Am. J. Bot. Cross-Journal Searching
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Alcalá, R. E.
Right arrow Articles by Domínguez, C. A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Alcalá, R. E.
Right arrow Articles by Domínguez, C. A.
Agricola
Right arrow Articles by Alcalá, R. E.
Right arrow Articles by Domínguez, C. A.
(American Journal of Botany. 2003;90:1341-1348.)
© 2003 Botanical Society of America, Inc.

Ecology

Patterns of prey capture and prey availability among populations of the carnivorous plant Pinguicula moranensis (Lentibulariaceae) along an environmental gradient1

Raúl E. Alcalá2 and César A. Domínguez2,3,4

2Departamento de Ecología Evolutiva, Instituto de Ecología, UNAM, A.P. 70-275, CP 04510, México, DF, Mexico; and 3Department of Ecology and Evolutionary Biology, University of California-Irvine, Irvine, California 92697-2525 USA

Received for publication January 21, 2003. Accepted for publication April 17, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
In this study we explored the effect of the physical environment and the availability of prey (biomass and taxonomic composition) on the patterns of prey capture and reproduction on five populations of Pinguicula moranensis (Lentibulariaceae) in areas ranging from pine–oak forests to desert scrublands. Environmental variation was summarized using principal factor analysis. Prey availability and prey capture increased toward the shadiest, most humid, and fertile population. The probability of reproduction and average bud production per population did not follow the same tendency because both fitness components peaked at the middle of the environmental gradient. These results suggest that the benefits derived from carnivory are maximized at sites fulfilling a trade-off between light, moisture, and prey availability. We also found that the taxonomic composition of both the available prey and that of the prey captured by plants varied among populations. The results also indicated that the prey captured by plants are not a random sample of prey available within populations. Overall, the results from this study revealed a marked amount of heterogeneity in the physical and biotic environment among the populations of P. moranensis, which has the potential to affect the outcome of the interaction between this carnivorous species and its prey.

Key Words: carnivorous habit • environmental heterogeneity • insectivorous plants • Lentibulariaceae • Pinguicula


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Carnivorous plants, a conspicuous group of species characterized by their ability to obtain nutrients through the capture of animal prey (Lloyd, 1942 ; Thompson, 1981 ; Benzing, 1987 ), have attracted the attention of biologists interested in the study of plant–animal interactions ever since Darwin (1875) . The evolutionary changes associated with the origin of carnivory involve modifications in the morphology, anatomy, and physiology of leaves that allow them to digest animal tissues. These changes generally have been accompanied by a reduction in the ability of roots to take nutrients from the soil (Lüttge, 1983 ; Givnish et al., 1984 ; Juniper et al., 1989 ; Adamec, 1997 ). Several studies have shown that carnivory may provide plants with a substantial amount of the mineral nutrients they require (Dixon et al., 1980 ; Schulze et al., 1991 ; Karlsson and Pate, 1992 ; Karlsson et al., 1994 ; Schulze et al., 1997 ; Ellison and Gotelli, 2001 ). Acquisition of nitrogen derived from prey capture also translates into higher survival, growth, and reproduction of carnivorous plants (Thum, 1988 ; Gibson, 1991 ; Zamora et al., 1997 , 1998 ; Thoren and Karlsson, 1998 ).

Because carnivorous plants usually occur in sites with nutrient scarcity, carnivory has been considered an adaptation to colonize nutrient-poor environments (Lüttge, 1983 ; Juniper et al., 1989 ; Ellison and Gotelli, 2001 ). The nitrogen obtained through prey capture must compensate not only for nutrient scarcity, but pay for the maintenance of traps and the costs associated with the digestion of animal tissues (Givnish et al., 1984 ; Benzing, 1987 ). Givnish et al. (1984) proposed that the maximum benefit of carnivory could only be achieved in wet, sunny habitats, where nutrient availability is the only factor limiting photosynthesis. Such a hypothesis assumes that different populations of carnivorous plants may be exposed to contrasting environments and make no predictions about the variation in prey availability. It is well known, however, that local insect abundance (the major prey of carnivorous plants) may vary strikingly from place to place and from one season to another (Karlsson et al., 1994 ; Zamora, 1995 ; Zamora et al., 1998 ). Thus, in addition to the limitations of the physical environment, populations of carnivorous plants may be exposed to contrasting availabilities of prey and to different prey communities. The interaction between these two sources of environmental variation may produce a complex ecological scenario that has the potential to affect the selective pressures operating on the evolution of carnivory.

In accordance, in this study we characterized the physical (i.e., light, water, nitrogen availability) and the biotic (prey availability) environmental variation in five populations of the carnivorous plant Pinguicula moranensis var. neovolcanica (Lentibulariaceae). We then determined the relationship between environmental heterogeneity and the patterns of prey capture among populations and evaluated the effect of these variables on two fitness components: the probability of reproduction and bud production.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 LITERATURE CITED
 
Plant species
Pinguicula moranensis var. neovolcanica (Zamudio) (P. moranensis, hereafter) is a prostrate rosette-forming perennial herb that captures and consumes small animals as a source of mineral nutrients. During the dry season (October–April), plants of this species aestivate by producing a resistance rosette composed of minute glandless leaves. Once the first rains begin (middle May), P. moranensis produces a capture rosette composed of 6–8 sticky leaves 35–95 mm in length (Zamudio, 1999 ). Like other species in this genus, the leaves of the capture rosette possess capture (stalked) and digestive (sessile) glands on their surface (Heslop-Harrison and Knox, 1971 ; Heslop-Harrison and Heslop-Harrison, 1980 ). The sticky mucilage secreted by capture glands retains prey that land on the surface of the leaf, and once the animals are trapped, the sessile glands digest them. The flowering season of this species occurs from June to August with a flowering peak in July. The hermaphroditic, purple flowers last for 8–10 d and are pollinated by small butterflies. Fertilized fruits ripen from July to September, producing hundreds of small wind-dispersed seeds. An embryological study of P. moranensis showed that the differentiation of floral buds and anthesis occur in the same reproductive season (S. Espinosa, Faculty of Sciences, UNAM, personal communication), suggesting that this species behaves as an income breeder in the sense of Thorén and Karlsson (1998) .

Study sites
Five populations of P. moranensis from central Mexico were chosen for this study, two from the state of Morelos and three from the state of Puebla (Fig. 1). The climate, substrate, and type of vegetation differ among the selected populations and include all the range of habitats and environmental conditions this species is exposed to in its area of distribution (Zamudio, 1999 ). Populations from Morelos were located within a pine–oak forest in the mountain range of Sierra de Tepoztlán (hereafter T1 and T2). The distance separating T1 and T2, which are both established on vertical sandstone cliffs, is about 100 m in a straight line. Populations from Puebla are established in a variety of environmental settings. The Lázaro Cárdenas population (LC) is located within an oak forest preserve near the city of Puebla. Plants of this population grow on the ground of an almost bare clay substrate. The fourth population is established in basaltic walls within a disturbed tropical dry forest, which is located around the Valsequillo Dam (VAL). Finally, plants from the Molcajac population (MOL) grow on small walls of limestone produced by water erosion. The area is a tropical arid scrub dominated by xerophytic vegetation. A detailed description of the climate and vegetation types can be found in Rzedowski (1981) .



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 1. Locations of the five studied populations of Pinguicula moranensis from central Mexico. T1 (Tepoztlán 1) and T2 (Tepoztlán 2) are located in the Sierra de Tepoztlán, in Morelos State, while LC (Lázaro Cárdenas), VAL (Valsequillo), and MOL (Molcajac) are in the state of Puebla

 
Data collection
Environmental variation
To characterize quantitatively the physical variation among populations, we measured light availability, air temperature, soil water content, and soil fertility for 1 wk in the middle of the growing season (July). Photosynthetic active radiation (PAR) and air temperature were recorded using quantum (LI-190SA) and temperature (LI-COR 1400-101) sensors connected to a data logger (Licor L-1000, Li-Cor, Lincoln, Nebraska, USA). Measurements were taken by sensors placed above the rosette of each individual plant between 1000 and 1400 on clear sunny days. The sampling order for each plant in a given population was randomly assigned. Once all the plants within a population were measured, we repeated the entire procedure until we completed five measurements for each plant. The five measurements of each variable were averaged for each individual and used in the analyses. For each population, we collected between 10 and 30 25-g samples of the substrate on which P. moranensis was growing. Samples were maintained in sealed plastic bags and weighed to the nearest 0.01 g to obtain fresh mass. Dry mass was then obtained by oven-drying the samples. Soil humidity was estimated for each sample by comparing of wet and dry soil mass. Because fertility analyses require a minimum of 100 g of soil, we pooled 4–5 samples thus rendering four replicates per population. Soil fertility was estimated by quantifying total nitrogen, phosphorus, and potassium (as parts per milliion) using Kjeldahl's method.

Patterns of prey capture and prey availability
To characterize taxonomically the composition of the diet of P. moranensis, as well as the among-population differences in the average amount of prey captured, we made monthly censuses of the animals captured by almost all the plants from each population (N = 21 in T1, N = 60 in T2, N = 80 in LC, N = 80 in VAL, and N = 70 in MOL). Although monthly censuses may underestimate the amount of prey captured by plants if some prey are completely digested before the next sampling, this method allows us to compare the patterns of prey capture among populations. Censuses started at the beginning of the growing season (May) of 1994 and finished in August (1994) when most plants were producing the resistance rosette. In each census, the number and taxonomic identity (order or higher) of the prey captured by each plant were recorded. All trapped prey were left on the plants. Because prey of P. moranensis belong to many taxa of different sizes, the number of trapped individuals is not a good descriptor of the amount of resources gained by each plant. Thus, prey capture for each plant was expressed in terms of dry mass. The average mass of the different animals this species consumes was estimated by collecting between 20 and 50 individuals of each taxon from unmarked plants. In all cases, we took care when collecting recently trapped animals to avoid mistakes in estimates of mass from partially digested prey. All samples were oven dried for 7 d at 40°C and weighed individually to the nearest 0.005 mg (Cahn model 4700 Electrobalance, Cahn Instruments, Cerritos, California, USA). These values were then used to calculate the average mass of each taxon (k) individual (AMk). The total amount of dry mass captured by each plant during the growing season was estimated as {Sigma}nk=1Nk x AMk, where Nk is the total number of prey of the k taxon captured by a given individual, and n is the number of taxa captured for all plants. Given that plant size (rosette area) is highly variable among the different individuals and that this attribute is probably associated with capture success, dry mass was standardized by the area of the rosette for each individual (in milligrams per square centimeter). In the middle of the growing season, all plants were filmed with a portable camera (Canon LP1, Canon, Tokyo, Japan), and rosette areas were estimated from the images using the Morphosys program (Meacham and Duncan, 1989 ) following the procedures described in Domínguez et al. (1998) .

The availability of prey in each population was estimated by means of artificial traps. Traps consisted of cardboard rectangles (8 x 10 cm) coated with an odorless glue (Tanglefoot, The Tanglefoot Company, Grand Rapids, Michigan, USA). Although Tanglefoot has a stronger adhesive power than plants, our method allowed us to estimate the availability of potential prey. Twenty traps were randomly placed in each population in the middle of the growing season (late July to early August) of 1994, when prey capture was previously observed to peak. All traps were placed directly on the substrate in a horizontal position and remained in the field for 1 wk, after which they were collected and the number and taxonomic identity of trapped animals were recorded. Data on the amount of biomass caught by artificial traps were expressed as milligrams per square centimeter using the same procedure for calculating the total biomass capture by plants (described earlier).

Reproductive responses
The reproductive responses of P. moranensis to its different environments were evaluated by measuring the probability of reproduction (proportion of reproductive plants per population) and the number of floral buds produced by each plant in every population. We used flower buds as a measure of the reproductive response of plants to environmental heterogeneity, because in doing so we avoided the variability brought about by possible differences in pollinator activity and composition among populations. Given that P. moranensis produces only one flower per stalk, bud production on every plant could be tallied in detail. Because we made monthly censuses of all plants in every population (see earlier), the presence of developing floral buds was annotated, and stalks were tagged with small plastic bands.

Data analyses
Environmental variation
Because environmental variables may be correlated, environmental variation among populations was analyzed by means of multivariate methods. Overall differences among populations were evaluated by means of a multivariate analysis of variance (PROC GLM, SAS, 1989 ). To summarize environmental differences among populations, a principal factor analysis was performed (PROC FACTOR, SAS, 1989 ). Principal factors were extracted through the maximum-likelihood method, and the number of factors to be retained was determined with a chi-square test. To maximize the amount of variance explained by the retained factors, the factors were rotated following the varimax method (Reyment and Jörekskog, 1993 ). Finally, once principal factors were rotated, factor scores were calculated using the PROC SCORE in SAS (1989) , and these scores were used as comprehensive descriptors of environmental variation among populations.

Patterns of prey capture and prey availability
The among-populations differences in the relative abundance of prey from different taxa were compared separately for plants and traps using chi-square tests. For these analyses, taxonomic groups of prey with low relative abundance were pooled into one category (Sokal and Rohlf, 1995 ). To compare the composition of the prey captured with the availability of prey from different taxa in each population, we performed a series of goodness-of-fit analyses (log-likelihood G test; Sokal and Rohlf, 1995 ). The expected number of prey from taxon i captured in population j was estimated as the product Neij = {phi}ij x NPj, where {phi}ij is the probability of capturing an individual of taxon i in population j ({phi}ij = NCij / NTj, where NCij = number of individuals from taxon i caught by the adhesive traps in population j, and NTj = total number of captures by adhesive traps in population j), and NPj = total number of prey captured by plants in population j. Each population was independently analyzed.

Environmental variation, prey availability, prey capture, and reproductive responses
We explored whether prey availability, prey capture, and bud production differ among populations and if they were related to environmental variation. Because each population contains several nonindependent observations (plants within populations), a series of regression models with replicates were performed (regression with more than one value of y for each x, Sokal and Rohlf, 1995 ). This method avoids the overestimation of the degrees of freedom brought about by the several measures of individual plants in each population and allows for the estimation of both the categorical (populations) and the continuous (the environmental gradient) variables. Environmental variation was characterized using the mean value of the scores derived from the first and second principal factors for each population. These values were then used as independent variables in the regression analyses.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 LITERATURE CITED
 
Physical environment
A multivariate analysis of variance revealed significant differences among populations in the overall physical environment (Wilks' lambda = 0.0003, F24,29 = 11.53, P = 0.0001). As shown by principal factor analysis, all physical variables were highly correlated. Two principal factors accounted for almost 70% of the variation in the physical environment (51 and 18%, factor 1 and 2, respectively). Soil humidity and total nitrogen had high positive loadings on factor 1, while the loading for air temperature and PAR was also high but negative. High positive scores of factor 1 are associated with sites with low temperatures and PAR, and higher levels of total nitrogen and soil humidity (T1 and T2, see Fig. 2A). In contrast, sites such as VAL and MOL had negative scores that described well-lit, warm, dry sites with nitrogen-poor soils (Fig. 2A). An analysis of variance performed on the scores derived from principal factor 1 revealed significant differences among sites (F4,13 = 46.69, P = 0.0001, R2 = 0.93). Accordingly, the environmental variation accounted for by factor 1 can be described as a gradient of temperature, soil fertility (N), humidity, and PAR (TNHL gradient hereafter, Fig. 2A). The higher factor loading on factor 2 was that of K, indicating that positive scores are associated with relatively high levels of this nutrient (Fig. 2B). There were significant differences among sites in the average score from factor two (F4,13 = 5.16, P = 0.01, R2; = 0.61). Significance, however, was only due to differences between LC and VAL (Fig. 2B). All the other comparisons between pairs of populations were not significant.



View larger version (11K):
[in this window]
[in a new window]
  		Fig. 2. Average values (±1 SE) for the scores derived from principal factor analysis describing the among-population variation in the physical environment of five populations of Pinguicula moranensis in central Mexico. Results for the two principal factors are presented. Significant differences among populations are indicated by different letters (Tukey's honestly significant difference for unequal sample sizes). See Fig. 1 for explanation of population abbreviations

 
Patterns of prey capture and prey availability
Pinguicula moranensis consumes a diverse array of animal species including insects, spiders, and gastropods. Diptera accounted for 80% of all captures, and its importance varied from 55 to 99% among the populations (Fig. 3). The second group in importance was Coleoptera, which accounted for 5.6% of the total prey captured, followed by Lepidoptera and Acarina. All the other groups contributed less than 2% of the total captures (Fig. 3A). Although Diptera was the most important source of prey in all populations, populations differed in the relative frequency of prey captured from each taxa ({chi}2(28) = 1258.01, P < 0.0001). Such differences are mainly due to the contrasting contributions of Lepidoptera, Acarina, and Coleoptera among the populations (see Fig. 3). Once the number of captured prey was corrected by the average amount of biomass contributed by an individual of a given taxon, Diptera continued to be the most important prey of P. moranensis (80.5%), Lepidoptera almost doubled its contribution (9.9%), and despite the relatively high number of Acarina individuals caught by plants, their contribution fell from 4.7 to 0.45%.



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 3. Percentage of prey from different taxa captured by plants and sticky traps in five populations of Pinguicula moranensis from central Mexico. Pooled data from all five populations (A) and separated tallies for each population are shown (B, T1; C, T2; D, LC; E, VAL; F, MOL see Fig. 1 for explanation of population abbreviations). Dp, Diptera; Lp, Lepidoptera; Ac, Acarina; Co, Collembola; Cl, Coleoptera; Hy, Hymenoptera; Ar, Araneae; Ot, Others

 
The taxonomic composition of the animals caught by the adhesive traps was similar to that of those captured by plants. Diptera was the most frequent group trapped in the adhesive traps, but their relative contribution (43.6) was only half of that observed in plants (Fig. 3A). In contrast, the relative abundance of other groups such as Acarina, Collembola, and Hymenoptera (and others) was higher. Again, populations differed significantly in the relative frequency of prey caught in adhesive traps from each taxa ({chi}2(28) = 233.24; P = 0.00001; see Fig. 3B–F).

The observed and expected number of prey from each taxon also differed markedly within each population (G test, P < 0.00001 for all populations). In all cases, plants trapped more Diptera than expected, while all the other groups of prey were underrepresented in most populations. Plants from the LC and VAL populations also trapped more Acarina and Coleoptera (LC) and Acarina and Lepidoptera (VAL) than expected by their relative abundances. These results indicate that the taxonomic composition of prey found in P. moranensis is not a random sample of the prey available within populations.

Prey capture was restricted from early June to mid-September (Fig. 4). For most populations, the median number of captures per plant increased until it reached a peak between August and September, followed by an abrupt decrease as capture leaves senesced.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 4. Temporal variation in median number of prey captured per plant in five populations of Pinguicula moranensis from central Mexico. Populations significantly differed in the median number of prey captured per plant in the entire season (Kruskal-Wallis test, {chi}24 = 51.14, P < 0.00001). See Fig. 1 for explanation of population abbreviations

 
Environmental variation, prey availability, prey capture, and reproductive responses
Populations showed marked differences in the probability of reproduction. The highest probability of reproduction was at LC (95% of the individuals reproduced), a population located in the middle of the TNHL gradient and with intermediate values of prey capture. Despite the fact that T2 was the population with the highest average value of prey capture (Fig. 5A), plants in this population failed to reproduce, suggesting that prey capture is not the main factor limiting reproduction in this species. In VAL, MOL, and T1, the percentage of reproducing individuals ranged from 40 to 60% (40, 42, and 62% for MOL, T1, and VAL, respectively).



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 5. Results from regression analyses with replicates on the effect of populations and environmental variation (scores derived from the two principal factors) on prey capture, prey availability, and bud production in five populations of Pinguicula moranensis from central Mexico. Scores from factor 1 describe a gradient of temperature, soil fertility (N), humidity, and photosynthetically active radiation (PAR) (TNHL gradient), while scores from factor 2 are associated with the variation in K. See Fig. 1 for explanation of population abbreviations

 
Regression analyses with replicates (see Materials and Methods) showed that prey capture, prey availability, and the number of buds produced per plant significantly differed among populations (Table 1; Fig. 5). Prey capture and prey availability were also linearly related to the TNHL gradient (Table 1, Fig. 5A and B), indicating that these two variables increased with soil humidity and total nitrogen and decreased in warmer, well-lit sites. These analyses also showed that the amount of biomass captured by P. moranensis increased from the prey-poor toward the prey-rich sites. Bud production was also related to the TNHL gradient, but in contrast to prey capture and prey availability, this relationship was quadratic (Fig. 5C). This result indicates that the reproductive response of P. moranensis is maximized in the mid-range of the environmental expanse this species inhabits and in intermediate values of prey availability and prey capture.


View this table:
[in this window]
[in a new window]
  		Table 1. Regression analyses with replicates (ANOVAs) of prey capture, prey availability, and bud production on environmental variation (scores from principal factor analysis) associated with Pinguicula moranensis populations. Separate analyses for the scores derived from the first and second principal factors are presented

 
There was a significant quadratic relationship between the scores from factor 2 and prey availability, an unexpected result, indicating that prey abundance peaked at populations with intermediate values of K (Fig. 5E). Neither prey capture nor bud production were related to the scores from factor 2 (Table 1, Fig. 5D and F).

In almost all cases, the effect of the deviations from the regression line was significant (Table 1), suggesting the existence of a large amount of random variation around the regression lines (see Sokal and Rohlf, 1995 ).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
LITERATURE CITED
 
Our results showed the presence of marked differences in the physical environment occupied by populations of P. moranensis. Most of this variation (51%) can be described in terms of one environmental gradient related to temperature, total nitrogen, soil humidity, and PAR (TNHL, scores from factor one). An additional 18% of the environmental variance was associated with variation in K (scores from factor two). The availability of prey also had striking differences among populations of P. moranensis, increasing toward the most shaded and humid populations. There was a significant linear relationship between prey availability and the TNHL gradient and a quadratic one with the scores from factor 2, indicating that prey abundance is strongly influenced by environmental heterogeneity (Karlsson et al., 1994 ; Zamora et al., 1998 ).

Differences in prey availability among populations were not restricted to the amount of prey (number of prey or dry mass), but also included the relative abundance of prey from different taxa. Although Diptera was the most common group caught in adhesive traps in all populations, there were marked differences in the relative abundance of almost all groups. For example, coleopterans were more abundant at the driest and sunny populations (VAL and MOL), whereas Diptera and Acarina had the opposite pattern. Thus, besides the physical heterogeneity associated with the different populations of P. moranensis, quantitative (prey availability) and qualitative (the relative abundance of prey from different taxa) differences in the biotic environment increased the amount of variation to which this species is exposed. Overall, our results revealed a marked amount of physical and biotic environmental heterogeneity among the populations of P. moranensis.

Given this heterogeneous scenario, it was not surprising that all our measurements of prey capture (the relative contribution of each group of prey, the median number of prey captured per plant, and the biomass of prey captured per individual) significantly differed among populations. In all cases, there was a marked difference between the taxonomic composition of the prey available within a population and that of prey actually caught by plants. Three other studies using different species of Pinguicula found evidence of an over-representation of certain taxa trapped by plants in comparison to artificial traps (Karlsson et al., 1987 ; Antor and García, 1994 ; Zamora, 1995 ). Although the over-representation of some groups of prey has been interpreted as a consequence of the color and smell of leaves, no mechanism of prey attraction has definitively been established for Pinguicula. According to Zamora (1990 , 1995 ), the taxonomic composition of the diet of Pinguicula nevadensis and P. vallisneriifolia was determined by the interaction between prey size and the retention capacity of the mucilage produced by plants. Because the retention capacity of mucilage depends on environmental conditions (Zamora, 1995 ) and because composition of prey caught among populations differed markedly, our results are better explained by the association between environmental variation and mucilage retention.

Environmental heterogeneity, expressed as the TNHL gradient, also affected the amount of biomass of prey captured per individual. Prey capture (dry mass) increased from the sunniest and relatively dry habitats (MOL and VAL) to the most humid, shady, and N- and prey-rich populations. Although prey capture had a positive relationship with the TNHL gradient, both the probability of reproduction and bud production were maximized at LC, a population characterized by intermediate values of PAR, prey abundance, and water availability. This result indicates that plants in prey-rich habitats (T1 and T2) failed to take advantage of the relatively high amount of prey they captured and suggests that carnivory cannot compensate for the diminished photosynthesis associated with the lower availability of PAR in the shadiest populations. These results agree with the experimental evidence provided by Zamora et al. (1998) that showed that plants of P. vallisneriifolia growing in deep shade were unable to benefit from experimental prey addition. Furthermore, although both prey availability and prey capture (in milligrams per square centimeter) peaked at T2, no sexual reproduction was observed in this population, suggesting that light, more than prey or water availability, is one of the main factors determining successful reproduction in P. moranensis.

Givnish et al. (1984) proposed that carnivory is adaptive only in nutrient-poor environments that are well lit and moist. It is assumed that the photosynthetic costs of carnivory exceed the benefits in either shady or dry habitats. Our results showed that carnivorous plants might colonize a broad range of environmental conditions ranging from pine–oak forests to tropical desert scrublands. Nonetheless, none of these habitats seems to attain the optimum combination of resources (well-lit, moist, high abundance of prey). The site with the poorest soil and relatively high levels of PAR was associated with the lowest availability of prey, high temperatures, and the driest soil. On the other extreme, populations with relatively nitrogen-rich and humid soils also had the highest availability of prey, low temperatures, and low levels of PAR. Thus, although water and prey availability were relatively high at the shade extreme of the environmental gradient, plants inhabiting these populations were probably limited by irradiance. The opposite extreme of the gradient had plenty of light, but both prey and water were scarce. In fact, the maximum reproductive output was attained in LC, a population with intermediate values of the relevant variables. In this respect, our results are more in accordance with the model proposed by Benzing (2000) that allows for explicit trade-offs between light, moisture, and nutrient (prey) availability. His model predicts that carnivory may occur in a variety of light regimes because of trade-offs associated with these two other niche axes.

We have shown in this study that different populations of P. moranensis face contrasting ecological scenarios, which may influence the interaction between carnivorous plants and their prey. Provided that such differences have fitness consequences, they could establish an ecological scenario favoring adaptive differentiation among populations of this carnivorous species (Bradshaw, 1972 ; Sork et al., 1993 ; Linhart and Grant, 1996 ).


   FOOTNOTES
 
1 The authors thank Arturo Galicia, Raúl Iván Martínez, Rubén Pérez Ishiwara for field assistance and logistic support. Aaron M. Ellison, Juan Fornoni, Rodolfo Dirzo, Juan Núñez-Farfán, Regino Zamora, and an anonymous reviewer made valuable comments on early drafts of the manuscript. This paper was written in partial fulfillment of the requirements for the PhD degree of R.E.A. at the Universidad Nacional Autónoma de México (UNAM). This study was supported by a PhD grant from CONACYT to R.E.A. and by a sabbatical fellowship to C.A.D. by CONACYT and UNAM. Back

4 Author for reprint requests (tejada{at}servidor.unam.mx ) Back


   LITERATURE CITED
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Adamec L. 1997 Mineral nutrition of carnivorous plants: a review. Botanical Review 63: 273-299

Antor R. J. M. B. García 1994 Prey capture by a carnivorous plant with hanging adhesive traps: Pinguicula longifolia. American Midland Naturalist 131: 128-135[CrossRef][ISI]

Benzing D. H. 1987 The origin and rarity of botanical carnivory. Trends in Ecology and Evolution 2: 364-369

Benzing D. H. 2000 Bromeliaceae: profile of and adaptive radiation. Cambridge University Press, New York, New York, USA

Bradshaw A. D. 1972 Some evolutionary consequences of being a plant. Evolutionary Biology 5: 25-47

Darwin C. 1875 Insectivorous plants. John Murray, London, UK

Dixon K. W. J. S. Pate W. J. Bailey 1980 Nitrogen nutrition of the tuberous sundew Drosera erythrorhiza Lindl. with special reference to catch of arthropoda fauna by its glandular leaves. Australian Journal of Botany 28: 283-297[CrossRef]

Domínguez C. A. L. E. Eguiarte J. Núñez-Farfán R. Dirzo 1998 Flower morphometry of Rhizophora mangle (Rhizophoraceae): geographical variation in Mexican populations. American Journal of Botany 85: 637-643[Abstract]

Ellison A. M. N. J. Gotelli 2001 Evolutionary ecology of carnivorous plants. Trends in Ecology and Evolution 11: 623-629

Gibson T. C. 1991 Competition among threadleaf sundews for limited insect resources. American Naturalist 138: 785-789[CrossRef][ISI]

Givnish T. J. E. L. Burkhardt R. E. Happel J. D. Weintraub 1984 Carnivory in the bromeliad Brocchinia reducta, with a cost/benefit model for the general restriction of carnivorous plants to sunny, moist, nutrient-poor habitats. American Naturalist 124: 479-497[CrossRef][ISI]

Heslop-Harrison Y. J. Heslop-Harrison 1980 Chloride ion movement and enzyme secretion from the digestive glands of Pinguicula. Annals of Botany 45: 729-731[Free Full Text]

Heslop-Harrison Y. R. B. Knox 1971 A cytochemical study of the leaf gland enzymes of insectivorous plants of the genus Pinguicula. Planta 96: 183-211[CrossRef][ISI]

Juniper B. E. R. J. Robins D. M. Joel 1989 The carnivorous plants. Academic Press, London, UK

Karlsson P. S. K. O. Nordell S. Eirefelt A. Svensson 1987 Trapping efficiency of three carnivorous Pinguicula species. Oecologia 73: 518-521[CrossRef][ISI]

Karlsson P. S. J. S. Pate 1992 Contrasting effects of supplementary feeding of insects or mineral nutrients on the growth and nitrogen and phosphorous economy of pygmy species of Drosera. Oecologia 92: 8-13[CrossRef][ISI]

Karlsson P. S. L. M. Thorén H. M. Hanslin 1994 Prey capture by three Pinguicula species in a subartic environment. Oecologia 99: 188-193[CrossRef][ISI]

Linhart Y. B. M. C. Grant 1996 Evolutionary significance of local genetic differentiation in plants. Annual Review of Ecology and Systematics 27: 237-277[CrossRef][ISI]

Lloyd F. E. 1942 The carnivorous plants. Chronica Botanica Ronald Press, New York, New York, USA

Lüttge U. 1983 Ecophysiology of carnivorous plants. In O. L. Lange, P. S. Nobel, C. B. Osmond, and H. Zeigler [eds.], Encyclopaedia of plant physiology, vol. 12C. Physiological Plant Ecology III, 489–517. Springer-Verlag, Berlin, Germany

Meacham C. A. T. D. Duncan 1989 MorphoSys ver. 1.26. University Herbarium, University of California, Berkeley, California, USA

Reyment R. A. K. G. Jöreskorg 1993 Applied factor analysis in the natural sciences. Cambridge University Press, Cambridge, New York

Rzedowski J. 1981 Vegetación de México. Limusa, México D.F., México

SAS. 1989 SAS/STAT user's guide, release 6.08. SAS Institute, Cary, North Carolina, USA

Schulze E. D. G. Gerbauer W. Schulze J. S. Pate 1991 The utilization of nitrogen from insect capture by different growth forms of Drosera from southwest Australia. Oecologia 87: 240-246[CrossRef][ISI]

Schulze W. E. D. Schulze J. S. Pate A. N. Gillison 1997 The nitrogen supply from soils and insects during growth of the pitcher plants Nepenthes mirabilis, Cephalotus follicularis and Darlingtonia californica. Oecologia 112: 464-471

Sokal R. R. F. J. Rohlf 1995 Biometry, 3rd ed. W. H. Freeman, New York, New York, USA

Sork V. L. K. A. Stowe C. Hochwender 1993 Evidence from local adaptation in closely adjacent subpopulations of northern red oak (Quercus rubra L.) expressed as resistance to leaf herbivores. American Naturalist 142: 929-936

Thompson J. N. 1981 Reversed animal–plant interactions: the evolution of insectivorous and ant-fed plants. Biological Journal of the Linnean Society 16: 147-155[CrossRef]

Thorén L. M. P. S. Karlsson 1998 Effects of supplementary feeding on growth and reproduction of three carnivorous plant species in a subartic environment. Journal of Ecology 86: 501-510[CrossRef]

Thum M. 1988 The significance of carnivory for the fitness of Drosera in its natural habita. 1. The reactions of Drosera intermedia and D. rotundifolia to supplementary feeding. Oecologia 75: 472-480[CrossRef][ISI]

Zamora R. 1990 The feeding ecology of a carnivorous plant (Pinguicula nevadense): prey analysis and capture constraint. Oecologia 84: 376-379[ISI]

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[CrossRef][ISI]

Zamora R. J. M. Gómez J. A. Hódar 1997 Responses of a carnivorous plant to prey and inorganic nutrients in a Mediterranean environment. Oecologia 111: 443-451[CrossRef][ISI]

Zamora R. J. M. Gómez J. A. Hódar 1998 Fitness responses of a carnivorous plant in contrasting ecological scenarios. Ecology 79: 1630-1644[ISI]

Zamudio S. 1999 Notas sobre la identidad de Pinguicula moranensis H. B. K., con la descripción de una variedad nueva. Acta Botanica Mexicana 49: 23-34





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Alcalá, R. E.
Right arrow Articles by Domínguez, C. A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Alcalá, R. E.
Right arrow Articles by Domínguez, C. A.
Agricola
Right arrow Articles by Alcalá, R. E.
Right arrow Articles by Domínguez, C. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS