Ecology, 93(1), 2012, pp. 35–45
Ó 2012 by the Ecological Society of America
Life history theory predicts fish assemblage response to
hydrologic regimes
MERYL C. MIMS1
AND
JULIAN D. OLDEN
School of Aquatic and Fishery Sciences, University of Washington, 1122 NE Boat Street, Seattle, Washington 98195 USA
Abstract. The hydrologic regime is regarded as the primary driver of freshwater
ecosystems, structuring the physical habitat template, providing connectivity, framing biotic
interactions, and ultimately selecting for specific life histories of aquatic organisms. In the
present study, we tested ecological theory predicting directional relationships between major
dimensions of the flow regime and life history composition of fish assemblages in perennial
free-flowing rivers throughout the continental United States. Using long-term discharge
records and fish trait and survey data for 109 stream locations, we found that 11 out of 18
relationships (61%) tested between the three life history strategies (opportunistic, periodic, and
equilibrium) and six hydrologic metrics (two each describing flow variability, predictability,
and seasonality) were statistically significant (P 0.05) according to quantile regression. Our
results largely support a priori hypotheses of relationships between specific flow indices and
relative prevalence of fish life history strategies, with 82% of all significant relationships
observed supporting predictions from life history theory. Specifically, we found that (1)
opportunistic strategists were positively related to measures of flow variability and negatively
related to predictability and seasonality, (2) periodic strategists were positively related to high
flow seasonality and negatively related to variability, and (3) the equilibrium strategists were
negatively related to flow variability and positively related to predictability. Our study
provides important empirical evidence illustrating the value of using life history theory to
understand both the patterns and processes by which fish assemblage structure is shaped by
adaptation to natural regimes of variability, predictability, and seasonality of critical flow
events over broad biogeographic scales.
Key words: disturbance; flow regime; freshwater fishes; life history; traits; United States.
INTRODUCTION
in large part, its response to environmental factors
describing the variability, predictability (stability), and
seasonality (duration, timing) of favorable habitat
conditions (Lytle 2001).
In freshwater environments, the importance of the
hydrologic regime for shaping the biophysical attributes
and function of rivers is well recognized (Poff et al. 1997,
Naiman et al. 2008). Intra- and interannual patterns in
river discharge define the habitat template and disturbance regimes of lotic ecosystems (Resh et al. 1988),
promote habitat heterogeneity (Tockner et al. 2000),
provide longitudinal and lateral access to spawning,
recruitment, and foraging habitat (Junk et al. 1989), and
select for life histories of aquatic and riparian organisms
via impacts on fitness such as mortality-inducing floods
and droughts (Lytle and Poff 2004). The consequences of
the hydrologic regime are manifested across both
ecological (growth rates, survival, and effects of fitness
on individuals) and evolutionary time scales during which
the environmental filtering of represented life history
strategies occurs through selection over many generations
(Southwood 1988, Townsend and Hildrew 1994).
Life history theory has sparked a renewed interest and
new perspectives in understanding the patterns and
drivers of animal and plant distributions in freshwater
ecosystems. Although considerable research has focused
Trade-offs among energetic investments in growth,
reproduction, and survivorship have resulted in the
evolution of life history strategies that enable an organism
to cope with ecological challenges (Stearns 1992). Early
investigators recognized a continuum in life history
strategies between what are referred to as r-selected
species (short-lived, small-bodied individuals with high
fecundity and low per capita investment per offspring)
and K-selected species (long-lived, large-bodied individuals with low fecundity and high per capita investment
per offspring) (MacArthur and Wilson 1967, Pianka
1970, Stearns 1976). The ecological theory resulting from
these pioneering efforts acknowledged the role of
environmental pressures in the evolution of optimal life
histories, which is summarized by Southwood’s concept
of the habitat template (‘‘templet’’) that identifies
ecological disturbance as a key determinant of trade-offs
between reproducing ‘‘here’’ vs. ‘‘there’’ and ‘‘now’’ vs.
‘‘later’’ (Southwood 1977, 1988). It is now widely
acknowledged that a species’ life history strategy dictates,
Manuscript received 25 February 2011; revised 13 June 2011;
accepted 22 July 2011. Corresponding Editor: D. R. Strong.
1 E-mail: [email protected]
35
36
MERYL C. MIMS AND JULIAN D. OLDEN
FIG. 1. Life history continuum model adapted from Winemiller (2005) and originally conceptualized in Winemiller and
Rose (1992). Inside arrows summarize fundamental trade-offs
between juvenile survivorship, generation time, and fecundity
that define the three end-point strategies. Outside arrows
summarize predicted relationships between flow dimensions
and life history strategies.
on macroinvertebrates (e.g., Townsend and Hildrew
1994, Statzner et al. 1997, Konrad et al. 2008; reviewed
by Verberk et al. 2008), life history theory has seen only
limited application to advancing our knowledge of
species–environment relationships for freshwater fishes.
This is despite the fact that many studies have used fish
traits as a currency of investigation and that relevant
ecological theory exists (Olden et al. 2010). Winemiller
and Rose (1992) identified three life history strategies in
both freshwater and marine fishes that represent the
essential trade-offs among the basic demographic
parameters of survival, fecundity, and onset and
duration of reproduction (Fig. 1). Opportunistic strategists are small-bodied species with early maturation
and low juvenile survivorship and are predicted to be
associated with habitats defined by frequent and intense
disturbance, mirroring the classic r strategy. Periodic
strategists are characterized by large body size, late
maturation, high fecundity, and low juvenile survivorship and are likely to be favored in highly periodic
(seasonal) environments (see Plate 1). Equilibrium
strategists are typically small to medium in body size
with intermediate times to maturity, low fecundity per
spawning event, and high juvenile survivorship largely
due to high parental care and small clutch size, closely
aligned with the K strategy. Equilibrium strategists are
predicted to be favored in more stable habitats with low
environmental variation. The three life history strategies
of the continuum are hypothesized to be adaptive with
respect to variability, predictability, and seasonality of
environmental regimes (Fig. 1; see also Winemiller
2005).
Convergence of trait composition along hydrologic
gradients has been demonstrated for freshwater fishes
(e.g., Poff and Allan 1995, Lamouroux et al. 2002, Erös
Ecology, Vol. 93, No. 1
2005, Blanck et al. 2007, Logez et al. 2010), but
empirical investigations that test predictions from life
history theory remain scant. Tedesco et al. (2008) found
that species exhibiting the periodic life history in western
Africa occurred predominantly in rivers with seasonal
(high duration, predictable timing) flow regimes; conversely, species with equilibrium strategies occurred in
stable, less-seasonal, drainage basins. Olden and Kennard (2010) found intercontinental life history convergence of fishes in drainage basins of Australia and the
southeast United States along similar gradients of
hydrologic variability, with an increasing prevalence of
opportunistic strategists and a decreasing prevalence of
periodic strategists concurrent with increasing hydrologic variability. Together, these studies have provided
initial support for the hypothesized response of fish life
history strategies to hydrologic conditions, but they
have investigated only single dimensions of the flow
regime (variability, predictability, or seasonality), and
hence only tested a subset of life history theory.
In this paper, we test life history theory by quantifying
relationships between variability, predictability, and
seasonality of natural flow regimes and the life history
composition of native fish assemblages throughout the
continental United States. Our study is novel in that it
spans a diverse hydrologic gradient (Poff 1996) involving a diverse taxonomy of fish assemblages whose
regional species pools differ due to unique zoogeographic histories (Mims et al. 2010). Drawing on life history
theory and the natural flow paradigm, we predict that
the prevalence of (1) opportunistic species will be
positively related to flow variability and negatively
related to flow predictability and seasonality, (2)
periodic species will be positively related to flow
seasonality and predictability and negatively related to
flow variability, and (3) equilibrium strategists will be
positively related to flow predictability (stability) and
negatively related to flow variability (Fig. 1, summarized
in Table 1). In addition to advancing our ecological
knowledge of patterns and processes in community
ecology, results from this study also help inform whether
a life history (traits-based) approach is useful in
predicting the response of fish assemblages to altered
flow regimes, a critical question of applied importance in
the emerging arena of ecologically sustainable flow
management.
MATERIAL
AND
METHODS
Species occurrence and life history strategies
Fish occurrence data for sites throughout the continental United States were synthesized from a number of
regional and national survey efforts, including the
Environmental Protection Agency’s (EPA) Environmental Monitoring and Assessment Program (EMAP) and
the regional counterpart (regional EMAP or REMAP;
Hughes et al. 2000), the United States Geological Survey
(USGS) National Water Quality Assessment Program
(NAWQA; Gilliom et al. 1995) and from various state
January 2012
FISH LIFE HISTORIES AND STREAM FLOW
37
TABLE 1. Relationships between the major dimensions of flow variability (V), predictability (P), and seasonality (S) described by
the hydrologic metrics and predicted prevalence of life history strategies according to ecological theory.
Predicted relationship with
proportional life histories Hydrologic Metrics
Abbreviation
Annual coefficient of
variation (V)
High pulse count (V)
AnnCV
Base flow index (P)
Flow predictability (P)
BFI
FPred
Constancy/predictability (S)
High pulse duration (S)
Const/Pred
HPD
HPC
Description
standard deviation of all daily flow values,
divided by the mean annual flow
number of high pulses within each water
year
7-d minimum flow/mean flow for year
constancy (C, a measure of temporal
invariance) þ contingency (M, a
measure of periodicity)
C/(C þ M )
median duration of high pulses (d)
Opportunistic Periodic Equilibrium
þ
þ
0
þ
þ
þ
0
þ
þ
0
Slope direction: positive (þ), negative (), or no slope (0).
agencies. This resulted in a national database of 5951
unique sampling sites (collected between 1989 and 2002)
represented by 530 fish species native to North America
(Herlihy et al. 2006). Because the data available to us
represented one-time surveys spanning a 14-year period
from potentially disparate (but effective and efficient)
survey techniques, we examined species presence/absence rather than abundance. Furthermore, we argue
that this is the appropriate ecological gain to examine
long-term functional responses of fish assemblages to
hydrology.
Fish life history strategies were evaluated according to
the model of Winemiller and Rose (1992) that positions
species along three primary life history axes defining the
opportunistic, periodic, and equilibrium endpoints. The
three axes of the model are (1) ln (fecundity), defined as
the number of eggs or offspring per female per spawning
season, (2) ln (length at maturation), defined as mean
female length at maturation (mm), and (3) ln (egg size
þ1) þ ln (parental care þ1) (termed juvenile investment),
defined as the mean diameter of mature, fully yolked,
ovarian oocytes (mm) and an index of parental care
following Winemiller (1989). Fish life history traits were
obtained from a comprehensive database for freshwater
fishes of the United States and Canada synthesized from
various literature, agency, and expert accounts (reported
in Mims et al. 2010). Because many species occupy
intermediate positions in life history space, we followed
Olden and Kennard (2010) by conducting a ‘‘soft’’
classification according to a species’ relative affinity
toward each strategy. This was accomplished by
calculating the Euclidean distance in trivariate life
history space between the species’ position and the
strategy endpoints, normalizing the distances between 0
and 1, and then computing the inverse of these values to
provide a ‘‘strategy weight’’ for each species. This
resulted in each fish species having a proportional
weight or affinity to the opportunistic, periodic, or
equilibrium strategies. For each site, we summed the
strategy weights of species present and computed the
proportional value for each strategy to represent relative
life history composition.
Study sites
Using the USGS national flow gage network we
identified gage stations representing natural (free-flowing or unimpounded) conditions by narrowing the
national network to 1659 stations included in the
Hydro-Climatic Data Network stations with strictly
defined criteria of measurement accuracy and contributing watersheds with minimal human activities (summarized in Vogel and Sankarasubramanian 2005). We
also identified 799 gages with no upstream large dams
and with minimal anthropogenic impacts to the
upstream drainage from the Geospatial Attributes of
Gages for Evaluating Stream flow, or GAGES, database
(Falcone et al. 2010).
Next, we identified 109 pairs of fish surveys and gage
stations in close geographic proximity by overlaying (in
a geographic information system) locations of surveys
and reference gages onto a national hydrography of
streams and rivers (Simley and Carswell 2009) (Fig. 2).
Guidelines for selecting pairs of gages and fish surveys
were designed to ensure that the hydrologic conditions
measured at the gage station reflect the conditions where
fish assemblages were surveyed (Larned et al. 2010).
Fish–gage pairs were required to be no more than 20
river kilometers apart, not separated by a dam or
diversion, and when separated by an inflowing tributary
for it to contribute to ,10% of the mean annual flow of
the mainstem. Data were also screened for quality of
hydrologic records. Only gages with sufficient records of
daily mean discharge (at least 15 years of continuous
data prior to the fish survey to ensure accurate and
precise estimates of hydrologic metrics), temporal
coverage (.70% overlap in flow records across gages
to ensure a common period of climatic conditions), and
quality (,10 days of missing data per year) were
included (following Kennard et al. 2010). The length
of gage records in our study ranged from 31 to 44 years
(mean ¼ 38.1 years). Notably, our study sites were
38
MERYL C. MIMS AND JULIAN D. OLDEN
Ecology, Vol. 93, No. 1
FIG. 2. Distribution throughout the United States of pairs of gages and fish surveys (n ¼ 109) used in our analysis.
dispersed widely across the United States, and only four
pairs of sites were located on the same mainstem river.
Of these four pairs, the average distance between the
members of each pair was .100 river kilometers. Spatial
autocorrelation was examined using Moran’s I and
Geary’s c (Legendre and Legendre 1998), which revealed
no clear or significant pattern of spatial dependence
(results not shown).
Hydrologic metrics
We quantified relationships between the proportional
life history composition at each site and six key hydrologic
metrics: annual coefficient of variation (AnnCV), high
pulse count (HPC), base flow index (BFI), flow predictability (FPred), flow constancy/flow predictability (Const/
Pred), and high pulse duration (HPD; Table 1). These
metrics were chosen out of a set of 69 metrics (reported in
Appendix A: Table A1) because they address fundamental
dimensions of the flow regime (variability, predictability,
and seasonality) hypothesized to be most important in
driving patterns in fish occurrence via the selection of life
history strategies (Bunn and Arthington 2002, Lytle and
Poff 2004) and represent the broader suite of hydrologic
metrics that are available from the literature (Olden and
Poff 2003) (Table 1). The metrics were calculated using the
software Indicators of Hydrologic Alteration (Mathews
and Richter 2007).
Statistical analysis
It is well recognized that the ecology and distribution
of fishes are strongly influenced by habitat volume
(water depth), current velocity, food availability, and
thermal regime, all of which are under hydrological
influence. Recent trait analyses have shown that
common determinants of fish species occurrence, such
as drainage area, are not significant predictors of trait
composition in the United States (Olden and Kennard
2010). However, a suite of other biotic and abiotic
factors not related to hydrology can shape the composition of freshwater fish assemblages (Jackson et al.
2001). Consequently, we deployed quantile regression
(Koenker and Bassett 1978), a method developed to
estimate rates of change in all parts of the distribution of
a response variable, which is well suited to examine
relationships between an environmental driver and a
biological response in cases where other influencing
factors are unmeasured and unaccounted for, as is often
the case with ecological data. Rather than focusing
exclusively on changes in the mean response which can
lead to underestimation, overestimation, or a failure to
detect changes in heterogeneous distributions (Rosenbaum 1995, Cade and Noon 2003), quantile regression
estimates multiple rates of change (slopes) in responses
with unequal variation.
January 2012
FISH LIFE HISTORIES AND STREAM FLOW
Relationships between the hydrologic metrics and
proportional life history composition as revealed by
quantile regression analysis can be characterized as strong
(significant for more than half the quantiles tested), weak
(significant for less than half the quantiles tested), or
limiting (significant only at the most extreme quantiles
examined). These three characterizations are important
for interpreting the importance of hydrology relative to
other potential drivers of life history composition.
Significant relationships across multiple quantiles suggest
that a particular flow parameter is an important predictor
of that life history. A weak association in which only one
or a few quantiles have significant relationships indicates
that a flow parameter is important in driving life history
composition across part of the distribution but that other
factors (biotic or abiotic) likely play an addition role.
Limiting relationships in which only the upper or lower
most quantiles are significant are different from weak
relationships in that they can identify flow parameters
acting as ‘‘ceilings’’ and ‘‘floors’’ (Konrad et al. 2008).
Ceilings and floors are environmental factors that
constrain the distribution of a response where other
environmental factors would have instead allowed for a
more extreme response (in this case, a higher or lower
proportional life history composition). Because quantile
regression measures differential effects between quantiles
and does not require a mean or median response, it is
particularly well suited for ecological investigations.
The choice of quantiles represents an arbitrary
selection both in terms of the number of quantiles and
the position of quantiles throughout the distribution. It
is likely that the chance of a Type I error increases with
the number of quantiles tested. Our analysis included
five quantiles distributed evenly throughout the distribution: 0.05 (the ‘‘floor’’), 0.25, 0.5 (the median), 0.75,
and 0.95 (the ‘‘ceiling’’). These five quantiles allowed us
to explore a range of quantiles throughout the distribution but limited the chance of a Type I error. The range
of reliable extreme quantile estimates is ultimately
limited by sample size (n). For our study, we used the
guidelines of n . 5/q and n . 5/(1 q), where q is the
quantile proposed by Rogers (1992) to determine the
limits of reliable extreme quantiles. All quantile regression analyses were performed using the QuantReg
package (Koenker 2005) in R 2.11.0 (R Development
Core Team 2005).
In addition to examining relationships between life
histories and hydrologic metrics for all 109 sites, we also
classified sites into six distinct hydrologic classes and
examined relationships between life histories and hydrologic metrics within the groundwater, perennial
runoff and perennial flashy classes for which adequate
sample sizes existed (Appendix A).
RESULTS
Relationships between hydrology and fish life histories
One hundred eighty-four native freshwater fish species
with available life history trait data were represented in
39
our final 109 gage–fish survey pairs (Fig. 2). Quantile
regression analysis of relationships between proportional life history composition and hydrologic metrics for all
gages revealed that 11 of 18 relationships (six metrics vs.
three life history strategy responses) were significant (P
0.05) for at least one quantile (Table 2), and 82% of
significant relationships supported predictions from life
history theory (Table 1). Relationships within hydrologic classes are reported in Appendix B. Of the
supportive relationships, two were considered strong
(over half of the quantiles significant), and seven were
weak (less than half of the quantiles significant; Fig.
3A). Across all statistically significant relationships, 21%
were floors (the extreme lower quantile, 0.05), 32% were
the 0.25 quantile, 32% were the median quantile (0.5),
15% were the 0.75 quantile, and ceilings (the extreme
upper quantile, 0.95) were never significant (Fig. 3B).
The rate of detection of floors and median responses
(recall no ceiling responses were detected) in significant
relationships is summarized in Fig. 3C. Note that 45% of
significant relationships (P 0.05) were not significant
for the median quantile and thus would not have been
detected using classic regression techniques examining
only the mean or median response.
We graphically display the quantile regression results
for six of the 18 relationships we tested (see Fig. 4); all
relationships are presented in Table 2, and those
relationships not shown in Fig. 4 are included in
Appendix B. As predicted by life history theory, the
proportion of opportunistic strategists was positively
related to the degree of flow variability both across
many quantiles (AnnCV, Fig. 4A) and in the form of a
floor, threshold-type response (HPC, Fig. 4B). The
proportion of opportunist strategists was also negatively
correlated with predictability for the lowest two
quantiles, potentially indicating a floor-type response
(FPred, Fig. 4C). These findings were corroborated by
significant relationships between the two other hydrologic metrics for which we made predictions for the
proportional opportunistic response (negative relationship with predictability, BFI; and negative relationship
with seasonality, HPD; Table 2). Periodic strategists
increased in proportion with seasonality across four of
five quantiles tested; a pattern consistent with predictions from life history theory (HPD: Fig. 4D). We also
found support (albeit weak) for a negative relationship
between the proportion of periodic strategists and
increased variability (AnnCV and HPC; Table 2).
Finally, the proportion of equilibrium strategists showed
a weakly significant negative correlation with one
surrogate of increased variability (AnnCV, Fig. 4E)
and a weakly significant positive correlation with
increased predictability (FPred, Fig. 4F). The equilibrium proportion was also weakly associated with an
increase in constancy of flows (Const/Pred, Table 2).
Contrary to predictions, the proportion of equilibrium
strategists was positively correlated with increased
frequency of disturbance (HPC, Table 2), and we
40
MERYL C. MIMS AND JULIAN D. OLDEN
Ecology, Vol. 93, No. 1
TABLE 2. Relationships between proportional life history strategies and six hydrologic metrics describing flow variability,
predictability, and seasonality throughout the United States (n ¼ 109 study sites).
Relationship
Opportunistic
AnnCV
HPC
BFI
FPred
Const/Pred
HPD
Periodic
AnnCV
HPC
BFI
FPred
Const/Pred
HPD
Equilibrium
AnnCV
HPC
BFI
FPred
Const/Pred
HPD
Model equations by quantile
Regression quantile
P value for slope
Slope direction (predicted)
¼
¼
¼
¼
¼
¼
¼
0.056(AnnCV) þ 0.234
0.028(AnnCV) þ 0.321
0.036(AnnCV) þ 0.338
0.030(HPC) þ 0.233
0.108(BFI) þ 0.341
0.215(FPred) þ 0.370
0.201(FPred) þ 0.392
NS
¼ 0.035(HPD) þ 0.345
¼ 0.016(HPD) þ 0.404
0.05
0.50
0.75
0.05
0.25
0.05
0.25
NS
0.05
0.75
0.001*
0.027*
0.048*
0.081
0.039*
0.018*
0.003*
NS
0.029*
0.072
þ
þ
þ
þ
NS
(þ)
(þ)
(þ)
(þ)
()
()
()
(0)
()
()
¼ 0.027(AnnCV) þ 0.357
¼ 0.027(HPC) þ 0.349
¼ 0.020(HPC) þ 0.362
NS
NS
NS
¼ 0.013(HPD) þ 0.255
¼ 0.028(HPD) þ 0.252
¼ 0.019(HPD) þ 0.291
¼ 0.029(HPD) þ 0.304
0.50
0.25
0.50
NS
NS
NS
0.05
0.25
0.50
0.75
0.027*
0.001*
0.013*
NS
NS
NS
0.042*
0.007*
0.049*
0.023*
NS
NS
NS
þ
þ
þ
þ
()
()
()
(0)
(þ)
()
(þ)
(þ)
(þ)
(þ)
¼ 0.024(AnnCV) þ 0.345
¼ 0.0153(HPC) þ 0.270
NS
¼ 0.143(FPred) þ 0.297
¼ 0.083(Const/Pred) þ 0.258
¼ 0.016(HPD) þ 0.327
¼ 0.020(HPD) þ 0.353
0.50
0.25
NS
0.75
0.25
0.25
0.50
0.050*
0.025*
NS
0.036*
0.085
0.030*
0.018*
þ
NS
þ
þ
()
()
(þ)
(þ)
(þ)
(0)
(0)
Notes: Life histories include opportunistic, periodic, and equilibrium strategies. Hydrologic metrics include annual coefficient of
variation (AnnCV), high pulse count (HPC), base flow index (BFI), flow predictability (FPred), and high pulse duration (HPD).
Relationships between proportional life history strategies and flow parameters are presented for all regression quantiles for which P
0.1. The direction of the slope of each relationship is given as positive or negative value followed by the slope direction as
predicted by life history theory in parentheses.
* P , 0.05; NS, not significant.
detected a weak but negative relationship between a
surrogate of seasonality (HPD, Table 2) and the
proportion of opportunistic species despite our prediction of no significant relationship.
DISCUSSION
Our study illustrates the fundamental role of the flow
regime as a key determinant of fish life history
composition across a broad biogeographical scale.
Despite regional disparities in zoogeography and phylogeny, we found that relationships between local fish
assemblage structure and the major dimensions of
hydrologic variability, predictability, and seasonality
corresponded to predictions according to life history
theory. By highlighting the functional response of fish
assemblages to the natural flow regime, our results have
implications for predicting the consequences of flow
alteration for species adapted to particular flow regimes
and for informing flow-management recommendations
(Poff et al. 2010).
Characterized by small-bodied species with early
maturation and low juvenile survivorship, opportunistic
strategists are predicted to be favored in environments
characterized by frequent and intense disturbance
(Winemiller and Rose 1992). Our results revealed that
as predicted, opportunistic strategists were positively
related to measures of flow variability (AnnCV and
HPC) and negatively related to flow predictability and
seasonality (BFI, FPred, and HPD). Quantile regression
also identified floors for opportunist strategists associated with frequency (HPC, positive) and duration
(HPD, negative) of high flow events, suggesting that
opportunistic strategists may not persist below certain
flood disturbance requirements. The Gila topminnow
(Poeciliopsis occidentalis) is a good example; this species
has historically lived ‘‘from feast to famine’’ in harsh
arid streams with highly variable population size and
geographic range supported by sporadic bursts of
reproductive superfetation (Minckley 1999). The longterm persistence of this species requires both intermittent and moderate floods to connect isolated (resourcerich) habitats after periods of prolonged drought and
allow for dispersal from secure refugia. Identification of
thresholds, such as the floor relationships found here, is
critical for informing water management actions that
January 2012
FISH LIFE HISTORIES AND STREAM FLOW
41
FIG. 3. Summary of quantile regression analysis of relationships between proportional life histories and six hydrologic metrics
describing flow variability, predictability, and seasonality. (A) The percentage of relationships with at least one significant quantile
(P 0.05) is shown along the y-axis for each of the three life histories along the x-axis (opportunistic [OPP], periodic [PER], and
equilibrium [EQU]). Agreement between pairs for which no significant relationship was predicted and no significant relationship
was observed were not included in summary statistics. Significant relationships were categorized as strongly supporting (.3
quantiles significant; slopes in agreement with predictions; in black), weakly supporting (in gray), or weakly opposing (in white) our
predictions. (B) A summary of the contribution of each quantile to the total number of quantiles detected as significant (n ¼ 19):
0.05, 21%; 0.25, 32%; 0.50, 32%; 0.75, 15%; 0.95, 0% (not pictured). (C) Representation of medians and floors in relationships for
which significant quantiles were detected (n ¼ 11): neither floor nor median detected, 27%; floor but not median detected, 18%;
median but not floor detected, 36%; both median and floor detected, 19%.
seek to achieve desired restoration or conservation
outcomes (Souchon et al. 2008).
Life history theory predicts that long-lived and largebodied periodic strategists cope with temporally heterogeneous habitats by displaying both the generation time
and energy reserves required to persist through extended
periods of adverse conditions (Winemiller 1989, Winemiller and Rose 1992). We found that the prevalence of
periodic strategists was positively related to high flow
seasonality (HPD) and negatively related to variability
(AnnCV and HPC), supporting theoretical expectations
that the periodic strategy is favored by predictable, highduration flow events. Seasonality of flows is particularly
important for temperate ecosystems in which other
variables such as temperature and light also fluctuate
seasonally and have direct impacts for physiology and
resource availability (Tockner et al. 2000). The observed
strong positive relationship between the prevalence of
periodic strategists and high flow seasonality (HPD) is
congruent with previous studies in other regions
(Tedesco et al. 2008) and emphasizes the importance
of predictable, high-duration flow events for periodic
strategists, which is likely important for spawning and
locomotive cues (Winemiller and Rose 1992) as well as
access to floodplain habitat and resources (Junk et al.
1989, Tockner et al. 2000). In support of this finding,
high-duration spring freshets resulting in floodplain
inundation on the lower Savannah River (southeast
United States) were positively correlated with increased
larval transport of periodic shad species (Alosa aestivalis, Dorosoma cepedianum, D. petenense) and common
carp (Cyprinus carpio) (Martin and Paller 2008).
Equilibrium strategists are characterized by high
parental care of few offspring and are predicted to be
associated with stable, predictable environments. We
found that the proportion of equilibrium strategists was
42
MERYL C. MIMS AND JULIAN D. OLDEN
Ecology, Vol. 93, No. 1
FIG. 4. Quantile regression results for relationships between hydrologic metrics (x-axis) and proportional life history
composition (y-axis) with significant quantiles regression lines identified to the right of the plot (solid lines represent quantiles
significant at P 0.05, and dashed lines represent quantiles significant at P 0.1). Hydrologic metrics are annual coefficient of
variation (AnnCV), high pulse count (HPC), flow predictability (FPred), and high pulse duration (HPD) (A) Proportion of
opportunistic strategists vs. log(AnnCV). (B) Proportion of opportunistic strategists vs. log(HPC). (C) Proportion of opportunistic
strategists vs. log(FPred). (D) Proportion of periodic strategists vs. log(HPD). (E) Proportion of equilibrium strategists vs.
log(AnnCV). (F) Proportion of equilibrium strategists vs. log(FPred).
negatively related to variability (AnnCV) and positively
related to predictability (Const/Pred and FPred, positive), providing support for both life history theory and
empirical evidence that equilibrium strategists are
associated with decreasing hydrologic variability and
increasing stability (Tedesco et al. 2008, Olden and
Kennard 2010). Species tending toward the equilibrium
life history endpoint strategy are notably rare from
native fish assemblages of the Colorado River Basin
(USA), historically characterized by flashy, highly
January 2012
FISH LIFE HISTORIES AND STREAM FLOW
43
PLATE 1. Ictiobus niger (black buffalo), one example of a periodic strategist, shown here in the Little Tennessee River System in
the southeastern United States. Photo credit: Dave Herasimtschuk.
variable flows; however, as flow alteration has led to
more predictable, less variable flows, many invasions of
nonnative fishes have now occurred into historically
vacant equilibrium life history space (Olden et al. 2006).
Nest building is characteristic of many equilibrium
strategists and is successful in stable environments where
hydrologic disturbance of stream-bed substrate is
minimal. For example, smallmouth bass (Micropterus
dolomieu) are equilibrium strategists with nest building
and guarding behavior and are frequent inhabitants of
regulated rivers with static, stable, flow regimes.
However, managing flows to produce a more variable
flow regime may make it difficult for this species to
persist. Experimental floods on the regulated Rapid
River in Maine during the summer of 2007 resulted in a
50% nest failure of nonnative smallmouth bass, and
recruitment of juvenile smallmouth bass was negatively
correlated with flow variability (Kleinschmidt Associates 2008).
Significant relationships between the proportion of
equilibrium strategists and flow dimensions were less
numerous, weaker, and less frequently agreed with
predictions than for both opportunistic and periodic
strategists. In examining relationships between stream
flow and life histories of fishes in the upper Missouri
River Basin, Kelly (2008) also found strong relationships between stream flow characteristics of variability
and low flow volume and the opportunistic and periodic
strategies; however, equilibrium strategists showed no
significant association with either stream flow characteristic. This may be explained in part because the
selective pressures of biotic factors such as competition
and predation increase in stable riverine environments
(Resh et al. 1988), thus promoting high juvenile density
dependence that favor equilibrium strategists (McCann
1998) and weakening relationships with flow conditions.
In the absence of human activities, flow regimes
interact with channel geology to shape the river’s
biophysical template upon which fish life history
strategies are formed and maintained (Lytle and Poff
2004). From an evolutionary perspective, floods and
droughts that are predictable over time can exert
primary selective pressures that favor certain fish life
histories, while at ecological time scales the flow regime
shape assemblage composition via the forces of environmental filtering on species persistence (Poff et al.
1997, Naiman et al. 2008). Our study focuses on
proximate mechanisms by which flow has shaped the
life history composition of contemporary assemblages
by examining biotic patterns at the site scale, characterizing natural flow regimes over recent decades, and
treating species traits as static in their attribution. Other
distal factors undoubtedly contribute to watershed
patterns of fish life histories at the continental scale
(Mims et al. 2010, Olden and Kennard 2010), but here
we focused on how the contemporary flow regime acts as
an important filter through which these other zoogeographic factors may operate (Naiman et al. 2008).
In the habitat template model, Southwood (1977)
postulates that if the only variables for reproductive
success were survival and growth rate, an optimal life
history strategy for a particular set of environmental
44
MERYL C. MIMS AND JULIAN D. OLDEN
characteristics would always exist. However, uncertainty
in the form of temporal and spatial variance in
environmental conditions can result in multiple favorable life histories, each resulting in equal chances of
successful reproduction. For example, high flow events
have been shown to differentially affect juvenile
abundance of closely related fishes within the same
assemblages on the Tallapoosa River, Alabama, where
some species endured high mortality in response to
floods while others thrived under high flow conditions
(Freeman et al. 2001). This highlights the importance of
examining traits-based responses to specific flow dimensions, such as done in the present study, and reveals that
these ‘‘response curves’’ to major dimensions of the
natural flow regime may be applicable across space and
time as well as taxonomies (Merritt et al. 2010).
Moreover, establishing baseline relationships between
freshwater organisms and hydrologic conditions is a
critical step in adopting environmental flow standards
for rivers currently affected by human activities such as
river regulation by dams (Richter et al. 2006, Poff et al.
2010). A traits-based approach is particularly useful
because it facilitates the transfer of scientific knowledge
between regions that naturally differ due to biogeographic constraints on regional species pools, but in
which life history strategies and trait adaptations are
hypothesized to converge across diverse taxonomies.
ACKNOWLEDGMENTS
We thank Alan Herlihy for providing the fish occurrence
data, LeRoy Poff and Zach Shattuck for their contributions to
the trait database, and Dave Lawrence for his assistance with
ensuring consistent taxonomy. Tim Essington, Josh Lawler,
Daniel Schindler, and two anonymous reviewers provided
valuable comments that greatly improved the manuscript.
Funding for M. C. Mims was provided by a University of
Washington Top Scholar Graduate Fellowship and a National
Science Foundation Graduate Research Fellowship (Grant No.
DGE-0718124), and by the John N. Cobb Scholarship in
Fisheries and the H. Mason Keeler Endowment for Excellence
through University of Washington’s School of Aquatic and
Fishery Sciences. J. D. Olden acknowledges funding support
from the U.S. Environmental Protection Agency Science to
Achieve Results (STAR) Program (Grant No. 833834).
LITERATURE CITED
Blanck, A., P. A. Tedesco, and N. Lamouroux. 2007.
Relationships between life-history strategies of European
freshwater fish species and their habitat preferences. Freshwater Biology 52:843–859.
Bunn, S. E., and A. H. Arthington. 2002. Basic principles and
ecological consequences of altered flow regimes for aquatic
biodiversity. Environmental Management 30:492–507.
Cade, B. S., and B. R. Noon. 2003. A gentle introduction to
quantile regression for ecologists. Frontiers in Ecology and
the Environment 1:412–420.
Erös, T. 2005. Life-history diversification in the Middle
Danubian fish fauna: a conservation perspective. Large
Rivers 16:289–304.
Falcone, J. A., D. M. Carlisle, D. M. Wolock, and M. R.
Meador. 2010. GAGES: a stream gage database for
evaluating natural and altered flow conditions in the
conterminous United States. Ecology 91:621.
Ecology, Vol. 93, No. 1
Freeman, M. C., Z. H. Bowen, K. D. Bovee, and E. R. Irwin.
2001. Flow and habitat effects on juvenile fish abundance in
natural and altered flow regimes. Ecological Applications
11:179–190.
Gilliom, R. J., W. M. Alley, and M. E. Gurtz. 1995. Design of
the national water-quality assessment program: occurrence
and distribution of water-quality conditions. U.S. Geological
Survey, Reston, Virginia, USA.
Herlihy, A. T., R. M. Hughes, and J. C. Sifneos. 2006.
Landscape clusters based on fish assemblages in the
conterminous USA and their relationship to existing
landscape classifications. American Fisheries Society Symposium 48:87–112.
Hughes, R. M., S. G. Paulsen, and J. L. Stoddard. 2000. EMAP
surface waters: a multi-assemblage, probability survey of
ecological integrity in the U.S.A. Hydrobiologia 423:429–
443.
Jackson, D. A., P. R. Peres-Neto, and J. D. Olden. 2001. What
controls who is where in freshwater fish communities—the
roles of biotic, abiotic, and spatial factors. Canadian Journal
of Fisheries and Aquatic Sciences 58:157–170.
Junk, W. J., P. B. Bayley, and R. E. Sparks. 1989. The flood
pulse concept in river–floodplain systems. Pages 110–127 in
D. P. Lodge, editor. Proceedings of the International Large
River Symposium. Canadian Special Publication of Fisheries
and Aquatic Sciences. Fisheries and Oceans Canada, Ottawa,
Ontario, Canada.
Kelly, V. 2008. Influence of streamflow regime and biotic
interactions on fish assemblage structure in rivers of the
Northern Great Plains. Dissertation. Oregon State University, Corvallis, Oregon, USA.
Kennard, M. J., S. J. MacKay, B. J. Pusey, J. D. Olden, and N.
Marsh. 2010. Quantifying uncertainty in estimation of
hydrologic metrics for ecohydrological studies. River Research and Applications 26:137–156.
Kleinschmidt Associates. 2008. Smallmouth bass/brook trout
habitat manipulation studies in Rapid River, Township C
and Upton, Oxford County, Maine. 2007 Progress Report.
http://www.maine.gov/ifw/fishing/reports/fishery_division/
rapidriver/rapidriver_flowstudy.pdf
Koenker, R. 2005. Quantreg: quantile regression. R package
version 4.54. http://cran.r-project.org/web/packages/quantreg/
index.html
Koenker, R., and G. Bassett. 1978. Regression quantiles.
Econometrica 46:33–50.
Konrad, C. P., A. M. D. Brasher, and J. T. May. 2008.
Assessing streamflow characteristics as limiting factors on
benthic invertebrate assemblages in streams across the
western United States. Freshwater Biology 53:1983–1998.
Lamouroux, N., N. L. Poff, and P. L. Angermeier. 2002.
Intercontinental convergence of stream fish community traits
along geomorphic and hydraulic gradients. Ecology 83:1792–
1807.
Larned, S. T., D. B. Arscott, J. Schmidt, and J. C. Diettrich.
2010. A framework for analyzing longitudinal and temporal
variation in river flow and developing flow–ecology relationships. Journal of the American Water Resources Association
46:541–553.
Legendre, P., and L. Legendre. 1998. Numerical ecology.
Second edition. Elsevier, Amsterdam, The Netherlands.
Logez, M., D. Pont, and M. T. Ferreira. 2010. Do Iberian and
European fish faunas exhibit convergent functional structure
along environmental gradients? Journal of the North
American Benthological Society 29:1310–1323.
Lytle, D. A. 2001. Disturbance regimes and life-history
evolution. American Naturalist 157:525–536.
Lytle, D. A., and N. L. Poff. 2004. Adaptation to natural flow
regimes. Trends in Ecology and Evolution 19:94–100.
MacArthur, R. H., and E. O. Wilson. 1967. The theory of
island biogeography. Princeton University, Princeton, New
Jersey, USA.
January 2012
FISH LIFE HISTORIES AND STREAM FLOW
Martin, F. D., and M. H. Paller. 2008. Ichthyoplankton
transport in relation to floodplain width and inundation and
tributary creek discharge in the lower Savannah River of
Georgia and South Carolina. Hydrobiologia 598:139–148.
Mathews, R., and B. D. Richter. 2007. Application of the
indicators of hydrologic alteration software in environmental
flow setting. Journal of the American Water Resources
Association 43:1400–1413.
McCann, K. 1998. Density-dependent coexistence in fish
communities. Ecology 79:2957–2967.
Merritt, D. M., M. L. Scott, N. L. Poff, G. T. Auble, and D. A.
Lytle. 2010. Theory, methods and tools for determining
environmental flows for riparian vegetation: riparian vegetation-flow response guilds. Freshwater Biology 55:206–225.
Mims, M. C., J. D. Olden, Z. R. Shattuck, and N. L. Poff. 2010.
Life history trait diversity of native freshwater fishes in North
America. Ecology of Freshwater Fish 19:390–400.
Minckley, W. L. 1999. Ecological review and management
recommendations for recovery of the endangered Gila
topminnow. Great Basin Naturalist 59(3):230–244.
Naiman, R. J., J. J. Latterell, N. E. Pettit, and J. D. Olden.
2008. Flow variability and the biophysical vitality of river
systems. Comptes Rendus Geoscience 340:629–643.
Olden, J. D., and M. J. Kennard. 2010. Intercontinental
comparison of fish life history strategies along a gradient of
hydrologic variability. Pages 83–107 in K. B. Gido and D. A.
Jackson, editors. Community ecology of stream fishes:
concepts, approaches, and techniques. American Fisheries
Society, Bethesda, Maryland, USA.
Olden, J. D., M. K. Kennard, F. Leprieur, P. A. Tedesco, K. O.
Winemiller, and E. Garcı́a-Berthou. 2010. Conservation
biogeography of freshwater fishes: past progress and future
directions. Diversity and Distributions 16:496–513.
Olden, J. D., and N. L. Poff. 2003. Redundancy and the choice
of hydrologic indices for characterizing streamflow regimes.
River Research and Applications 19:101–121.
Olden, J. D., N. L. Poff, and K. R. Bestgen. 2006. Life-history
strategies predict fish invasions and extirpations in the
Colorado River Basin. Ecological Monographs 76:25–40.
Pianka, E. R. 1970. On r- and K-selection. American Naturalist
102:592–597.
Poff, N. L. 1996. A hydrogeography of unregulated streams in
the United States and an examination of scale-dependence in
some hydrological descriptors. Freshwater Biology 36:71–91.
Poff, N. L., and J. D. Allan. 1995. Functional organization of
stream fish assemblages in relation to hydrological variability. Ecology 76:606–627.
Poff, N. L., J. D. Allan, M. B. Bain, J. R. Karr, K. L.
Prestegaard, B. D. Richter, R. E. Sparks, and J. C.
Stromberg. 1997. The natural flow regime. BioScience
47:769–784.
Poff, N. L., et al. 2010. The ecological limits of hydrologic
alteration (ELOHA): a new framework for developing
regional environmental flow standards. Freshwater Biology
55:147–170.
R Development Core Team. 2005. R 2.11.0. R Project for
Statistical Computing, Vienna, Austria. www.r-project.com
Resh, V. H., A. V. Brown, A. P. Covich, M. E. Gurtz, H. W. Li,
G. W. Minshall, S. R. Reice, A. L. Sheldon, J. B. Wallace,
and R. C. Wissmar. 1988. The role of disturbance in stream
45
ecology. Journal of the North American Benthological
Society 7:433–455.
Richter, B. D., A. T. Warner, J. L. Meyer, and K. Lutz. 2006. A
collaborative and adaptive process for developing environmental flow regulations. River Research and Applications
22:297–318.
Rogers, W. H. 1992. Quantile regression standard errors. Stata
Technical Bulletin 9:16–19.
Rosenbaum, P. R. 1995. Quantiles in nonrandom samples and
observational studies. Journal of the American Statistical
Association 90:1423–1431.
Simley, J. D., and W. J. Carswell. 2009. The national map—
hydrography. Fact sheet 2009-3054.1-4. U.S. Geological
Survey, Reston, Virginia, USA.
Souchon, Y., C. Sabaton, R. Deibel, D. Reiser, J. Kershner, M.
Gard, C. Katopodis, P. Leonard, N. L. Poff, W. J. Miller,
and B. L. Lamb. 2008. Detecting biological responses to flow
management: missed opportunities; future directions. River
Research and Applications 24:506–518.
Southwood, T. R. E. 1977. Habitat, the templet for ecological
strategies? Journal of Animal Ecology 46:337–365.
Southwood, T. R. E. 1988. Tactics, strategies and templets.
Oikos 52:3–18.
Statzner, B., K. Hoppenhaus, M. F. Arens, and P. Richoux.
1997. Reproductive traits, habitat use and templet theory: a
synthesis of world-wide data on aquatic insects. Freshwater
Biology 38:109–135.
Stearns, S. C. 1976. Life-history tactics: review of ideas.
Quarterly Review of Biology 51:3–47.
Stearns, S. C. 1992. The evolution of life histories. Oxford
University Press, Oxford, UK.
Tedesco, P. A., B. Hugueny, T. Oberdorff, H. H. Durr, S.
Merigoux, and B. d. Merona. 2008. River hydrological
seasonality influences life history strategies of tropical
riverine fishes. Oceologia 156:691–702.
Tockner, K., F. Malard, and J. V. Ward. 2000. An extension of
the flood pulse concept. Hydrological Processes 14:2861–
2883.
Townsend, C. R., and A. G. Hildrew. 1994. Species traits in
relation to a habitat templet for river systems. Freshwater
Biology 31:265–275.
Verberk, W. C. E. P., H. Siepel, and H. Esselink. 2008. Lifehistory strategies in freshwater macroinvertebrates. Freshwater Biology 53:1722–1738.
Vogel, R. M., and A. Sankarasubramanian. 2005. USGS
hydro-climatic data network (HCDN): monthly climate
database, 1951–1990. Data set. Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge,
Tennessee, USA. http://dx.doi.org/10.3334/ORNLDAAC/
810
Winemiller, K. O. 1989. Patterns of variation in life-history
among South American fishes in seasonal environments.
Oecologia 81:225–241.
Winemiller, K. O. 2005. Life history strategies, population
regulation, and implications for fisheries management.
Canadian Journal of Fisheries and Aquatic Sciences 52:875.
Winemiller, K. O., and K. A. Rose. 1992. Patterns of lifehistory diversification in North American fishes: implications
for population regulation. Canadian Journal of Fisheries and
Aquatic Sciences 49:2196–2218.
SUPPLEMENTAL MATERIAL
Appendix A
Hydrologic methods and classification (Ecological Archives E093-004-A1).
Appendix B
Life history and hydrologic metric relationships (Ecological Archives E093-004-A2).