Directory UMM :Data Elmu:jurnal:A:American Journal of Botany:Vol88.Issue1.Jan2001:
American Journal of Botany 88(1): 170–176. 2001.
BLADDER
FUNCTION IN
(LENTIBULARIACEAE):
UTRICULARIA PURPUREA
IS CARNIVORY IMPORTANT?1
JENNIFER H. RICHARDS
Department of Biological Sciences, Florida International University, Miami, Florida 33199 USA
Utricularia purpurea is a rootless, free-floating, aquatic, carnivorous plant. I quantified biomass investment in U. purpurea traps
and determined when traps begin to function and what they trap in natural habitats. In the Everglades of south Florida, plants invest
an average of 26% of their biomass in bladders, although bladder number varies among sites and over time. Leaves begin trapping as
they mature, and on leaves one whorl older than the most recently matured leaves, almost 100% of bladders have allochthonous
material. Despite the substantial investment in their biomass, bladders capture few aquatic microinvertebrates. Almost all mature
bladders, however, have living communities of algae, zooplankton, and associated debris. These results support the hypotheses that
the important association in U. purpurea bladders is a mutualism rather than a predator–prey interaction and that the major benefit to
the plants from bladders is derived from this community.
Key words:
purpurea.
bladderworts; Lentibulariaceae; periphyton; plant carnivory; plant nutrients; trapping rates; Utricularia; Utricularia
The morphological and physiological adaptations of carnivorous plants are among the most complex in the plant kingdom
(Juniper, Robins, and Joel, 1989). The complexity of these
adaptations encourages hypotheses about function on the basis
of structures. A critical evaluation of function, however, requires quantification of the numbers and types of prey captured
in natural environments, as well analysis of the prey’s contribution to the plant’s nutrient budget. The number of prey captured by Drosera and Pinguicula species under natural conditions has been reported (Karlsson et al., 1987; Thum, 1988;
Zamora, 1995), but comparable data are lacking for Utricularia, the most species-rich genus of carnivorous plants.
Carnivorous plants are defined as plants that (1) absorb nutrients from dead animals next to their surfaces and thus obtain
increased fitness and (2) have some morphological, physiological, or behavioral feature to attract, capture, and/or digest prey
(Givnish, 1989; Juniper, Robins, and Joel, 1989). The carnivorous habit has arisen independently at least six times (Albert,
Williams, and Chase, 1992). Forty-two percentage of the species classified as carnivorous occur in Utricularia (Juniper,
Robins, and Joel, 1989), and Utricularia has the widest geographical distribution of any carnivorous genus, with species
found worldwide from the tropics to arctic regions (Taylor,
1989). Habit varies in the genus, including free-floating and
affixed aquatics (;27%), lithophytes and epiphytes (;13%),
and terrestrial species that grow in seasonally wet or moist
environments (;60%) (Taylor, 1989). Members of the genus
are rootless and have small bladders that can trap allochthonous material. Each bladder has a door that flexes inward. Internal glands in the bladders pump water out, so that ‘‘set’’
traps are under tension with the door lodged against a lip of
cells. Mechanical stimulation causes the bladder door to flex
open, allowing water and material to be sucked into the bladder.
A cost-benefit model for the evolution of carnivory in plants
predicts that carnivory will evolve in sunny, moist, nutrientpoor environments (Givnish, 1989). This description typifies
Utricularia habitats in the Everglades of south Florida. Several
recent studies of Utricularia have examined the costs of carnivory, measuring investment in carnivory as investment in
bladders, quantified as bladder number, size, and/or biomass
(Friday, 1991, 1992; Knight and Frost, 1991; Knight, 1992).
Bladders are known to trap small aquatic animal prey, such as
rotifers, copepods, ostracods, cladocerans, and chironomids
(Friday, 1989; Knight and Frost, 1991), and they absorb N and
P from these prey (Friday and Quarmby, 1994). Thus, in Utricularia species the benefit for investing in bladders is assumed
to be nutrients derived from trapping and digesting aquatic
organisms. Studies of Utricularia that quantify trapping rates
and prey in natural environments, however, are limited. This
study quantifies investment in bladders and natural trapping
rates in Utricularia purpurea growing in the Everglades of
south Florida. In order to understand plasticity in investment
in carnivory, it also examines morphological variation in this
species.
Terminology—Although the morphological homology of the
trap-bearing structures in U. purpurea has been debated (Rutishauser and Sattler, 1989; Taylor, 1989), here they are referred to as leaves. These leaves occur in whorls along the
stem, are subdivided into photosynthetic filaments (the ‘‘capillary filaments’’ of Taylor, 1989) and bear bladders at their
tips (Fig. 1). The time between the initiation of successive leaf
whorls is referred to as a plastochron, a developmentally defined unit of time (Erickson and Michelini, 1957; Larson and
Isebrands, 1971).
1
Manuscript received 21 September 1999; revision accepted 28 March
2000.
The author thanks Dr. Evelyn Gaiser for help identifying algae and microinvertebrates, Tina Ugarte for technical assistance, and Dr. David Lee, Dr.
Tom Frost and Dr. Rolf Rutishauser for constructive reviews. This research
was completed as an ancillary project to ongoing research at Florida International University’s Southeast Environmental Research Center (FIU SERC).
Access to field sites was provided by the SERC ecosystem project ‘‘Numerical
interpretation of Class III narrative nutrient water criteria for Everglades Wetlands,’’ funded through Cooperative Agreement CA5000–3–9030. Plants were
collected under USDI National Park Service collecting permit numbers
19970026 and 19980065.
MATERIALS AND METHODS
Materials—Utricularia purpurea Walter is a free-floating, rootless aquatic
macrophyte that is native from Canada to Central America and the Antilles
(Taylor, 1989). This study was conducted in south Florida, where U. purpurea
170
January 2001]
RICHARDS—U. PURPUREA
BLADDER FUNCTION
171
Fig. 1. A diagram of Utricularia purpurea vegetative morphology; brackets and curved lines delimit morphological units that were measured (bracket) or
counted (curved line). (A) Whole plant; bladders are shown only at the end of each leaf, but they can be borne at the tips of all leaf filaments. (B) Individual
leaf from a leaf whorl; four whorls of photosynthetic filaments (1–4) are illustrated; these filaments can have two additional orders of branching.
occurs as isolated individuals or as an integral part of a floating periphyton
mat. All of the data reported in this study were collected on plants that were
free-floating, rather than those in the periphyton mat. These plants grow
throughout the water column from the surface to the benthic muck. Plants
were collected from Shark River Slough in Everglades National Park at three
sites (sites A, B, and C) that are part of a long-term experimental study on
the effects of phosphorus additions on the Everglades ecosystem (Childers et
al., in press). The three sites are located in long-hydroperiod marshes dominated by Eleocharis cellulosa, Panicum hemitomon, Nymphaea odorata, Nymphoides aquatica, Sagittaria lancifolia, and the U. purpurea–periphyton mat.
Two other species of Utricularia, U. foliosa L. and U. gibba L., are also
found at these sites. Water column depth in the study area ranges from ;0.25
to 1 m. Growth rates were measured on plants in flume channels prior to P
Fig. 2. Percentage of occupied bladders by leaf age; 0 5 first mature
whorl of leaves. Younger leaf whorls are numbered negatively, older leaf
whorls are positive. N 5 30 plants, with 4–6 leaf whorls and one leaf per
whorl examined per plant; 20 bladders were analyzed per leaf. Error bars 5
1 SD.
dosing. Plants for morphological studies and bladder-content analyses were
collected adjacent to the experimental flumes.
Methods—Morphological measurements and biomass allocation—Plants
for morphological measurements, bladder content data, and biomass allocation
studies were collected during the wet season, which occurs from May to
October in south Florida. Plants from site A were collected on 23 October
and 28 October 1998, from site B on 13 July and 22 July 1998, and from
site C on 22 July 1998. Plants from sites B and C were also processed for
biomass allocation data; the material for biomass allocation from site A was
lost, so an additional collection from site A was made during the dry season
on 21 March 1999. Plastic collecting bags were submerged and filled with
water, and plants were then moved gently from the water column into the
bags, so they were not exposed to air during collection. Plants were brought
back to the laboratory, refrigerated, and processed within 4 d of collection.
Similar morphological data were taken on all plants. Traits measured were
internode length and diameter, leaf number per whorl, leaf length, number of
primary photosynthetic whorls per leaf, and bladder number. For morphological data a single leaf was measured per node from the two most recently
matured nodes per plant. For biomass allocation studies, three leaves from
the most recently matured node were measured.
Since each whorl of leaves is a developmental unit and the plant then
consists of units that are produced at the apex and sloughed off at the base,
I analyzed biomass investment in trapping on the basis of investment in a
nodal unit, equal to the whorl of leaves and the subtending internode. The
first mature nodal unit of plants was removed. Internode length and leaf length
were measured with a ruler to the nearest millimetre (Fig. 1). Internode diameter was measured with electronic digital calipers (MAX-CAL, Fowler Co.,
Inc., Newton, Massachusetts, USA). Leaf number per node and number of
whorls of primary photosynthetic filaments per leaf were counted on three
leaves at the node. On each of these three leaves, bladders were counted,
removed, and dried in an 808C oven to constant mass. Each leaf minus its
bladders was dried separately, as was the subtending internode. These parts
were weighed on a microbalance (Mettler Toledo AB54, Mettler Toledo Inc.,
Westerville, Ohio, USA). Percentage investment in bladders and support struc-
172
AMERICAN JOURNAL
tures was calculated from these data. Biomasses of leaves and bladders per
leaf for a whorl were averaged and multiplied by the number of leaves per
whorl to obtain biomass per leaf whorl. This was added to the dry mass of
the subtending internode to determine biomass per nodal unit, and relative
biomass allocations to bladders, leaves, and stem (internode) were calculated.
Growth-rate determinations—Growth rates were determined for free-floating plants; initial measurements indicated that plants in mats had much slower
growth rates than those reported here. I marked 12–15 plants at four permanent quadrats at each flume site. The free end of a string tied to a styrofoam
packing peanut was tied above the most recently matured node of an individual U. purpurea plant. This node was defined based on internode length, leaf
length, amount of leaf reflexion, and bladder expansion. The floating styrofoam peanut enabled me to relocate marked plants after several weeks of
growth in situ. Plants grew for 4–6 wk, and then the number of mature nodes
above the marked node was counted. New plants were marked and followed
for every sampling date. Growth was measured every 2–4 mo from October
1997 to November 1998.
Leaf development and bladder contents—Plants used for morphological
measurements from sites A, B, and C were analyzed for bladder contents.
Leaf whorls were numbered based on relative developmental stage. The most
recently matured leaf whorl, as determined by morphological criteria (see
description above), was designated as 0. Older leaf whorls were numbered
positively, and younger whorls numbered negatively from the 0 whorl. For
developmental and bladder content analysis, a single leaf was sampled from
each node beginning with node 1 and sampling toward the apex to the 23 to
24 whorl. Leaf length and length of the subtending internode were measured
for each whorl, while leaf number and the number of whorls of leaf subdivisions were counted. Bladder number was also counted for all but the youngest leaves. Twenty bladders from each leaf were examined under a Leitz
Dialux 20 compound light microscope. Bladders were analyzed by focusing
down through the bladder and recording presence or absence of allochthonous
material. The bladder walls are a single cell thick and relatively transparent,
as illustrated in Figs. 7 and 9, which were taken through bladder walls, and
as can be seen in Figs. 5, 6, and 8, which show dark contents through transparent bladder walls. Internal and external walls of bladders were identified
by the different types of glands present on the inside and outside of bladders
(Juniper, Robins, and Joel, 1989; Taylor, 1989). The types of multicellular
invertebrates present per bladder were quantified for samples from sites A
and C.
Statistical analyses—Data were checked for normality; nonnormal data
were transformed prior to analysis or were analyzed with nonparametric tests.
Site differences were analyzed using one-way ANOVA or Kruskal-Wallis tests
in the JMP version 3 for MacIntosh computers. Pairwise comparisons were
made using Student’s t test. Variation in bladder number among leaves, plants,
and sites was analyzed with nested ANOVA. Averages are reported 61 SD.
RESULTS
Morphology—Free-floating individuals of U. purpurea are
;0.5 m in total length in Shark River Slough. The plants produce whorls of leaves as they grow at the tip and rot at the
base (Fig. 1). Every second node can produce a branch bud
and an inflorescence bud. Inflorescences grow above the water
and produce one or two flowers (Fig. 1; for additional details
of development, see Rutishauser and Sattler, 1989). Internodes
of plants used for the bladder content study averaged 29 mm
in length and 1.1 mm in diameter (Table 1). A typical plant
in Shark River Slough tends to produce five leaves per whorl
(Table 1; 63% of nodes had five leaves, 34% had six leaves,
and 3% had four leaves). Each leaf produced whorls of four
or fewer photosynthetic filaments, which could themselves be
subdivided (Figs. 1B, 3). Most Shark River Slough plants had
five of these whorls per leaf (Table 1; 65% had five whorls,
OF
BOTANY
[Vol. 88
27% had four whorls, and 8% had six whorls). Individual leaf
length averaged 39 mm (Table 1). Bladders of U. purpurea
are borne at the tips of all orders of leaf filaments.
Site and seasonal variation in morphology—Variation in
bladder number/leaf between leaves at a single node was not
significant (P 5 0.3670) nor was variation between bladder
number/leaf at successive nodes (P 5 0.9514). Bladder number/leaf, however, did vary significantly among plants (P ,
0.0001).
Neither internode diameter nor leaf length varied significantly among sites, and plants had similar numbers of leaves
per nodal whorl and similar numbers of whorls of photosynthetic filaments per leaf (Table 1). Plants did differ significantly among sites with respect to internode length and the
number of bladders. In general, plants at sites A and B tended
to resemble each other and to differ from plants at site C.
Plants at site C produced more bladders per leaf and node than
plants at the other two sites (Table 1).
Plants at site A that were collected in the wet season and
the dry season were not significantly different in leaf length
(40 6 9 mm in October 1998, vs. 44 6 8 mm in March 1999)
but did differ significantly in bladder number (55 6 31 bladders per leaf in October 1999 vs. 87 6 34 bladders per leaf
in March 1999).
Investment in bladders—Individual leaves in a whorl produced an average of 74 bladders per leaf (Table 2). Given an
average of five leaves per whorl, each whorl was estimated to
have 370 bladders.
Utricularia purpurea plants invested an average of 26 6
10% of their biomass in bladders; investment varied from 9
to 45%. Most plants allocated the majority of their biomass to
the leaf filaments that support the bladders. The biomass of
these filaments, which are the major photosynthetic portion of
the plant, averaged 64% of the total biomass of a nodal unit
(bladders 1 supporting filaments 1 subtending internode), or
2.3 times the biomass of the bladders themselves (Table 2).
The internode supporting the whorl contributed 10% of the
total biomass of the leaf whorl.
Leaf and internode dry masses did not vary significantly
among sites, but bladder dry mass per leaf and individual bladder biomass did (Table 2). Plants at site C produced significantly more bladders per leaf than those at site B during the
wet season. The dry-season plants from site A were similar to
wet-season site C plants in their investment in bladders and
differed significantly from wet-season site B plants (Table 2).
Bladder number was negatively correlated with individual
bladder biomass, so that plants with more bladders tended to
have lighter bladders, although the correlation was not strong
(R2 5 0.09, P 5 0.0052, N 5 86).
Trapping efficiency—Bladders on very young leaves of U.
purpurea did not have allochthonous material inside them
(Fig. 2). These bladders, which were small, anthocyanic, and
had bladder doors that were not fully developed (Figs. 4, 5),
were on leaf whorls 2–4 plastochrons younger than the first
whorl of mature leaves. The youngest leaf with bladder contents was two plastochrons younger than the first mature leaves
(Fig. 2). From this leaf whorl the number of occupied bladders
increased exponentially with leaf age until bladders in the leaf
one plastochron older than the first mature leaf were almost
100% occupied (Figs. 2, 4–9). The amount of material in each
January 2001]
RICHARDS—U. PURPUREA
BLADDER FUNCTION
173
Figs. 3–9. Utricularia purpurea leaves and bladders. 3. Portion of a leaf photosynthetic filament. 4. Bladder from 22 leaf. 5. Bladder from 21 leaf. 6.
Bladder from 0 leaf. 7. Interior of bladder in Fig. 6; arrow points to a green alga. 8. Bladder from 11 leaf. 9. Interior of bladder in Fig. 8; arrows point to
living algae (desmids and diatoms). Bar 5 20 mm in Figs. 3, 4, 7, 9; bar 5 100 mm in Figs. 5, 6, 8. D 5 bladder door; H 5 hairs on bladder door; P 5
photosynthetic epidermis.
174
AMERICAN JOURNAL
OF
BOTANY
[Vol. 88
TABLE 1. Morphological measurements from the most recently matured nodes of U. purpurea. N 5 10 plants per site, with a single leaf from the
most recently matured node and the next older node measured on each plant.
Characteristic
Site A
Internode length (mm)
Internode diameter (mm)
Leaf no./node1
Leaf length (mm)
No. of filament whorls/leaf1
No. of bladders/leaf
Bladder no./leaf length
32.0 6 6.0
1.1 6 0.2
5 (4–6)
39.7 6 9.1
5 (4–6)
55 6 31a
1.3 6 0.5a
1
2
Site B
Site C
28.9 6 4.1
1.3 6 0.3
5 (4–6)
38.5 6 4.8
5 (4–6)
59 6 28ab
1.5 6 0.6ab
a
26.9 6 6.5
1.0 6 0.3
5 (5–6)
37.9 6 4.9
5 (4–5)
77 6 32b
1.9 6 0.7b
ab
b
Average
Significant?2
29.4 6 6.1
1.1 6 0.3
5 (4–6)
38.7 6 6.6
5 (4–6)
64 6 31
1.6 6 0.7
Y
N
N
Y
Y
Mode (Range).
For characteristics with significant differences (P # 0.05) among them, measurements followed by the same letter are not significantly different.
bladder also increased with increasing leaf age (Figs. 5, 6, 8).
Occupied bladders contained a mixture of living and detrital
material. All bladders with contents had some type of photosynthetic tenants, primarily blue–green algae, diatoms, green
algae, and/or photosynthetic protists (Figs. 7, 9). Many bladders also had living rotifers. A variety of larger organisms,
such as copepods, ostracods, and cladocerans, were found
more rarely. At sites A and C, 880 of 1400 bladders examined
from different-aged leaves had contents. Of these, 20.2% had
living rotifers, 5.1% had cladocerans, 1.3% had chironomids,
0.6% had ostracods, 0.5% had copepods, and 0.9% had other
multicellular invertebrates. The rotifers in these samples were
swimming around in the bladders, whereas the larger invertebrates were usually dead.
Growth rates—Utricularia purpurea expanded 1–2 nodes
per week (Fig. 10). Growth varied seasonally, being greater in
the summer wet season and slower in the winter dry season
(Fig. 10). Growth rates were greatest in May and June at the
beginning of the wet season and began to decline later in the
summer.
DISCUSSION
Are these results general?—The results presented here
question the importance of trapping small aquatic animals for
growth of Utricularia purpurea. The data were obtained in the
Everglades of south Florida, however, where invertebrate
standing stocks are remarkably low as compared to other wetlands (Turner et al., 1999). Are the trapping rates and trap
contents reported here unique to Utricularia species in the
Everglades? Whether the trapping rates and bladder communities found in the Everglades are representative of Utricularia
species in other environments or of terrestrial species of Utricularia needs further study. Previous researchers, however,
have reported the presence of living algae and ciliates in blad-
ders of Utricularia species (Botta, 1976; Mosto, 1979; Juniper,
Robins, and Joel, 1989).
Investment in bladders—Species of Utricularia invest a
large proportion of their biomass in bladders. While U. purpurea invested approximately one-quarter of its biomass in
bladders, U. macrorhiza growing in lakes in northern Wisconsin invested 38–48% (Knight and Frost, 1991), and U. vulgaris
growing in disused clay pits in Cambridgeshire, UK, invested
;50% biomass in bladders (Friday, 1992). I have also found
an average of 29% investment in bladders from plants collected from a pond on the Florida International University
campus (J. H. Richards, Florida International University, unpublished data). This percentage investment is similar to what
terrestrial plants in moderately low-nutrient environments invest in root biomass (Tilman, 1988)
The percentage of biomass invested in bladders is plastic
and can vary with environmental conditions. In all studies that
have considered the question, the largest variation in bladder
number is between sites. In U. purpurea, as in other species
of Utricularia (Knight and Frost, 1991; Friday, 1992), variation among leaves on a plant sampled at a single time is usually minor. Variation among plants at a single site is more
complex and probably represents the degree of habitat heterogeneity displayed at a particular site. Whereas Knight and
Frost (1991) found minor variation in bladder number among
plants in a lake, Friday (1992) found large variation among
plants at a single site. In the data reported here U. purpurea
bladder number varied among plants both within and between
sites.
Benefits from bladders—The conventional view of bladderwort carnivory is that these plants benefit from nutrients
derived by digestion of trapped microinvertebrates (Friday,
1989; Juniper, Robins, and Joel, 1989; Knight and Frost, 1991;
Ulanowicz, 1995). In south Florida bladders of U. purpurea
TABLE 2. Biomass allocation to bladders, filaments, and internode in the first mature nodal unit (5 leaf whorl 1 subtending internode) of U.
purpurea. N 5 8–12 plants per site, with 3 leaves from the most recently-matured node measured on each plant. Samples for site A were
collected in March 1999; samples for sites B and C were collected in July 1998.
Characteristic
Site A
Nodal unit dry mass (mg)
Bladder no.
Bladder dry mass (mg)
Bladder dry mass/nodal unit (mg)
Leaf filament dry mass/nodal unit (mg)
Internode dry mass/nodal unit (mg)
6
6
6
6
6
6
1
25
86
20
8
15
2.3
Site B
a
8
31a
6a
3a
6
1.1
17
52
12
3
14
2.0
6
6
6
6
6
6
Site C
b
7
34b
5b
1b
9
1.1
20
83
18
6
14
1.6
6
6
6
6
6
6
Average
b
3
34a
10a
3a
6
0.3
21
74
17
6
14
2.0
6
6
6
6
6
6
8
36
8
3
7
1
Significant?1
Y
Y
Y
Y
N
N
For characteristics with significant differences (P # 0.05) among them, measurements followed by the same letter are not significantly different.
January 2001]
RICHARDS—U. PURPUREA
BLADDER FUNCTION
175
stantly augmented by material from new trapping events or
whether bladder contents increase through reproduction inside
the bladders is unknown.
Although the importance of the Utricularia-periphyton interaction has been recognized (Friday, 1989; Ulanowicz,
1995), the possible importance of the direct interaction has
been overshadowed by concepts derived from models of carnivory. The interpretation of the benefit of carnivory presented
here provides an explanation for Knight and Frost’s (1989)
otherwise anomalous result that U. macrorhiza plants in more
nutrient-rich environments invested more in bladders. If the
benefit of the bladders depends on the productivity of the microcommunity and that productivity increases with increased
nutrients, plants would be predicted to produce more bladders
under increased nutrient levels.
Fig. 10. Average number of new leaf whorls produced by U. purpurea
per week measured at 1–4 mo intervals between June 1997 and November
1998. Data from sites A, B, and C combined. N $ 80 for each point. Error
bars 5 61 SD.
on plants in natural habitats trap very low percentages of these
microinvertebrates, except for rotifers, and the rotifers observed inside bladders were alive. Similarly low invertebrate
trapping rates were found for U. foliosa in the Everglades
(Bern, 1997; A. Bern, J. H. Richards, and B. Fry, Florida
International University, unpublished data). Calculations of
how much nitrogen and phosphorus plants could derive from
these trapping rates, if the only benefit is from digestion of
microinvertebrates, suggest that the return on a 25–50% investment in bladders is ,1% of the plant N and P (Bern, 1997;
A. Bern, J. H. Richards, and B. Fry, Florida International University, unpublished data).
Although plants had few bladders with dead microinvertebrates, almost 100% of mature U. purpurea bladders supported
living communities of microorganisms and associated detritus.
These communities were derived from the external environment, as very young bladders lacked them, and the number of
inhabited bladders, as well as inhabitant density and diversity,
increased over time. The ubiquitous presence of these communities supports the hypothesis that Utricularia plants derive
more benefit from by-products of this community than from
carnivory, i.e., that the important association in Utricularia
bladders is a mutualism rather than a predator–prey interaction. The bladders in Utricularia, therefore, may provide benefit through a detrital food web rather than a carnivorous interaction.
This view is a simplification of Ulanowicz’s (1995) positive
feedback model for Utricularia carnivory, in which the benefit
funneled to Utricularia from periphyton comes through ingestion of animal grazers. In the model this benefit allows Utricularia carnivory to be advantageous in oligotrophic environments (Ulanowicz, 1995). I suggest that the carnivory part of
the periphyton-Utricularia association may be incidental to the
direct interaction.
In the summer plants expand a leaf whorl about every 5 d,
and they begin trapping in 22 leaves, so they become 100%
occupied after 15 d. Free-floating plants generally have at least
three whorls of mature leaves before they become completely
covered with algae and appear to be senescent. Thus, after
bladders are fully occupied, they may support these communities for another 15 d. After a bladder has been tripped, it can
reset itself (Lloyd, 1933; Sydenham and Findlay, 1975; Juniper, Robins, and Joel, 1989). Whether communities are con-
Currency of the ‘‘carnivorous’’ interaction in aquatic
plants—Utricularia is the most diverse genus of carnivorous
plants, accounting for a little less than half of all carnivorous
species (Juniper, Robins, and Joel, 1989), but the genus is
unusual in having submerged aquatic species. The interaction
in an aquatic species, such as Utricularia purpurea, may differ
substantially from carnivory in terrestrial species of Utricularia or in species of other carnivorous plants. If the community interaction inside bladders provides the benefit for investment of biomass in bladders in U. purpurea, the currency
in the interaction may not be solely mineral nutrients. For example, carbon dioxide concentrations potentially limit photosynthesis in freshwater aquatic plants (Falkowski and Raven,
1997). Carbon dioxide derived from respiratory processes in
the bladders could be a significant benefit in the slow-moving
aquatic environments typical for some Utricularia species,
similar to the benefit obtained from CO2 uptake through the
roots in isoetids (Raven et al., 1988). This benefit could help
to explain Utricularia’s success in the aquatic environment.
The endangered Aldrovanda vesiculosa (Droseraceae), a miniature aquatic version of Dionea, is the only other free-floating
aquatic carnivorous species. A recent study found that the
most important condition for rapid growth in A. vesiculosa was
a high-CO2 concentration in the water (Adamec, 1997).
In addition to the possible benefit of CO2 supplements, bladders in aquatic species could also be advantageous because
they provide a contained microenvironment to enhance breakdown of organic matter, prevent diffusion of the breakdown
products, and provide Utricularia plants with a monopoly on
released nutrients.
Bladder mechanics and community dynamics—The absence of microinvertebrates from many bladders coupled with
the presence of some type of material in all of the bladders
means either that trapping is very inefficient or that bladders
do not require animals to trip them under natural conditions.
Alternative possibilities that need to be investigated are (1)
debris carried by currents or the current itself is sufficient to
cause opening or (2) endogenous factors, such as internal tension, can cause bladder opening. Lloyd (1933) reports that
movement of U. purpurea plants in water often releases the
traps.
LITERATURE CITED
ADAMEC, L. 1997. Photosynthetic characteristics of the aquatic carnivorous
plant Aldrovanda vesiculosa. Aquatic Botany 59: 297–306.
176
AMERICAN JOURNAL
ALBERT, V. A., S. E. WILLIAMS, AND M. W. CHASE. 1992. Carnivorous
plants: phylogeny and structural evolution. Science 257: 1491–1495.
BERN, A. L. 1997. Studies on nitrogen and phosphorus uptake by the carnivorous bladderwort, Utricularia foliosa, in south Florida wetlands.
Master’s thesis, Florida International University, Miami, Florida, USA.
BOTTA, S. M. 1976. Sobre las trampas y las victimas o presas de algunas
especies argentinas del genero Utricularia. Darwiniana 20: 127–154.
CHILDERS, D. L., R. D. JONES, J. TREXLER, C. BUZZELLI, S. DAILEY, A.
EDWARDS, E. GAISER, K. JAYACHANDARAN, A. KENNE, D. LEE, J.
MEEDER, M. K. NAIR, J. H. K. PECHMANN, A. RENSHAW, J. RICHARDS,
M. RUGGE, L. J. SCINTO, P. STERLING, AND W. VAN GELDER. In press.
Quantifying the effects of low-level phosphorus enrichment on unenriched Everglades wetlands with in situ flumes and phosphorus dosing.
In K. Porter and J. Porter [eds.], The Everglades hydroscape. St. Lucie
Press, Delray Beach, Florida, USA.
ERICKSON, R. O., AND F. J. MICHELINI. 1957. The plastochron index. American Journal of Botany 44: 297–305.
FALKOWSKI, P. G., AND J. A. RAVEN. 1997. Aquatic photosynthesis. Blackwell Science, Malden, Massachusetts, USA.
FRIDAY, L. E. 1989. Rapid turnover of traps in Utricularia vulgaris L. Oecologia 80: 272–277.
———. 1991. The size and shape of traps of Utricularia vulgaris L. Functional Ecology 5: 602–607.
———. 1992. Measuring investment in carnivory: seasonal and individual
variation in trap number and biomass in Utricularia vulgaris L. New
Phytologist 121: 439–445.
———, AND C. QUARMBY. 1994. Uptake and translocation of prey-derived
15
N and 32P in Utricularia vulgaris L. New Phytologist 126: 273–281.
GIVNISH, T. J. 1989. Ecology and evolution of carnivorous plants. In W. G.
Abrahamson [ed.], Plant-animal interactions, 243–290. McGraw–Hill,
New York, New York, USA.
JUNIPER, B. E., R. J. ROBINS, AND D. M. JOEL. 1989. The carnivorous plants.
Academic Press, London, UK.
KARLSSON, P. S., K. O. NORDELL, S. EIREFELT, AND A. SVENSSON. 1987.
Trapping efficiency of three carnivorous Pinguicula species. Oecologia
73: 518–521.
KNIGHT, S. E. 1992. Costs of carnivory in the common bladderwort, Utricularia macrorhiza. Oecologia 89: 348–355.
OF
BOTANY
[Vol. 88
———, AND T. M. FROST. 1991. Bladder control in Utricularia macrorhiza:
lake-specific variation in plant investment in carnivory. Ecology 72: 728–
734.
LARSON, P. R., AND J. G. ISEBRANDS. The plastochron index as applied to
developmental studies of cottonwood. Canadian Journal of Forest Research 1: 1–11.
LLOYD, F. E. 1933. The structure and behavior of Utricularia purpurea. Canadian Journal of Research 8: 234–252.
MOSTO, P. 1979. Algas en trampas de Utricularia oligosperma St. Hill. Boletin. Sociedad Argentina de Botanica 18: 89–100.
RAVEN, J. A., L. L. HANDLEY, J. J. MACFARLANE, S. MCINROY, L. MCKENZIE, J. H. RICHARDS, AND G. SAMUELSSON. 1988. The role of CO2
uptake by roots and CAM in acquisition of inorganic C by plants of the
isoetid lifeform: a review, with new data on Eriocaulon decangulare L.
New Phytologist 108: 125–148.
RUTISHAUSER, R., AND R. SATTLER. 1989. Complementarity and heuristic
value of contrasting models in structural botany. III. Case study on shootlike ‘‘leaves’’ and leaf-like ‘‘shoots’’ in Utricularia macrorhiza and U.
purpurea (Lentibulariaceae). Botanische Jahrbu¨cher Systematic fu¨r
Pflanzengeschichte und Pflanzengeographie 111: 121–137.
SYDENHAM, P. H., AND G. P. FINDLAY. 1975. Transport of solutes and water
by resetting bladders of Utricularia. Australian Journal of Plant Physiology 2: 225–351.
TAYLOR, P. 1989. The genus Utricularia. Kew Bulletin Additional Series
XIV. HMSO, London, UK.
THUM, M. 1988. The significance of carnivory for the fitness of Drosera in
its natural habitat. 2. The amount of captured prey and its effect on
Drosera intermedia and Drosera rotundifolia. Oecologia 81: 401–411.
TILMAN, D. 1988. Plant strategies and the dynamics and structure of plant
communities. Princeton University Press, Princeton, New Jersey, USA.
TURNER, A. M., J. C. TREXLER, C. F. JORDAN, S. J. SLACK, P. GEDDES, J. H.
CHICK, AND W. F. LOFTUS. 1999. Targeting ecosystem features for conservation: standing crops in the Florida Everglades. Conservation Biology 13: 898–911.
ULANOWICZ, R. E. 1995. Utricularia’s secret: the advantage of positive feedback in oligotrophic environments. Ecological Modelling 79: 49–57.
ZAMORA, R. 1995. The trapping success of a carnivorous plant, Pinguicula
vallisneriifolia: the cumulative effects of availability, attraction, retention
and robbery of prey. Oikos 73: 309–322.
BLADDER
FUNCTION IN
(LENTIBULARIACEAE):
UTRICULARIA PURPUREA
IS CARNIVORY IMPORTANT?1
JENNIFER H. RICHARDS
Department of Biological Sciences, Florida International University, Miami, Florida 33199 USA
Utricularia purpurea is a rootless, free-floating, aquatic, carnivorous plant. I quantified biomass investment in U. purpurea traps
and determined when traps begin to function and what they trap in natural habitats. In the Everglades of south Florida, plants invest
an average of 26% of their biomass in bladders, although bladder number varies among sites and over time. Leaves begin trapping as
they mature, and on leaves one whorl older than the most recently matured leaves, almost 100% of bladders have allochthonous
material. Despite the substantial investment in their biomass, bladders capture few aquatic microinvertebrates. Almost all mature
bladders, however, have living communities of algae, zooplankton, and associated debris. These results support the hypotheses that
the important association in U. purpurea bladders is a mutualism rather than a predator–prey interaction and that the major benefit to
the plants from bladders is derived from this community.
Key words:
purpurea.
bladderworts; Lentibulariaceae; periphyton; plant carnivory; plant nutrients; trapping rates; Utricularia; Utricularia
The morphological and physiological adaptations of carnivorous plants are among the most complex in the plant kingdom
(Juniper, Robins, and Joel, 1989). The complexity of these
adaptations encourages hypotheses about function on the basis
of structures. A critical evaluation of function, however, requires quantification of the numbers and types of prey captured
in natural environments, as well analysis of the prey’s contribution to the plant’s nutrient budget. The number of prey captured by Drosera and Pinguicula species under natural conditions has been reported (Karlsson et al., 1987; Thum, 1988;
Zamora, 1995), but comparable data are lacking for Utricularia, the most species-rich genus of carnivorous plants.
Carnivorous plants are defined as plants that (1) absorb nutrients from dead animals next to their surfaces and thus obtain
increased fitness and (2) have some morphological, physiological, or behavioral feature to attract, capture, and/or digest prey
(Givnish, 1989; Juniper, Robins, and Joel, 1989). The carnivorous habit has arisen independently at least six times (Albert,
Williams, and Chase, 1992). Forty-two percentage of the species classified as carnivorous occur in Utricularia (Juniper,
Robins, and Joel, 1989), and Utricularia has the widest geographical distribution of any carnivorous genus, with species
found worldwide from the tropics to arctic regions (Taylor,
1989). Habit varies in the genus, including free-floating and
affixed aquatics (;27%), lithophytes and epiphytes (;13%),
and terrestrial species that grow in seasonally wet or moist
environments (;60%) (Taylor, 1989). Members of the genus
are rootless and have small bladders that can trap allochthonous material. Each bladder has a door that flexes inward. Internal glands in the bladders pump water out, so that ‘‘set’’
traps are under tension with the door lodged against a lip of
cells. Mechanical stimulation causes the bladder door to flex
open, allowing water and material to be sucked into the bladder.
A cost-benefit model for the evolution of carnivory in plants
predicts that carnivory will evolve in sunny, moist, nutrientpoor environments (Givnish, 1989). This description typifies
Utricularia habitats in the Everglades of south Florida. Several
recent studies of Utricularia have examined the costs of carnivory, measuring investment in carnivory as investment in
bladders, quantified as bladder number, size, and/or biomass
(Friday, 1991, 1992; Knight and Frost, 1991; Knight, 1992).
Bladders are known to trap small aquatic animal prey, such as
rotifers, copepods, ostracods, cladocerans, and chironomids
(Friday, 1989; Knight and Frost, 1991), and they absorb N and
P from these prey (Friday and Quarmby, 1994). Thus, in Utricularia species the benefit for investing in bladders is assumed
to be nutrients derived from trapping and digesting aquatic
organisms. Studies of Utricularia that quantify trapping rates
and prey in natural environments, however, are limited. This
study quantifies investment in bladders and natural trapping
rates in Utricularia purpurea growing in the Everglades of
south Florida. In order to understand plasticity in investment
in carnivory, it also examines morphological variation in this
species.
Terminology—Although the morphological homology of the
trap-bearing structures in U. purpurea has been debated (Rutishauser and Sattler, 1989; Taylor, 1989), here they are referred to as leaves. These leaves occur in whorls along the
stem, are subdivided into photosynthetic filaments (the ‘‘capillary filaments’’ of Taylor, 1989) and bear bladders at their
tips (Fig. 1). The time between the initiation of successive leaf
whorls is referred to as a plastochron, a developmentally defined unit of time (Erickson and Michelini, 1957; Larson and
Isebrands, 1971).
1
Manuscript received 21 September 1999; revision accepted 28 March
2000.
The author thanks Dr. Evelyn Gaiser for help identifying algae and microinvertebrates, Tina Ugarte for technical assistance, and Dr. David Lee, Dr.
Tom Frost and Dr. Rolf Rutishauser for constructive reviews. This research
was completed as an ancillary project to ongoing research at Florida International University’s Southeast Environmental Research Center (FIU SERC).
Access to field sites was provided by the SERC ecosystem project ‘‘Numerical
interpretation of Class III narrative nutrient water criteria for Everglades Wetlands,’’ funded through Cooperative Agreement CA5000–3–9030. Plants were
collected under USDI National Park Service collecting permit numbers
19970026 and 19980065.
MATERIALS AND METHODS
Materials—Utricularia purpurea Walter is a free-floating, rootless aquatic
macrophyte that is native from Canada to Central America and the Antilles
(Taylor, 1989). This study was conducted in south Florida, where U. purpurea
170
January 2001]
RICHARDS—U. PURPUREA
BLADDER FUNCTION
171
Fig. 1. A diagram of Utricularia purpurea vegetative morphology; brackets and curved lines delimit morphological units that were measured (bracket) or
counted (curved line). (A) Whole plant; bladders are shown only at the end of each leaf, but they can be borne at the tips of all leaf filaments. (B) Individual
leaf from a leaf whorl; four whorls of photosynthetic filaments (1–4) are illustrated; these filaments can have two additional orders of branching.
occurs as isolated individuals or as an integral part of a floating periphyton
mat. All of the data reported in this study were collected on plants that were
free-floating, rather than those in the periphyton mat. These plants grow
throughout the water column from the surface to the benthic muck. Plants
were collected from Shark River Slough in Everglades National Park at three
sites (sites A, B, and C) that are part of a long-term experimental study on
the effects of phosphorus additions on the Everglades ecosystem (Childers et
al., in press). The three sites are located in long-hydroperiod marshes dominated by Eleocharis cellulosa, Panicum hemitomon, Nymphaea odorata, Nymphoides aquatica, Sagittaria lancifolia, and the U. purpurea–periphyton mat.
Two other species of Utricularia, U. foliosa L. and U. gibba L., are also
found at these sites. Water column depth in the study area ranges from ;0.25
to 1 m. Growth rates were measured on plants in flume channels prior to P
Fig. 2. Percentage of occupied bladders by leaf age; 0 5 first mature
whorl of leaves. Younger leaf whorls are numbered negatively, older leaf
whorls are positive. N 5 30 plants, with 4–6 leaf whorls and one leaf per
whorl examined per plant; 20 bladders were analyzed per leaf. Error bars 5
1 SD.
dosing. Plants for morphological studies and bladder-content analyses were
collected adjacent to the experimental flumes.
Methods—Morphological measurements and biomass allocation—Plants
for morphological measurements, bladder content data, and biomass allocation
studies were collected during the wet season, which occurs from May to
October in south Florida. Plants from site A were collected on 23 October
and 28 October 1998, from site B on 13 July and 22 July 1998, and from
site C on 22 July 1998. Plants from sites B and C were also processed for
biomass allocation data; the material for biomass allocation from site A was
lost, so an additional collection from site A was made during the dry season
on 21 March 1999. Plastic collecting bags were submerged and filled with
water, and plants were then moved gently from the water column into the
bags, so they were not exposed to air during collection. Plants were brought
back to the laboratory, refrigerated, and processed within 4 d of collection.
Similar morphological data were taken on all plants. Traits measured were
internode length and diameter, leaf number per whorl, leaf length, number of
primary photosynthetic whorls per leaf, and bladder number. For morphological data a single leaf was measured per node from the two most recently
matured nodes per plant. For biomass allocation studies, three leaves from
the most recently matured node were measured.
Since each whorl of leaves is a developmental unit and the plant then
consists of units that are produced at the apex and sloughed off at the base,
I analyzed biomass investment in trapping on the basis of investment in a
nodal unit, equal to the whorl of leaves and the subtending internode. The
first mature nodal unit of plants was removed. Internode length and leaf length
were measured with a ruler to the nearest millimetre (Fig. 1). Internode diameter was measured with electronic digital calipers (MAX-CAL, Fowler Co.,
Inc., Newton, Massachusetts, USA). Leaf number per node and number of
whorls of primary photosynthetic filaments per leaf were counted on three
leaves at the node. On each of these three leaves, bladders were counted,
removed, and dried in an 808C oven to constant mass. Each leaf minus its
bladders was dried separately, as was the subtending internode. These parts
were weighed on a microbalance (Mettler Toledo AB54, Mettler Toledo Inc.,
Westerville, Ohio, USA). Percentage investment in bladders and support struc-
172
AMERICAN JOURNAL
tures was calculated from these data. Biomasses of leaves and bladders per
leaf for a whorl were averaged and multiplied by the number of leaves per
whorl to obtain biomass per leaf whorl. This was added to the dry mass of
the subtending internode to determine biomass per nodal unit, and relative
biomass allocations to bladders, leaves, and stem (internode) were calculated.
Growth-rate determinations—Growth rates were determined for free-floating plants; initial measurements indicated that plants in mats had much slower
growth rates than those reported here. I marked 12–15 plants at four permanent quadrats at each flume site. The free end of a string tied to a styrofoam
packing peanut was tied above the most recently matured node of an individual U. purpurea plant. This node was defined based on internode length, leaf
length, amount of leaf reflexion, and bladder expansion. The floating styrofoam peanut enabled me to relocate marked plants after several weeks of
growth in situ. Plants grew for 4–6 wk, and then the number of mature nodes
above the marked node was counted. New plants were marked and followed
for every sampling date. Growth was measured every 2–4 mo from October
1997 to November 1998.
Leaf development and bladder contents—Plants used for morphological
measurements from sites A, B, and C were analyzed for bladder contents.
Leaf whorls were numbered based on relative developmental stage. The most
recently matured leaf whorl, as determined by morphological criteria (see
description above), was designated as 0. Older leaf whorls were numbered
positively, and younger whorls numbered negatively from the 0 whorl. For
developmental and bladder content analysis, a single leaf was sampled from
each node beginning with node 1 and sampling toward the apex to the 23 to
24 whorl. Leaf length and length of the subtending internode were measured
for each whorl, while leaf number and the number of whorls of leaf subdivisions were counted. Bladder number was also counted for all but the youngest leaves. Twenty bladders from each leaf were examined under a Leitz
Dialux 20 compound light microscope. Bladders were analyzed by focusing
down through the bladder and recording presence or absence of allochthonous
material. The bladder walls are a single cell thick and relatively transparent,
as illustrated in Figs. 7 and 9, which were taken through bladder walls, and
as can be seen in Figs. 5, 6, and 8, which show dark contents through transparent bladder walls. Internal and external walls of bladders were identified
by the different types of glands present on the inside and outside of bladders
(Juniper, Robins, and Joel, 1989; Taylor, 1989). The types of multicellular
invertebrates present per bladder were quantified for samples from sites A
and C.
Statistical analyses—Data were checked for normality; nonnormal data
were transformed prior to analysis or were analyzed with nonparametric tests.
Site differences were analyzed using one-way ANOVA or Kruskal-Wallis tests
in the JMP version 3 for MacIntosh computers. Pairwise comparisons were
made using Student’s t test. Variation in bladder number among leaves, plants,
and sites was analyzed with nested ANOVA. Averages are reported 61 SD.
RESULTS
Morphology—Free-floating individuals of U. purpurea are
;0.5 m in total length in Shark River Slough. The plants produce whorls of leaves as they grow at the tip and rot at the
base (Fig. 1). Every second node can produce a branch bud
and an inflorescence bud. Inflorescences grow above the water
and produce one or two flowers (Fig. 1; for additional details
of development, see Rutishauser and Sattler, 1989). Internodes
of plants used for the bladder content study averaged 29 mm
in length and 1.1 mm in diameter (Table 1). A typical plant
in Shark River Slough tends to produce five leaves per whorl
(Table 1; 63% of nodes had five leaves, 34% had six leaves,
and 3% had four leaves). Each leaf produced whorls of four
or fewer photosynthetic filaments, which could themselves be
subdivided (Figs. 1B, 3). Most Shark River Slough plants had
five of these whorls per leaf (Table 1; 65% had five whorls,
OF
BOTANY
[Vol. 88
27% had four whorls, and 8% had six whorls). Individual leaf
length averaged 39 mm (Table 1). Bladders of U. purpurea
are borne at the tips of all orders of leaf filaments.
Site and seasonal variation in morphology—Variation in
bladder number/leaf between leaves at a single node was not
significant (P 5 0.3670) nor was variation between bladder
number/leaf at successive nodes (P 5 0.9514). Bladder number/leaf, however, did vary significantly among plants (P ,
0.0001).
Neither internode diameter nor leaf length varied significantly among sites, and plants had similar numbers of leaves
per nodal whorl and similar numbers of whorls of photosynthetic filaments per leaf (Table 1). Plants did differ significantly among sites with respect to internode length and the
number of bladders. In general, plants at sites A and B tended
to resemble each other and to differ from plants at site C.
Plants at site C produced more bladders per leaf and node than
plants at the other two sites (Table 1).
Plants at site A that were collected in the wet season and
the dry season were not significantly different in leaf length
(40 6 9 mm in October 1998, vs. 44 6 8 mm in March 1999)
but did differ significantly in bladder number (55 6 31 bladders per leaf in October 1999 vs. 87 6 34 bladders per leaf
in March 1999).
Investment in bladders—Individual leaves in a whorl produced an average of 74 bladders per leaf (Table 2). Given an
average of five leaves per whorl, each whorl was estimated to
have 370 bladders.
Utricularia purpurea plants invested an average of 26 6
10% of their biomass in bladders; investment varied from 9
to 45%. Most plants allocated the majority of their biomass to
the leaf filaments that support the bladders. The biomass of
these filaments, which are the major photosynthetic portion of
the plant, averaged 64% of the total biomass of a nodal unit
(bladders 1 supporting filaments 1 subtending internode), or
2.3 times the biomass of the bladders themselves (Table 2).
The internode supporting the whorl contributed 10% of the
total biomass of the leaf whorl.
Leaf and internode dry masses did not vary significantly
among sites, but bladder dry mass per leaf and individual bladder biomass did (Table 2). Plants at site C produced significantly more bladders per leaf than those at site B during the
wet season. The dry-season plants from site A were similar to
wet-season site C plants in their investment in bladders and
differed significantly from wet-season site B plants (Table 2).
Bladder number was negatively correlated with individual
bladder biomass, so that plants with more bladders tended to
have lighter bladders, although the correlation was not strong
(R2 5 0.09, P 5 0.0052, N 5 86).
Trapping efficiency—Bladders on very young leaves of U.
purpurea did not have allochthonous material inside them
(Fig. 2). These bladders, which were small, anthocyanic, and
had bladder doors that were not fully developed (Figs. 4, 5),
were on leaf whorls 2–4 plastochrons younger than the first
whorl of mature leaves. The youngest leaf with bladder contents was two plastochrons younger than the first mature leaves
(Fig. 2). From this leaf whorl the number of occupied bladders
increased exponentially with leaf age until bladders in the leaf
one plastochron older than the first mature leaf were almost
100% occupied (Figs. 2, 4–9). The amount of material in each
January 2001]
RICHARDS—U. PURPUREA
BLADDER FUNCTION
173
Figs. 3–9. Utricularia purpurea leaves and bladders. 3. Portion of a leaf photosynthetic filament. 4. Bladder from 22 leaf. 5. Bladder from 21 leaf. 6.
Bladder from 0 leaf. 7. Interior of bladder in Fig. 6; arrow points to a green alga. 8. Bladder from 11 leaf. 9. Interior of bladder in Fig. 8; arrows point to
living algae (desmids and diatoms). Bar 5 20 mm in Figs. 3, 4, 7, 9; bar 5 100 mm in Figs. 5, 6, 8. D 5 bladder door; H 5 hairs on bladder door; P 5
photosynthetic epidermis.
174
AMERICAN JOURNAL
OF
BOTANY
[Vol. 88
TABLE 1. Morphological measurements from the most recently matured nodes of U. purpurea. N 5 10 plants per site, with a single leaf from the
most recently matured node and the next older node measured on each plant.
Characteristic
Site A
Internode length (mm)
Internode diameter (mm)
Leaf no./node1
Leaf length (mm)
No. of filament whorls/leaf1
No. of bladders/leaf
Bladder no./leaf length
32.0 6 6.0
1.1 6 0.2
5 (4–6)
39.7 6 9.1
5 (4–6)
55 6 31a
1.3 6 0.5a
1
2
Site B
Site C
28.9 6 4.1
1.3 6 0.3
5 (4–6)
38.5 6 4.8
5 (4–6)
59 6 28ab
1.5 6 0.6ab
a
26.9 6 6.5
1.0 6 0.3
5 (5–6)
37.9 6 4.9
5 (4–5)
77 6 32b
1.9 6 0.7b
ab
b
Average
Significant?2
29.4 6 6.1
1.1 6 0.3
5 (4–6)
38.7 6 6.6
5 (4–6)
64 6 31
1.6 6 0.7
Y
N
N
Y
Y
Mode (Range).
For characteristics with significant differences (P # 0.05) among them, measurements followed by the same letter are not significantly different.
bladder also increased with increasing leaf age (Figs. 5, 6, 8).
Occupied bladders contained a mixture of living and detrital
material. All bladders with contents had some type of photosynthetic tenants, primarily blue–green algae, diatoms, green
algae, and/or photosynthetic protists (Figs. 7, 9). Many bladders also had living rotifers. A variety of larger organisms,
such as copepods, ostracods, and cladocerans, were found
more rarely. At sites A and C, 880 of 1400 bladders examined
from different-aged leaves had contents. Of these, 20.2% had
living rotifers, 5.1% had cladocerans, 1.3% had chironomids,
0.6% had ostracods, 0.5% had copepods, and 0.9% had other
multicellular invertebrates. The rotifers in these samples were
swimming around in the bladders, whereas the larger invertebrates were usually dead.
Growth rates—Utricularia purpurea expanded 1–2 nodes
per week (Fig. 10). Growth varied seasonally, being greater in
the summer wet season and slower in the winter dry season
(Fig. 10). Growth rates were greatest in May and June at the
beginning of the wet season and began to decline later in the
summer.
DISCUSSION
Are these results general?—The results presented here
question the importance of trapping small aquatic animals for
growth of Utricularia purpurea. The data were obtained in the
Everglades of south Florida, however, where invertebrate
standing stocks are remarkably low as compared to other wetlands (Turner et al., 1999). Are the trapping rates and trap
contents reported here unique to Utricularia species in the
Everglades? Whether the trapping rates and bladder communities found in the Everglades are representative of Utricularia
species in other environments or of terrestrial species of Utricularia needs further study. Previous researchers, however,
have reported the presence of living algae and ciliates in blad-
ders of Utricularia species (Botta, 1976; Mosto, 1979; Juniper,
Robins, and Joel, 1989).
Investment in bladders—Species of Utricularia invest a
large proportion of their biomass in bladders. While U. purpurea invested approximately one-quarter of its biomass in
bladders, U. macrorhiza growing in lakes in northern Wisconsin invested 38–48% (Knight and Frost, 1991), and U. vulgaris
growing in disused clay pits in Cambridgeshire, UK, invested
;50% biomass in bladders (Friday, 1992). I have also found
an average of 29% investment in bladders from plants collected from a pond on the Florida International University
campus (J. H. Richards, Florida International University, unpublished data). This percentage investment is similar to what
terrestrial plants in moderately low-nutrient environments invest in root biomass (Tilman, 1988)
The percentage of biomass invested in bladders is plastic
and can vary with environmental conditions. In all studies that
have considered the question, the largest variation in bladder
number is between sites. In U. purpurea, as in other species
of Utricularia (Knight and Frost, 1991; Friday, 1992), variation among leaves on a plant sampled at a single time is usually minor. Variation among plants at a single site is more
complex and probably represents the degree of habitat heterogeneity displayed at a particular site. Whereas Knight and
Frost (1991) found minor variation in bladder number among
plants in a lake, Friday (1992) found large variation among
plants at a single site. In the data reported here U. purpurea
bladder number varied among plants both within and between
sites.
Benefits from bladders—The conventional view of bladderwort carnivory is that these plants benefit from nutrients
derived by digestion of trapped microinvertebrates (Friday,
1989; Juniper, Robins, and Joel, 1989; Knight and Frost, 1991;
Ulanowicz, 1995). In south Florida bladders of U. purpurea
TABLE 2. Biomass allocation to bladders, filaments, and internode in the first mature nodal unit (5 leaf whorl 1 subtending internode) of U.
purpurea. N 5 8–12 plants per site, with 3 leaves from the most recently-matured node measured on each plant. Samples for site A were
collected in March 1999; samples for sites B and C were collected in July 1998.
Characteristic
Site A
Nodal unit dry mass (mg)
Bladder no.
Bladder dry mass (mg)
Bladder dry mass/nodal unit (mg)
Leaf filament dry mass/nodal unit (mg)
Internode dry mass/nodal unit (mg)
6
6
6
6
6
6
1
25
86
20
8
15
2.3
Site B
a
8
31a
6a
3a
6
1.1
17
52
12
3
14
2.0
6
6
6
6
6
6
Site C
b
7
34b
5b
1b
9
1.1
20
83
18
6
14
1.6
6
6
6
6
6
6
Average
b
3
34a
10a
3a
6
0.3
21
74
17
6
14
2.0
6
6
6
6
6
6
8
36
8
3
7
1
Significant?1
Y
Y
Y
Y
N
N
For characteristics with significant differences (P # 0.05) among them, measurements followed by the same letter are not significantly different.
January 2001]
RICHARDS—U. PURPUREA
BLADDER FUNCTION
175
stantly augmented by material from new trapping events or
whether bladder contents increase through reproduction inside
the bladders is unknown.
Although the importance of the Utricularia-periphyton interaction has been recognized (Friday, 1989; Ulanowicz,
1995), the possible importance of the direct interaction has
been overshadowed by concepts derived from models of carnivory. The interpretation of the benefit of carnivory presented
here provides an explanation for Knight and Frost’s (1989)
otherwise anomalous result that U. macrorhiza plants in more
nutrient-rich environments invested more in bladders. If the
benefit of the bladders depends on the productivity of the microcommunity and that productivity increases with increased
nutrients, plants would be predicted to produce more bladders
under increased nutrient levels.
Fig. 10. Average number of new leaf whorls produced by U. purpurea
per week measured at 1–4 mo intervals between June 1997 and November
1998. Data from sites A, B, and C combined. N $ 80 for each point. Error
bars 5 61 SD.
on plants in natural habitats trap very low percentages of these
microinvertebrates, except for rotifers, and the rotifers observed inside bladders were alive. Similarly low invertebrate
trapping rates were found for U. foliosa in the Everglades
(Bern, 1997; A. Bern, J. H. Richards, and B. Fry, Florida
International University, unpublished data). Calculations of
how much nitrogen and phosphorus plants could derive from
these trapping rates, if the only benefit is from digestion of
microinvertebrates, suggest that the return on a 25–50% investment in bladders is ,1% of the plant N and P (Bern, 1997;
A. Bern, J. H. Richards, and B. Fry, Florida International University, unpublished data).
Although plants had few bladders with dead microinvertebrates, almost 100% of mature U. purpurea bladders supported
living communities of microorganisms and associated detritus.
These communities were derived from the external environment, as very young bladders lacked them, and the number of
inhabited bladders, as well as inhabitant density and diversity,
increased over time. The ubiquitous presence of these communities supports the hypothesis that Utricularia plants derive
more benefit from by-products of this community than from
carnivory, i.e., that the important association in Utricularia
bladders is a mutualism rather than a predator–prey interaction. The bladders in Utricularia, therefore, may provide benefit through a detrital food web rather than a carnivorous interaction.
This view is a simplification of Ulanowicz’s (1995) positive
feedback model for Utricularia carnivory, in which the benefit
funneled to Utricularia from periphyton comes through ingestion of animal grazers. In the model this benefit allows Utricularia carnivory to be advantageous in oligotrophic environments (Ulanowicz, 1995). I suggest that the carnivory part of
the periphyton-Utricularia association may be incidental to the
direct interaction.
In the summer plants expand a leaf whorl about every 5 d,
and they begin trapping in 22 leaves, so they become 100%
occupied after 15 d. Free-floating plants generally have at least
three whorls of mature leaves before they become completely
covered with algae and appear to be senescent. Thus, after
bladders are fully occupied, they may support these communities for another 15 d. After a bladder has been tripped, it can
reset itself (Lloyd, 1933; Sydenham and Findlay, 1975; Juniper, Robins, and Joel, 1989). Whether communities are con-
Currency of the ‘‘carnivorous’’ interaction in aquatic
plants—Utricularia is the most diverse genus of carnivorous
plants, accounting for a little less than half of all carnivorous
species (Juniper, Robins, and Joel, 1989), but the genus is
unusual in having submerged aquatic species. The interaction
in an aquatic species, such as Utricularia purpurea, may differ
substantially from carnivory in terrestrial species of Utricularia or in species of other carnivorous plants. If the community interaction inside bladders provides the benefit for investment of biomass in bladders in U. purpurea, the currency
in the interaction may not be solely mineral nutrients. For example, carbon dioxide concentrations potentially limit photosynthesis in freshwater aquatic plants (Falkowski and Raven,
1997). Carbon dioxide derived from respiratory processes in
the bladders could be a significant benefit in the slow-moving
aquatic environments typical for some Utricularia species,
similar to the benefit obtained from CO2 uptake through the
roots in isoetids (Raven et al., 1988). This benefit could help
to explain Utricularia’s success in the aquatic environment.
The endangered Aldrovanda vesiculosa (Droseraceae), a miniature aquatic version of Dionea, is the only other free-floating
aquatic carnivorous species. A recent study found that the
most important condition for rapid growth in A. vesiculosa was
a high-CO2 concentration in the water (Adamec, 1997).
In addition to the possible benefit of CO2 supplements, bladders in aquatic species could also be advantageous because
they provide a contained microenvironment to enhance breakdown of organic matter, prevent diffusion of the breakdown
products, and provide Utricularia plants with a monopoly on
released nutrients.
Bladder mechanics and community dynamics—The absence of microinvertebrates from many bladders coupled with
the presence of some type of material in all of the bladders
means either that trapping is very inefficient or that bladders
do not require animals to trip them under natural conditions.
Alternative possibilities that need to be investigated are (1)
debris carried by currents or the current itself is sufficient to
cause opening or (2) endogenous factors, such as internal tension, can cause bladder opening. Lloyd (1933) reports that
movement of U. purpurea plants in water often releases the
traps.
LITERATURE CITED
ADAMEC, L. 1997. Photosynthetic characteristics of the aquatic carnivorous
plant Aldrovanda vesiculosa. Aquatic Botany 59: 297–306.
176
AMERICAN JOURNAL
ALBERT, V. A., S. E. WILLIAMS, AND M. W. CHASE. 1992. Carnivorous
plants: phylogeny and structural evolution. Science 257: 1491–1495.
BERN, A. L. 1997. Studies on nitrogen and phosphorus uptake by the carnivorous bladderwort, Utricularia foliosa, in south Florida wetlands.
Master’s thesis, Florida International University, Miami, Florida, USA.
BOTTA, S. M. 1976. Sobre las trampas y las victimas o presas de algunas
especies argentinas del genero Utricularia. Darwiniana 20: 127–154.
CHILDERS, D. L., R. D. JONES, J. TREXLER, C. BUZZELLI, S. DAILEY, A.
EDWARDS, E. GAISER, K. JAYACHANDARAN, A. KENNE, D. LEE, J.
MEEDER, M. K. NAIR, J. H. K. PECHMANN, A. RENSHAW, J. RICHARDS,
M. RUGGE, L. J. SCINTO, P. STERLING, AND W. VAN GELDER. In press.
Quantifying the effects of low-level phosphorus enrichment on unenriched Everglades wetlands with in situ flumes and phosphorus dosing.
In K. Porter and J. Porter [eds.], The Everglades hydroscape. St. Lucie
Press, Delray Beach, Florida, USA.
ERICKSON, R. O., AND F. J. MICHELINI. 1957. The plastochron index. American Journal of Botany 44: 297–305.
FALKOWSKI, P. G., AND J. A. RAVEN. 1997. Aquatic photosynthesis. Blackwell Science, Malden, Massachusetts, USA.
FRIDAY, L. E. 1989. Rapid turnover of traps in Utricularia vulgaris L. Oecologia 80: 272–277.
———. 1991. The size and shape of traps of Utricularia vulgaris L. Functional Ecology 5: 602–607.
———. 1992. Measuring investment in carnivory: seasonal and individual
variation in trap number and biomass in Utricularia vulgaris L. New
Phytologist 121: 439–445.
———, AND C. QUARMBY. 1994. Uptake and translocation of prey-derived
15
N and 32P in Utricularia vulgaris L. New Phytologist 126: 273–281.
GIVNISH, T. J. 1989. Ecology and evolution of carnivorous plants. In W. G.
Abrahamson [ed.], Plant-animal interactions, 243–290. McGraw–Hill,
New York, New York, USA.
JUNIPER, B. E., R. J. ROBINS, AND D. M. JOEL. 1989. The carnivorous plants.
Academic Press, London, UK.
KARLSSON, P. S., K. O. NORDELL, S. EIREFELT, AND A. SVENSSON. 1987.
Trapping efficiency of three carnivorous Pinguicula species. Oecologia
73: 518–521.
KNIGHT, S. E. 1992. Costs of carnivory in the common bladderwort, Utricularia macrorhiza. Oecologia 89: 348–355.
OF
BOTANY
[Vol. 88
———, AND T. M. FROST. 1991. Bladder control in Utricularia macrorhiza:
lake-specific variation in plant investment in carnivory. Ecology 72: 728–
734.
LARSON, P. R., AND J. G. ISEBRANDS. The plastochron index as applied to
developmental studies of cottonwood. Canadian Journal of Forest Research 1: 1–11.
LLOYD, F. E. 1933. The structure and behavior of Utricularia purpurea. Canadian Journal of Research 8: 234–252.
MOSTO, P. 1979. Algas en trampas de Utricularia oligosperma St. Hill. Boletin. Sociedad Argentina de Botanica 18: 89–100.
RAVEN, J. A., L. L. HANDLEY, J. J. MACFARLANE, S. MCINROY, L. MCKENZIE, J. H. RICHARDS, AND G. SAMUELSSON. 1988. The role of CO2
uptake by roots and CAM in acquisition of inorganic C by plants of the
isoetid lifeform: a review, with new data on Eriocaulon decangulare L.
New Phytologist 108: 125–148.
RUTISHAUSER, R., AND R. SATTLER. 1989. Complementarity and heuristic
value of contrasting models in structural botany. III. Case study on shootlike ‘‘leaves’’ and leaf-like ‘‘shoots’’ in Utricularia macrorhiza and U.
purpurea (Lentibulariaceae). Botanische Jahrbu¨cher Systematic fu¨r
Pflanzengeschichte und Pflanzengeographie 111: 121–137.
SYDENHAM, P. H., AND G. P. FINDLAY. 1975. Transport of solutes and water
by resetting bladders of Utricularia. Australian Journal of Plant Physiology 2: 225–351.
TAYLOR, P. 1989. The genus Utricularia. Kew Bulletin Additional Series
XIV. HMSO, London, UK.
THUM, M. 1988. The significance of carnivory for the fitness of Drosera in
its natural habitat. 2. The amount of captured prey and its effect on
Drosera intermedia and Drosera rotundifolia. Oecologia 81: 401–411.
TILMAN, D. 1988. Plant strategies and the dynamics and structure of plant
communities. Princeton University Press, Princeton, New Jersey, USA.
TURNER, A. M., J. C. TREXLER, C. F. JORDAN, S. J. SLACK, P. GEDDES, J. H.
CHICK, AND W. F. LOFTUS. 1999. Targeting ecosystem features for conservation: standing crops in the Florida Everglades. Conservation Biology 13: 898–911.
ULANOWICZ, R. E. 1995. Utricularia’s secret: the advantage of positive feedback in oligotrophic environments. Ecological Modelling 79: 49–57.
ZAMORA, R. 1995. The trapping success of a carnivorous plant, Pinguicula
vallisneriifolia: the cumulative effects of availability, attraction, retention
and robbery of prey. Oikos 73: 309–322.