (2-68)
2001 2002
 52
TABLE 5.  LIFE HISTORY TABLE FOR Etheostoma boschungi.  This table uses one-batch 
fecundity.  All parameters are the same with the two-batch fecundity.  See text for 
sources and definitions. 
 
 
 
 
 
 
 
 
 
 
Age (yr) 
(i )
2001 
(n
i
)
P
i
L
i
d
i
q
i
T
i
e
i 
(yr) 
f
i
F
i
0-
1 209 130 357 0.074 112 194 0.926 129.00 0.62 92
2 15 10 26 0.571 12 7 0.429 16.54 1.07 92
3 9 5 15 0 4 9 1.000 4.41 0.50 92
? 234 145 398
95% CL
 53
TABLE 6.  SPECIES CLOSELY RELATED TO Etheostoma boschungi.  P
i
 = survival rates for 
each age class, f
i
 = mature ova or fecundity, % = assumption that each age class was 
collected in proportion to its relative number in the population, that the population was 
neither increasing nor decreasing, and that the number of juveniles entering the 
population each year was constant. 
 
 
 
Subgenus Species f
i
P
0
P
1
P
2
P
3
P
4
References
Ozarka Etheostoma cragini 380 - 0.047 0.429 0 a, e, f, g, i
94% 4% 2%
Ozarka Etheostoma pallididorsum 78 - 0.539 0 a, c, e, g
65% 35%
Ozarka Etheostoma punctulatum 524 0.549 0.855 0.264 0 a, d, e, g, j
47% 26% 22% 6%
Ozarka Etheostoma trisella 102 - 0.396 0 a, e, g, h
Fuscatelum Etheostoma parvipinne 0 b, e, g
Psycromaster Etheostoma tuscumbia 78 0 a, e, g
d
 Hotalling and  Taber, 1986
i
 Taber, Taber, and Topping, 1986
j
 Vives, 1987
Survival and Population Proportion
e
 Kuehne and Barbour, 1983
f
 Labbe and Fausch, 2000
g
 Page, 1983
h
 Ryon, 1985
a
 Bart, Jr. and Page, 1992
b
 Etnier and Starnes, 1993
c
 Hambrick and Robison, 1979
 54
TABLE 7.  ESTIMATED ANNUAL SURVIVAL RATES FROM FAST WITH Etheostoma 
boschungi  DATA.  F
i
 = adult fertility rates and P
0
 = juvenile survival rate (derived from 
one-batch and two-batch fecundities). 
 
 
 
 
 
Method F
i
one-batch two-batch
Chen/Watanabe, 1989 0.226 0.783 0.017 0.008
Hoenig, 1983 0.242 0.769 0.017 0.008
Jensen, 1996 0.153 0.850 0.018 0.009
Pauly, 1980 0.199 0.808 0.018 0.008
Peterson/Wroblewski, 1984 0.207 0.800 0.017 0.008
Quinn/Deriso, 1999 (used 1%) 0.215 0.792 0.017 0.008
Quinn/Deriso, 1999 (used 5%) 0.368 0.665 0.014 0.007
P
0
Estimated 
Survival
 55
TABLE 8.  ESTIMATED POPULATION SIZE PROJECTIONS FOR Etheostoma boschungi.   
F
i
 = adult fertility rates, f
i
 = one-batch (92 ova) and two-batch (197 ova) fecundity, P
i
 = 
survival rates, ? = population growth rate, and ? = total population size. 
 
 
P
1
P
2
P
3
P
0
f
i
F
i
?
? < 1 after 
i yrs
?  > 1000 
in i  yrs
0.074 0.571 0 0.004 92 0.184 0.309 5
0.074 0.571 0 0.002 197 0.197 0.322 5
0.074 0.571 0 0.005 92 0.230 0.355 6
0.074 0.571 0 0.006 92 0.276 0.400 6
0.074 0.571 0 0.007 92 0.322 0.444 7
0.074 0.571 0 0.003 197 0.296 0.419 7
0.074 0.571 0 0.008 92 0.368 0.488 8
0.074 0.571 0 0.009 92 0.414 0.533 9
0.074 0.571 0 0.010 92 0.460 0.577 9
0.074 0.571 0 0.004 197 0.394 0.513 9
0.074 0.571 0 0.005 197 0.493 0.608 10
0.074 0.571 0 0.011 92 0.506 0.621 11
0.074 0.571 0 0.012 92 0.552 0.666 13
0.074 0.571 0 0.013 92 0.598 0.710 15
0.074 0.571 0 0.006 197 0.591 0.703 15
0.074 0.571 0 0.014 92 0.644 0.755 19
0.074 0.571 0 0.015 92 0.690 0.799 24
0.074 0.571 0 0.007 197 0.690 0.799 24
0.074 0.571 0 0.016 92 0.736 0.844 32
0.074 0.571 0 0.017 92 0.782 0.889 46
0.074 0.571 0 0.008 197 0.788 0.894 48
0.074 0.571 0 0.018 92 0.828 0.933 79
0.074 0.571 0 0.019 92 0.874 0.978 >100
0.074 0.571 0 0.009 197 0.887 0.990 >100
0.074 0.571 0 0.020 92 0.920 1.023 64
0.074 0.571 0 0.021 92 0.966 1.068 22
0.074 0.571 0 0.010 197 0.985 1.087 18
0.074 0.571 0 0.022 92 1.012 1.113 14
0.074 0.571 0 0.023 92 1.058 1.158 10
0.074 0.571 0 0.011 197 1.084 1.184 9
0.074 0.571 0 0.024 92 1.104 1.204 8
0.074 0.571 0 0.025 92 1.150 1.249 7
0.074 0.571 0 0.012 197 1.182 1.280 6
0.074 0.571 0 0.013 197 1.281 1.377 5
 56
TABLE 9.  LIFE HISTORY TABLE FOR Etheostoma brevirostrum.  This table uses two-batch 
fecundity.  US = upstream of Highrock Lake, DS = downstream of Highrock Lake, 
Combo = upstream and downstream of Highrock Lake combined.  See text for sources 
and definitions. 
 
 
 
 
 
Age (yr) 
(i )
2001 
(n
i
)
P
i
L
i
d
i
q
i
T
i
e
i 
(yr) 
f
i
F
i
0-
1 539 521 557 0.286 347 385 0.714 462.20 0.86 175
2 154 142 166 0.250 96 116 0.750 115.55 0.75 175
3 39 3 74 0 19 39 1.000 19.26 0.50 175
? 732 666 798
Age (yr) 
(i )
2001 
(n
i
)
P
i
L
i
d
i
q
i
T
i
e
i 
(yr) 
f
i
F
i
0-
1 371 361 381 0.077 200 342 0.923 213.83 0.58 175
2 29 10 47 0 14 29 1.000 14.26 0.50 175
3 0 0 0 175
? 399 371 427
Age (yr) 
(i )
2001 
(n
i
)
P
i
L
i
d
i
q
i
T
i
e
i 
(yr) 
f
i
F
i
0-
1 948 938 959 0.186 562 772 0.814 684.68 0.72 175
2 176 165 186 0.197 105 141 0.803 122.68 0.70 175
3 35 2 68 0 17 35 1.000 17.34 0.50 175
? 1159 1105 1213
Combo 95% CL
95% CL
95% CLDS
US
 57
TABLE 10.  ELASTICITY VALUES FOR Etheostoma brevirostrum.  See text for sources and 
definitions. 
 
 
 
F
i
P
i
1 0.286 0.737 0.560 0.200
US 2 0.250 0.737 0.160 0.040
3 0 0.737 0.040 0
1 0.077 0.929 0.867 0.067
DS 2 0 0.929 0.067 0
3 0.929 0 0
1 0.186 0.818 0.675 0.150
Combo 2 0.197 0.818 0.125 0.025
3 0 0.818 0.025 0
F
i
n
i
P
i
Elasticity
 58
TABLE 11.  ESTIMATED ANNUAL SURVIVAL RATES FROM FAST WITH Etheostoma 
brevirostrum  DATA.  F
i
 = adult fertility rates and P
0
 = juvenile survival rate (derived 
from two-batch fecundities). 
 
 
 
 
 
 
Method
Estimated 
Survival
F
i
P
0
Chen/Watanabe, 1989 0.378 0.658 0.008
Hoenig, 1983 0.242 0.769 0.009
Jensen, 1996 0.348 0.681 0.008
Pauly, 1980 0.425 0.623 0.007
Peterson/Wroblewski, 1984 0.177 0.827 0.009
Quinn/Deriso, 1999 (used 1%) 0.215 0.792 0.009
Quinn/Deriso, 1999 (used 5%) 0.368 0.665 0.008
 59
TABLE 12.  SPECIES CLOSELY RELATED TO Etheostoma brevirostrum.  P
i
 = survival rates 
for each age class, f
i
 = mature ova or fecundity, % = assumption that each age class was 
collected in proportion to its relative number in the population, that the population was 
neither increasing nor decreasing, and that the number of juveniles entering the 
population each year was constant. 
 
 
 
 
 
 
 
Subgenus Species f
i
P
0
P
1
P
2
P
3
P
4
References
Ulocentra Etheostoma coosae 88 0.556 0.186 0 a, g, h, i, k
60% 33% 6%
Ulocentra Etheostoma pyrrhogaster 0.761 0.286 0 a, b, e, g, k
51% 38% 11%
Ulocentra Etheostoma raneyi 52 - 0.104 0.104 0 f, k
47% 6%
Ulocentra Etheostoma simoterum 152 0.123 0 a, g, i, j, k
89% 11%
Ulocentra Etheostoma zonale 40 - 0.180 0.182 - 0 a, c, e, g, i, k
82% 15% 3%
Ulocentra Etheostoma zonistium 0.728 0.181 0 a, b, e, k
54% 39% 7%
Ulocentra Etheostoma baileyi 35 0 e, g, k
Ulocentra Etheostoma barrenense 800 0 e, g, k
Ulocentra Etheostoma duryi 0e, g, i, k
Ulocentra Etheostoma etnieri , g, i, k
Ulocentra Etheostoma flavum 0d, e g, k
1=28.4,   
2,3=144
1=77,    
2,3=128
k
 Porter, Cavender and Fuerst, 2002
c
 Erickson and Mahan, 1982
e
 Etnier and Starnes, 1993
d
 Etnier and Bailey, 1989
f
 Johnston and Haag, 1996
a
 Bart, Jr. and Page, 1992
b
 Carney and Burr, 1989
i
 Page, 1983
j
 Page and Mayden, 1981
g
 Kuehne and Barbour, 1983
h
 O'Neil, 1981
Survival and Population Proportion
 60
TABLE 13.  ESTIMATED FERTILITY RATES AND GROWTH RATE FROM CLOSELY RELATED 
SPECIES WITH Etheostoma brevirostrum.  P
0
 = juvenile survival, F
i
 = estimated adult 
fertility rate, ? = growth rate, SAD = stable age distribution and Reprod = reproductive 
values. 
 
US DS COMBO US DS COMBO US DS COMBO
P
0
0.56 0.56 0.56 0.76 0.76 0.76 0.12 0.12 0.12
F
i
48.61 48.61 48.61 66.58 66.58 66.58 10.76 10.76 10.76
? 48.90 48.69 48.80 66.86 66.65 66.76 11.04 10.83 10.94
?  after 1 yr 35,767   19,432 56,536  48,914 26,602 77,351 8,066 4,323 12,676
1 0.994 0.994 0.996 0.996 0.999 0.997 0.974 0.993 0.983
SAD 2 0.006 0.006 0.004 0.004 0.001 0.003 0.025 0.007 0.017
3 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.000 0.000
1 0.334 0.334 0.334 0.334 0.334 0.334 0.337 0.335 0.34
Reprod 2 0.334 0.334 0.334 0.334 0.333 0.334 0.335 0.333 0.34
3 0.332 0.332 0.332 0.332 0.333 0.333 0.328 0.333 0.33
Etheostoma coosae Etheostoma pyrrhogaster Etheostoma simoterum
 61
TABLE 14.  ESTIMATED POPULATION SIZE PROJECTIONS FOR Etheostoma brevirostrum.  F
i
 
= adult fertility rates, f
i
 = two-batch fecundity, P
i
 = survival rates, ? = population growth 
rate, and ? = total population size. 
 
 
P
1
P
2
P
3
P
0
f
i
F
i
?
? < 1 after i 
yrs
?  > 1000 in i 
yrs
Combo 0.186 0.197 0 0.001 175 0.088 0.224 4
DS 0.077 0 0.004 175 0.350 0.415 6
US 0.286 0.250 0 0.001 175 0.079 0.254 6
US 0.286 0.250 0 0.001 175 0.088 0.268 7
Combo 0.186 0.197 0 0.003 175 0.263 0.429 8
US 0.286 0.250 0 0.002 175 0.175 0.387 8
DS 0.077 0 0.005 175 0.438 0.504 9
US 0.286 0.250 0 0.003 175 0.263 0.492 9
DS 0.077 0 0.006 175 0.525 0.593 11
US 0.286 0.250 0 0.004 175 0.350 0.591 12
Combo 0.186 0.197 0 0.005 175 0.438 0.613 14
DS 0.077 0 0.007 175 0.613 0.682 15
US 0.286 0.250 0 0.005 175 0.438 0.686 17
DS 0.077 0 0.008 175 0.700 0.770 22
US 0.286 0.250 0 0.006 175 0.525 0.779 26
Combo 0.186 0.197 0 0.007 175 0.613 0.792 30
DS 0.077 0 0.009 175 0.788 0.858 39
US 0.286 0.250 0 0.007 175 0.613 0.871 47
Combo 0.186 0.197 0 0.008 175 0.700 0.881 >50
DS 0.077 0 0.010 175 0.875 0.946 >50
DS 0.077 0 0.012 175 1.050 1.122 8
US 0.286 0.250 0 0.010 175 0.875 1.142 3
Combo 0.186 0.197 0 0.010 175 0.875 1.057 0
 62
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 1.  Historical locations (circles) and sampling site (star) of Etheostoma boschungi 
within tributaries of the Tennessee River drainage, USA. Letters on the map correspond 
to the following: a) South Fork Buffalo River and Chief Creek of Buffalo River, b) Shoal 
Creek, c) Cypress Creek, d) Swan Creek, and e) Flint River. 
a 
b
c 
d
e 
TENNESSEE 
ALABAMA 
 63
 
 
 Fig. 2.  Historical locations (circles) and sampling sites (star) of Etheostoma 
brevirostrum within tributaries of the Coosa River drainage, USA.  Letters on the map 
correspond to the following: a) Conasauga River, b) Coosawattee River, c) Etowah River, 
and d) Choccolocco Creek. 
 
c 
a 
d
TENNESSEE 
 
ALABAMA  GEORGIA 
b 
 64
 Fig. 3.  Map of the Dodd Site, Middle Cypress Creek on Dodd Road, Wayne County, 
Tennessee (S = small seepage stream, M = area where the larger seepage stream joins 
with Middle Cypress Creek) (TERRASERVER-USA: An online provider of USGS digital 
maps [web application], http://terraservice.net/, 2004, unpubl.). 
 
 
 
 65
 
 
 
 Fig. 4.  Sampling locations (circles) of E. brevirostrum within Shoal Creek of the 
Coosa River drainage in Alabama, USA. 
 
 
 66
 
 
 
 
 Fig. 5.  A three-stage, birth-pulsed, pre-breeding model for adult populations of E. 
boschungi and E. brevirostrum.  Circles denote stages in the age-structured model and 
arrows indicate transition probabilities. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 67
 
 Fig. 6.  Length-frequency distribution of collected Etheostoma boschungi from 
selected years.  A = University of Alabama museum specimens, B = (McGregor and 
Shepard, 1995) C = 2001-2002 census data. 
B 
C 
A 
 68
 Fig. 7. Length-frequency distribution of collected Etheostoma brevirostrum from 
2001 and 2002 in Shoal Creek. 
 
0
2
4
6
8
10
12
21 24 27 30 33 36 39 42 45 48 51
Length
Fr
eq
uenc
y
2001 Combined
2002 Combined
 69
 
 
 Fig. 8. Example of projected population size per age class of Etheostoma boschungi 
when P
0
 = 0.004, P
1
 = 0.074, P
2
 = 0.571, P
3
 = 0, and F
i
 = 0.184, using a one-batch 
fecundity. 
 
 
 
0
50
100
150
200
250
0 5 10 15 20 25 30 35 40 45 50
Time
E
s
t
i
m
a
te
d Si
z
e
age 1 age 2 age 3 N'
 70
APPENDIX I: Description and map of sampling sites for each species 
 71
Study sites of E. boschungi (circles). 
 72
Code Location
Longitude/Latitude       
(Decimal degree)
System County State
1 Cemetary Branch @ Hwy 5 -87.824980/ 34.981177 Cypress Lauderdale AL
2 Dulin Branch @ Hwy 157 -87.813611/ 35.005278 Cypress Lauderdale AL
3 Greenbriar Branch @ CR 8/CR 139 -87.757863/ 34.941673 Cypress Lauderdale AL
4 Lindsey Ck @ CR152/Wallace Ridge Rd - Z site -87.864201/ 34.961195 Cypress Lauderdale AL
5 Lindsey Ck @ CR8/CR10 -87.870651/ 34.967825 Cypress Lauderdale AL
6 Lindsey Ck @ CR81/154 -87.815310/ 34.925086 Cypress Lauderdale AL
7 Lindsey Ck @ Murphy's Chapel/CR 8 -87.890841/ 34.976175 Cypress Lauderdale AL
8 Lindsey Ck @ Natchez Trace/CR5 -87.831700/ 34.945085 Cypress Lauderdale AL
9
North Fork Cypress Ck, Cemetery Br @ Natchez 
Trace Pkwy - D Austin site
-87.822409/ 34.969915 Cypress Lauderdale AL
10 Threet Creek @ Natchez Trace -87.821567/ 34.956233 Cypress Lauderdale AL
11 Middle Cypress Ck @ CR139 -87.734198/ 34.974254 Cypress Lauderdale AL
12
Middle Cypress Ck @ Dodd Rd (stream crosses 
under road)
-87.770027/ 35.055643 Cypress Wayne TN
13 Middle Cypress Ck @ Dodd Rd -on farm -87.771527/ 35.061713 Cypress Wayne TN
14 Middle Cypress Ck @ Gilchrist Rd -87.764747/ 35.049803 Cypress Wayne TN
15 Middle Cypress Ck @ Middle Cypress Ck Rd -87.750307/ 35.039523 Cypress Wayne TN
16
Middle Cypress Ck @ Pumping Station 
Rd/HWY227
-87.736967/ 35.029253 Cypress Wayne TN
17
Cooper's Br trib @ Natchez Trace Pkwy-seep, 1mi 
from picnic area
-87.823150/ 35.015767 Cypress Wayne TN
18 Cypress Ck @ Trace Pkwy (by park) -87.820859/ 35.029254 Cypress Wayne TN
19 Dulin Branch Trib @ N of Hwy 227 -87.815556/ 35.014444 Cypress Wayne TN
20 Chief Ck @ Hwy 240 -87.425400/ 35.372783 Buffalo Lawrence TN
21 Little Shoal Ck @ Lawrenceburg, Hwy 43 -87.296125/ 35.320359 Shoal Lawrence TN
22
Little Shoal Creek @ Speedway, Dooley Rd off Hwy 
43
-87.285074/ 35.327870 Shoal Lawrence TN
23 Swan Ck @ Elkton Rd/ CR86 -86.951677/ 34.830918 Swan Limestone AL
24 Swan Ck @ Piney Chapel Rd/CR81 - L&N site -86.960567/ 34.848417 Swan Limestone AL
25 Briar Fork trib @ Scott Road -86.677500/ 34.964722 Flint Madison AL
26 Briar Fork trib @ Scott Rd/State Line Rd -86.681667/ 34.992778 Flint Madison AL
27 Briar Fork @ Scott Rd-Scott Orchard -86.675536/ 34.991691 Flint Madison AL
Summary of E. boschungi sampling sites.  Latitude and longitude for each site is shown 
in decimal degrees.  
 
 
 
 73
Study sites of E. brevirostrum located above and below Highrock Lake (circles). 
 
 
 74
Site
US/DS of 
Highrock
Site Description
1US
Shoal Creek, Pine Glenn Recreation Area off Forest Road 531, 8.1 mi. N Hwy 78, Cleburne Co., AL 
(N33.725470/W-85.602710)
3US
Shoal Creek, 5 mi. N Heflin, 1 mi. downstream from Pine Glenn Recreation Area, Cleburne Co., AL 
(N33.722222/W-85.609370)
4DS
Shoal Creek, off Forest Road 531, below Highrock Lake, near Cleburne-Calhoun County Line, Cleburne Co., AL 
(N33.714630/W-85.631850)
17 DS Shoal Creek, 0.5 mi. from end of FS rd. 530, Cleburne Co., AL (N33.710730/W-85.637580)
Summary of E. brevirostrum sampling sites for 2001 and 2002 movement studies (US = 
Upstream site of Highrock Lake Dam, DS = Downstream site of Highrock Lake Dam).  
Latitude and longitude for each site is shown in decimal degrees.