A rolling stone gathers no eggs Jordt and Taylor 2021 Ecology

A rolling stone gathers no eggs: the
importance of stream insect egg
laying natural history for
stream restoration
SAMANTHA JORDT 1 AND BRAD W. T AYLOR
Manuscript received 26 October 2020; revised 11 December
2020; accepted 11 January 2021. Corresponding Editor: John
Pastor.
Department of Applied Ecology, North Carolina State
University, Raleigh, North Carolina 27695 USA.
1E-mail: [e-mail]
Citation:Jordt S., and B. W. Taylor. 2021. A rolling stone gath-
ers no eggs: the importance of stream insect egg laying natural
history for stream restoration. Ecology 00(00):e03331. 10.1002/
ecy.3331
Key words: natural channel design; oviposition; particle size;
recruitment; substrate stability.
The degradation of waterways has led to the wide-
spread practice of stream restoration aimed at accelerat-
ing recovery from damage by land-use change and other
disturbances. However, the biological recovery of
restored streams often lags behind their physical and
chemical recovery (Louhi et al. 2011). Improving stream
restoration is important for many reasons, including
mitigating the impacts of development and because
annually billions of dollars are invested in stream
restoration worldwide. Slow biological recovery is not
surprising given the emphasis on restoring physical
stream features (e.g., riffle-pool sequence, meanders,
bank stability) and the expectation that biological recov-
ery will naturally follow (Palmer et al. 2014). Restoration
typically improves physical habitats for fish and larval
insects (Bernard et al. 2007); yet, habitats for adult aqua-
tic insects are overlooked. Here, we report observations
of egg laying in restored and reference streams that
reveal how natural history science can be used to
enhance stream restoration design, particularly for the
recovery of stream insects.
Stream insects have complex life cycles that involve
developing in an aquatic habitat, emerging to reproduce
terrestrially, then returning to water to lay eggs. Females
have various egg-laying strategies; some release their
eggs while flying over the water surface, but an estimated
75% of aquatic insect species attach their eggs to objects
within the stream such as rocks (Fig. 1A–F) or wood
(Statzner and B^eche 2010). For example, mayflies from
the widespread genus Baetis (Ephemeroptera) select
rocks partially extending above the water’s surface
(hereafter emergent rocks, Fig. 1A–C) to land on, crawl
underneath, and attach tombstone-shaped egg masses
(Peckarsky et al. 2000). Caddisflies (Trichoptera) lay
eggs in much the same way (Fig. 1D; Smith and Storey
2018). Adult waterpenny beetles (Coleoptera) use emer-
gent rocks as both mating sites and access points for
females to crawl underwater and lay their bright yellow
eggs (Fig. 1E). Gelatinous egg masses of aquatic true
flies such as midges (Diptera) are commonly found on
emergent rocks (Fig. 1F) and are diverse in structure
and attachment mechanisms (Reich 2004).
Egg-laying females can be highly selective (Smith and
Storey 2018), often preferring emergent rocks that are
large, unembedded, and in fast and constant flow. For
many species, egg masses containing hundreds to thou-
sands of eggs are highly aggregated, with a few emergent
rocks receiving >90% of egg masses (Peckarsky et al.
2000). This selective behavior increases egg survival,
hatching success (Bovill et al. 2013), and larval recruit-
ment (Encalada and Peckarsky 2012). Moreover, avail-
ability of rocks suitable for egg laying can limit stream
insect populations (Alp et al. 2013) and determine
stream insect community composition (Kennedy et al.
2016). Thus, egg laying on emergent rocks represents a
linkage between stream physical characteristics and
insect natural history that could have applications for
stream conservation and restoration to either maintain
or accelerate recovery of stream insect diversity and
abundance.
During the summer of 2019, the abundance of rocks
suitable for stream insect egg laying and the number of
eggs masses they contained was observed throughout the
egg-laying period (June to October) in 10 restored and
three reference streams in western North Carolina
(Appendix S1: Fig. S1). Streams were restored ≥7yr
ago, in agriculturally dominated watersheds containing
cropland and cattle grazing, and watersheds had <5%
impervious surface area (Appendix S1: Table S1).
Restoration involved riverbed and bank reconstruction,
riparian buffer planting, and cattle exclusion. Restored
streams in North Carolina (Tullos et al. 2009) as else-
where (Louhi et al. 2011) have slow biological recovery
and low insect diversity and abundance. Reference
streams of similar size with predominantly forested
Article e03331; page 1
The Scientific Naturalist
Ecology, 0(0), 2021, e03331
©2021 The Authors.Ecology published by Wiley Periodicals LLC on behalf of Ecological Society of America
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and
distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
watersheds were located nearby (Appendix S1: Fig. S1).
Suitable egg-laying habitats were identified as emergent
rocks with submerged spaces that adult insects could
access along the sides and underneath (i.e., unembed-
ded). Emergent rocks were numbered using a nontoxic
paint pen and removed from the water to identify and
count egg masses, species or morphotype, and sketch
mass locations. New masses were identified by compar-
ing sketches to those made previously. Rocks were
replaced in their original location and position. We
returned every 1–2 weeks to census these and newly
emergent rocks for egg masses. Mayfly, caddisfly, and
true fly taxa constituted ~90 % of the masses in refer-
ence and restored streams, with mayflies (45%) domi-
nant in restored streams and mayflies (39%) and
caddisflies (36%) codominant in reference streams.
Densities of emergent rocks (number/m
2 streambed)
were 103% greater in reference than restored streams,
and as water levels declined emergent rocks increased in
reference but not restored streams throughout the sum-
mer (repeated-measures ANOVA, time 9 stream type
F1,24.1 =6.210,P =0.02; Fig. 2A; Appendix S1:
FIG. 1. (A) A restored stream in western North Carolina with examples of emergent rocks. (B) Underside of an emergent rock
that female mayflies (Ephemeroptera) and caddisflies (Trichoptera) crawled underneath and attached egg masses. (C) Female may-
fly (Ephemeroptera: Baetidae) and egg mass. (D) Egg masses of a caddisfly (Trichoptera: Hydropsychidae) taxon. (E) Female water-
penny beetle (Coleoptera: Psephenidae) and egg mass. (F) Egg masses of a midge (Diptera: Chironomidae) taxon. Scale bar in
panels C–F is 0.5 cm.
Article e03331; page 2 THE SCIENTIFIC NATURALIST Ecology, Vol. xx, No. xx
Fig. S2). Emergent rocks were larger in reference than
restored streams (repeated-measures ANOVA, stream
type F1,14.1 =7.907,P =0.01; Fig. 2B). As a result of
more egg-laying habitat, the number of egg masses stan-
dardized for streambed area was 78% higher in reference
than restored streams (repeated-measures ANOVA,
stream type F1,8.4 =4.343,P =0.04; Fig. 2C). However,
densities of egg masses on individual rocks (i.e., number
masses per m
2 of individual rocks) were 72% higher in
restored than reference streams (repeated-measures
FIG. 2. Reference (blue bars) and restored (orange bars) streams differed in (A) number of emergent rocks, (B) size of emergent
rocks, (C) number of egg masses standardized for wetted stream area, (D) number of egg masses standardized for individual rock
area, (E) the distribution of egg mass densities among individual emergent rocks (for display purposes, the restored stream distribu-
tion is truncated with bin 3,600 including five rocks: three with 3,600, one with 3,900, and one with 7,200 egg masses per rock area),
and (F) percentage of emergent rocks that rolled or were not recovered. Means in panels A–D and F are weighted by the number of
weeks sampled for each site from 29 May to 13 October 2019. In panels A–D, means are back transformed and asymmetric error
bars are 95% confidence intervals back transformed from the natural-log scale. In panels A, B, and D–F, emergent rocks include
only those that were unembedded. The histogram bin width is 300.
Xxxxx 2021 THE SCIENTIFIC NATURALIST Article e03331; page 3
ANOVA, stream type F1,14.6 =5.867,P =0.04;
Fig. 2D). The distribution of egg masses among rocks
was highly aggregated (observed distribution differed
from expected Poisson (random) distribution, reference
P <0.0001 and restored P <0.0001) and egg masses
were more aggregated in restored than reference streams
(Kolmogorov-Smirnov test,D =0.1597,P =0.0001),
with reference streams containing many rocks with few
masses and restored streams containing more rocks with
many masses (Fig. 2E). These data suggest sufficient
numbers of adults were able to disperse to restored
streams but availability of suitable rocks could limit egg
laying and thus limit insect recovery. Moreover, greater
egg mass densities on emergent rocks in restored streams
is notable because the percentage of emergent rocks that
rolled or were unrecoverable (rolled out of the study
reach or buried in sediment) was 91% higher in restored
than in reference streams (generalized linear model,
stream type v2 =5.36, df =12,P =0.02; Fig. 2F;
Appendix S1: Fig. S2), and differences in emergent rock
stability were greater during late summer (generalized
linear model, time 9 stream type v2 =3.90,P =0.04),
when egg laying was high in reference streams. Rolled or
missing rocks increased following rainstorms in restored
streams and, in one site, no submerged or emergent
rocks remained after a storm. We could not determine
egg survival on rocks that rolled or went missing because
some rocks could not be recovered or were recovered
more than 2 weeks later and their eggs could have
hatched during this time. However, prior to rolling, 26%
(16 of 62) of the rolled and recovered rocks had egg
masses (n =140 masses), indicating insects lay eggs on
unstable substrates. Although survival of eggs on buried
or rolled rocks is unknown, mortality due to desiccation
can occur in less than 2 h for some mayflies and caddis-
flies (Miller et al. 2020) and we observed on multiple
occasions that rolling resulted in exposure of the eggs
above the water. In reference streams, emergent rocks
remained intact, except during heavy rainstorms in early
June. These results indicate that, in moderate-gradient,
rocky-bottomed, restored streams, insect egg laying is
limited by the number of emergent rocks. Further,
because emergent rocks in restored streams are unstable,
egg mortality may be higher due to emergent rocks roll-
ing or being buried in sediment. Thus, even if adult
females can disperse to restored streams and all other
habitat conditions are suitable, larval insect populations
may fail to establish because adult females lack stable
egg-laying habitat.
In small rocky-bottomed streams the characteristics of
suitable egg-laying habitat (i.e., large, stable, emergent,
unlikely to dry rocks in fast-flowing water) for many
stream insects are known and could be readily incorpo-
rated into restoration projects to increase stream insect
recruitment (Storey et al. 2017). However, to maximize
cost-effectiveness and achieve biological recovery,
restoration should consider the site- and taxon-specific
natural history and multivariate nature of egg-laying
habitat selection. Moreover, restoration projects could
be used to test the degree to which insects in restored
streams are limited by egg-laying habitat or other fac-
tors. For instance, adult dispersal from adjacent refer-
ence streams in degraded and disconnected landscapes,
or post-recruitment processes such as larval food
resources and mortality, factors that may negate the pos-
itive effects of restored egg-laying habitat.
One of the most vexing problems in stream restoration
is that the return of biological conditions lags behind the
return of physical habitat and water quality. Our obser-
vations suggest that, in addition to the number of suit-
able egg-laying sites, the stability of sites may limit insect
recruitment and, hence, recovery of insect diversity and
abundance in these and similar restored streams. Indeed,
many of the taxa that use the underside of emergent
rocks as egg-laying habitat are used as indicators of
stream health (Ephemeroptera, Plecoptera, and Tri-
choptera). Thus, considering the natural history of
stream insect egg laying during the hydrologic and geo-
morphologic engineering of restored streams is likely to
improve stream restoration and accelerate biological
recovery, as the importance of natural history has
informed other types of restoration projects (Tewksbury
et. al. 2014).
ACKNOWLEDGMENTS
This research was supported by the North Carolina Depart-
ment of Environmental Quality and Division of Mitigation Ser-
vices. We thank Periann Russell, Barbara Doll, and Eric Fleek
for advice selecting sites, Blake Dilworth for assistance with R,
Mick Demi and Jared Balik for suggestions that improved early
drafts, and the anonymous reviewers for their kind words and
valuable insight.
LITERATURE CITED
Alp, M., L. Indermaur, and C. T. Robinson. 2013. Environmen-
tal constraints on oviposition of aquatic invertebrates with
contrasting life cycles in two human-modified streams. Fresh-
water Biology 58:1932–1945.
Bernard, J. M., J. Fripp, and K. Robinson, editors. 2007. Stream
restoration design. National Engineering Handbook Part 654
(210-VINEH). U.S. Department of Agriculture, Natural
Resources Conservation Service, Washington, D.C.
Bovill, W. D., B. J. Downes, and J. Lancaster. 2013. A test of the
preference-performance hypothesis with stream insects: selec-
tive oviposition affects the hatching success of caddisfly eggs.
Freshwater Biology 58:2287–2298.
Encalada, A. C., and B. L. Peckarsky. 2012. Large-scale manip-
ulation of mayfly recruitment affects population size. Oecolo-
gia 168:967–976.
Kennedy, T. A., J. D. Muehlbauer, C. B. Yackulic, D. A. Lytle,
S. W. Miller, K. L. Dibble, E. W. Kortenhoeven, A. N. Met-
calfe, and C. V. Baxter. 2016. Flow management for hydro-
power extirpates aquatic insects, undermining river food
webs. BioScience 66:561–575.
Article e03331; page 4 THE SCIENTIFIC NATURALIST Ecology, Vol. xx, No. xx
Louhi, P., H. Mykr
€a, R. Paavola, A. Huusko, T. Vehanen, A.
M€aki-Pet€ays, and T. Muotka. 2011. Twenty years of stream
restoration in Finland: little response by benthic macroin-
vertebrate communities. Ecological Applications 21:1950–
1961.
Miller, S. W., M. Schroer, J. R. Fleri, and T. A. Kennedy. 2020.
Macroinvertebrate oviposition habitat selectivity and egg-
mass desiccation tolerances: implications for population
dynamics in large regulated rivers. Freshwater Science
39:584–599.
Palmer, M. A., K. L. Hondula, and B. J. Koch. 2014. Ecological
restoration of streams and rivers: shifting strategies and shift-
ing goals. Annual Review of Ecology, Evolution, and System-
atics 45:247–269.
Peckarsky, B. L., B. W. Taylor, and C. C. Caudill. 2000. Hydro-
logic and behavioral constraints on oviposition of stream
insects: implications for adult dispersal. Oecologia 125:186–
200.
Reich, P. 2004. Patterns of composition and abundance in
macroinvertebrate egg masses from temperate Australian
streams. Marine and Freshwater Research 55:39–56.
Smith, B. J., and R. G. Storey. 2018. Egg characteristics and
oviposition behaviour of the aquatic insect orders
Ephemeroptera, Plecoptera and Trichoptera in New Zealand:
a review. New Zealand Journal of Zoology 45:287–325.
Statzner, B., and L. A. B^eche. 2010. Can biological invertebrate
traits resolve effects of multiple stressors on running water
ecosystems? Freshwater Biology 55:80–119.
Storey, R. G., D. R. Reid, and B. J. Smith. 2017. Oviposition
site selectivity of some New Zealand aquatic macroinverte-
brate taxa and implications for stream restoration. New Zeal-
and Journal of Marine and Freshwater Research 51:165–181.
Tewksbury, J. J., et al. 2014. Natural history’s place in science
and society. BioScience 64:300–310.
Tullos, D. D., D. L. Penrose, G. D. Jennings, and W. G. Cope.
2009. Analysis of functional traits in reconfigured channels:
implications for the bioassessment and disturbance of river
restoration. Journal of the North American Benthological
Society 28:80–92.
SUPPORTING INFORMATION
Additional supporting information may be found in the
online version of this article at http://onlinelibrary.wiley.com/
doi/10.1002/ecy.3331/suppinfo
Xxxxx 2021 THE SCIENTIFIC NATURALIST Article e03331; page 5