Management Experiments on Lake Erie Flowering Rush

Bowling Green State University ScholarWorks@BGSU

Honors Projects Honors College

Spring 4-29-2013

Alyssa Dietz

Follow this and additional works at: https://scholarworks.bgsu.edu/honorsprojects

Repository Citation Dietz, Alyssa, "Management Experiments on Lake Erie Flowering Rush" (2013). Honors Projects. 19. https://scholarworks.bgsu.edu/honorsprojects/19

This work is brought to you for free and open access by the Honors College at ScholarWorks@BGSU. It has been accepted for inclusion in Honors Projects by an authorized administrator of ScholarWorks@BGSU. 1

Management Experiments on Lake Erie Flowering Rush

Alyssa Dietz

HONORS PROJECT

Submitted to the University Honors Program

at Bowling Green State University in partial

fulfillment of the requirements for graduation with

UNIVERSITY HONORS

April 29 th , 2013

Advisors:

Dr. Helen J. Michaels, Department of Biological Sciences

Dr. Enrique Gomezdelcampo, Department of Environment and Sustainability

Summary:

The issue of invasive species has been a major concern for many Great Lakes ecosystems over the past several decades. Since the first noted invasive animal, the sea lamprey, these fresh water ecosystems have been invaded by over 180 different plants and animals that create over

$100 million worth of damage a year (Rosaen et al . 2012). Because the five Great Lakes, along with their rivers, streams, and wetlands are such an interconnected series, the introduction of any invasive species to one of the Great Lakes, such as Lake Erie, means that invaders are likely to spread to all other freshwater ecosystems in the area (Rosaen et al . 2012). With the various ways that many introduced species can alter the nutrient availability, habitat quality, and species richness, many wildlife managers are continuing to develop management strategies that remove or greatly limit the further spread of invasive species. Of major concern are invading aquatic plants whose ranges continue to expand. The purpose of this study was to understand the response of an invasive aquatic plant, Butomus umbellatus L. in Lake Erie wetlands to the combined management strategies of clipping and flooding. These strategies are hoped to be shared with local wetlands managers in better hopes of limiting the impacts and concerns that this invasive has created in local ecosystems.

Introduction:

Given their diverse morphology, invasive macrophytes, or aquatic plants, have the potential to alter community complexity, change oxygen levels within aquatic systems, and facilitate in the invasion of other exotic species (Schultz and Dibble 2012). Though significantly detrimental to native ecosystems, macrophytes receive very little attention compared to the issues of invasive aquatic animals. One particularly understudied species found in Lake Erie is 3

Butomus umbellatus , commonly known as flowering rush. Originally introduced from Europe for aquatic gardens, B. umbellatus has now spread throughout along the northern United States and

Canadian border (Brown and Eckert 2005 ) (Figure 1).

Figure 1. The Invasive Range of B. umbellatus spans most of the U.S./Canadian border.

Characterized by triangular leaves, B. umbellatus is not a true rush, but is rather a single species belonging to the p lant family Butomaceae ( Brown and Eckert 2005). Though little is known about its specific origin and impacts, the reproduction and life history of B. umbellatus are well known. B. umbellatus has three methods of reproduction and can reproduce sexually or asexually. B. umbellatus can reproduce sexual ly through seed, but there is som e argument over the viability of seeds produced by Nort h American populations ( Kliber and Eckert 2005 ) (Picture

1). It is thought that most populations in this region reproduce dominantly via asexual methods through rhizomes (Picture 2) and vegetative prop agules (Picture 3) (Brown and Eckert 2005 ;

Parkinson et al . 2010). These vegetative propagules are pieces of rhizomes that break off from adult plants and form roots while floating. When marsh water levels drop, the vegetative 4 propagules are likely able to establish quickly, potentially giving B. umbellatus a great advantage over smaller native seedlings.

Picture 1. Flower of B. umebllatus produce seeds Picture 2. Below ground rhizomes through sexual reproduction. allow for asexual reproduction.

Picture 3 . Vegetative propagule growth on the rhizome of an established plants are produced by asexual reproduction.

As an invasive, B. umbellatus can alter native vegetation and macroinvertebrate communities (Shultz and Dibble 2012). In Canadian areas, B. umebellatus has begun to threaten

native populations of Zizania aquatica (wild rice). With further spread throughout the Great

Lakes Region, Canada has designated B. umbellatus as a “principal invasive alien” (White et al .

1993 in Brown and Eckert 2005). In the United States, B. umbellatus has been found in Lake

Erie wetlands since the early twentieth century (Brown and Eckert 2005). One particular regional

population is located at Ottawa National Wildlife Refuge (ONWR) in Oak Harbor, Ohio. ONWR

is a diked wetland ecosystem manager by altering flooding levels necessary for habitat for native

shorebird populations. As pilot studies have demonstrated, B. umbellatus is capable of changing

macroinvertebrate community richness, greatly reducing this vital food resource for local fish

and bird populations (Dietz, unpublished data). With such an impact to the major taxa of the

region, ONWR and other local wetland refuges have begun to implement management strategies

to control B. umbellatus .

Most strategies to remove B. umbellatus have had minimal success in restoring invaded

marshes. Common strategies include manually digging out plants, cutting, and the application of

herbicides (Parkinson et al . 2010). Digging plants has had mixed results due to the further spread

of B. umbellatus after disturbances to the rhizomes (Parkinson et al . 2010). In addition, the

application of herbicides has proven inefficient due to the thin emergent leaves of B. umbellatus

that reduce the chance of chemicals adhering to the plant (Parkinson et al . 2010). Applied

chemicals could wash away and potentially alter the chemical composition of water where they

could impact non-target species, such as amphibians. These inefficiencies in current management

strategies of B. umbellatus have caused many wildlife managers to search for new ways to

successfully manage or eradicate the invasive species. 6

Some of the effective treatments developed for the removal of other aquatic plants include inundation (flooding) and clipping (Kercher and Zedler 2004). Clipping removes leaves and stem tissue critical for photosynthesis and the transportation of oxygen to roots, limiting the future growth of the plant (Mayence et al . 2010). The inclusion of secondary clipping treatments is believed to further limit the development of plants. This second cutting is intended to remove additional growth material after the plant has used its carbohydrate and nutrient reserves but before new leaves have begun to conduct photosynthesis and begin to restore the resources that it took to grow them.

Inundation limits both light levels for photosynthesis and oxygen for root respiration. As water depth increases, both of these vital environmental conditions decrease, reducing growth and plant survival (Kercher and Zedler 2004). While some programs treating B. umbellatus with clipping have had little impact on the eradication of the species, a combination of inundation and clipping may be a more effective method in preventing the species’ spread into more southern wetlands (Parkinson et al . 2010). Previously conducted pilot studies investigating the impact of flooding levels alone have suggested that B. umbellatus tended to have lower biomass in higher water depths (Dietz, unpublished data).

To develop better management strategies of B. umbellatus in Lake Erie wetlands, this experiment investigated the effects of these alternative management strategies of clipping and inundation through the following experimental questions: 1) How does flooding effect the development and growth of B. umbellatus ? and 2) Do combined treatments of flooding and clipping effectively reduce the growth of B. umbellatus ?

Materials and Methods: 7

Two hundred B. umbellatus plants were collected from Ottawa National Wildlife Refuge

(ONWR) in Oak Harbor, Ohio in late October 2012. Twenty-four hours following collection, all plants were washed to remove algae and herbivorous invertebrates that could damage the leaves.

Senescing leaves were trimmed before the plants were planted in natural clay based cat litter with no chemical additives (Special Kitty Natural Clay Cat Litter). Cat litter was chosen because of its similarity in composition to the clay-based soils of natural marshes. After washing, all plants were vernalized for eleven weeks in a growth chamber at Bowling Green State University.

All plants were stored at 8 ˚C with 8 hour photoperiods and were watered every two days to maintain at least one to two inches of standing water throughout dormancy.

After eleven weeks of vernalization, plants were placed in the Bowling Green State

University greenhouse under supplemental lighting in order to break dormancy. After emergence in early February 2013, plants were again washed and senesced or dead leaves were removed before planting in 11 cm (4.5 in) clear plastic pots containing 300 g of 100% dry clay.

Rhizomes of larger plants that exceeded 6 cm were trimmed in order to allow further growth in pots through the course of the experiment.

After planting, 180 pots were placed into twenty 208 L (55 gallon white) drums lined with two clear plastic bags so that each barrel contained nine plants each (Pictures 4 and 5). An original shallow water depth of 3 cm above sediment was maintained for one week following transplant. This 3 cm water depth was chosen so as to maintain water above the lip of pots in each barrel to ensure sediment saturation due to the lack of drainage from the plastic pots. 8

Picture 4. (Left) Plant arrangement within barrels prior to flooding and clipping.

Picture 5. (Below) Barrel a rrangement in the BGSU greenhouse.

After one week, ten barrels were flooded to a depth of 30 cm ab ove the sediment level while the remaining barrels were left at the shallow depth of 3 cm. These two depths simulated both flooded and shallow water conditions of local wetlands at ONWR and were rand omly assigned to barrels. Plants within each barrel were also randomly assigned to one of th ree clipping treatments (control, single cut, double cut) . Control plants were not cut during the experiment. Single cut plants were cut only once at the 3 rd week of the experime nt and double cut plants were cut alongside single cut plants at the 3 rd week and again at the 7 th week of the experiment. After assignment of water depth and cutting , six groups of cross treatments were created (Table 1). Group one an d two served as control treatments and remained uncut.

However, while group one plants remained at a shallow depth of 3 cm , group two pl ants were subjected to flooded conditions at 30 cm. Group 3 and 4 plants were only cut once at the third week of growth and were separated between shallow and flooded conditions respectively.

Groups 5 and 6 were cut twice. All t hree of the cut treatments w ere represented in a single barrel subjected to one of the two flooding treatments. For optimal growth over the course of the 9

experiment, a twelve-hour photoperiod was applied to replicate early summer light conditions. In

addition, water in each barrel was drained and refilled every other week to limit algal growth that

could alter light levels in the deeper water treatments.

Table 1. Assignment and Treatment Application of Plant Groups

Treatment Description Group 1: Non-flooded Control Water added to 3 cm above sediment Group 2: Flooded Control Water added to 30 cm above sediment Group 3: Single Clipping Non-flooded Water added to 3 cm about sediment and plants clipped once at 3 rd week Group 4: Single Clipping Flooded Water added to 30 cm above sediment and plants cut once at 3 rd week Group 5: Double Clipping Non-flooded Water added to 3 cm above sediment and plants cut at 3 rd and 7 th week Group 6: Double Clipping Flooded Water added to 30 cm above sediment and plants cut at 3 rd and 7 th week

Clipping treatments occurred at the third and seventh week of the experiment. At the

third week, both single cut and double cut plants were clipped. Double cut plants were clipped for a second time at the seventh week of the experiment following regrowth from the initial first cut at the third week. For all clipping treatments plants were cut at a height of 8 cm above sediment level in each pot. Clipped plant material was kept and dried in a 60 °C drying oven for two weeks. After drying, the dried biomass of each clipping was recorded. In addition to removed tissue biomass, the height and width of each remaining leaf was recorded three weeks after each clipping treatment to record the level of regrowth. This height and width was used to approximate the total remaining leaf area of each plant following regrowth from clipping. Only leaves that showed regrowth above the 8 cm cut height were recorded. 10

We wanted to look at the effects of clipping and flooding treatments individually, as well

as at the combined effects of flooding and clipping. We ran a generalized linear model fit to test

for these treatment effects on the leaf area of B. umbellatus plants. We inspected data for normality. A transformation of Log 10 (x + 0.1) was performed to meet the assumptions of normality. Plants showing zero centimeters of regrowth were assumed dead and were not included in the analysis. Initial weight of each plant was also taken into consideration as a covariate to also consider how original starting size may have impacted leaf regrowth. Analyses were performed in JMP 10 (SAS Institute Inc., Cary, NC).

Results:

Clipping treatments had a significant effect on the total area of plant leaves for both single and double cut plants, as clipped plants had less leaf area than unclipped plants (p <

0.0001) (Graph 1). However, the leaf areas between single and double cut plants did not significantly differ (p = 0.3319). Leaf area of control plants in both flooded and unflooded treatments did not significantly differ (p = 0.4655) (Graph 2). A combination of depth and clipping treatments had a significant effect on the average plant area across both cutting treatments, as the clipped plants in cut and flooded treatments which had less leaf area than control plants (p = 0.0315) (Graph 3). However, the differences in leaf area of single and double clipping treated plants were again insignificant even when combined with depth treatment

(p=0.8234). 11

4.5 * Effect of Clipping Alone 4 ) 2 3.5 3 2.5 2 1.5 1 log(total area)leaf (cm 0.5 0 0 1 2 Number of Cuts

Graph 1. Uncut plants had a higher total leaf area than both single and double cut plants.

4.4 Effect of Flooding Alone 4.3 ) 2 4.2 4.1 4 3.9 3.8 3.7 log(total area)leaf (cm 3.6 3.5 3 30 Depth (cm)

Graph 2. Leaf area of flooded plants did not differ significantly from area of non-flooded plants. 12

5 4.5 Effect of Combined Treatments

) 2 4 3.5 3 2.5 2 1.5 1 log(total area)leaf (cm 0.5 0 low high low high low high control 1 cut 2 cut Treatment

Graph 3. Leaf area among cut and flooded plants was lower than the leaf area of uncut plants. The leaf area of double cuts did not differ from single cuts significantly. Discussion:

It was originally thought that plants in the deeper water experiments would have the lowest amount of leaf regrowth at the time of analysis due to the impacts of flooding on dissolved oxygen for root respiration (Kercher and Zedler 2004). However, this experiment suggests that flooding alone is not an effective management method to control B. umbellatus .

Because the growth and resource allocation of many macrophyte species depend on flooding and inundation, it appears that flooded B. umbellatus plants reallocated a majority of their resources to the production of new leaves (Miao and Zou 2012). By producing leaves that reached above the water surface, these plants were able to get sufficient oxygen below to their roots, effectively reducing the impact of inundation on plant growth (Miao and Zou 2012). However, resource allocation of plant nutrients to the production of aboveground biomass is dependent upon the available resources from belowground biomass (Miao and Zou 2012). It is possible that B. umbellatus in this experiment already had adequate belowground biomass that could provide 13 nutrients for the regrowth of plants that exceeded water depth. The adequate belowground biomass that provides B. umbellatus its greater flood tolerance is most likely from the early growth and allocation of resources typical of macrophyte species (Kercher and Zedler 2004).

Further cutting treatments or deeper water depths than those presented here may be needed to better exhaust plants of below ground biomass and oxygen to more effectively limit the growth of above ground tissue.

However, because lower water depth plants were more impacted by clipping treatments, it may be ideal for management at local wetlands to cut or mow B. umbellatus during prescribed or seasonal draw-downs. It should also be noted that given the tapering shape of B. umbellatus leaves, estimates of area were based on the widest area of each leaf resulting in an overestimate of true leaf area. While it is likely that this estimate is close to accurate, could underestimate differences among treatments.

The combined effects of clipping and flooding demonstrate that multifaceted management strategies may be the most successful in preventing the further growth and spread of invasive marsh plants. Due to the lack of difference in the response between single and double cut plants, it also appears that the number of times a plant is cut does not alter response. Rather, the application of a single cut is enough to sufficiently limit the growth of B. umbellatus . As current literature suggests, the success of rapid invasion and expansion of invasive ranges highly favors the allocation of nutrients and resources to the production of reproductive tissues (Brown and Eckert 2005). Further investigation of treatment effects on reproductive material such as seeds and vegetative propagules may prove insightful in how to limit the spread of B. umbellatus into non-invaded wetlands. 14

In addition to altered plant growth following the application of cutting treatments, it was observed that plants that remained uncut throughout the experiment had begun to produce reproductive tissue in the form of vegetative propagules. This observation follows past studies that have suggested that B. umebllatus will only produce reproductive tissues, such as flowers, with the appropriate shallow water depth (Brown and Eckert 2005). While the production of vegetative propagules produced by experimental plants was not statistically analyzed in this experiment, the failure of cut plants to produce reproductive tissue of any kind suggest that clipping may not entirely eradicate an invasive stand of B. umbellatus , but could be efficient in limiting spread to new areas.

Based on the results of this experiment, additional research on the ecology and life history of B. umbellatus is warranted. Further analysis will investigate the changes in overall biomass between varying treatments and may demonstrate better differences between double and single cut plants. Analysis on rhizome size between the beginning and conclusion of this experiment as well as the production of vegetative propagules may also provide additional information on the resource allocation of plants among various treatments.

Literature Cited:

Brown JS, Eckert CG. 2005. Evolutionary increase in sexual and clonal reproductive capacity during biological invasion in an aquatic plant Butomus umbellatus (Butimaceae). American Journal of Botany 92(3):495-502. Kercher SM, Zedler JB. 2004. Multiple disturbances accelerate invasion of reed canary grass (Phalaris arundinacea L.) in a mesocosm study. Oceologia 138:455-465. Kliber A, Eckert CG. 2005. Interaction between founder effect and selection during biological invasion in an aquatic plant. Evolution 59(9):1900-1913. Mayence CE, Marshall DJ, Godfree RC. 2010. Hydrologic and mechanical control for an invasive wetland plants, Juncus ingens, and implications for rehabilitating and managing Murray River floodplain wetlands, Australia. Wetlands Ecology Management 18:717- 730. Miao SL, Zou CB. 2012. Effects on inundation on growth and nutrient allocation of six major macrophytes in the Florida everglades. Ecological Engineering 42:10-18. Parkinson H, Mangold J, Dupuis V, Rice P. 2010. Biology, ecology, and management of flowering rush (Butomus umbellatus): Montana State University Extension . EB0201 Rosaen AL, Grover EA, Spencer CW. 2012. The costs of aquatic invasive species to great lakes states: Anderson Economics Group. Russell IA, Kraaij T. 2008. Effects of cutting Phragmites australis along inundation gradient, with implications for managing reed encroachment in a South African estuarine lake system, Wetlands Ecology Management 16: 383-393. Schultz R, Dibble E. 2012. Effects of invasive macrophytes on freshwater fish and macroinvertebrate communities: the role of invasive plant traits. Hydrobiologia 684: 1- 14. Photo Credits: USDA, NRCS. 2012. The PLANTS Database (http://plants.usda.gov, 18 July 2012). National Plant Data Team, Greensboro, NC 27401-4901 USA.