USF Libraries
USF Digital Collections

Bottom-up and top-down effects on insects herbivores along a natural salinity gradient in a florida salt marsh

MISSING IMAGE

Material Information

Title:
Bottom-up and top-down effects on insects herbivores along a natural salinity gradient in a florida salt marsh
Physical Description:
Book
Language:
English
Creator:
Albarracin, Maria Teresa
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Gall makers
Sap suckers
Borrichia frutescens
Fertilizer
Sticky traps
Parasitoids
Herbivory
Dissertations, Academic -- Biology -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: I compared the strength of bottom- up and top-down effects on insect herbivores along a natural salinity gradient in salt marsh communities in West - Central, Florida. I used a 2x2 factorial design with plots divided into four different treatments: 1) fertilizer applied to increase plant quality 2) sticky traps added to remove natural enemies (parasitoids) 3) fertilizer applied and sticky traps added and 4) control plots. These plots were placed on 7 different sites containing the salt marsh plant Borrichia frutescens along a natural stress salinity gradient. In each plot I determined the abundance of the sap sucker Pissonotus quadripustulatus, the gall maker Asphondylia borrichiae, spiders and the number of chewed leaves and bored stems. I also recorded leaf area, plant density, plant height and foliar nitrogen. Plants in fertilized plots exhibited increased height, density and leaf area.
Thesis:
Thesis (M.S.)--University of South Florida, 2005.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Maria Teresa Albarracin.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 42 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001681112
oclc - 62777566
usfldc doi - E14-SFE0001002
usfldc handle - e14.1002
System ID:
SFS0025323:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001681112
003 fts
005 20060215071246.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 060104s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001002
035
(OCoLC)62777566
SFE0001002
040
FHM
c FHM
049
FHMM
090
QH307.2 (Online)
1 100
Albarracin, Maria Teresa.
0 245
Bottom-up and top-down effects on insects herbivores along a natural salinity gradient in a florida salt marsh
h [electronic resource] /
by Maria Teresa Albarracin.
260
[Tampa, Fla.] :
b University of South Florida,
2005.
502
Thesis (M.S.)--University of South Florida, 2005.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 42 pages.
3 520
ABSTRACT: I compared the strength of bottom- up and top-down effects on insect herbivores along a natural salinity gradient in salt marsh communities in West Central, Florida. I used a 2x2 factorial design with plots divided into four different treatments: 1) fertilizer applied to increase plant quality 2) sticky traps added to remove natural enemies (parasitoids) 3) fertilizer applied and sticky traps added and 4) control plots. These plots were placed on 7 different sites containing the salt marsh plant Borrichia frutescens along a natural stress salinity gradient. In each plot I determined the abundance of the sap sucker Pissonotus quadripustulatus, the gall maker Asphondylia borrichiae, spiders and the number of chewed leaves and bored stems. I also recorded leaf area, plant density, plant height and foliar nitrogen. Plants in fertilized plots exhibited increased height, density and leaf area.
590
Adviser: Peter Stiling.
653
Gall makers.
Sap suckers.
Borrichia frutescens.
Fertilizer.
Sticky traps.
Parasitoids.
Herbivory.
690
Dissertations, Academic
z USF
x Biology
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1002



PAGE 1

Bottom-Up and Top-Down Effects on Insects Herbivores Along a Natural Salinity Gradient in a Florida Salt Marsh by Maria Teresa Albarracin A thesis submitted in partial fulfillment of the requirement s for the degree of Master of Science Department of Biology College of Arts and Science University of South Florida Major Professor: Peter Stiling, Ph.D. Susan Bell, Ph.D. Gary Huxel, Ph.D. Date of Approval: March 24, 2005 Keywords: gall makers, sap suckers, Borrichia frutescens fertilizer, sticky traps, parasitoids, herbivory Copyright 2005, Maria Teresa Albarracin

PAGE 2

Acknowledgments I would like to thank my adviser Peter Stiling for all his help and guidance during this process. For their valuable comments I am thankful with the member of my comittee Susan Bell and Gary Huxel. I also want to thank the Stiling’s lab (Tatiana Cornelissen, Mark Barrett, Amanda Baker, Laura Altfield and Rebecca Forkhner) for their friendship and all their help reviewing this thesis. For their hel p in the field I want to thank Tatiana, Celina Bellanceau, Gwen Oberholtzer, Ivan Gualtero, Samuel Albarracin and Claudia Bernal. Thank you to my family in special my sister Claudia and German for their support and to my parents for inspired me. Final ly I want to thank Ivan for all this years of friendship and love To all of you Gracias!

PAGE 3

i Table of Contents List of T ables .......................................................................................................v List of Fi gures....................................................................................................... vi Abstract ..............................................................................................................v iii Introducti on...........................................................................................................1 Rationale and spec ific te sts.......................................................................3 Methods ................................................................................................................5 Study syst em.............................................................................................5 Experimental design..................................................................................6 Plant and soil va riables..............................................................................8 Herbivores and parasit oid respon ses........................................................9 Data anal ysis.............................................................................................9 Results ...............................................................................................................10 Plant and soil va riables............................................................................10 Herbivores and par asitoids ......................................................................11 Discussio n..........................................................................................................27 Referenc es.........................................................................................................30

PAGE 4

ii List of Tables Table 1 Summary of ANOVA results ( df, F -ratios and P -values) for the effect of site, fertilization and traps for plant variables measured on Borrichia in Marc h 2003. .............................................23 Table 2 Summary of ANOVA results ( df, F -ratios and P -values) for the effect of site, fertilization and tr aps for plant variables measured on Borrichia in June 2003..................................................................24 Table 3 Summary of ANOVA results ( df, F -ratios and P -values) for the effect of site, fertilization and traps for Asphondylia galls, Pissonotus abundance and percentage of parasitism on Pissonotus eggs................................................................................25 Table 4 Summary of ANOVA results ( df, F -ratios and P -values) for the effect of site, fertilization and traps for number of chewed leaves and stem bor ers abundance ..................................................26

PAGE 5

iii List of Figures Figure 1. Proposed factorial design to test the effects of bottom-up and top-down forces on Borrichia frutescens Treatments to be imposed are fertilization (Y = ye s, N = no) and exclusion of parasitoids (Y = yes, N = no). Nu mbers 1 to 7 represent the sites over the salinit y gradien t........................................................................4 Figure 2. Soil salinity (mean and standard e rror) of the 7 sites surveyed. Sites are arranged from the lowest to t he highest soil salinity. Means with the same letters are not si gnificant diffe rent........................................14 Figure 3. Mean ( SE) leaf area for every site surveyed, for fertilized and control plots among the 7 sites survey ed............................................15 Figure 4. Mean ( SE) stem height of Borrichia plants among the 7 sites surveyed .............................................................................................15 Figure 5. Mean ( SE) foliar nitrogen after fertilization of Borrichia plants..........16 Figure 6. Mean ( SE) per centage of green stems of Borrichia plants in fertilizer and control plots am ong the 7 sites surveyed........................16 Figure 7. Pearson’s correlations for plant variables measured on Borrichia’ monoculture among the 7 sites studied. Each symbol represents a different site (n=72) ..........................................................................17 Figure 8. Mean ( SE) galls abundance counted in the 7 sites surveyed ..........18

PAGE 6

iv Figure 9. Mean ( SE) gall size (mm) measured on the 7 sites surveyed.........18 Figure 10. Mean ( SE) gall parasitis m on the 7 sites surveyed. .......................19 Figure 11. Mean ( SE) Pissonotus abundance (Number of Pissonotus counted on 20 stems/plot) found in fertilized and control plots on each of the 7 st udied site s .................................................................19 Figure 12. Mean ( SE) percentage of Pissonotus egg parasitism on stems collected on fertilizer and control plots.....................................20 Figure 13. Mean ( SE) number of c hewed leaves (number of chewed leaves/50 stems) counted on fert ilized and contro l plots....................20 Figure 14. Mean ( SE) number of stem borers counted on fertilized and control pl ots........................................................................................21 Figure 15. Pearson’s correlations for herbivores abundances counted on Borrichia ’ monoculture among the 7 sites studied. Each symbol represents a different site (n=72)........................................................22

PAGE 7

v Bottom-Up and Top-Down Effects on Insect Herbivores Along a Natural Salinity Gradient in a Florida Salt Marsh Maria Teresa Albarracn ABSTRACT I compared the strength of bottomup and top-down effects on insect herbivores along a natural salinity gradient in salt marsh communities in West Central, Florida. I used a 2x2 factorial design with plots divided into four different treatments: 1) fertilizer applied to incr ease plant quality 2) sticky traps added to remove natural enemies (parasitoids) 3) fertilizer applied and sticky traps added and 4) control plots. These plots were pl aced on 7 different sites containing the salt marsh plant Borrichia frutescens along a natural stress salinity gradient. In each plot I determined t he abundance of the sap sucker Pissonotus quadripustulatus the gall maker Asphondylia borrichiae spiders and the number of chewed leaves and bored stems. I also recorded leaf area, plant density, plant height and foliar nitrogen. Plants in fert ilized plots exhibited increased height, density and leaf area. In fertilized plot s, galls were more abundant than in nonfertilized plots (approximately 63% in crease), and the same pattern was observed for Pissonotus (55% increase). For chewed leaves and spiders there

PAGE 8

vi were non significant increases on fertiliz ed as compared to unfertilized plots. There were no significant effects of nat ural enemy (parasitoid) removal. Gall density and Pissonotus density were both significantly more abundant in the sites with lower soil salinity, but there was no interaction of either treatment with salinity level. My results suggest that in this system bottom up effects are stronger than top down effect s but there is no change in the strength of bottom up and or top down effects along an env ironmental salinity gradient.

PAGE 9

1 Introduction The understanding of which factors in fluence the densities of herbivore populations has been pursued by ecologists for decades. Hairston et al. (1960) argued for the logical primacy of top-down fo rces in their widely known “world is green” hypothesis (also known as t he HSS hypothesis), suggesting that herbivores are limited by their natural enemies and not by their host plants. Fretwell (1977) and Oksanen et al. (1981) extended the HSS theory, suggesting the idea that productivity controls the dynam ics of trophic levels in the food chain. They predicted that bottom-up forces ma y often limit herbivore abundance in food webs which two trophic levels and that top-down forces could control herbivore densities only in food webs with three trophic levels. In contrast, Menge and Sutherland (1976, 1987) proposed a model in which the strength of biotic factors, such as herbiv ory or predation, would be strongly influenced by environmental stress. In stressful environments, such as a wave battered shore, natural enemies may not be present so that herbivores would be more influenced by competition or by abiotic factors. In more benign conditions, natural enemies would be more common and have greater effects. This study aims to investigate effects of both top-down and bottom-up factors on herbivores along a nat ural salinity gradient in a salt marsh community

PAGE 10

2 in West –Central Florida. Salt mars hes are considered stressful ecosystems because they are found in highly saline in tertidal zones in which nitrogen levels can be low (Adam 1990). Salinity and ni trogen can induce changes in plant quality that in turn affect the strength of top-down and bottom-up factors. Stiling and Moon (2005 b) performed an ex tensive factorial experiment in a salt marsh whereby they altered the strength of top-down and bottom-up factors along an experimentally generated sa linity gradient (by adding salt pellets) in a large Borrichia monoculture. Bottom-up effects, added fertilizer, were greater in more saline plots because nitrogen was more lim iting than in less saline plots, and increases of herbivores were greater. The strength of top-down effects also varied, but not in a predi ctable way. For gall makers top-down effects from parasitoids were greater in more salin e areas because gall gr owth was limited enabling parasitoids greater access to fly larvae inside and increasing parasitism rates. For sap suckers, which laid their eggs inside stems, top-down effects were less in saline conditions because stems were tougher and less easy to penetrate by parasitic mymarid wasps. Thus, in this system environmental stress did appear to change the effects of top-down and bottom-up factors, but not in a predictable way. However, in nature Borrichia occurs in many different clones and in many different sites, each with va rying salinity (Stiling and Rossi 1996). We wanted to know if results from an experimentally generat ed range of salinities in a large Borrichia monoculture held true under natural changes of salinity between different sites c ontaining different clones.

PAGE 11

3 Rationale and specific tests This study used a 2X2 factorial desi gn (Figure 1) along a natural salinity gradient to address the importance of t he combined effects of plant nutrients and natural enemies on herbivores on Borrichia frutescens The following hypotheses were tested: 1. Herbivore densities should decrease al ong a natural salinity gradient due to the negative impacts of salinity on plant quality and its subsequent effects on herbivore populations. 2. Herbivore densities should be higher in fertilized plots compared to unfertilized plots due to the posit ive effects of fertilization on Borrichia nutritional quality. 3. Herbivore densities and survivorship should be higher in plots in which natural enemies are excluded due to t he strong impacts of parasitoids on survivorship of herbivore eggs and larvae. 4. Bottom-up effects should be greatest at more saline sites where nitrogen is more limiting as evidenced by an interaction of fertilizer and salinity on herbivore densities. 5. Top-down effects should vary in strength along a salinity gradient according to the type of herbivore involved. Thus, there should be an interaction between parasitoid removal and salinity on herbivores densities

PAGE 12

4 but the direction of change in herbi vore populations along the salinity gradient would be different for galls than for sap suckers. Figure 1 Factorial design to test the effe cts of bottom-up and top-down forces on Borrichia frutescens Treatments are fertilizat ion (Y = yes, N = no) and exclusion of parasitoids (Y = yes, N = no). Numbers 1 to 7 represent the sites over the salinity gradient. 1234567Treatments TreatmentsFertilizer Parasites ExcludedYN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, NSalinity GradientLowHigh 1234567 1234567Treatments TreatmentsFertilizer Parasites ExcludedYN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, N YN Y, NY, NSalinity GradientLowHigh

PAGE 13

5 Methods Study System Borrichia frutescens, commonly known as sea oxeye daisy, is a rhizomatous shrub typically less than a meter in height. It is common in the intertidal zones of salt marshes, and can be found in association with Salicornia virginica Batis maritima and Distichlis spicata (Antlfinger 1981). An important characteristic of Borrichia is that it can be found over a wide range of salinities (Richards et al 2004). Asphondylia borrichia (Diptera, Cecidomyiidae) is a gall-making fly that oviposits in the apical meristems of Borrichia stems, inducing the formation of galls. Each gall has on average four chambers and each chamber has one larva (Rossi and Stiling 1995). Asphondylia is multivoltine and has around five to seven overlapping generations in a year. It occurs on Borrichia throughout the year, but highest densities are usually found in spring and early summer (Rossi et al. 1992). Asphondylia larvae have four species of hymenopteran parasitoids: Rileya cecidomyiae and Tenuipetiolus teredon (Eurytomidae), Torymus umbilicatus (Torymidae), and Galeopsomyia haemon (Eulophidae). Parasitism is the greatest cause of mortality for Asphondylia and is relatively easy to observe

PAGE 14

6 in the field by examini ng the emergence holes of the galls. The parasitoids usually make small round holes while the emergence hole of Asphondylia is bigger with the puparia attac hed (Stiling and Rossi 1997). Pissonotus quadrispustulatus (Delphacidae) is a phloem feeder, monophagous on Borrichia (Denno 1978, Stiling 1994). The adults are brachypterous, a characteristic that ma kes these insects easy to observe and survey. They are also multivoltine wit h overlapping generations, and their eggs are laid in the stems of Borrichia Eggs are attacked by the fairy fly Anagrus sp. nr armatus (Hymenoptera: Mymaridae) and this constitutes by far the greatest cause of mortality (Moon and Sti ling 2000). The dryinid parasitoid Pseudogonatopus arizonicus also attacks Pissonotus although levels of attack by this parasitoid are generally le ss than 1% (Stiling and Moon 2005 a). Another herbivore feeding on Borrichia is the stem borer Argyresthia sp. (Lepidoptera: Argyresthiidae). In the imma ture stages this stem borer feeds on the mesoderm beneath the epidermis of the stems. The damage produced is recognizable because the stems have a papery appearance (Moon and Stiling 2002 a). Spiders and damselflies are al so found in natural patches of Borrichia and may be predators that feed on herbivores present in this system.

PAGE 15

7 Experimental Design This study was conducted in seven separate salt marshes around Tampa Bay, Florida. Study sites were locat ed in Hillsborough (Simmons Park-Ruskin, Upper Tampa Bay) and Pinellas Countie s (Fort de Soto, St. Petersburg, Honeymoon Island and two islands in the intracoastal waterway, called CW6 and CW5). These sites were chosen for their variability in Borrichia size and location along a natural salinity gradient. The areas wi th high salinity tend to have shorter Borrichia plants (Richards et al 2004) with smaller leaf area and lower densities of herbivores (Honeymoon Island, Simmons Park, Upper Tampa Bay), while the areas with low salinity have taller Borrichia with higher densities of herbivores (CW5, CW6, Fort de Soto). In Fort de So to there were two di stinctly different sites, one with relatively high soil sa linity, and one with low soil salinity. At each site, top down effects were manipulated by reducing parasitism levels using 13x8 cm sticky traps, wh ich are yellow cards, covered with tanglefoot adhesive (Gempler’s Belleville, Wisconsin, USA). Five traps were placed per plot (four in each corner and one in the middle). Moon and Stiling (2002 a) demonstrated that five sticky tr aps on each plot significantly reduced parasitism levels of both gall makers and sap suckers. Bottom up effects were manipulated by adding urea-based nitrogen fertilizer (46:0:0 NPK) poured over the entire area of the plo t. Traps were changed monthl y and 50g of fertilizer was added every two months, to avoid burning t he plants. Each experimental site

PAGE 16

8 contained three replicates of each treatm ent with the exception of the islands CW5 and CW6 in which I placed just two rep licates, due to their small size, for a total of 72 plots. Surveys of herbivores were made twice every month from March through October 2003. The following treatment s were assigned randomly to each plot: 1) control plots; 2) plots with st icky traps added (to reduce top-down effects); 3) plots with fertilizer added (to increase bottom-up ef fects); and 4) plots with both traps and fertilizer added. Plant and soil response variables At the beginning and end of the experim ent I measured plant height in each plot by measuring the height of 10 plant s and plant density, by randomly placing a 0.25 m2 quadrant in each plot and counting t he stems inside this area. Because increase salinity can change pl ant morphology, inducing tough woody stems, and this can affect herbivory density, perc entage of green stems, was measured by counting 100 stems in each plot and scoring them as “green” and soft or “woody” and hard (Stiling and Moon 2005 b). Leaf area was measured by randomly collecting 5 leaves from each plot, digita lizing pictures of each leaf and using a software UTHSCA Image tool (Wilcox et al. 1996). Foliar nitrogen of Borrichia plants were quantified using a CHN analyzer. The five collected leaves were oven dried, ground in a Wiley Mill, and analyzed for foliar nitrogen (% dry mass) using CE Instruments NC2100 CN Anal yzer (CE Elantech, Lakewood, N.J., USA). To measure soil salinity a techni que from Pennings and Moore (2001) was

PAGE 17

9 used. Samples of soil were collected from every plot, taken into the laboratory and weighed. Samples were then air dri ed and reweighed, then placed into an Erlenmeyer flask with ten ml of distill ed water. Then, an amount of distillated water equal to the amount of water origi nally in the sample, was added. After another day, when the sediments settled down, the salinity was recorded using a refractometer (VEE GEE A366ATC, salinity 0-100‰). Herbivore and parasitoid responses In order to study how bottom-up and top-down factors affect the herbivores of this system, all herbivores were counted monthly. Asphondylia galls were counted on 200 stems on each plot, Pissonotus on 20 stems, chewed leaves on 50 stems, stem borers ( Argyresthia sp.) on 50 stems, and spiders on 50 plants. Parasitism levels of Pissonotus eggs were estimated by sub-sampling 10 stems per plot in June and dissecting them under a microscope. Parasitized eggs can be distinguished because t hey appear black or orange while nonparasitized eggs are white with red spots (Moon and Sti ling 2000). Parasitism levels of Asphondylia galls were estimated by counting the number of fly and parasitoid emergence holes on galls on each plot at every census.

PAGE 18

10 Data Analysis Treatments effects on total herbivore densities summed over the year were tested using a 3-way factorial AN OVA with site, fertilizer addition and parasitoid removal as factors. Data was transformed using log10 (x+1) when necessary. When data could not be transfo rmed to meet normality assumptions, a Kruskall–Wallis one–way Anova or Mann – Whitney was used. Means and standard errors are reported untransformed. In addition Pearson’s correlations were sometimes used to correlate her bivore densities to environmental data using all 72 plots. All statistical an alysis were performed using SYSTAT 9.0 (Wilkinson 1999)

PAGE 19

11 Results Plant and soil variables Soil salinity among the seven studi ed sites ranged between 4 and 18 ppt with a mean value of 10.33 0.827 (Figure 2), and were significantly different between sites (Kuskall-Wallis = 58.564, P<0.001). Most of the treatments and interacti ons between treatments had no effect on plant variables (Table 2, Figures 3-6). The two exceptions were that fertilizer significantly increased both leaf area and percentage foliar nitrogen. In contrast site had a significant effect on all pl ant variables both pre and post treatment (Table 1). Total mean density of Borrichia stems, plant height and leaf area were not significantly different between fertiliz er and removal of par asitism treatments. However, site had a significant effect on all plant variables (Figures 3-6). Correlations revealed that in general most plant variables significantly decreased with increases in soil salinit y, except stem density, wh ich increased (figure 7). June data on percentage of green stems co uld not be transformed to meet ANOVA assumptions, so the three factors were analyzed separate using a Kruskal -Wallis test. Site was again the onl y factor that had a significant effect on

PAGE 20

12 the mean percentage of green stems (Krusk al-Wallis= 45.759, P< 0.001, (Figure 6). Herbivore and parasitoid responses Herbivore densities were significantly different among sites (Tables 3-4, Figures 8-14). Furthermore the effect of site differed according to the type of herbivore so that one site was bes t for galls (CW6), another for Pissonotus (Fort de Soto), another for stem borers (C W5) and yet another for chewed leaves (Simmons and Honeymoon island). Site also significantly affected gall size (KW=31.276, d.f.=6, P < 0. 001), gall parasitism (F6,40= 19.148, P<0.001) and Pissonotus parasitism (F6,40= 4.855, P=0.001). Significant negative correlations were observed for Pissonotus abundance (P=0.001), stem borer abundance (P=0.008), gall size (P=0.054) and gall par asitism (P=0.028) with an increase in soil salinity (Figure 14). The only guild with a positive correlation with increasing salt was the chewers (P<0.001) (Figures 16). Site had a significant effect, (P < 0. 001) on the density of both the spiders and the damselflies. However, none of the other treatment s had a significant effect on the densities of these predat ors. I was not expecting differences between treatments with thes e organisms, because they are predators that move easily among plots, and the sticky traps may be not effi cient enough to trap this class of predators.

PAGE 21

13 Fertilizer had a significant effect on gall abundance and gall size (MannWhitney U= 754.5, P=0.028). In general, galls on the fertilized plots exhibited bigger sizes than the ones on the control plots, with th e exception of the ones found at Upper Tampa Bay. As a result, fertilizer significantly reduced gall parasitism. On the other hand because of an increase in green stems, fertilizer significantly increased parasitism rates of Pissonotus eggs. The amount of this increase in egg parasitism on fertilized pl ots varied among sites resulting in a significant site x fertilizer inte raction (P=0.018). Despite this, Pissonotus densities showed an increase on fertilized plots. There were no other effects of fertilizer or other interactions wit h other treatments. Traps failed to influence herbivore abundance or parasitism levels of either Asphondylia or Pissonotus There were also no significant interactions between traps and any other treatment (Tables 3-4).

PAGE 22

14 Soil Salinity (ppt) 0 5 10 15 20 25 FSoto 1 FSoto 2 UPTB CW5CW6Simmons HMI LOWHIGH Salinity Gradient a a a,b a,b b b b Figure 2. Soil salinity (mean and standard erro r) of the 7 sites surveyed. Sites are arranged from the lowest to the highest so il salinity. Means wit h the same letters are not significant different.

PAGE 23

15 Height June (cm) 0 20 40 60 80 100 FSoto 1 CW5UPTBFSoto 2 CW6Simmons HMI LOW HIGH Salinity Gradient Fertlized Control Leaf area(cm2) 0 2 4 6 8 10 FSoto1CW5UPTBFSoto 2CW6SimmonsHMI Fertilized Control Figure 3. Mean ( SE) leaf area of Borrichia plants in fertilized and control plots among the 7 sites surveyed. Figure 4. Mean ( SE) stem height of Borrichia plants in fertilized and control plots among the 7 sites surveyed.

PAGE 24

16 % Green stems (june) 0 20 40 60 80 100 FSoto 1 CW5UPTBFSoto 2 CW6SimmonsHMI LOW HIGH Salinity Gradient Fertlized Control Foliar Nitrogen 0.0 0.5 1.0 1.5 2.0 2.5 FSoto 1 CW5 UPTBFSoto 2CW6SimmonsHMI Fertilizer Control Figure 5. Mean ( SE) foliar nitrogen of Borrichia plants in fertilized and control plots among the 7 sites surveyed. Figure 6. Mean ( SE) perc entage of green stems of Borrichia plants in fertilized and control plots among the 7 sites surveyed.

PAGE 25

17 Figure 7. Pearson’s correlations for plant variables measured on Borrichia among the 7 sites studied. Each symbol r epresents a different site (n=72). Salinity (ppt) 0510152025303540 Percentage of green stems 0 20 40 60 80 100 120 r=-0.225 p=0.057 For de Soto 1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI Salinity (ppt) 0510152025303540 Foliar nitrogen 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 r=-0.210 p=0.076 Fort de Soto 1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI Salinity (ppt) 0510152025303540 Leaf area (cm 2 ) 0 2 4 6 8 10 12 r=-0.459 p<0.001 Fort de Soto 1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI Salinity (ppt) 0510152025303540 Stem density 0 20 40 60 80 100 r=0.465 p<0.001 Fort de Soto 1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI Salinity (ppt) 0510152025303540 Stem Height (cm) 0 20 40 60 80 100 r = -0.635 P<0.001 Fort de Soto 1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI

PAGE 26

18 gall size (mm) 0 2 4 6 8 10 12 14 16 FSoto 1 CW5UPTBFSoto 2 CW6Simmons HMI LOW HIGH Salinity Gradient Fertlized Control Total Asphondylia galls 0 50 100 150 200 250 FSoto 1 CW5UPTB FSoto 2 CW6 Simmons HMI LOW HIGH Salinity Gradient Fertlized Control Figure 8. Mean ( SE) gall abundance in fertilized and control plots among the 7 sites surveyed. Figure 9. Mean ( SE) gall size (mm) measured in fertilized and control plots among the 7 sites surveyed.

PAGE 27

19 Pissonotus abundance 0 20 40 60 80 Simmons UPTB CW6 CW5 FSoto 1 HMI Fertilized Control FSoto 2 Figure 10. Mean ( SE) gall parasitism in fertilized and control plots among the 7 sites surveyed. Figure 11. Mean ( SE) Pissonotus abundance (Number of Pissonotus counted on 20 stems/plot) found in fertilized and control plots on each of the 7 studied sites. Gall Parasitism 0 10 20 30 40 50 60 70 Simmons UPTBCW6 CW5 FSoto 1HMI Fertilized Control FSoto 2

PAGE 28

20 % of Egg Parasitism ( Pissonotus ) 0 20 40 60 80 100 FSoto 1 CW5 UPTBFSoto 2 CW6Simmons HMI LOW HIGH Salinity Gradient Fertlized Control Number of chewed leaves 0 20 40 60 80 FSoto 1 CW5UPTBFSoto 2 CW6Simmons HMI LOW HIGH Salinity Gradient Fertlized Control Figure 12. Mean ( SE) percentage parasitism on Pissonotus eggs in fertilized and control plots among the 7 sites surveyed. Figure 13. Mean ( SE) number of chewed leaves (number of chewed leaves/50 stems) counted on fertilized and control plots among the 7 sites surveyed.

PAGE 29

21 Stem borer abundance 0 10 20 30 40 50 FSoto 1 CW5UPTB FSoto 2 CW6Simmons HMILOW HIGHSalinity Gradient Fertlized Control Figure 14. Mean ( SE) number of stem borers counted on fertilized and control plots among the 7 sites surveyed.

PAGE 30

22 Figure 15. Pearson’s correlations fo r herbivores abundances counted on Borrichia monoculture among the 7 sites st udied. Each symbol represents a different site (n=72). Salinity (ppt) 0510152025303540 Pissonotus abundance 0 20 40 60 80 100 120 r=-0.392 p=0.001 Fort de Soto 1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI Salinity (ppt) 0510152025303540 Stem borer abundance 0 10 20 30 40 50 60 70 r=-0.310 p=0.008 Fort de Soto 1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI Salinity (ppt) 0510152025303540 Gall size (mm) 0 5 10 15 20 r=-0.228 p=0.054 Fort de Soto1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI Salinity (ppt) 0510152025303540 Asphondilya galls 0 50 100 150 200 250 r=0.127 p=0.289 Fort de Soto 1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI Salinity (ppt) 0510152025303540 Asphondilya gall parasitism 0 20 40 60 80 r=--0.259 p=0.028 Fort de Soto 1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI Salinity (ppt) 0510152025303540 chewed leaves 0 20 40 60 80 100 120 r=0.468 p<0.001 Fort de Soto 1 CW5 UPTB Fort de Soto 2 CW6 Simmons HMI

PAGE 31

23 Table 1. Summary of ANOVA results ( df, F -ratios and P -values) for the effects of site, fertilization and traps for plant variables measured on Borrichia in March 2003. Values in bold indicate stat istically significant results. Source df Leaf area F P Stem density F P % of green stems F P Site 6,4414.198 <0.001 5.686 <0.001 14.301 < 0.001 Fertilization 1,440.004 0. 952 0.6000.443 1.510 0.226 Traps 1,440.953 0.334 0.0080.930 0.219 0.642 Site x Fertilization 6,44 0.568 0.754 0.3760.890 0.453 0.839 Site x Traps 6,440.430 0.855 0.1620.985 0.388 0.883 Fertilization x traps 1,440. 025 0.874 0.00 0.997 0.383 0.539 Sitextrapsxfertilization 6,44 0.165 0.985 1.1270.362 0.485 0.816

PAGE 32

24 Table 2. Summary of ANOVA results ( df, F -ratios and P -values) for the effect of site fertilization and traps for plant variables measured on Borrichia in June 2003. Values in bold indicate statistically significant results. Source df Leaf area F P Stem density F P Height F P Foliar nitrogen F P Site 6,44 8.530 <0.001 21.6 <0.001 39.564 <0.001 8.801 <0.001 Fertilization 1,44 6.671 0.013 0.749 0.392 0.035 0.852 19.05 <0.001 Traps 1,44 0.109 0.743 1.126 0.294 0.092 0.763 0.178 0.675 Site x Fertilization 6,44 0.144 0.989 0.887 0.512 0.666 0.677 1.570 0.179 Site x Traps 6,44 0.214 0.971 0.354 0.904 0.330 0.917 0.605 0.725 Fertilization x traps 1,44 0.675 0.416 2.042 0.160 0.034 0.855 0.255 0.616 Site x Fertilizer x Traps 6,44 0.317 0.925 0.528 0. 784 0.446 0.844 0.992 0.443

PAGE 33

25 Table 3. Summary of ANOVA results ( df, F -ratios and P -values) for the effect of site, fertilization and traps on Asphondylia galls and gall parasitism, Pissonotus abundance and percentage of parasitism on Pissonotus eggs. Values in bold indicate statistically significant results. Source df Gall abundance F P Gall parasitism F P Pissonotus abundance F P % of egg parasitism ( Pissonotus ) F P Site 6,40 16.38 <0.001 19.19 <0.001 5.99 <0.001 4.855 <0.001 Fertilization 1,40 4.18 0.048 6.10 0.018 3.43 0.071 4.085 0.050 Traps 1,40 0.24 0.63 0.29 0. 59 0.02 0.882 1.079 0.305 Site x Fertilizati on 6,40 0.39 0.88 3.89 0.004 0.55 0.768 0.450 0.841 Site x Traps 6,40 0.21 0.97 1. 43 0.23 0.24 0.962 2.087 0.076 Fertilization x traps 1,40 0.00 0.99 0.06 0.81 0.00 0.996 0.117 0.734 Site x traps x fertilization 6,40 1.42 0.23 1.17 0.34 0. 36 0.902 1.232 0.311

PAGE 34

26 Table 4 Summary of ANOVA results ( df, F -ratios and P -values) for the effect of site, fertilization and traps for number of chewed leaves and stem borers abundance. Values in bold indicate st atistically significant results. Source df Number of chewed leaves F P Stem borers F P Site 6,44 21.202 <0.001 5.581 <0.001 Fertilization 1,44 0. 345 0.560 1.447 0.235 Traps 1,44 0.543 0.465 1.729 0.195 Site x Fertilization 6, 44 0.344 0.910 0.936 0.479 Site x Traps 6,44 0. 408 0.870 2.179 0.063 Fertilization x traps 1, 44 1.057 0.309 0.038 0.846 Site x traps x fertilizati on 6,44 0.886 0.514 1.722 0.138

PAGE 35

27 Discussion In this study I found that plant mo rphology of the salt marsh plant Borrichia frutescens varied dramatically among sites. At the extreme, plants at CW5 were more than twice the height of those at Honeymoon Island and their leaves were also more than twice as big. Borrichia height is greatly influenced by interstitial soil salinity (Richards et al 2004) at these Florida sites and plant height was significantly negatively correlated with so il salinity. Foliar nitrogen was also variable between sites, and was generally less in higher salinity plots. Borrichia herbivores also varied dramatically in density between sites. Galls at one site, CW6, were over three times as abundant as at any other site, while densities of Pissonotus were twice as great at Fort de Soto as at Simmons and Honeymoon island. Stem borers were very common at CW5 but not at any other site. Such patterns could be influenced by competition such that if galls are common at CW6 then there are fewer res ources available for other herbivores. Moon and Stiling (2002 b) showed that hi gh densities of stem borers tended to reduce the density of Pissonotus eggs which they kill as they bore through the stems, but in the pres ent study the number of Pissonotus was low at Simmons and Honeymoon Island despite there being fe w stem borers. In general, densities

PAGE 36

28 of Borrichia herbivores are so low as to not influence the densities of one another and there is an abundance of unoccupied stems at each site (Stiling et al 1999). Of the applied treatments, the only consistent effects were those due to fertilizer. Increased plant quality in creased the density of galls and Pissonotus both directly, and in the case of Pissonotus indirectly, by lowering egg parasitism. Significant trap effects were not detected in the present study. Nor were there many interactions between treatments. It appears bottom-up effects are much stronger at all sites than are top-down ones, a conclusion supported by previous experiments in this system (Moon and Stiling 2000, Stiling and Moon 2005 a, b). Previous, but recent, experim ental manipulations of bottom-up and top-down effects along an experimentally generated sa linity gradient showed significant interactions between salt level and both top-down and bottom-up effects.Bottom-up effects were much w eaker in more saline plots possibly because increased salinity reduces the plant’s ability to uptake additional nutrients (Jefferies and Perkins 1977). Se condarily, plant toughness increases in saline plots and this reduces the strength of top-down effects of parasitoids which cannot so easily penetr ate stems to parasite Pissonotus eggs. For Asphondylia on the other hand, more saline plots r educe gall size, and fertilization cannot increase gall size enough to reduce gall par asitism as it can on less saline plots. However these experiment s were performed on large Borrichia monocultures at a single site which probably consisted of re latively few clones, or at least similar ones. Stiling and Rossi (1996) showed that clonal variation among sites may

PAGE 37

29 alter the resistance of plants to herbivo re attack. Under such uniform conditions and where clone was controlled for, it may have been easier to detect significant interactions of effects with salinity (St iling and Moon 2005 b). In the “real world” of large site to site variation in soil salinity, soil nutrients, plant size, plant highness and the clonal identity of plants, as experienced in this study, it is harder to detect these subtle interacti ons between treatments. Only the strongest effects are detected, in this case the effe ct of fertilizer. Thus, as previous studies have noted in this system (Moon and St iling 2002 a,c, Stiling and Moon 2005 a,b) and in general (Hunter and Price 1992, Pr ezler and Boecklen 1996, Price 2002) the relative strength of bottom-up effect s are strong. Only as evidenced by experimental manipulations in which t op-down effects are not well controlled.

PAGE 38

30 References Adam, P. 1990. Salt marsh ecology.Camb ridge University Press, Cambridge,UK. Antlfinger, A. 1981. The genetic basis of micro differentiation in natural and experimental populations of Borrichia frutescens in relation to salinity. Evolution 35: 1056 – 1068 Denno, R.F. 1978. The optimum popul ation strategy for plant hoppers (Homoptera: Delphacidae) in st ables marsh habitats. Canadian Entomologist 110: 135 – 142 Fretwell, S.D. 1977. The regul ation of plant communities by food chain exploring them. Persepct. Biol. Med. 20: 169 -185. Hairston, N.G., F.E. Smith, and L.B. Slobodkin.1960. Community structure, population control, and co mpetition. American Na turalist 44: 421 – 425.

PAGE 39

31 Hunter, M.D. and P.W. Price. 1992. Pl aying chutes and ladders: heterogeneity and the relative roles of bottom-up and top-down forces in natural communities. Ecology 73: 724 – 732. Jefferies, R.L., and N. Perkins. 1977. The effects on the vegetation on the vegetation of the additions of the inorgani c nutrients to salt marsh soils at Stiffkey, Norfolk. Journal of Ecology 65:867-882 Menge, B.A., and J.P. Sutherland. 1976. Spec ies diversity gradients: synthesis of the roles of predation, competition, and temporal heterogeneity. American Naturalist 110: 351-369 Menge, B.A., and J. P. Sut herland. 1987. Community r egulation: variation in disturbance, competition, and predation in relation to environmental stress and recruitment. American Naturalist 130: 730 – 757. Moon, D., and P. Stiling. 2000. Relative importance of abi otically induced direct and indirect effects on a salt – marsh herbivore. Ecology 81: 470 – 481 Moon, D. and P. Stiling. 2002 a. The effects of salinity and nutrients on a tritrophic salt –marsh syst em. Ecology 83: 2465 -2476.

PAGE 40

32 Moon, D. and P. Stiling. 2002 b. Top-down, bottom-up, or side to side? Withintrophic-level interactions modify trophic dynamics of a salt marsh herbivore. Oikos 98: 480 – 490. Moon, D. and P. Stiling. 2004. The influenc e of a salinity and nutrient gradient on coastal vs. upland tritrophic comp lexes. Ecology 85: 2709 2716. Oksanen, L. S.D.Fretwell, J. Aruda, and P.Niemela.1981. Exploitation ecosystems in gradients of productivi ty. American Naturalist 118: 240-261. Preszler, R.W. and W.J. Boecklen. 1996. The influence of elevation on tri – trophic interactions: Opposing gradients of top – down and bottom – up effects on a leaf – mining mo th. Ecoscience 3: 75 – 80. Price, P.W (2002) Resource-driven terre strial interaction webs. Ecological Research 17: 241-247. Pennings, S.C. and D.J.Moor e. 2001. Zonation of shr ubs in western Atlantic saltmarshes. Oecologia 126: 587-594 Richards, C. L., J.L. Hamrick, L.A. Donovan and R. Mauricio. 2004. Unexpectedly high clonal diversity of two salt marsh perennials across a severe environmental gradient. Ecological letters 7: 1155 – 1162.

PAGE 41

33 Rossi, A.M, P. Stiling, D.R. Strong and D.M. Jhonson. 1992. Does gall diameter affect the parasitism rate of Asphondylia borrichiae (Diptera: Cecidomyiidae)? Ecological Entomology 17: 149 – 154 Rossi, A.M., P. Stiling .1995. Intraspecific variation in growth-rate, size, and parasitism of galls induced by asphondylia-borrichiae (diptera, cecidomyiidae) on three host species. Annals of the Entomological Society of America 88: 39-44. Rossi, A.M. and P. Stiling. 1998. The intera ction of plant clone and abiotic factors on a gall-making midge. Oecologia 116: 170 – 176 Stiling, P. 1994. Coastal insect herbivor e populations are strongly influenced by environmental variation. Ecol ogical Entomology 19: 39-44. Stiling, P and A.M. Rossi. 1995. Coastal insect herbivore communities are affected more by local environmental -conditions than by plant genotype. Ecological Entomology 20: 184-190. Stiling, P. and A. M. Rossi. 1996. Comple x effects of genotype and environment on insect herbivores and their enemies. Ecology 77: 2212-2218.

PAGE 42

34 Stiling, P. and A.M. Rossi. 1997. Experim ental manipulations of top-down and bottom-up factors in a tri-trophi c system. Ecology 78: 1602 -1606. Stiling, P., A.M. Rossi, M. Cattell, and T. Bowdish. 1999. Weak competition between coastal insect herbivores. Florida entomologist 82: 599-608. Stiling, P. and D. Moon. 2005 a. Quality or quantity: the di rect and indirect effects of host plants on herbivores and t heir natural enemies. Oecologia 142:413-420. Stiling, P. and D. Moon. 2005 b. Are trophodynamics models worth their salt?The relative roles of top-down and bottomup effects along a salinity gradient in a Florida Salt marsh. Ecology. Wilcox, C. Donald, S. Brent Dove, W. Doss, McDavid and David B. Greer.1996. UTHSCSA Image Tool. Version 3.0. Department Of Dental Diagnostic Science. University of Texas Health Science Center, San Antonio, Texas. Wilkinson, L. 1999. SYSTAT: The system for statistics. Version 9. – SYSTAT, Inc., Evanston, Illinois, USA.